A 7000-year history of coastal environmental changes from Mexico’s Pacific coast: A multi-proxy record from Laguna Mitla,Guerrero

Abstract

The lack of multi-millennial multi-proxy paleoenvironmental reconstructions from Mexico’s Pacific coast has limited our understanding of the regional response to climate change and sea-level rise. A 479-cm core covering the last 6900 years was extracted from Laguna Mitla in the state of Guerrero on Mexico’s Pacific coast. Beginning as a Rhizophora-dominated salt pan ~6900 yr BP, at ~6500 yr BP, the site transitioned to a mangrove swamp dominated by Laguncularia, which lasted about 300 years. The beach barrier formed from ~6200 to 5200 yr BP, during which time, the site existed as an intermittently sheltered bay, the result of large, rapid changes in wave energy associated with the shifting barrier location and changes in stability. After the beach barrier was stabilized at ~5200 yr BP, water level at the coring site became a function of precipitation rather than sea level. Since that time, deposition has alternated between peat, laid down in a mangrove swamp, and clay intervals characterized by high concentrations of titanium and a predominantly regional pollen signal, representing open-water lagoon phases. Seven periods of increased water level, varying in duration, occurred during the backbarrier period, with El Niño-Southern Oscillation (ENSO) likely the main climatic mechanism causing these periodic shifts in the paleo-precipitation levels. We suggest that the deepest water levels detected over the last ~3200 years correlate with periods of increased ENSO activity. The spatial distribution of tropical cyclone rainfall, which represents a significant percentage of total annual precipitation along Mexico’s Pacific coast, may explain the inconsistencies between our record and paleoclimatic records from Mexico’s interior, but more work is needed to test this hypothesis.

Keywords

El Niño-Southern Oscillation Laguna Mitla mangroves Pacific coast of Mexico paleoclimate paleoenvironmental reconstruction pollen sediment tropical cyclones x-ray fluorescence

Introduction

Paleoenvironmental reconstructions are of great importance for understanding the mechanisms and impacts of climatic changes in the seasonal tropics, where precipitation variability may drive significant changes to biophysical systems (e.g. Baker et al., 2001; Hodell et al., 2005a; Peterson and Haug, 2006, and references therein). Multi-millennial paleoclimatological and paleoenvironmental research has been conducted throughout Mexico, focused in the Sierra Madre highlands (Bernal et al., 2011; Berrio et al., 2006; Figueroa-Rangel et al., 2008; Lachniet et al., 2013; Metcalfe et al., 2010; Roy et al., 2013; Watts and Bradbury, 1982), the Yucatan peninsula (Hodell et al., 2005b; Islebe and Sánchez, 2002), and the arid northern region (Roy et al., 2014). Paleoenvironmental studies conducted along the Pacific mainland margins have provided a glimpse into coastal hazards and geological histories (Goman et al., 2005, 2010; Gonzalez-Quintero, 1980; Ramírez-Herrera et al., 2007, 2009, 2012), but they have not focused on precipitation changes. Given the general aridity of Mexico’s Pacific coast and the important role precipitation plays in controlling vegetation patterns, the paucity of records presents a serious obstacle to understanding the regional environmental history.

Here, we present a ~7000-year multi-proxy paleoenvironmental reconstruction from Laguna Mitla, located ~40 km west-northwest (WNW) of Acapulco, Guerrero, which reveals significant regional paleoclimatological and hydrological change throughout the late-Holocene. The analyses of sediment cores previously extracted from Laguna Mitla (Ramírez-Herrera et al., 2007, 2009) focused on lagoon formation and marine incursion events. Our study focuses instead on the geomorphic and biological changes that occurred within this basin in response to precipitation changes, sea-level rise, and the occurrence of extreme events in this climatically sensitive region.

Geographical setting

Mexico is physiographically dominated by the Central Mexican Plateau. It is sandwiched between two longitudinally oriented mountain ranges: the Sierra Madre Occidental to the west and the Sierra Madre Oriental to the east. The Sierra Madre del Sur mountain range lies along the Pacific coast, south of the Sierra Madre Occidental.

Mexico lies on the western edge of the North American Plate, overriding the oceanic Cocos Plate (Pardo and Suárez, 1995). The subduction zone is marked by the Middle America Trench, located <100 km offshore of the Pacific coast. The tectonic setting results in a narrow continental shelf, steep topographic relief, and a large sediment supply driven by longshore currents. The associated coastal deformation is highly variable along the southern section of Mexico’s Pacific coast. For example, the Jalisco coast, to the north of our study region, is subject to slow, continuous (interseismic) uplift and episodic coseismic subsidence, resulting in overall uplift, as evidenced by the large number of marine terraces and wave-cut platforms (Ramírez-Herrera et al., 2004). To the south in Guerrero, the situation is the reverse with interseismic subsidence predominating over coseismic uplift, resulting in an estimated long-term subsidence rate of 0.3–0.4 cm/yr (Ramírez-Herrera et al., 2011), with slow-slip events also common (Alva and Kostoglodov, 2007; Cavalié et al., 2013). Rare larger megathrust events are posited as resulting in subsidence (Ramírez-Herrera et al., 2011).

Common features of the Pacific coastal plain include prograding beach ridges (Curray et al., 1969; Ramírez-Herrera and Urrutia-Fucugauchi, 1999), relict Pleistocene beach ridges farther inland (Curray et al., 1969; Lankford, 1977), and barred inner-shelf lagoons (Lankford, 1977). The latter typically formed ~5000–6000 years ago when the rate of sea-level rise slowed and coastal basins, cut off from the sea by the formation of wide beach barriers, filled with meteoritic water (Lankford, 1977). Typically these lagoons, fed by short streams originating in the highlands, connect to the sea through a single, narrow outlet channel (Lankford, 1977). Because of the large sediment supply, these channels usually close seasonally and often for much longer periods. Water level in these barred inner-shelf lagoons is often higher than sea level. The connections to the sea can open, often explosively, when rising water levels exert pressure on the inlet, eventually breaching the beach barrier. Such events can result from sudden increases in river discharge following large precipitation events (Lankford, 1977).

Regional vegetation varies with altitude. Pine-oak forests dominate the wet, higher elevations, with tropical deciduous forest at lower elevations and thorn forest, shrubland, and grassland communities at the lowest elevations (CANABIO, 2008; Leopold, 1950). Mangrove swamps, mainly consisting of red mangrove (Rhizophora mangle), are common along the coast.

Western Mexico mostly experiences a tropical wet/dry climate (Aw in the Köppen Climate Classification; Blouet, 2002). Dry westerly flow dominates much of Mexico’s Pacific coast and interior during the winter (Blouet, 2002; Mosiño-Aleman and Garcia, 1974). Precipitation occurs mainly during summer when the Intertropical Convergence Zone (ITCZ) moves to the north, with some rainfall, originating from the eastern North Pacific (ENP), contributing to the predominant flow from the Gulf of Mexico (Blouet, 2002; Metcalfe et al., 2000; Mosiño-Aleman and Garcia, 1974). Rain shadows are prevalent throughout central and western Mexico.

Averaging about 15 named storms annually (National Hurricane Center, 2011), the ENP is one of the most active basins for tropical cyclones (TCs). The ENP TC season spans May to November (Jáuregui, 2003), with most storms originating south of the mainland and tracking to the northwest before dissipating in the central Pacific Ocean. Direct TC landfalls are rare along western Mexico south of the Baja Peninsula, as most storms, deflected to the west by the Sierra Madre del Sur, assume coast-parallel tracks (Zavala Sansón, 2004). The impact of TCs on regional climatology cannot be overstated, with their associated rainfall contributing ~40% of coastal Guerrero’s warm season totals, decreasing to 9% in Guerrero’s interior (Englehart and Douglas, 2001). El Niño periods tend to increase ENP TC frequency (Jáuregui, 2003; Jien et al., 2015), intensity (Jin et al., 2014), and wetness, largely from the influence of the monsoon trough and warm sea surface temperatures (Rodgers et al., 2000). Larson et al. (2005) indicated that TCs bring more precipitation to western Mexico (including Guerrero) during El Niño periods than during La Niña periods.

Study site

Guerrero occupies the section of the Pacific coast from ~16.3°Ν to 17.9°N, draining the western and southern edges of the Sierra Madre del Sur, which reach elevations of 3500 m (Figure 1). The climate is tropical monsoonal with a rainy season from May to October. Coastal precipitation is ~1300 mm/yr (National Meteorological Service of Mexico, 2015), increasing landward orographically up the western face of the Sierra Madre del Sur until reaching the first ridge, which attains elevations of 2800 m. Several lagoons occur along the narrow coastal plain, including the irregularly shaped Laguna Mitla (Figure 1), ~22 km long with a maximum width of ~4 km. Water depth in the western end of the lagoon varies from 0.5 to 1 m, with a surface level an estimated 1 m above mean sea level (MSL). Fluvial input is from several perennial streams draining the Sierra Madre del Sur. A narrow former tidal channel exists at its extreme eastern tip. Linear islands Magueyes and El Conejo and several submerged sandy shoals occur in the western end of the lagoon. They represent remnant beach ridges (Ramírez-Herrera et al., 2007, 2009) from a Pleistocene highstand, common features on Mexico’s Pacific coast (Lankford, 1977). The lagoon is oligohaline, with salinity levels ~3.5 ppt (Contreras-Espinosa and Warner, 2004). Laguna Mitla is fringed by mangroves and Typha (cattail) patches, with deciduous thorn forest at slightly higher elevations to the north. The lagoon is separated from the ocean by a ~900-m-wide sandy beach barrier, consisting of 16 beach ridges (Ramírez-Herrera et al., 2007), some reaching ~8–10 m in height. A 194-cm core (ACA-04-01) extracted from the narrow mangrove swamp located between the southern, seaward edge of Laguna Mitla and the beach ridge plain yielded a date of 3166 ¹⁴C yr BP (calibrated age range: ~3300–~3450 yr BP) from 155 cm depth (Ramírez-Herrera et al., 2007). The dominance of sand in the core suggests deposition from the beach ridges (Ramírez-Herrera et al., 2007), with ~3300 yr BP a conservative estimate of the onset of beach ridge construction, given the ~39 cm of sand located below the radiocarbon sample.

Figure 1.

Study site. Basemap of the study region ((a) Google Earth, Landsat) showing the locations of Laguna Mitla and the Sierra Madre del Sur Mountains in the state of Guerrero. Laguna Mitla is shown ((b) Google Earth, Digital Globe), with the location of the coring site (core 5) pinpointed. Islands Magueyes (M) and El Conejo (EC) are labeled.

Methods

Here, we present a multi-proxy examination of a 479-cm core (Mitla 5) extracted ~10 m from the western, seaward edge of the lagoon at a water depth of ~1 m in December, 2009 (Figure 1). Core 5 consists of twelve ~50-cm Russian peat borer segments, with ~10 cm overlap between segments.

In the laboratory, core stratigraphy and physical characteristics were described. Water content and percentages of organics, carbonates and residual (mainly siliciclastics) were determined by loss on ignition (LOI; Liu and Fearn, 2000) at 1-cm resolution. Elemental concentrations were determined at 2-cm resolution using a Delta Innov-X handheld x-ray fluorescence (XRF) unit. This device contains a factory calibration (Compton Normalization) to provide accurate output; validation was certified by standards NIST 2702 and 2781. The concentrations of 27 elements were measured: here, we display the concentrations of titanium (Ti), chlorine (Cl), strontium (Sr), calcium (Ca), and sulfur (S). A total of nine organic samples were sent to the NOSAMS laboratory at the Woods Hole Oceanographic Institution for accelerator mass spectrometry (AMS) radiocarbon dating, with the results calibrated to calendar years by Calib 7.0 (Stuiver and Reimer, 1993) and the IntCal 13 curve (Reimer et al., 2013; Table 1). A total of 48 samples taken at ~10-cm intervals were processed for pollen, fungal spores, charcoal, and dinoflagellates using standard procedures (Faegri and Iversen, 1989). At least 300 pollen grains were counted for each sample. All pollen taxa were included in the pollen sum used to determine pollen percentages; a total pollen sum (instead of an arboreal or terrestrial-only pollen sum) is justified because the pollen assemblages are dominated by mangrove and wetland pollen taxa with only relatively minor contributions from upland forest types. A reference collection at the Global Change and Coastal Paleoecology Laboratory (Louisiana State University) was consulted to aid identification, in addition to many publications (Lozano-García and Martínez-Hernández, 1990; Roubik and Moreno, 1991; Van Geel et al., 2003, 2011; Willard et al., 2004). All fungal spore species were grouped into a single category. Charcoal fragments >10 µm were totaled for each sample until at least 300 pollen grains were counted (Liu et al., 2008).

Table 1.

Radiocarbon dating results for core 5. Radiocarbon years were calibrated to calendar years using Calib 7.0 (Stuiver and Reimer, 1993) and IntCal 13 curve (Reimer et al., 2013).

Sample ID	Depth	Laboratory no.	Material	¹⁴C_age	Cal. BP (2σ)	Relative area under probability distribution	Median probability
Mitla 5AAA 2	2	OS-90761	Plant/wood	>Modern	N/A	N/A	N/A
Mitla 5AAA 34	34	OS-90770	Plant/wood	1840 ± 45	1629–1655	0.039	1775
					1660–1666	0.005
					1693–1880	0.955
Mitla 5 61	61	OS-84452	Plant/wood	2710 ± 110	2491–2602	0.061	2836
					2607–2642	0.019
					2678–3083	0.896
					3090–3145	0.022
					3152–3155	0.001
Mitla 5 125	125	OS-83891	Plant/wood	3160 ± 30	3273–3284	0.026	3388
Mitla 5 125	125	OS-83891	Plant/wood	3160 ± 30	3340–3450	0.974	3388
Mitla 5 207	207	OS-83892	Plant/wood	4070 ± 30	4440–4486	0.159	4558
					4499–4644	0.704
					4677–4692	0.015
					4761–4800	0.122
Mitla 5 290	290	OS-83942	Plant/wood	4500 ± 30	5046–5205	0.65	5167
Mitla 5 290	290	OS-83942	Plant/wood	4500 ± 30	5210–5296	0.35	5167
Mitla 5H 45	324	OS-93066	Plant/wood	4760 ± 30	5333–5348	0.041	5521
					5353–5371	0.029
					5463–5587	0.93
Mitla 5 414	414	OS-83941	Plant/wood	6050 ± 35	6795–6990	1	6903
Mitla 5L 44	475	OS-90763	Plant/wood	6000 ± 60	6678–6708	0.031	6842
Mitla 5L 44	475	OS-90763	Plant/wood	6000 ± 60	6712–6984	0.969	6842

Results

Above an 8-cm basal sand layer, core 5 is predominately dark muddy peat (~85% water content, ~80% organics, ~2–3% carbonates, and ~10–40% residual material; Figure 2). The peat is interbedded with clastic layers of varying thicknesses. The clastic layers have lower percentages of water (~40–80%) and organics (~10–50%) and higher percentages of residues (~50–85%) than the peat. Core 5 is divided into six stratigraphic zones on the basis of LOI, XRF, and palynological analyses (Figures 2 and 3). The core chronology was developed from the median probability value for each sample (Telford et al., 2004). The date of 6050 ± 35 ¹⁴C yr BP at 414 cm was rejected as it overlaps with the date of 6000 ± 60 ¹⁴C yr BP at 475 cm.

Figure 2.

Sedimentary data. Displayed is the core litholog (left), with LOI curves showing percentage of water (wet weight), organics (dry weight), carbonates (dry weight), and residual (dry weight; middle). Radiocarbon dates for nine organic samples are displayed in their stratigraphic position (middle left). Sedimentation rates between each pair of dates were calculated on the assumption of constant sedimentation between dates. A total of six zones were identified based on sedimentary, microfossil, and elemental analyses (left).

Figure 3.

Elemental concentrations and microfossils. The concentration of selected elements (middle), microfossils (right), and inferred calendar ages of the zone boundaries (far right) is displayed next to the LOI curves (left). Median calendar dates for the nine radiocarbon samples are displayed on the far left. Pollen types are graphed as percentage of total pollen. Pollen concentration, charcoal, fungal spores, and dinoflagellates are graphed as grains or particles per cubic centimeter. A 5× exaggeration line was added to the charcoal and pollen concentration curves.

A total of 27 pollen taxa were identified, with mangrove pollen taxa (Rhizophora, Laguncularia, and Conocarpus) dominating most samples. Pinus, Quercus, Ulmus, Alnus, and Ostrya/Carpinus, grouped together as ‘Highland Forest’, occur in relatively high percentages at several levels. Because of their relatively low percentages when compared with other taxa, a ‘Lowland Tropical Forest’ group was created that consists of Flacourtiaceae, Verbenaceae, Anacardiaceae, Labiatae, Euphorbiaceae, Fabaceae, Rutaceae, Burseraceae, Apocynaceae, Guttiferae, and Moraceae-Urticaceae, along with the group Sapindaceae/Loranthaceae which contains morphological similarities. Amaranthaceae, Poaceae, Ambrosia, and Asteraceae were grouped as ‘Disturbance’ taxa. With the exception of an episodically occurring granulate-tricolporate (GTC) grain that achieves 27% in one sample at 24 cm, unknown taxa were <1% of the pollen sum. Charcoal concentrations were generally stable except for a large spike at 381 cm.

Zone 6 (479–441 cm)

The bottom 8 cm (479–472 cm) of the core consists of gray sand (Figure 2). At 471 cm, peat abruptly replaces the sand, with organic content increasing from 13% to 36% within 3 cm. Above 465 cm, the sediments become darker, marked by increases in organic and water contents, reaching 81% and 50%, respectively, at the top of the zone. Carbonate contents are <5% throughout the zone. The concentrations of most elements, which are relatively high for the bottom sand layer, gradually decrease above the sand throughout the zone, although Cl concentrations remain relatively elevated (Figure 3). Rhizophora dominates much of zone 6, ranging from 32% to 88% (Figure 3). Fungal spore concentrations are relatively high, at over 5900 spores per cubic centimeter.

Zone 5 (440–401 cm)

Organic and water contents increase upcore in the lower part of this interval, with organics reaching a maximum value of 87% at 415 cm before dropping to <70% at the top of the zone. Carbonate contents are <4%. Compared with zone 6, concentrations drop for most elements, reaching negligible values at ~425 cm. An exception is Cl, which decreases only gradually until ~425 cm, above which the concentration increases through the top of the zone. Laguncularia gradually replaces Rhizophora within this zone, increasing from 10% to 40% as Rhizophora decreases from 48% to 20%. Fungal spore concentrations range from 10,000 to 41,000 spores per cubic centimeter.

Zone 4 (400–291 cm)

Zone 4 is dominated by a black, woody peat interrupted by multiple clastic bands. Clastic layers vary in thickness (<1 to ~15 cm), organic percentages, and color, characterized by light brown clay in the bottom half of the zone, transitioning to bluish-gray clay at ~340 cm. The clastic/peat contacts are abrupt, with organic percentages typically fluctuating by ~20% within a single centimeter at the interfaces and by >30% at three levels (356–355, 319–318, and 292–291 cm). Carbonate percentages spike at 394–392 cm (6%), 327–325 cm (14%), 309–308 cm (8%), and 294–292 cm (14%). Ti concentrations display episodic secular increases in the lower part of the zone but decrease rapidly above 315 cm. Sr, Ca, and S concentrations increase dramatically at ~330–291 cm, matching carbonate spikes.

Zone 4 contains minimum and maximum Rhizophora percentages (0–90%), which display an inverse relationship with Laguncularia (50% at 303 cm) and Conocarpus (60% at 353 cm). Spikes exist in Batis and Typha, mostly from 391 to 340 cm during a period of dominance by Conocarpus. Compared with percentages throughout zones 5 and 6, Highland Forest taxa percentages are relatively moderate and stable. Total pollen concentration fluctuates dramatically, spiking at 381 cm. Fungal spore concentrations are highly variable. Dinoflagellate concentrations spike at the top of zone 4, extending into the bottom of zone 3.

Zone 3 (290–211 cm)

Zone 3 is dominated by a highly decomposed black peat at 290–230 cm. Above ~230 cm, the material becomes more fibrous, and some large plant fragments are present. Water and organic contents vary from 77% to 91% and 58% to 87%, respectively, while carbonates never exceed 3.9%. The interval from 273 to 262 cm exhibits slightly lower water and organic contents than the remaining zone. Most elemental concentrations are relatively low in this zone compared with zones 4–6.

Rhizophora and Laguncularia dominate zone 3. At the base of the zone, Rhizophora percentages increase rapidly to 88% while Laguncularia drops to <1%. From 270 to 220 cm, Laguncularia percentages increase to >30% as Rhizophora drops to <40%. At the zone top, this trend reverses, with Rhizophora at 69% and Laguncularia at 10%. Highland Forest taxa percentages are relatively high and stable when compared with zones 4–6, while Batis percentages are low, peaking at 2% of the total pollen sum. Fungal spore concentrations are moderate when compared with zones 4–6, decreasing upcore in zone 3.

Zone 2 (210–101 cm)

Zone 2 contains black woody peat interrupted by four clastic layers with lower water and organic percentages. The two thin clastic intervals (166–163 and 174–173 cm), marked by sharp bottom and top contacts, consist of brown clay; the two thicker intervals (143–132 and 198–187 cm) are brown-gray clayey peat. Elemental concentrations are generally low in zone 2 compared with zones 3–6, with low spikes occurring largely in or adjacent to the clastic intervals. Cl concentrations decrease throughout zone 2. Rhizophora and Laguncularia dominate the peat. Conocarpus, Batis, Typha, fern spores, and pollen concentrations spike in the clay intervals. Compared with zones 3–6, the Highland Forest percentages are moderate and the fungal spore concentrations are low, with small spikes occurring in peat intervals. Charcoal achieves its highest abundance for the core in the clay layer at 140 cm.

Zone 1 (100–0 cm)

Zone 1 contains black woody peat interrupted by three thick tan-gray clay intervals at 99–89, 79–65, and 32–0 cm. Ti concentration spikes in the clay bands while dropping to background levels in the peat sections. Combined percentages of Rhizophora and Laguncularia are distinctly lower than in other zones. Within this zone, mangrove percentages are relatively high in the peat section but very low in the clay sections, the latter being generally dominated by Batis and Typha along with the Disturbance taxa. Fungal spores are either absent or rare throughout this zone.

Discussion

Paleoenvironmental reconstruction

This ~7000-year record yields one of the longest and best-dated paleoenvironmental records from Mexico’s Pacific coast. The core 5 stratigraphy, ranging from basal sand, blue-gray clay layers, peat, thin brown clay layers, clayey peat, and thick clay deposits, indicates a dynamic, rapidly changing physical environment. The ability to differentiate sediments of marine and terrestrial origin using XRF data is of particular importance for this study. Titanium is commonly analyzed to detect sediments of terrestrial origin (Kirwan et al., 2011) and is a reliable proxy for determining runoff (Haug et al., 2001; Woodruff et al., 2008). Elements associated with marine or high salinity conditions include Sr, Ca, Cl, and S (Chen et al., 1997; Nichol et al., 2007; Ramírez-Herrera et al., 2007; Schofield et al., 2010). Marine intrusions from any number of causes (TC storm surge, tsunami runup, collapse/breaching of the beach barrier) would result in the deposition of marine material, commonly characterized by lower water and organic matter contents relative to limnic or lagoonal sediments; higher carbonate content and Sr, Ca, Cl, and S concentrations; and the presence or increased abundance of marine microfossils such as broken marine shells, dinoflagellates, and foraminifera tests.

In paleoenvironmental reconstructions, the relative elevational differences between the coring site and contemporaneous sea level are important. However, the sea-level history for the Pacific coast of Mexico is not well constrained, consisting of only a single curve from Nayarit produced >45 years ago (Curray et al., 1969). This record indicates a 40-m sea-level rise over the last 10,000 years, with the rate beginning to slow ~7500 yr BP when sea level was ~14 m below present. The rate of rise is shown to have further slowed over time, with only ~1 m of rise occurring over the last 3000 years. The changing relationship between sea level and the elevation of the core top, which is a control over sedimentation at the core 5 site, is presented in Figure 4. While we recognize the uncertainties associated with this curve, it is in general agreement with relative sea-level studies from the central and southern coasts of California, all of which lack evidence of highstands over the last 10,000 years (Engelhart et al., 2015; Reynolds and Simms, 2015). Therefore, acknowledging the possible errors, we present the Nayarit curve as the best data currently available for estimating millennial-scale regional sea level, particularly as the curve provides useful information concerning the onset of sea-level stabilization. While slow-slip events are capable of causing long-term uplift or subsidence throughout Guerrero (Alva and Kostoglodov, 2007), there is no evidence to support the occurrence of sudden or short-term vertical movements for this 7000-year record, especially since the presence of mangroves throughout the core suggests that this site has been at or near sea level throughout this period.

Figure 4.

Summary diagram. Proxy information is aligned along a timescale displaying the regional sea-level curve (Curray et al., 1969; dark curve), and the depth of the core top (dashed curve) plotted against present sea level, and LOI curves. The sea-level curve is shown with a thick and fuzzy line because of its potential imprecision from its relatively old age. The concentrations of Ti, a representative terrestrial metal, are shown below, followed by the pollen percentages (shown as bars) of four important taxa (Batis, Conocarpus, Laguncularia, and Rhizophora), and the concentration of fungal spores.

Saline pond/salt pan (zone 6, 6880–6540 yr BP)

Although the Nayarit sea-level record suggests that the site was several meters above sea level during this period (Figure 4), the pollen assemblage is dominated by Rhizophora, which typically occurs at or near sea level. The low organic content indicates a sparsely vegetated environment, suggesting that the site may have existed as an isolated saline pond or salt pan, perched a few meters above sea level and fringed by patchy clumps of mangroves. Such elevated mangrove environments occasionally occur in arid locations, even without a clear connection to the sea (Spaulding et al., 2010). The high concentrations of Ti suggest runoff of meteoric water, which when ponded under highly evaporative conditions would become increasingly saline, perhaps augmented by occasional marine flooding events or sea spray. An alternative explanation, that an earthquake uplifted the then-coastal site at a later date, is possible, although no sedimentological evidence for such an event exists within the core.

Mangrove swamp (zone 5, 6540–6180 yr BP)

A decrease in Rhizophora, coupled with the increase in Laguncularia, which commonly occupies less saline environments (Hogarth, 2007), indicates a decrease in salinity. Increases in organic contents indicate a lusher and/or more extensive swamp, perhaps resulting from the colonization of surrounding mudflats by Languncularia. An alternative possibility is that raising water levels resulted in the better preservation of organic material. The decrease in Ti concentrations suggests that this shift in mangrove type did not necessarily result from increased precipitation, which would have been marked by increased concentrations of Ti because of increased slopewash. These fresher conditions likely represent a rise in the water table, driven upward as a freshwater lens by the rapidly rising sea level. This environmental change has been observed in many coastal locations (Hatcher et al., 1982; McCloskey and Liu, 2013; Peros et al., 2007).

Coastal beach ridge development/sheltered bay (zone 4, 6180–5170 yr BP)

At the beginning of the zone, when the core top was >2 m above MSL (Figure 4), eroded terrestrial material was probably washed into the basin and drained out to the sea, leaving relatively indistinct terrestrial signals marked by increases in clay content and Ti concentrations. Simultaneously, the rising sea level began to transport both offshore and longshore sediments landward, isolating the coring site behind the emerging beach barrier. This probably led to a less saline environment and distinct terrestrial signatures, as both runoff and eroded materials became trapped within the basin. This is supported by a decrease in Rhizophora percentages and an increase in Conocarpus at the zone bottom. The continuous environmental change exhibited by the highly fluctuating LOI contents suggests that these barriers were low and probably highly variable, both spatially and temporally. These dynamic coastal processes and the semi-continuous creation and destruction of mudflats likely explain the spikes in Batis, a plant commonly found following disturbances in mangrove swamps (Lonard et al., 2011; Rogers, 1982).

By ~5600 yr BP, sea level began directly influencing deposition at the coring site, marked by gradual increases in marine element concentrations, carbonates, and Rhizophora, the latter likely representing mangrove expansion into the lower, marine-flooded sections of the basin (Figures 3 and 4). At this point, the clastic layers became increasingly mixed, consisting of bluish-gray marine clay and dark peat, with Ti concentrations dropping dramatically while marine elements (Sr, Ca, and S) reach their highest concentrations in the core. At the zone top, clay bands with dinoflagellates and high concentrations of marine elements indicate the occurrence of marine intrusions, possibly TC or tsunami driven, which due to the rising sea level were able to inundate the core 5 site. Similar marine environments during this time frame have been documented from the Pacific coast in Guatemala (Neff et al., 2006) and the state of Nayarit (Curray et al., 1969)

After these intrusions, the concentration of marine elements decreases, while Laguncularia increases at the expense of Rhizophora. Stabilizing sea level and the consequent consolidation of the beach barrier, beginning ~5200 yr BP, resulted in a lower energy environment and a highly productive wetland. A similar, though earlier (7300–7200 yr BP) transition, from subareal to backbarrier environment, has been documented from coastal Sonora (Caballero et al., 2005).

Backbarrier mangrove swamp (zone 3, 5170–4580 yr BP)

The beach barrier continued to grow, prograding seaward and likely increasing in height. Because of the increasing strength of the beach barrier, the site was disconnected from the sea and became a backbarrier environment. Ramírez-Herrera et al. (2007) suggested that Laguna Mitla was a lagoon beginning at least ~5300 yr BP, because of fresh to brackish peat with a radiocarbon date of 4630 ¹⁴C yr BP in core ACA-04-06, extracted 5 km from the coast. The Rhizophora-rich and Laguncularia-rich peat indicates a mangrove swamp, benefitting from the reduced wave energy associated with the protection provided by the barrier. Rhizophora dominated at the beginning of this zone because of the higher salinity of the trapped marine water. By ~4800 yr BP, the swamp began to freshen, with Laguncularia increasing at the expense of Rhizophora, likely because of increased precipitation, possibly augmented by the final closing of all tidal channels and the consolidation of the beach barrier. The high concentrations of fungal spores suggest organic decomposition, further indicating minimal water levels (e.g. Berrio et al., 2006). The presence of dinoflagellates at ~5100 yr BP suggests the existence of small and/or intermittent tidal channels, or alternatively, colonization from a variety of factors, such as specie reproduction (Kremp and Heiskanen, 1999), or dispersals by animals and/or wind. Overall, the lack of marine sand, low concentrations of marine elements, and diminished dinoflagellate concentrations through much of zone 3 suggests a robust coastal barrier, and decreased marine influence, likely the result of the reduction in the rate of sea-level rise.

Mangrove swamp/shallow lagoons (zone 2, 4580–3180 yr BP)

Two distinct sediment types occur in zone 2. The first is a high-organic peat with minimal Ti concentrations and high abundance of mangrove pollen, which we interpret as representing a Rhizophora-rich and Laguncularia-rich swamp, similar to that occurring in zone 3. The second sediment type is a brown-gray clay, either in the form of thick clayey peat layers or thin clay layers, both of which are characterized by moderate Ti concentrations, decreased percentages of Rhizophora and Laguncularia pollen, diminished fungal spore concentrations, and spikes in both local pollen (e.g. Conocarpus and Batis) and regional pollen (e.g. Pinus; Figure 3). We interpret the four clay intervals as indicating periods of higher backbarrier water levels, marking transformations from a swamp into either a deeply flooded wetland (thick clayey peat) or a brackish/oligohaline lagoon (thin clay layers). The peaks in pollen concentration in the clastic sections suggest that they are not instantaneous, high-energy sedimentary deposits, as these are often characterized by low pollen concentration (Liu et al., 2011). The increase in pollen of Batis, a genus commonly found in open wetlands adjacent to tropical lagoons (Peros et al., 2007), indicates extensive marshes and swamps along the lagoon fringe, with the pollen likely delivered to the site via the ephemeral stream in the northwest corner of the lagoon, or by lagoon currents as increasing water levels encroached on Batis communities on the remnant beach ridges.

Because the two thick clayey peat layers contain in situ plant fibers, water depths were likely shallower during these intervals than during the periods covered by the two thin clay layers. The deeper water episodes, each lasting ~200 years, occurring at ~4430–4270 and ~3680–3490 yr BP, bracket two shorter open-water periods (centered at ~4080 and ~3950 yr BP). All four of these intervals likely represent times of increased precipitation into a closed backbarrier system with no tidal connection to the sea. The ends of three of the lagoon phases are marked by small spikes in marine elements (Ca and S), perhaps suggesting that the lagoon phases terminated with explosive blowouts, followed by short-lived marine incursions. Alternatively, these small spikes could indicate increased salinity during a period of extreme aridity and/or high evaporation (e.g. Neff et al., 2006). Decreased salinity is evidenced by the general decline in Cl and the increase in Conocarpus throughout the zone. High charcoal concentration and increases in the Disturbance taxa in the top clastic layer (3680–3490 yr BP) are likely related to anthropogenic burning and land degradation from increased human populations, as has been documented from the Michoacán uplands at this time (Watts and Bradbury, 1982).

Mangrove swamp/mature lagoons (zone 1, 3180 yr BP–present)

Zone 1 also consists of alternating peat and clay intervals. Ti concentrations are especially high in the clay intervals, which are thicker and more distinct than in zone 2. The geochemical signatures of these layers resemble those occurring in the lower clastic bands in zone 4 (terrestrially derived), while the upper bands (marine derived) show the reverse pattern. As in zone 2, we interpret these two sediment types as representing swamp (peat) and open-water lagoon (clay) environments. The high pollen concentrations in the clay layers suggest that they did not result from instantaneous, high-energy events. The lagoon phases in zone 1 appear to be the freshest in the backbarrier record, as Cl concentrations plummet while Typha and Laguncularia increase. In the top clay unit, the large percentages of Highland Forest taxa represent the regional pollen signal, which typically exhibit a positive relationship with water body size (Peros et al., 2007). The three lagoon phases, occurring during ~3170–3080, 2990–2870, and 1680–present, are generally longer and more pronounced than in zone 2, probably because of a wetter climate and the increasing stability of the beach barrier. The lack of plant macrofossils or in situ peat in these layers suggests deeper water levels than during the zone 2 lagoon phases, likely the result of a wider and stronger beach barrier. Comparing core depth with the sea-level curve indicates that the elevation of the land surface (core top) may have dropped by ~1 m relative to sea level during zone 2 (Figure 4). This implies that over time, more extensive and/or intense dry periods would be required to drive the lagoon-to-swamp transitions. It should be noted that because of the position of core 5 near the edge of the present lagoon, the lagoon-to-swamp transitions do not necessarily indicate drying of the entire lagoon or that all lagoon phases identified in our core covered the same spatial extent.

The increase in pollen percentages for the Disturbance taxa and elevated Ti concentrations in the clastic layers probably records anthropogenic activity, possibly land clearing, occurring within the watershed during this period. This corresponds with evidence of increasing land use and disturbance through the late-Holocene in the highlands of Michoacán (Watts and Bradbury, 1982), Guerrero (Berrio et al., 2006), and Oaxaca (Joyce and Mueller, 1992). Other indicators include the presence of dinoflagellates, possibly resulting from human-induced algal blooms (Pospelova et al., 2002; Watts and Bradbury, 1982). Unidentified microorganisms discovered in recent sediments from Laguna Tetitlan, 10 km west of Mitla, have similarly been attributed to human activities (Gonzalez-Quintero, 1980).

Tectonic controls

Because Laguna Mitla is located near a convergent plate boundary over a subduction zone, earthquakes are frequent and have the potential to have caused sudden land-level changes at the site, thereby complicating the identification of the climatic signal (e.g. Caffrey et al., 2015). This raises the question as to whether regional tectonics may have been a significant driver of environmental and sedimentary changes at Laguna Mitla, and that the lagoon phases may have resulted from tectonically driven subsidence, during which earthquakes, by abruptly dropping the surface at the coring site, transformed the swamp into a lagoon. Under this scenario, interseismic uplift would have gradually raised the surface back to the original height, driving a slow lagoon-to-swamp transformation.

In core ACA-04-06, extracted on the landward edge of Laguna Mitla and ~5 km inland, Ramírez-Herrera et al. (2007) documented a 60-cm-thick interval of interbedded sands and clay/mud marked by high concentrations of marine elements, overlain by at least 45 cm of bluish silt and clay rich in sulfides and carbonates. The authors interpreted the lower sand and clay/mud layer as resulting from a sudden rise in relative sea level, likely a tsunami at ~3400 yr BP (Ramírez-Herrera et al., 2011), with the overlying blue silts and clays indicating a post-event marine environment. The stratigraphy of a second core, ACA-03-02, also from the northern edge of the lagoon, ~4 km inland, contained a ~93-cm section dominated by relatively organic-poor clays and silts containing shells and increased marine element contents, characteristics that were attributed to a marine intrusion associated with coastal subsidence ~3400 yr BP (Ramírez-Herrera et al., 2009). We find these interpretations unlikely for a number of reasons. In regard to this specific event, there is no matching clastic deposit in core 5, which is located closer to the ocean than the cores from the northern edge of the lagoon. Although tsunamis can occasionally result in areas of erosion found seaward of the area of deposition (e.g. Richmond et al., 2012), it seems improbable that such a process would be sufficient to explain the lack of large-grained clastics at our coring site, only ~ 1 km from the sea. A possible explanation is that these clastic units found north of the lagoon may have originated from interior sources and been deposited by processes unrelated to extreme events (Bianchette et al., 2016). More broadly, the posited regional tectonic processes (coseismic subsidence and interseismic uplift) are the reverse of that exhibited by historical events, which are characterized by coseismic uplift and interseismic subsidence. This inconsistency was suggested to have resulted from a megathrust earthquake so much larger than normal that it ruptured the plate interface, thereby reversing the normal process (Ramírez-Herrera et al., 2009, 2011), but no evidence of such a cataclysmic event is found in any of our stratigraphic records from Laguna Mitla (Bianchette et al., 2016).

In addition, the clay to peat transitions in core 5 are likely too frequent to be explained by tectonic activity. The topmost peat at the coring site, which was likely at or near water level when being formed, is now covered by 30 cm of clay and 1 m of water, meaning that an earthquake capable of producing ~130 cm of vertical movement would be needed to explain its present elevation. According to Ortiz et al. (2000), two historic magnitude 7 earthquakes generated only 15 ± 3 and 7 ± 3 cm of vertical displacement at Acapulco. The occurrence of seven events, genetically different from and larger than any historic events over the last 5200 years seems unlikely. This is supported by the lack of clear sedimentary evidence (erosional unconformities, reworked sediments, sand layers, and/or mud caps; Atwater, 1987; Cochran et al., 2006; Darienzo and Peterson, 1990; Garduño-Monroy et al., 2011) that such large and frequent events should have left in core 5.

The secular trend toward deeper, fresher, and long-lasting lagoon phases over the last 5200 years also argues against tectonically driven vertical movements as the mechanism forcing the sedimentary changes. Since the beach barrier has likely been widening, the expected pattern for marine-generated clastic layers (resulting from either tsunami waves or subsidence) would have been thicker and more frequent deposition during the early stage of the backbarrier record, becoming thinner and/or less frequent toward the present. Since core 5 reveals the opposite pattern, with thicker clastic deposition toward the core top, this further suggests that these layers are controlled by backbarrier water depth, rather than as marine deposits formed by tectonically driven vertical movements.

Lagoonal blowouts as geomorphic controls on sedimentary record

One control over lagoon water level is the propensity for blowouts when the pressure exerted by high water levels within surpasses the structural strength of the beach barrier. These blowouts, a common occurrence for regional barred inner-shelf lagoons (Lankford, 1977), are capable of drastically lowering surface levels if the lagoon surface is above MSL, and emptying a lagoon if the lagoon bottom is above MSL. The return intervals of such events can vary, depending on barrier morphology, size of catchment basin, vegetation cover, and the amount and intensity of rainfall. However, the regional sediment supply is so large that such breaches can close within weeks (e.g. Cochran et al., 2009). The sedimentary effects of such events may also vary. During wet periods, such breaches may leave little or no sedimentary signal, as freshwater lagoon conditions can quickly reestablish from rainfall. However, blowouts during dry periods may leave a dramatic sedimentary signal because with insufficient precipitation to refill the lagoon, the location can rapidly transform from lagoon to swamp. This may explain the abrupt transitions from clay to peat layers exhibited throughout zones 2 and 1.

Paleoclimatic reconstruction – The role of El Niño-Southern Oscillation and ITCZ

Two climatic factors are expected to exert important controls over precipitation patterns across the Neotropics, including our study site – El Niño-Southern Oscillation (ENSO) and the ITCZ. Correlations between Laguna Mitla’s 6900-year paleoenvironmental record and the long-term ENSO and ITCZ records provide insight into the relative influence of these two climatic drivers on regional precipitation patterns (Figure 5).

Figure 5.

Paleoclimate records, with locations shown in top left. Core 5 (site 1) residual content, marking lagoon phases during the backbarrier period, is compared with long-term climate records from (listed left to right) El Junco, Galápagos Islands (2 – Conroy et al., 2008); Laguna Pallacacocha, Ecuador (3 – Moy et al., 2002); Alfonso Basin, Mexico (4 – Staines-Urias et al., 2015); and Cariaco Basin, Venezuela (5 – Haug et al., 2001). Throughout the backbarrier period, the seven lagoon phases (high residual contents) correspond to periods of increased ENSO activity, suggesting more frequent, and possibly more intense, El Niño periods. Wet conditions at Laguna Mitla, beginning in zone 2, correlate to increased ENSO activity beginning around 4000 yr BP, and a large-scale climatic reorganization related to the southward migration of the ITCZ. Throughout zone 1, heightened ENSO activity can be correlated with more frequent and deeper lagoon phases at Laguna Mitla.

Our interpretation of zone 6 as reflecting a salt pan surrounded by Rhizophora patches suggests a period of low precipitation during ~6900–6500 yr BP. If the water adjacent to the Rhizophora came solely from precipitation, the required salinity levels could only have resulted from high evaporation. Even if this water was storm driven from the sea, moderate rainfall would have reduced the salinity levels over time. Long-term ENSO records indicate fairly low levels of activity during this period (Conroy et al., 2008; Moy et al., 2002; Staines-Urias et al., 2015), when the ITCZ occupied a far northern position (Haug et al., 2001; Figure 5). The identification of precipitation trends during zone 5 is uncertain, as ecological succession was likely largely driven by a rise in the water table. The limited soil erosion marked by low Ti concentrations coupled with high Cl concentrations, however, suggests that this period was likely dry. Zone 4 similarly exhibits limited sensitivity to climatic fluctuations since environmental changes were mainly driven by rising sea level and the highly variable coastal geomorphology. ENSO activity was variable, though mostly low during zone 4, while the ITCZ maintained a northerly position, with a slow southward drift beginning near the end of the period (Figure 5).

The correlation between climate and depositional environment becomes clearer after the establishment of the coastal barrier by the beginning of zone 3. The peat in the lower part of the zone contains high concentrations of fungal spores and significant Rhizophora percentages, indicating low water level and dry conditions. Although the peat-dominated sediments and high fungal spore concentrations continue to indicate low water, rising Laguncularia percentages suggest a decrease in salinity toward the end of the zone, possibly from an increase in precipitation. Dry conditions throughout zone 3 correlate with the minimal ENSO activity (infrequent El Niño periods) at the Galápagos Islands (Conroy et al., 2008; Riedinger et al., 2002; Figure 5) and the eastern tropical Pacific Ocean (Koutavas and Joanides, 2012). The ITCZ was still in a northerly position, slowly migrating to the south (Figure 5).

In zones 2 and 1, the swamp-to-lagoon transitions were most likely driven by dramatic changes in precipitation levels, with the seven occurrences of deeper water, or lagoon phases, representing increased precipitation at 4430–4270, 4080, 3950, 3680–3490, 3170–3080, 2990–2870, and 1680 yr BP to present. Toward the beginning of zone 2, a major climatic reorganization began. At ~4000 yr BP, the increased variability and southward migration of the ITCZ led to its coupling with ENSO (Figure 5) and the subsequent increase in ENSO activity (Koutavas et al., 2006; Toth et al., 2012). ENSO activity increased at ~4000 yr BP at El Junco (Conroy et al., 2008), ~3500 yr BP at Laguna Pallacacocha (Moy et al., 2002), and ~3000 yr BP in the Alfonso Basin (Staines-Urias et al., 2015), approximately matching the seven periods of increased water level during the backbarrier period. Lagoon phases marked by the occurrence of thick clay layers lacking peat indicate deeper water levels after ~3200 yr BP, corresponding to a period of increased ENSO activity at all three regional locations (Figure 5). The close correspondence between increased ENSO activity and wetter climate at Laguna Mitla during the late-Holocene is consistent with observations of increased seasonal precipitation (based on river discharge) during El Niño periods in central and western Mexico (Berrio et al., 2006; Heine, 1987), even though ENSO’s effects can vary spatially in this region (Bhattacharya and Chiang, 2014; Magaña et al., 2003; Pavia et al., 2006). This increase in seasonal precipitation is due to a combination of the northward migration of the mid-tropospheric subtropical ridge (Metcalfe et al., 2000), heightened influence of the tropical westerlies (Kao and Yu, 2009), and increased TC activity resulting from warmer surface waters and reduced vertical shear (Rodgers et al., 2000).

Proxy records from several sites across the Neotropics have indicated that long-term latitudinal shifts of the ITCZ are a major driver of Holocene precipitation changes (Haug et al., 2001; Pérez-Cruz, 2013). However, this relationship does not seem to apply to the Holocene precipitation record from Laguna Mitla. Since Guerrero lies just north of the current average summer position of the ITCZ (Pérez-Cruz, 2013; Philander et al., 1996), it could be expected that more northerly ITCZ positions would result in increased precipitation and southern positions in increased aridity. However, the Laguna Mitla record suggests drier conditions during the middle-Holocene when the ITCZ was positioned further north and increasing precipitation and lake levels after 4000 yr BP when the ITCZ was shifting to a more southerly position. We suggest that the decrease in precipitation resulting from the southward retreat of the ITCZ may have been compensated by increased precipitation associated with increased ENSO activity. While this relationship between the ITCZ and ENSO requires additional attention (Koutavas et al., 2006), a link has been established between the southward migration of the ITCZ and an increase in ENSO frequency and intensity in the ENP, the probable mechanism being the weakening of the easterly winds (Clement et al., 2000; Haug et al., 2001).

Regional paleoclimatological correlations – The role of TC precipitation

Our findings that the climate at Laguna Mitla exhibited progressively wetter conditions during the last ~4000 years can be evaluated in the context of other late-Holocene paleoclimatic records in the region. A review of the regional paleoclimatic literature suggests that the late-Holocene climatic conditions in central, southern, and western Mexico were characterized by considerable spatial and temporal variability, with remarkably incongruent or contradictory trends reported between coastal and inland sites. For example, a pollen record from a coastal site, Laguna Tetitlan, located just 10 km west of Laguna Mitla, identified dry periods from ~2000 to ~1500 yr BP separated by wet periods from ~3000 to ~2000 and ~1500 to ~450 yr BP (Gonzalez-Quintero, 1980), generally corresponding with the timing of the lagoon phases in zone 1 (~3170–3080, 2990–2870, 1680 yr BP–present) at Laguna Mitla. Similarly, a pollen record from the Middle America Trench (Habib et al., 1970), another coastal site, indicated moist conditions from 4000 to 3300 yr BP, dry conditions from ~3300 to 1700 yr BP, and moisture increasing after 1700 yr BP (Habib et al., 1970). By contrast, the uplands of central and southern Mexico were mostly dry during the late-Holocene (Metcalfe et al., 2015), despite some localities exhibiting evidence of periodic wetness (e.g. Arnauld et al., 1997; Jones et al., 2015; Park et al., 2010). Two recent stalagmite records from Diablo (Bernal et al., 2011) and Juxtlahuaca (Lachniet et al., 2013), located >100 km inland from Laguna Mitla, also exhibited a general drying trend over the last 7000 years. It is remarkable that the Laguna Mitla paleoclimatic record shows a trend of increasing wetness during the late-Holocene, which correlates better with adjacent coastal sites than with sites farther inland. We hypothesize that heightened ENSO activity (i.e. more frequent El Niño) during the late-Holocene led to an increase in TC activity along the Pacific coast of Mexico, resulting in more TC rainfall. Because of the orographic and rain shadow effects of the Sierra Madre del Sur, TC-generated rainfall tends to be mostly concentrated along the coast and decreases sharply landward. The steep coast-to-inland gradient in TC-generated rainfall amounts can be illustrated by two recent coast-paralleling TCs, Pauline (1997) and John (2006), which brought 7.5–10 times more precipitation to coastal Acapulco than Apaxtla, only 120 km inland (National Meteorological Service of Mexico, 2015; Figure 6). We suggest that the heightened ENSO activity after 4000 yr BP, which resulted in an increase in ENP TC activity and consequently an increase in coastal TC-induced rainfall, could have contributed to the apparent incongruence in precipitation patterns between coastal and inland sites during the late-Holocene. More work is needed to test this hypothesis.

Figure 6.

TC precipitation amounts measured from stations at Acapulco, Chilpancingo, and Apaxtla during coast-paralleling hurricanes Pauline (1997) and John (2008), showing the contrast between coastal and inland locations. Coastal sites receive the bulk of precipitation from these events, as Acapulco received over 4× and 10× the rainfall totals during Pauline compared with the interior Chilpancingo and Apaxtla sites, respectively. Similarly, during John, Acapulco received approximately 3× and 7.5× the rainfall totals in Chilpancingo and Apaxtla. Note that Acapulco is situated near our Mitla site, whereas Chilpancingo and Apaxtla are located near other regional paleoclimatic sites (Bernal et al., 2011; Lachniet et al., 2013) that exhibit contradictory precipitation trends to that of Mitla during the late-Holocene. Precipitation data are from the National Meteorological Service of Mexico.

Conclusion

The multi-proxy reconstruction shows that the basin surrounding the coring site has exhibited significant environmental changes over the last ~6900 years. At the beginning of the record the site was an isolated salt pan fringed by Rhizophora, sitting above MSL, with the high salinity as a result of arid conditions. Less saline conditions, marked by the replacement of Rhizophora by Laguncularia and then Conocarpus, began ~6500 yr BP as the freshwater aquifer, driven upward by the rising sea level, approached the ground surface. These conditions lasted until ~5600 yr BP, when the rising sea directly influenced conditions at the coring site. During the episodic formation of the beach barrier from ~6200 to 5200 yr BP, all proxies fluctuated dramatically as environmental conditions repeatedly alternated between marine and freshwater dominance. By ~5200 yr BP, the stabilizing beach barrier created a physical barrier between the coring site and the sea, disconnecting the lagoon’s water level from sea level, and making precipitation the primary control over backbarrier water depth. As a result, after this period, swamp/lagoon transitions served as a reliable proxy for regional precipitation. A total of seven wet phases were detected during the backbarrier period (4430–4270, 4080, 3950, 3680–3490, 3170–3080, 2990–2870, and 1680 yr BP–present), varying in intensity and duration. These lagoon phases were progressively deeper and long-lasting during the late-Holocene, likely the result of increased ENSO activity, while the effect of the latitudinal position of the ITCZ diminished. Our results also highlight the environmental processes leading to the formation of the backbarrier environment around 5000 yr BP, and the role of TCs in producing the spatial variability of paleo-precipitation patterns throughout western and central Mexico during the late-Holocene.

Footnotes

Acknowledgements

We would like to thank Dr Blanca Figueroa-Rangel and Dr Miguel Olvera Vargas for their logistical support during fieldwork. We thank Mr Ulises Cruz-Valera for field assistance.

Funding

This research was supported by grants from the US National Science Foundation (BCS-1003654), Geological Society of America (Graduate Student Research Grant), Inter-American Institute for Global Change Research (IAI-SGP-CRA2050), and Association of American Geographers (Dissertation Research Grant).

References

Alva

Kostoglodov

(2007) Aseismic slow slip events in Mexico from tide gauge records. Geos 27: 119.

Arnauld

Metcalfe

Petrequin

(1997) Holocene climatic change in the Zacapu lake basin, Michoacan: Synthesis of results. Quaternary International 43–44: 173–179.

Atwater

(1987) Evidence for great Holocene earthquakes along the outer coast of Washington state. Science 236: 942–944.

Baker

Seltzer

Fritz

et al . (2001) The history of South American tropical precipitation for the past 25,000 years. Science 291: 640–643.

Bernal

Lachniet

McCulloch

et al . (2011) A speleothem record of Holocene climate variability from southwestern Mexico. Quaternary Research 75: 104–113.

Berrio

Hooghiemstra

van Geel

et al . (2006) Environmental history of the dry forest biome of Guerrero, Mexico, and human impact during the last c. 2700 years. The Holocene 16: 63–80.

Bhattacharya

Chiang

JCH

(2014) Spatial variability and mechanisms underlying El Nino-induced droughts in Mexico. Climate Dynamics 43: 3309–3326.

Bianchette

McCloskey

Liu

(2016) Re-evaluating the geological evidence for Late-Holocene marine incursion events along the Guerrero Seismic Gap on the Pacific Coast of Mexico. PloS ONE 11: e0161568.

Blouet

(2002) The environment. In: Blouet

Blouet

(eds) Latin America and the Caribbean: A Systematic and Regional Survey. New York: John Wiley & Sons, pp. 13–50.

10.

Caballero

Peñalba

Martínez

et al . (2005) A Holocene record from a former coastal lagoon in Bahía Kino, Gulf of California, Mexico. The Holocene 15: 1236–1244.

11.

Caffrey

Horn

Orvis

et al . (2015) Holocene environmental change at Laguna Saladilla, coastal north Hispaniola. Palaeogeography Palaeoclimatology Palaeoecology 436: 9–22.

12.

CANABIO (2008) Main vegetation types: Michoacan case. Available at: http://www.conabio.gob.mx/conocimiento/cambios_veg/doctos_ing/tipos.html.

13.

Cavalié

Pathier

Radiguet

et al . (2013) Slow slip event in the Mexican subduction zone: Evidence of shallower slip in the Guerrero seismic gap for the 2006 event revealed by the joint inversion of InSAR and GPS data. Earth and Planetary Science Letters 367: 52–60.

14.

Chen

Zhang

(1997) Quaternary stratigraphy and trace-element indices of the Yangtze Delta, eastern China, with special reference to marine transgressions. Quaternary Research 47: 181–191.

15.

Clement

Seager

Cane

(2000) Suppression of El Nino during the mid-Holocene by changes in the Earth’s orbit. Paleoceanography 15: 731–737.

16.

Cochran

Reese

Liu

(2009) Tropical Storm Gamma and the Mosquitia of eastern Honduras: A little-known story from the 2005 hurricane season. Area 41: 425–434.

17.

Cochran

Berryman

Zachariasen

et al . (2006) Paleoecological insights into subduction zone earthquake occurrence, eastern North Island, New Zealand. Geological Society of America Bulletin 118: 1051–1074.

18.

Conroy

Overpeck

Cole

et al . (2008) Holocene changes in eastern tropical Pacific climate inferred from a Galapagos lake sediment record. Quaternary Science Reviews 27: 1166–1180. Available at: http://www.ncdc.noaa.gov/paleo/study/6109 (accessed 11 May 2014).

19.

Contreras-Espinosa

Warner

(2004) Ecosystem characteristics and management considerations for coastal wetlands in Mexico. Hydrobiologia 511: 233–245.

20.

Curray

Emmel

Crampton

(1969) Holocene history of a strand plain Lagoonal Coast, Nayarit, Mexico. In: Ayala-Castanares

(ed.) Coastal Lagoons: A Symposium. Mexico City: Universidad Nacional Autonoma de Mexico, pp. 63–100.

21.

Darienzo

Peterson

(1990) Episodic tectonic subsidence of late-Holocene salt marshes, northern Oregon Central Cascadia Margin. Tectonics 9: 1–22.

22.

Engelhart

Vacchi

Horton

et al . (2015) A sea-level database for the Pacific coast of central North America. Quaternary Science Reviews 113: 78–92.

23.

Englehart

Douglas

(2001) The role of eastern North Pacific tropical storms in the rainfall climatology of western Mexico. International Journal of Climatology 21: 1357–1370.

24.

Faegri

Iversen

(1989) Textbook of Pollen Analysis. New York: John Wiley & Sons.

25.

Figueroa-Rangel

Willis

Olvera-Vargas

(2008) 4200 years of pine-dominated upland forest dynamics in west-central Mexico: Human or natural legacy? Ecology 89: 1893–1907.

26.

Garduño-Monroy

Soria-Caballero

Israde-Alcántara

et al . (2011) Evidence of tsunami events in the paleolimnological record of Lake Patzcuaro, Michoacan, Mexico. Geofisica Internacional 50: 147–161.

27.

Goman

Joyce

Mueller

(2005) Stratigraphic evidence for anthropogenically induced coastal environmental change from Oaxaca, Mexico. Quaternary Research 63: 250–260.

28.

Goman

Joyce

Mueller

et al . (2010) Multiproxy paleoecological reconstruction of prehistoric land-use history in the western region of the lower Rio Verde Valley, Oaxaca, Mexico. The Holocene 20: 761–772.

29.

Gonzalez-Quintero

(1980) Paleoecologia de un sector costero de Guerrero, Mexico (3000 anos). Memorias 86: 133–157.

30.

Habib

Thurber

Ross

et al . (1970) Holocene palynology of the Middle American Trench near Tehuantepec, Mexico. Memoirs of the Geological Society of America 126: 233–261.

31.

Hatcher

Simoneit

BRT

Mackenzie

et al . (1982) Organic geochemistry and pore water chemistry of sediments from Mangrove Lake, Bermuda. Organic Geochemistry 4: 93–112.

32.

Haug

Hughen

Sigman

et al . (2001) Southward migration of the Intertropical Convergence Zone through the Holocene. Science 293: 1304–1308. Available at: https://www.ncdc.noaa.gov/paleo/study/2560 (accessed 11 May 2014).

33.

Heine

(1987) Anthropogenic sedimentological changes during the Holocene in Mexico and Central America. Striae: A Monograph Series for Quaternary Studies 26: 51–63.

34.

Hodell

Brenner

Curtis

(2005a) Terminal classic drought in the northern Maya lowlands inferred from multiple sediment cores in Lake Chichancanab (Mexico). Quaternary Science Reviews 24: 1413–1427.

35.

Hodell

Brenner

Curtis

et al . (2005b) Climate change on the Yucatan Peninsula during the Little Ice Age. Quaternary Research 63: 109–121.

36.

Hogarth

(2007) The Biology of Mangroves. New York: Oxford University Press.

37.

Islebe

Sánchez

(2002) History of Late-Holocene vegetation at Quintana Roo, Caribbean coast of Mexico. Plant Ecology 160: 187–192.

38.

Jáuregui

(2003) Climatology of landfalling hurricanes and tropical storms in Mexico. Atmósfera 16: 193–204.

39.

Jien

Gough

Butler

(2015) The influence of El Niño-Southern Oscillation on tropical cyclone activity in the Eastern North Pacific Basin. Journal of Climate 28: 2459–2474.

40.

Jin

Boucharel

Lin

(2014) Eastern Pacific tropical cyclones intensified by El Niño delivery of subsurface ocean heat. Nature 516: 82–85.

41.

Jones

Metcalfe

Davies

et al . (2015) Late-Holocene climate reorganisation and the North American Monsoon. Quaternary Science Reviews 124: 290–295.

42.

Joyce

Mueller

(1992) The social impact of anthropogenic landscape modification in the Rio Verde drainage basin, Oaxaca, Mexico. Geoarcheology 7: 503–526.

43.

Kao

H-Y

J-Y

(2009) Contrasting Eastern-Pacific and Central-Pacific types of ENSO. Journal of Climate 22: 615–632.

44.

Kirwan

Murray

Donnelly

et al . (2011) Rapid wetland expansion during European settlement and its implication for marsh survival under modern sediment delivery rates. Geology 39: 507–510.

45.

Koutavas

Joanides

(2012) El Niño-Southern Oscillation extrema in the Holocene and Last Glacial Maximum. Paleoceanography 27: PA4208.

46.

Koutavas

deMenocal

Lynch-Stieglitz

(2006) Holocene trends in tropical Pacific sea surface temperatures and the El Nino-Southern Oscillation. PAGES News 13: 22–23.

47.

Kremp

Heiskanen

A-S

(1999) Sexuality and cyst formation of the spring-bloom dinoflagellate Scrippsiella hangoei in the coastal northern Baltic Sea. Marine Biology 134: 771–777.

48.

Lachniet

Asmerom

Bernal

et al . (2013) Orbital pacing and ocean circulation-induced collapses of the Mesoamerican monsoon over the past 22,000 y. Proceedings of the National Academy of Sciences 110: 9255–9260.

49.

Lankford

(1977) Coastal Lagoons of Mexico: Their origin and classification. In: Wiley

(ed.) Estuarine Processes. Cambridge, MA: Academic Press, pp. 182–215.

50.

Larson

Zhou

Higgins

(2005) Characteristics of landfalling tropical cyclones in the United States and Mexico: Climatology and interannual variability. Journal of Climate 18: 1247–1262.

51.

Leopold

(1950) Vegetation zones of Mexico. Ecology 31: 507–518.

52.

Liu

Fearn

(2000) Reconstruction of prehistoric landfall frequencies of catastrophic hurricanes in northwestern Florida from lake sediment records. Quaternary Research 54: 238–245.

53.

Liu

Shen

(2008) A 1200-year proxy record of hurricanes and fires from the Gulf of Mexico coast: Testing the hypothesis of hurricane-fire interactions. Quaternary Research 69: 29–41.

54.

Liu

Bianchette

et al . (2011) Storm deposition in a coastal backbarrier lake in Louisiana caused by hurricanes Gustav and Ike. Journal of Coastal Research SI64: 1866–1870.

55.

Lonard

Judd

Stalter

(2011) The biological flora of coastal dunes and wetlands: Batis maritima C. Linnaeus. Journal of Coastal Research 27: 441–449.

56.

Lozano-García

MDS

Martínez-Hernández

(1990) Palinología de la Estación de Biología Tropical Los Tuxtlas. Mexico City: Universidad Nacional Autonoma de Mexico.

57.

McCloskey

Liu

(2013) A 7000 year record of paleohurricane activity from a coastal wetland in Belize. The Holocene 23: 278–291.

58.

Magaña

Vazquez

Perez

et al . (2003) Impact of El Niño on precipitation in Mexico. Geofisica Internacional 42: 313–330.

59.

Metcalfe

Barron

Davies

(2015) The Holocene history of the North American Monsoon: ‘known knowns’ and ‘known unknowns’ in understanding its spatial and temporal complexity. Quaternary Science Reviews 120: 1–27.

60.

Metcalfe

Jones

Davies

et al . (2010) Climate variability over the last two millennia in the North American Monsoon region, recorded in laminated lake sediments from Laguna de Juanacatlan, Mexico. The Holocene 20: 1195–1206.

61.

Metcalfe

O’Hara

Caballero

et al . (2000) Records of Late Pleistocene-Holocene climatic change in Mexico: A review. Quaternary Science Reviews 19: 699–721.

62.

Mosiño-Aleman

Garcia

(1974) The climate of Mexico. In: Landsberg

(ed.) Climates of North America. Amsterdam: Elsevier, pp. 345–391.

63.

Moy

Seltzer

Rodbell

et al . (2002) Variability of El Nino/Southern Oscillation activity at millennial timescales during the Holocene epoch. Nature 420: 162–165. Available at: http://www.ncdc.noaa.gov/paleo/study/5488 (accessed 4 November 2013).

64.

National Hurricane Center (2011) Tropical cyclone climatology. Available at: http://www.nhc.noaa.gov/climo (accessed 29 June 2011).

65.

National Meteorological Service of Mexico (2015) Normales climatologicas. Available at: http://smn.cna.gob.mx/index.php?option=com_content&view=article&id=42&Itemid=75 (accessed 6 January 2015).

66.

Neff

Pearsall

Jones

et al . (2006) Climate change and population history in the Pacific Lowlands of Southern Mesoamerica. Quaternary Research 65: 390–400.

67.

Nichol

Goff

Devoy

RJN

et al . (2007) Lagoon subsidence and tsunami on the west coast of New Zealand. Sedimentary Geology 200: 248–262.

68.

Ortiz

Singh

Kostoglodov

et al . (2000) Source areas of the Acapulco-San Marcos, Mexico earthquakes of 1962 (M 7.1;7.0) and 1957 (M 7.7), as constrained by tsunami and uplift records. Geofisica Internacional 39: 337–348.

69.

Pardo

Suárez

(1995) Shape of the subducted Rivera and Cocos plates in southern Mexico: Seismic and tectonic implications. Journal of Geophysical Research 100: 12357–12373.

70.

Park

Byrne

Boehnel

et al . (2010) Holocene climate change and human impact, central Mexico: A record based on maar lake pollen and sediment chemistry. Quaternary Science Reviews 29: 618–632.

71.

Pavia

Graef

Reyes

(2006) PDO-ENSO effects in the climate of Mexico. Journal of Climate 19: 6433–6438.

72.

Pérez-Cruz

(2013) Hydrological changes and paleoproductivity in the Gulf of California during middle and late-Holocene and their relationship with ITCZ and North American Monsoon variability. Quaternary Research 79: 138–151.

73.

Peros

Reinhardt

Davis

(2007) A 6000-year record of ecological and hydrological changes from Laguna de la Leche, north coastal Cuba. Quaternary Research 67: 69–82.

74.

Peterson

Haug

(2006) Variability in the mean latitude of the Atlantic Intertropical Convergence Zone as recorded by riverine input of sediments to the Cariaco Basin (Venezuela). Palaeogeography Palaeoclimatology Palaeoecology 234: 97–113.

75.

Philander

SGH

Halpern

et al . (1996) The role of low-level stratus clouds in keeping the ITCZ mostly north of the equator. Journal of Climate 9: 2958–2972.

76.

Pospelova

Chmura

Boothman

et al . (2002) Dinoflagellate cyst records and human disturbance in two neighboring estuaries, New Bedford Harbor and Apponagansett Bay, Massachusetts (USA). Science of the Total Environment 298: 81–102.

77.

Ramírez-Herrera

Urrutia-Fucugauchi

(1999) Morphotectonic zones along the coast of the Pacific continental margin, southern Mexico. Geomorphology 28: 237–250.

78.

Ramírez-Herrera

Kostoglodov

Urrutia-Fucugauchi

(2004) Holocene-emerged notches and tectonic uplift along the Jalisco coast, Southwest Mexico. Geomorphology 58: 291–304.

79.

Ramírez-Herrera

Kostoglodov

Urrutia-Fucugauchi

(2011) Overview of recent coastal tectonic deformation in the Mexican Subduction Zone. Pure and Applied Geophysics 168: 1415–1433.

80.

Ramírez-Herrera

Cundy

Kostoglodov

et al . (2007) Sedimentary record of late-Holocene relative sea-level change and tectonic deformation from the Guerrero Seismic Gap, Mexican Pacific Coast. The Holocene 17: 1211–1220.

81.

Ramírez-Herrera

Cundy

Kostoglodov

et al . (2009) Late-Holocene tectonic land-level changes and tsunamis at Mitla lagoon, Guerrero, Mexico. Geofisica Internacional 48: 195–209.

82.

Ramírez-Herrera

Lagos

Hutchinson

et al . (2012) Extreme wave deposits on the Pacific coast of Mexico: Tsunamis or storms? A multi-proxy approach. Geomorphology 360–371.

83.

Reimer

Bard

Bayliss

et al . (2013) Intcal13 and Marine13 radiocarbon age calibration curves 0-50,000 years cal BP. Radiocarbon 55: 1869–1887.

84.

Reynolds

Simms

(2015) Late quaternary relative sea level in Southern California and Monterey Bay. Quaternary Science Reviews 126: 57–66.

85.

Richmond

Szczuciński

Chagué-Goff

et al . (2012) Erosion, deposition, and landscape change on the Sendai coastal plain, Japan, resulting from the March 11, 2011 Tohoku-oki tsunami. Sedimentary Geology 282: 27–39.

86.

Riedinger

Steinitz-Kannan

Last

et al . (2002) A similar to 6100 C-14 yr record of El Niño activity from the Galapagos Islands. Journal of Paleolimnology 27: 1–7.

87.

Rodgers

Adler

Pierce

(2000) Contributions of tropical cyclones to the North Pacific climatological rainfall as observed from satellites. Journal of Applied Meteorology 39: 1658–1678.

88.

Rogers

(1982) The Bataceae in the southeastern United States. Journal of the Arnold Arboretum 63: 375–386.

89.

Roubik

Moreno

(1991) Pollen and Spores of Barro Colorado Island. St. Louis, MO: Missouri Botanical Garden.

90.

Roy

Quiroz-Jimenez

Chavez-Lara

et al . (2014) Humid Pleistocene-Holocene transition and early Holocene in sub-tropical northern Mexico and possible Gulf of California forcing. Boreas 43: 577–587.

91.

Roy

Rivero-Navarette

Lopez-Balbiaux

et al . (2013) A record of Holocene summer-season palaeohydrological changes from the southern margin of Chihuahua Desert (Mexico) and possible forcings. The Holocene 23: 1105–1114.

92.

Schofield

Edwards

Mighall

et al . (2010) An integrated geochemical and palynological study of human impacts, soil erosion and storminess from southern Greenland since c. AD 1000. Palaeogeography Palaeoclimatology Palaeoecology 295: 19–30.

93.

Spaulding

Kainuma

Collins

(2010) World Atlas of Mangroves. A collaborative project of ITTO, ISME, FAO, UNEP-WCMC, UNESCO-MAB, UNU-INWEH and TNC. London: Earthscan, p. 319.

94.

Staines-Urias

Gonzalez-Yajimovich

Beaufort

(2015) Reconstruction of past climate variability and ENSO-like fluctuations in the southern Gulf of California (Alfonso Basin) since the last glacial maximum. Quaternary Research 83: 488–501.

95.

Stuiver

Reimer

(1993) Extended 14C data base and revised CALIB 3.0 14C age calibration program. Radiocarbon 35: 215–230.

96.

Telford

Heegaard

Birks

HJB

(2004) The intercept is a poor estimate of a calibrated radiocarbon age. The Holocene 14: 296–298.

97.

Toth

Aronson

Vollmer

et al . (2012) ENSO drove 2500-year collapse of Eastern Pacific Coral Reefs. Science 337: 81–84.

98.

Van Geel

Buurman

Brinkkemper

et al . (2003) Environmental reconstruction of a Roman Period settlement site in Uitgeest (The Netherlands), with special reference to coprophilous fungi. Journal of Archaeological Science 30: 873–883.

99.

Van Geel

Gelorini

Lyaruu

et al . (2011) Diversity and ecology of tropical African fungal spores from a 25,000-year palaeoenvironmental record in southeastern Kenya. Review of Palaeobotany and Palynology 164: 174–190.

100.

Watts

Bradbury

(1982) Paleoecological studies at Lake Patzcuaro on the west-central Mexican Plateau and at Chalco in the Basin of Mexico. Quaternary Research 17: 56–70.

101.

Willard

Cooper

Gamez

et al . (2004) Atlas of pollen and spores of the Florida Everglades. Palynology 28: 175–227.

102.

Woodruff

Donnelly

Mohrig

(2008) Reconstructing relative flooding intensities responsible for hurricane-induced deposits from Laguna Playa Grande, Vieques, Puerto Rico. Geology 36: 391–394.

103.

Zavala Sansón

(2004) The mechanical influence of continental topography on the trajectories of tropical cyclones near the west coast of Mexico. Atmósfera 17: 151–170.