Journal of Coastal Research 470-481 Royal Palm Beach, Florida Spring 2000

Selective Sorting, Storage and Progressive Dilution of Sediment in Two Tropical Deltas, Veracruz, Mexico Zhongyuan Chent, Daniel Jean Stanley], and Eric E. Wright§ tState Key Laboratory for :j:Deltas-GlobalChange §Marine Science Department Estuarine and Coastal Program Coastal Carolina University Research E-206 NMNH Paleobiology Conway, SC 29528, U.S.A. East China Normal Smithsonian Institution University Washington, D.C. 20560, Shanghai 200062, China P.R. U.S.A.

. ~ .. CHEN, Z.; STANLEY, D.J., and WRIGHT, E.E., 2000. Selective Sorting, Storage and Progressive Dilution of Sediment .tllllllll:. in Two Modern Deltas, Veracruz, Mexico. Journal of Coastal Research, 16(2), 470-481. Royal Palm Beach (Florida), ~ ISSN 0749-0208.

~ ~ Compositional and textural attributes of the surficial sediment in deltaic settings typically differ from the sediment susuTyJ -+; load carried by their source rivers. This phenomenon is evaluated here by semiquantitative study of the variability 1&r--- of petrologic characteristics in different delta subenvironments of the Nautla and Tecolutla deltas along Mexico's Veracruz coast. Sediment composition and texture in the different environments of both deltas are related to processes that involve selective sorting during seasonal flooding and temporary storage after deposition. Displacement of par­ ticles of different size, density and shape onto the two tropical delta plains is controlled by fluctuating fluvial discharge that responds to a highly variable seasonal rainfall pattern. These discharge variations induce changes in hydraulic equivalence and, consequently, the release of substantially different proportions of volcanic, mica, heavy mineral and grain-size fractions in the Veracruz delta settings. These phenomena, together with progressive dilution of sediment upon reaching the coast, result in significant modifications to the original petrologic attributes of fluvial loads as they are transported seaward. Comparable variations of composition and texture in subsurface Holocene sections would likely record former large­ scale climate, rainfall and fluvial discharge fluctuations. Petrologic analysis of core sections recovered at Prehispanic sites on the two deltas could be used to detect past long-term flooding pulses and droughts, natural events that archaeologists suggest caused deterioration of subsistence resources, stratigraphic hiatuses, cultural discontinuities and human migrations in this region.

ADDITIONAL INDEX WORDS: Archaeological sites, delta sedimentation, fluvial discharge, GulfofMexico, Holocene, particle composition, rain pattern, sediment load, texture, transport processes.

INTRODUCTION by the high Sierra Madre Oriental chain of eastern Mexico, provides an appropriate setting to gain further insight on the A river's unidirectional flow carries a sediment load that is nature of depositional variability in deltas. Preliminary sur­ generally representative of materials transported from upslo­ veys were made of two small tropical deltas, Nautla (SELF, pe source terrains. However, petrologic analyses of modern 1971, 1975) and Tecolutla (LENTELL, 1975), positioned -----35 deposits positioned in deltas and contiguous coasts indicate km apart on the tropical Veracruz margin of the Gulf of Mex­ that the proportions of compositional components accumulat­ ico (Figure 1). Building upon these earlier studies, the cur­ ed in such coastal settings usually differ substantially from rent investigation focuses specifically on the semiquantita­ those of their major fluvial channels in more landward and tive changes of composition and texture in these two deltas upslope sectors. Several petrologic investigations, such as to determine the relation that may exist among petrology, that of the Burdekin delta in Australia (ETHRIDGE et al., delta subenvironment and sedimentation process. Also con­ 1975), have detailed the compositional and textural variabil­ sidered are depositional processes that modify the fluvial ity in delta systems that are proximal to source terrains. loads of the Nautla and Tecolutla rivers as they reach deltaic Most such studies of modern deltaic deposits, however, have and contiguous coastal environments. Definition of petrologic been made in settings that are at a considerable distance variability in the two lower river systems in this investiga­ from terrigenous source areas. Examples include the Nile tion provides a means to interpret the past geological and (STANLEY and CHEN, 1991; STANLEY et al., 1992) and Yang­ human history of this region. tze (YAN and Xu, 1987) where composition and texture of the sediment released in various deltaic subenvironments differ DESCRIPTION OF STUDY AREA markedly from that of the sediment load carried by their main river channels. The Nautla and Tecolutla rivers are short (-----160 km and The southwestern Gulf of Mexico coastal margin, flanked -----220 km, respectively) torrential rivers, with headwaters in the proximal Sierra Madre Oriental mountain chain (Figures 99057 received and accepted in revised 20 July 1999. 1, 2). Tributaries of these two rivers reach elevations in ex- Sediments in Tropical Deltas 471

GULF OF MEXICO

o 20km , , i o 20mi Legend IIAlluvial and L-..J delta plain

. c..• Highland ... ':..' (200-3000 m) Highland (>3000 m) 19° -----....------..... 00'

Figure 1. Map (modified from several sources) showing the two delta study areas on the Veracruz coastal margin, southwestern Gulf of Mexico.

cess of 3000 m, and channels flow down steep eastern moun­ vation headlands (TAMAYO, 1962; SELF, 1971). Rain falls pri­ tain flanks toward the northeast. Beyond the base of moun­ marily in summer and early fall, and is heaviest duringAugust tain slopes, the rivers traverse piedmont hills, alluvial fans and September. Differences in transport regime and sediment and terraces (Figures 3, 4A) and extend across the lower, nar­ supply to the two deltas are mainly a result of dissimilarities row ("-'15 km), near-horizontal coastal plain. Sediment source in fluvial basin area and consequent amounts of discharge (TA­ areas for the Nautla and Tecolutla deltas are primarily vol­ MAYO, 1962). The area of the Nautla drainage basin is ap­ canic terrains widely exposed in the Neovolcanic Cordillera proximately 2270 km-; the average river flow is 1,086,000 m'' of the Sierra. Peaks include the Cofre de Perote (4282 m) and per km 2 of basin area, accounting for a mean annual discharge Pico de Orizaba (5747 m), andesitic cones active during the of 2465 million m", In contrast, the Tecolutla drainage basin Cenozoic (GARFIAS and CHAPIN, 1949; TAMAYO, 1962; MEX­ area (8080 km-) is about 3.5 times larger, while the average ICO, DIRECCION GENERAL DE GEOGRAFIA, 1984; MORAN­ river flow per basin area (931,800 m" per km") is nearly the ZENTENO and Collaborators, 1994). same as the Nautla. This accounts for a mean total annual Sediment in both drainage basins is transported to the sea river discharge of the Tecolutla that is at least 3 times larger by rivers that are directly subject to seasonal rainfall and (7529 million m") than in the Nautla. During dry periods, flow flooding. Mean annual rainfall ranges regionally from "-'1500 in the lower Nautla is reduced to "-'20 mvsec, while in the mm on the coastal plain to as much as 3000 mm in high-ele- lower Tecolutla it is at least 30 m-/sec.

Journal of Coastal Research, Vol. 16, No.2, 2000 472 Chen, Stanley and Wright

TECOLUTLA RIVER o Gulf of Mexico

o

20°30' o Sample Location

o Dune and Beach Marsh • Mangrove Archeological o Site ~ Approximate *'\. Delta Boundary

20°25'

I II 97°05' 97°00' ~lometers o NAUTLA RIVER "

Gulf o of Mexico

20°15' o Nt o kilometers o 5 o

20°10'

96°50' 96°45' Figure 2. Maps of the Tecolutla (upper) and Nautla (lower) deltas, positioned on the Veracruz margin, showing surficial sediment sample sites in diverse environments.

The Nautla and Tecolutla Holocene deltas are among sev­ are constrained by adjacent higher terrains formed by Ter­ eral wave-dominated and sharply truncated depocenters tiary and Quaternary deposits (Figure 6A). This coastal re­ along the NW-SE trending Veracruz margin (Figure 5D-F). gion, characterized by few, low magnitude, shallow earth­ The distance between apex and coast is '"'"'9 km in the Nautla quake epicenters, has remained relatively stable during the delta and '"'"'7 km in the Tecolutla delta, and both have a Holocene. Some faulting and neotectonic deformation (LEN­ coastal margin length of '"'"' 14 km. The trapezoidal Nautla del­ TELL, 1975), however, appear to control the configuration of ta area is '"'"' 115 km", and that of the triangular Tecolutla is the two depocenters, including angular paths of their fluvial '"'"'50 km". Delta margins and distributary channel patterns channels and distributaries (Figures 2, 3). The savannah veg-

Journal of Coastal Research, Vol. 16, No.2, 2000 Sediments in Tropical Deltas 473

Figure 3. Space Shuttle image showing the base of the Sierra Madre Oriental (SMO) and adjacent piedmont, on the Tecolutla coastal plain and delta, including main river channel (T), marsh (M), mangrove (MA), and coastal barrier (B). In winter, wave-driven coastal currents (small arrows) drive sediment toward the SE. Near-vertical photograph was taken 118 nautical miles above earth surface on November 18, 1990.

etation that now prevails over large parts of this coastal area (following the major Gulfcirculation pattern), with flow rang­ (Figure 4A-D ) results primarily from recent human activity, ing from 6 to 12 nautical miles per day. As a response to these including deforestation of tropical stands, cattle ranching and seasonally variable physical conditions, fluvial sediment from plantation agriculture (WILKERSON, 1976, 1983; HEBDA et the Nautla and Tecolutla rivers is dispersed along the coast a1., 1991). by wave-driven currents toward the SE in winter (Figures 3, Fluvial discharge reaches the Veracruz coast where the 5D-F), and NW in summer. mean air temperature ranges from ...... 21°C to 29.5°C annually (LEIPPER, 1954a), mean annual surface temperature of Gulf METHODS water ranges from 21°C to > 25.5°C, and mean tidal range A total of 122 surficial sediment samples on and adjacent approximates 1.3 m (TAMAYO, 1962). Most pertinent in this to the Nautla (62 samples) and Tecolutla (60 samples) deltas study are wind and surface current patterns that vary sea­ were collected in February 1996 (Figure 2). These samples sonally. In winter (J anuary), wind velocity (10 to 12 knots) were recovered in the following 11 environments (using nu­ and direction are variable, most from easterly directions, merical codes 1-11), from land to sea: river (1), natural levee with fewer southerlies but more northerlies, and surface (2), delta plain (3), marsh (4), red mangrove (Rhizophora ocean currents are oriented primarily toward the southeast mangle) thicket (5), upper estuary (tidal creeks in Nautla, (as part of a Gulf counter eddy), with flow approximating 4 tidal ponds in Tecoutla) (6), lower estuary (where channels nautical miles per day (LEIPPER, 1954b). Conditions differ in widen markedly and are affected by the sea) (7), dune (8), summer (July), when wind velocities diminish (6 to 8 knots) beach (9), breaker zone (10), and inner shelf(11). At least five and directions are less variable, primarily from the south­ to eight samples were collected in each of the above environ­ east, and surface currents are oriented toward the northwest ments in the two deltas.

Journal of Coastal Research, Vol. 16, No.2, 2000 474 Chen, Stanley and Wright

Figure 4. (A) Plantation agriculture on piedmont base of the Sierra Madre Oriental in the vicinity of the study area. (B) Ranchland near apex of Nautla delt a, with view of one of the pyr amids at the EIPital site (arrow). (C) Tecolutla river flowing across the delt a flood plain. (D) Coring operation on the Tecolutla delta flood plain on which Prehi spanic drainage patterns are pr eserved (noted by gently undulating surface). (E, F) Small, fres hwater to brackish, ponded areas surrounded by den se swamp and marsh vegetation, Tecolutla delta. Photos taken in Febru ary 1996 (D.J. S.).

The amount of organic matter (by weight percentage) was particles of coarse size (>1 mm ) were separated by sieving, determined for a representative cut of each sample by com­ and organic matter was removed by using bleach. More than bustion at 450°C for 4 hours. A separate sample fraction was 300 grains in the sand-size residue (63 to 1000 urn ) were selected to determine grain size and identify sand-size com­ identified and counted in each sample. The following 14 com­ positional components; in this cut, plant fragments and other ponents were recorded: volcanic (rock fragments, glass), mica,

Journal of Coastal Research, Vol. 16, No.2, 2000 Sediments in Tropical Deltas 475

Figure 5. (A) Red mangrove thicket on Tecolutla delta plain. (B) Mouth of small distributary, flowing along NW margin of Tecolutla delta. (C) Sandbar in the Nautla's lower estuary; coast in background.(D) Sampling next to coastal barrier island in the lower Nautla estuary where it opens to the sea. (E) Dunes and beach on the NW Tecolutla coastal margin. (F) Eroded delta coast, near the mouth of the Nautla river. Photos taken in February 1996 (D.J .S.). light minerals, lithic aggregate, calcareous fragment (lime­ alyzer (LS 200). Moment measures and percentages of clay stone fragment, unidentified biogenic material), heavy min­ «4 urn), silt (4-63 urn), and sand (>63 urn) were obtained erals (opaque, transparent), iron nodule, foraminifera, ostra­ for each sample. cod, pelecypod, gastropod, echinoderm, diatom, and 'other'. Averaged percentages of sand-size compositional compo­ Grain size was determined for the fraction ranging from nents, organic matter and textural parameters for each en­ 0.4 to 1000 urn, using a Coulter laser diffraction particle an- vironment in the Nautla and Tecolutla deltas are listed in

Journal of Coastal Research, Vol. 16, No.2, 2000 476 Chen, Stanley and Wright

A 4000 ---1"!!1,.. -- ntermediate seg.-roastal seg.

:[3000 of: ~ 2000 ! w1000 o . 150

8 100 -

80 -~ CD 60 en S e oCD 40 ~ 20

o I 150 135 120 105 90 75 ._------Distance (km) ClOD

-~ 75 CD en S f: 50 CD U a.. ~ 25

I '- -- ~ - ~ .... ~ .... ~ - ~ .... ~ ...... , ...... , '- ~ ... ~ ...... , ... ~ 'i v ...... X ...... , ...... ,:- v '- v 'r x bX" o i 1 2 3 4 5 6 7 8 9 10 11 r=-:-=l ~ ,~~ :~I Iv Lv LI ~,..., vLv 1+++1++ ~ ~ l...:...=-.:J L:...:....:..J r2Za Volcanics Granite Carbonate Clastics Alluvium Light mineral Light miner. Others (Quartz) (Feldspar) (Undiffer.) Deltaic (1 - 7) and Contiguous Coastal (8 - 11) Environments 1 - River 2 - Natural levee 3 - Delta plain 4 - Marsh 5 - Mangrove 6 - Upper estuary 7 - Lower estuary 8 - Dune 9 - Beach 10 - Breaker zone 11 - Inner shelf

Figure 6. (A) Nautla downslope profile, from headwaters to Gulf coast, showing 3 major physiographic segments and position of geological exposures. (B) Major compositional components distributed along the Nautla river between headwater and coastal segments. A and B are modified from SELF (1975). (C) Lower Nautla profile, across the Nautla delta proper, showing variable proportions of compositional components (data obtained in this study) in each of 11 deltaic and contiguous coastal settings.

Figures 7 and 8. Complete petrologic database listings that collected between headwaters and coast in the Nautla drain­ record non-averaged numerical information for each individ­ age basin. Compositional components are dominated by vol­ ual sample are available from the authors. canic (rock fragment, glass), quartz, feldspar, limestone frag­ ment, and a heavy mineral suite comprising largely magne­ OBSERVATIONS tite, pyroxene, and amphibole. These are derived from expo­ Petrologic data of our Nautla delta study are compared to sures in the eastern and southeast flank of the Sierra Madre those of SELF (1971, 1975) who examined sediment samples Oriental (Figure 6A): Cenozoic volcanics, including andesite,

Journal of Coastal Research, Vol. 16, No.2, 2000 Sediments in Tropical Deltas 477

Compositional Component (0/0 Sand) Sediment Texture

Q) ci> Q) ~ co Q) N co Nautla "'C ..c en o o u: :5 C CJ) U 'c (J) :.::i co c Environments co ::::J o 'ro "'C o o "'C C 'c C (Number of Samples (5 ~ C ~ Q) t5 co co co 0> c "'C Examined) > o (J) >­ o co C > Q) co ro :E co co CJ) o 0> Q) ~ :.::i I ?f? 1=River (8) 21.8 1.1 8.8 61.3 6.6 0.4 0.5 398.6 3.7 8.4 88.0 2=Natural Levee (5) 32.6 0.5 2.5 59.4 4.6 0.3 1.1 71.5 22.5 43.7 34.2 3=Flood Plain (5) 32.4 0.8 1.8 58.9 5.6 0.5 1.4 42.6 29.0 51.1 19.2 4=Marsh (5) 13.7 1.5 4.5 70.5 9.6 0.2 15.0 96.1 26.3 39.1 34.3 5=Mangrove (6) 56.5 0.8 0.0 40.9 1.6 0.2 5.1 49.1 26.7 50.3 23.0 6=Upper Estuary (6) 54.8 1.0 3.0 31.1 3.8 6.3 6.9 30.2 30.2 52.9 16.9 7=Lower Estuary (7) 11.3 0.4 13.9 62.6 11.0 1.1 2.0 281.9 9.9 19.0 71.1 8=Dune (5) 5.8 0.1 24.7 38.6 29.6 1.2 0.2 263.0 0.4 0.7 98.9 9=Beach (5) 6.6 0.0 30.9 48.3 12.8 1.3 0.1 279.6 0.1 0.2 99.7 1O=Breaker Zone (5) 6.5 0.0 36.9 48.3 6.6 1.7 0.1 323.8 0.0 0.1 99.8 11=Inner Shelf (5) 12.2 0.2 16.5 65.1 4.3 1.8 0.7 149.9 3.0 9.6 87.4

Nautla Composition

100.0

90.0

[[I] 80.0 Biogenic ~ Shell ~ 'E 70.0 • Heavy Min. Q) c: and Iron o 60.0 a. E II Lights and <3 50.0 Lith. Agg. "'i D Calcareous 5 40.0 Fragment .iii:t:i &. 30.0 .Mica E <3 20.0 ~Volcanic

10.0

0.0 4 6 9 10 11 Environments

Figure 7. Table listing averaged sand-size compositional components, organic matter (% by weight), and sediment texture from 62 surficial samples in 11 environments of the Nautla delta. Compositional data are depicted in graph below.

ash and tuff; granite of undetermined age; Cretaceous lime­ as a response to the altered fluvial regime where slope de­ stone; and Tertiary and Quaternary clastics in piedmont, al­ creases on the lower part of alluvial fans (SELF, 1975, his luvial fan and terrace deposits at the base of the steep Sierra Figure 3). The simplified, downslope-trending mineralogical Madre slope. pattern that he compiled shows a general seaward increase SELF (1971, 1975) defined three fluvial segments of the in volcanic and limestone rock fragments, and somewhat de­ Nautla (Figure 6A,B): headwater (from -----3750 m, on the creased proportion of light minerals (quartz, feldspar) toward flanks of the Cofre de Perote volcano, where slope exceeds 50 the coast (Figure 6B). His findings indicate that Nautla sed­ m/km), intermediate (where slope decreases to -----5 m/km), iment reaching the Gulf coastal margin comprises to 48% vol­ and coastal (where the Nautla flows on a slope of <1 m/km), canic material (33% rock fragments, 15% ash), 34% light min­ That author recorded a sharp decrease in percent gravel, erals (22% quartz, 12% feldspar), 7% calcareous fragments, mean grain size, and some heavy mineral species at the base and to 11% heavy minerals plus 'other' components (SELF, of the intermediate segment, and interpreted these changes 1975).

Journal of Coastal Research, Vol. 16, No.2, 2000 478 Chen, Stanley and Wright

Compositional Component (0/0 Sand) Sediment Texture 0) C Q) ci> 0) 0 Q) (J) ~ « .= Q) N Co "0 2 U5 o Tecolutla o u, C .c :5 (f) co C o 'c ro (J) :.:J ro ~ >. Environments ro o ::J o 'co ro "0 o o "0 'c o <3 U5 C (Number of Samples ~ C Q) ro (5 ~ ro 'c t5 cf2. #. co 0) ro C "0 Examined) > o (J) o ~ ro C Q) ro co 1:: iii o (f) o 0) ~ :.:J cf2. 1=River (7) I 15.5 0.3 1.5 I 358.5 84.9 2=Natural Levee (5) I 27.1 0.2 3.5 I 153.0 I 42.9 3=Flood Plain (5) I 20.6 0.2 4.1 I 76.0 36.3 4=Marsh (6) I 15.2 0.2 13.0 I 26.3 12.2 5=Mangrove (6) I 56.5 0.2 26.0 I 28.5 13.0 6=Upper Est. (5) I 9.9 8.7 28.5 I 37.5 13.4 7=Lower Est. (6) I 16.9 4.2 3.4 I 152.0 60.8 8=Dune (5) I 6.2 0.8 0.4 I 226.9 98.0 9=Beach (5) I 9.8 1.9 0.5 I 227.7 100.0 10=Breaker Zone (5) I 8.4 2.6 0.6 I 276.0 99.7 11=Inner Shelf (5) I 9.8 1.5 0.8 I 177.9 96.9

Tecolutla Composition

100.0

90.0

80.0 [llJBiogenic ';!...., Shell a3 70.0 e • Heavy Min. o a. and Iron E 60.0 o D Lights and U (ij 50.0 Lith. Agg. c: o D Calcareous E 40.0 (J)o Fragment a. E 30.0 .Mica o U 20.0 ~Volcanic

10.0

0.0 10 11 Environments

Figure 8. Table listing averaged sand-size compositional components, organic matter (% by weight), and sediment texture from 60 surficial samples in 11 environments of the Tecolutla delta. Compositional data are depicted in graph below.

Using sample data obtained in the current study, we com­ while smaller proportions are present in mangrove, lower es­ pile a similar graph for the ~ 15 km wide base of slope-Nautla tuary and coastal environments. Recent whole shells of ma­ delta-coastal sector (Figure 6C). Shown is the percentage of rine origin accounts for high values of 'other' in coastal sam­ compositional components for each of the 11 Nautla suben­ ples; mica and heavy minerals, also in the 'other' category, vironments (1-7, deltaic, 8-11, contiguous coastal). Of note occur in marsh, river channel and lower estuary, but are are particularly high proportions of volcanic fragments in the scarce in mangrove. High proportions of marine shells char­ mangrove and upper estuary, and very low values of these acterize dune, beach, and offshore environments. It is appar­ components in coastal environments. Reworked carbonate ent that there is as much, and in some cases more, compo­ fragments and shells are present in the lower estuary, but sitional variation along the ~ 15 km wide lower Nautla-coast­ scarce in the marsh and mangrove. Light mineral suites are al plain sector (Figure 6C) as there is along the entire ~ 160 particularly abundant along the river channel, on delta nat­ km-Iong fluvial stretch between Sierra Madre headwaters ural levees and flood plain surfaces along the river channel, and the shore (Figure 6B).

Journal of Coastal Research, Vol. 16, No.2, 2000 Sediments in Tropical Deltas 479

Table 1. Ranking of compositional parameters by decreasing order ofimportance for the top 4 parameters for each delta environment (percentage oftotal sediment as calculated from Figures 7 and 8).

2 = 3 = 6= 7 = 10 = 1 = Natural Flood 4= 5 = Upper Lower 8 = 9= Breaker 11 = Components River Levee Plain Marsh Mangrove Estuary Estuary Dune Beach Zone Offshore NAUTLA DELTA Volcanic Mica 2 3 3 4 2 2 4 4 4 4 4 Calcareous Frags. 4 3 3 2 2 2 Light & Lith. Aggregate 1 2 2 2 3 4 1 1 1 1 1 Heavy Min. & Iron 4 2 3 3 Biogenic Shell Organic Matter 4 3 4 3 Inorganic Mud 3 1 1 1 1 2 3 TECOLUTLA DELTA Volcanic Mica 4 3 3 4 3 4 4 4 4 3 3 Calcareous Frags. 2 4 3 3 2 2 2 Light & Lith. Aggregate 1 2 2 3 4 3 2 1 1 1 1 Heavy Min. & Iron 2 3 4 Biogenic Shell Organic Matter 4 2 2 2 Inorganic Mud 3 1 1 1 1 4

To better define the relationship between petrology and whole provides a higher ranking for organic matter in the delta subenvironment, it is useful to consider a somewhat marsh, mangrove and upper estuary, and calcareous frag­ different and larger number of components. Thus, for each of ment in the river channel. Also in contrast with the Nautla the 11 Nautla environments, the numerical data in Figure 7 is the lower ranking of volcanics in river, mangrove and up­ is listed within 6 compositional groups (these incorporate all per estuary. 14 components described in the above Methods section), plus average organic matter and textural data. Our database (Fig­ DISCUSSION ure 7) shows that composition of sand-size components in riv­ er channel, levee, and adjacent flood plain is similar to that The database presented here indicates that (1) the propor­ of the Nautla above the piedmont base of slope. In contrast, tion of compositional components, amount of organic matter the other environments record a much greater variation. and texture all vary extensively from environment-to-environ­ The 11 Nautla environments can be readily differentiated ment in each delta, and (2) these parameters vary similarly in by using the various proportions of volcanic, mica, carbonate the comparable settings of the two deltas. Both Nautla and fragment and biogenic shell, and heavy minerals (Figure 7). Tecolutla rivers carry large sediment loads, with concentration Noted is the high content of sand-size volcanics in mangrove of suspended solids to > 1.0 part/l000 (TAMAYO, 1962) and, as and upper estuary, mica in marsh, heavy minerals in dune, noted earlier, drain similar geological terrains. Moreover, both and biogenic shell in estuarine and coastal environments. deltas contain large proportions of volcanic material in man­ Also of significance are high proportions of organic matter in groves, and high amounts of organic matter and finer grain the marsh, mangrove and upper estuary, and considerably size in marsh, mangrove and upper estuary. However, an im­ coarser mean grain-size in the lower estuary and coastal en­ portant clue to understanding the dynamics of sediment trans­ vironments. Mud-rich sediment prevails on all delta plain en­ port is provided by the major difference that exists between vironments. the two deltas: there is a much lower proportion of volcanic Considering the Nautla delta sediment as a whole, we iden­ material in the upper estuary (tidal ponds) of the Tecolutla tify the 4 most important parameters in each of the suben­ (10%) than in the upper estuary (tidal creeks) of the Nautla vironments (Table 1). Assessment of the dominant 2 param­ (55%). Although the Nautla river carries only a somewhat eters shows that inorganic mud-size sediment prevails in del­ higher proportion of volcanic material (----22%) as compared to ta flood plain and estuarine environments, while lights and the Tecolutla (----16%), there is more than five times the lithic aggregates are the dominant components in river, lower amount of volcanics in the upper estuary of the Nautla (Figure estuary and coastal environments. Volcanics are also impor­ 7) than that in the Tecolutla (Figure 8). While Tecolutla ponds tant in river, mangrove and upper estuary, calcareous frag­ may serve as more effective traps (Figure 4E,F) for mud-size ments in marine settings, and heavy minerals in dune. material, organic matter and volcanic particles, their markedly The dataset for the Tecolutla delta system (Figure 8) is increased proportions in the Nautla is a function of smaller strikingly similar to that of the Nautla. Among significant drainage basin area and lower overall discharge, as well as of differences recorded (Table 1) are the much lower volcanics input from more local flood plain runoff. in upper estuary, and relatively high mica only in marsh and Overbank flow prevails in both rivers at times of flood, pro­ mangrove. An evaluation of the Tecolutla sediment as a viding large contributions of smaller particles (silt, clay), high

Journal of Coastal Research, Vol. 16, No.2, 2000 480 Chen, Stanley and Wright proportions of volcanic, and some organic matter; these com­ attributes of sediment reaching the sea. In these high-energy ponents become selectively trapped in mangroves, particular­ settings, selected components are removed, and proportions ly by tidal forces. Flooding also flushes clay and silt fractions of original particles are masked by introduction of both mod­ seaward from the estuaries of both delta systems. The envi­ ern and relict terrestrial particles from other coastal areas ronment-to-environment variations of composition (such as and in situ carbonate shell material. The progressive dilution proportion of volcanic particles) and texture and differences effect (cf PETTIJOHN, 1957) results in a surficial sediment between Nautla and Tecolutla recorded here can best be in­ cover on the inner shelf that comprises a mix of reworked terpreted in light of different hydrologic attributes of the two modern and relict components, one that bears little resem­ fluvial systems. The larger Tecolutla discharge results in blance to the Nautla and Tecolutla terrigenous material car­ more efficient flushing of its upper estuary and marshes, and ried to the coast. reduced temporary storage of sediment therein (cf LEOPOLD The Veracruz margin and ecologically rich environments of et al., 1964; MEADE, 1988). This decreased retention accounts the Nautla and Tecolutla deltas have long served as bread­ for lower amounts of volcanic particles and mica in these two baskets for human occupants. Both delta surfaces include im­ environments than in the Tecolutla river system, and is also portant vestiges of Prehispanic cultural centers and agricul­ lower than in comparable settings of the Nautla. tural complexes. There is, for example, evidence of pre-16th We attribute some petrologic differences between suben­ Century patterned wetlands, such as raised planting plat­ vironments in the two deltas to effects of selective sorting (cf forms, canals and drainage systems (Figure 4D) developed to STANLEY, 1978) that occur primarily at times of seasonal enhance agricultural utilization of marginal soils (WILKER­ flooding and overflow onto the delta plain. In settings of var­ SON, 1976, 1983; HEBDA et al., 1991). Human centers, includ­ iable energy levels beyond the base of slope, particles re­ ing Santa Luisa in the Tecolutla and EI Pital in the Nautla leased from fluvial flow are segregated on the basis of size, (Figures 2, 4B), waxed and waned as populations shifted at specific gravity and particle shape (FREUNDT and ROSI, different times during the Holocene (WILKERSON, 1983, 1998). This is a function of mineral grains of different com­ 1994b). Correlation of compositional and textural character­ position and/or texture and/or shape that are hydraulically istics with fluvial regimes and specific depositional environ­ equivalent, i.e. although displaying different attributes, these ments in the two deltas can help archaeologists understand grains have similar fall velocities and are deposited together causes for suggested changes of human occupation in this re­ (PETTIJOHN, 1957; PETTIJOHNet al., 1973). The phenomenon gion since at least 6000 years before present. of hydraulic equivalence plays a determinant role in the seg­ In other deltas, such as the Yangtze, it has been shown regation of sediment particles by variable energy levels of that long-term floods and droughts cause stratigraphic hia­ flow across a delta plain. This is recorded by the distribution tuses and environmental change that induce cultural disrup­ of less dense volcanic material (specific gravity ranging from tion and human relocation (cf STANLEY et al., 1999). In sub­ 1.0 to 2.0), relative to that of light minerals and denser an­ surface sections of the Nautla and Tecolutla, changes in pro­ desite and rhyolite fragments (specific gravities from -----2.0 to portions of grain size, volcanics, mica, and heavy mineral 2.8), heavy mineral suites (>2.9), and basalt (-----3.1). The dis­ suites, as determined semiquantitatively in this study, would tribution pattern of volcanic particles, especially glass, is con­ signal modified fluvial discharge induced by altered climate­ trolled by shape, i.e. grains are less spherical (values of -----0.5­ rain pattern. It is anticipated that the record of petrologic 0.7) than most light and heavy mineral particles with which variations in radiocarbon-dated sediment borings obtained in they are associated. In similar fashion, mica serves as a dis­ and adjacent to archaeological sites in the Veracruz deltaic tinct marker of transport conditions. This component, with areas will provide new insight on the timing and nature of typical flake shape and densities ranging from -----2.8 to 3.1, is altered climate fluctuation. Marked changes in fluvial dis­ often associated with elongate and less dense plant frag­ charge regimes and deterioration in local subsistence resourc­ ments; mica also commonly occurs with light minerals and es likely had a direct impact on Prehispanic Veracruz cultural lithic aggregates of somewhat smaller size in the wetlands centers (WILKERSON, 1976, 1994a,b). It is also possible that and on the flood plain of the delta. The above and other trans­ the petrologic approach used in this study can help determine port-related petrologic associations are determined from Ta­ if such climatic and hydrologic fluctuations occurred random­ ble 1 and Figures 7 and 8. ly or cyclically, and perhaps if they were related to larger­ Upon reaching the coast, composition and texture of the scale phenomena such as EI Nino events (cf MEGGERS, fluvial sediment is laterally displaced by moderate to high­ 1994a,b). energy marine processes (Figures 5D-F). The higher dis­ charge of Tecolutla fluvial runoff and sediment has caused a CONCLUSIONS greater seaward progradation, and a consequent gentle cus­ pate shape, of its coastal delta margin (Figures 2, 3). In con­ There is close correlation between petrologic attributes and trast, the Nautla's smaller discharge has resulted in a trun­ delta subenvironments in the Nautla and Tecolutla deltas on cated and coast-parallel delta shoreline (Figures 2, 5F). Lag­ the southwestern Gulfof Mexico margin. Although positioned concentrate sands rich in heavy minerals accumulate in re­ close to the base of highlands and receiving sediment directly sponse to the winnowing and selective removal of less dense from adjacent mountain sources, we show that composition (such as volcanics) and fine-grained particles. Wave-driven and texture of surficial deposits in some deltaic subenviron­ coastal currents that prevail along the shoreline and inner ments differ markedly from those of fluvial loads derived fur­ shelf(Figure 2) completely modify any remaining fluvial load ther upslope. Findings here indicate that transport processes,

Journal of Coastal Research, Vol. 16, No.2, 2000 Sediments in Tropical Deltas 481 responsible for distributing sediment onto the two Veracruz Waters, and Marine Life. Fishery Bulletin of the Fish and Wildlife deltas, induce more than the simple, seaward, grain-size fin­ Service, Volume 55, 89-98. LEIPPER, D.F., 1954b. Physical oceanography of the Gulf of Mexico. ing expected in a unidirectional flow system. In: GALTSOFF, P.S. (ed.), Gulf of Mexico; Its Origin, Waters, and Important and consistent variations of compositional and Marine Life. GALTSOFF, P.S. (ed.) Fishery Bulletin of the Fish textural characteristics in specific deltaic and contiguous and Wildlife Service, Volume 55, 119-137. coastal settings are largely a response to the different energy LENTELL" R.L., 1975. Depositional History of the Rio Tecolutla Es­ tuary, Mexico. MSc thesis (unpublished), University of Florida, conditions that prevail in such geographic environments. Ma­ 43p. jor controls of sediment displacement and deposition on delta LEOPOLD,L.B.; WOLMAN, M.G., and MILLER, J.P., 1964. Fluvial Pro­ plains include selective sorting and temporary storage of par­ cesses in Geomorphology. San Francisco: Freeman, 522p. ticles on deltas as related to flood discharge. In addition, pro­ MEADE, R.H., 1988. Movement and storage of sediment in river sys­ gressive sediment dilution occurs as downslope transported tems. In: LERMAN, A., and MEYBECK, M. (eds.) Physical and Chemical Weathering Geochemical Cycles. Dordrecht, Holland: river material reaches the sea and is dispersed by high-en­ Kluwer Academic, pp. 165-179. ergy wind and coastal current conditions. We expect that MEGGERS, B.J., 1994a. Archaeological evidence for the impact of these processes, causing significant modifications in the pe­ mega-Nino events on Amazonia during the past two millennia. trology of the original fluvial load, prevail in most deltas. As Climatic Change, 28, 321-338. MEGGERS, B.J., 1994b. Biogeographical approaches to reconstruct­ exemplified in other delta investigations, definition of petro­ ing the prehistory of Amazonia. Biogeographica, 70, 97-110. logic attributes in the surficial cover is essential to interpret MEXICO, DIRECCION GENERAL DE GEOGRAFIA, 1984. Carta de Ter­ subsurface Holocene facies recovered in sediment borings, renos y Conjuntos Estratotect6nicos de la Republica Mexicana. and reconstruct a delta's paleogeographic evolution through Scale, 1: 2,000,000. Instituto Mexicano del Petroleo and Instituto time and space. Nacional de Estadistica Geografia y Informatica (Geological map, in 2 sheets). Major compositional and textural variations of the type re­ MORAN-ZENTENO, D.J., and COLLABORATORS, 1994. Geologia de la corded in this study, if identified and correlated laterally in Republica Mexicana. (In English, translated and with additional dated subsurface sections in the Nautla and Tecolutla, would annotated bibliography by WILSON, J.C. and SANCHEZ-BARREDA, help detect significant phases of Holocene flooding and L.). Tulsa, Oklahoma: American Association of Petroleum Geolo­ gists, 160p. droughts in the past. Such events inevitably affected the evo­ PETTIJOHN, F.J., 1957. Sedimentary Rocks. New York: Harper, 718p. lution of Prehispanic cultural centers on the two delta plains PETTIJOHN, F.J.; POTTER, P.E., and SIEVER, R., 1973. Sand and in this Veracruz region by altering essential subsistence re­ Sandstone. New York: Springer-Verlag, 618p. sources and inducing human migrations. SELF, R.P., 1971. Petrology of Holocene Sediments in the Rio Nautla Drainage Basin and the Adjacent Beaches. Ph.D. dissertation (Un­ ACKNOWLEDGEMENTS publ.), Rice University, Houston, Texas, 112 p. SELF, R.P., 1975. Petrologic changes in fluvial sediments in the Rio We thank Dr. S.J.K. Wilkerson, director of The Institute Nautla drainage basin, Veracruz, Mexico. Journal ofSedimentary for Cultural Ecology of the Tropics, Veracruz, Mexico for en­ Petrology, 45, 140-149. STANLEY, D.J., 1978. Pumice gravels in the Riviere Claire, Marti­ couraging this study, providing logistical support and kindly nique: selective sorting by fluvial processes. Sedimentary Geology, hosting us at his Institute during our fieldwork. Laboratory 21, 161-168. assistance was provided by Messrs. W. Boykins and J. Win­ STANLEY, D.J., and CHEN, Z., 1991. Distinguishing sand facies in gerath, and technical assistance with manuscript preparation the Nile Delta, Egypt, by stained grain and compositional com­ ponent analyses. Journal of Coastal Research, 7, 863-877. by Mr. N. Tandon, all with the Department of Paleobiology, STANLEY, D.J.; CHEN, Z., and SONG, J., 1999. Inundation, sea-level NMNH. Our appreciation is also expressed to Drs. S. Soter rise and transition from Neolithic to Bronze Age cultures, Yangtze and A.G. Warne for their useful reviews. The project was Delta, China. Geoarchaeology, 14, 15-26. funded by grants from the National Geographic Society, STANLEY, D.J.; WARNE, A.G.; DAVIS, H.R.; BERNASCONI, M.P., and Smithsonian Institution (Scholarly Studies Program, Office CHEN, Z., 1992. The late Quaternary north-central Nile Delta from Manzala to Burullus lagoons, Egypt. National Geographic Re­ of the Fellowships and Grants, National Museum of Natural search & Exploration, 8, 22-51. History Walcott Fund), TCTPF-China, and China NSF TAMAYO, J.L., 1962. Geografia General de Mexico, Geografia Fisica, (Grant No. 49971011). Vol. 2 (2n d Edition). Instituto Mexicano de Investigaciones Econ­ arnicas, Mexico, 648p. liTERATURE CITED WILKERSON,S.J.K., 1976. Huastec presence and cultural chronology in North-Central Veracruz, Mexico. Symposium on the Huasteca, ETHRIDGE, F.G.; GOPINATH, T.R., and DAVIES, D.K., 1975. Recog­ Congres International des Americanistes, Paris, 22p. nition of deltaic environments from small samples. In: BROUS­ WILKERSON, S.J.K., 1983. So green and like a garden: intensive ag­ SARD, M.L. (ed.), Deltas, Models for Exploration. Houston, Texas: riculture in ancient Veracruz. In: DARCH,J.P. (ed.), Drained Field Houston Geological Society, pp. 151-164. Agriculture in Central and South America. Manchester: BAR In­ FREUNDT, A., and ROSI, M. (eds.), 1998. From Magma to Tephra. ternational Series 189, 55-90. Elsevier, Amsterdam: 318p. WILKERSON, S.J.K., 1994a. El Pital & the Prehispanic Settlements GARFIAS, V.R., and CHAPIN, T.C., 1949. Geologia de Mexico. Edito­ of the Lower Nautla River Drainage, veracruz, Mexico. Veracruz: rial Jus, Mexico, 202p. Institute for Cultural Ecology of the Tropics, 116p. HEBDA, R.J.; SIEMENS, S.H., and ROBERTSON, A., 1991. Stratigra­ WILKERSON, S.J.K., 1994b. The Garden City of El Pital: the genesis phy, depositional environment, and cultural significance of Holo­ of classic civilization in Eastern Mesoamerica. National Geograph­ cene sediments in patterned wetlands of Central Veracruz, Mex­ ic Research & Exploration, 10, 56-71. ico. Geoarchaeology, 6, 61-84. YAN, Q., and Xu, S.Y. (Eds.), 1987. Recent Yangtze Delta Deposits (in LEIPPER, D.F., 1954a. Marine meteorology of the Gulf of Mexico, a Chinese with English summary). East China Normal University, brief review. In: GALTSOFF, P.S. (ed.), Gulf of Mexico; Its Origin, Shanghai, 438p.

Journal of Coastal Research, Vol. 16, No.2, 2000