THE HYDROGRAPHY and SEDIMENTS of the GULF of VENEZUELA1 John M

THE HYDROGRAPHY AND SEDIMENTS OF THE GULF OF VENEZUELA1 John M. Zeigler Woods Hole Oceanographic Institution, Woods Hole, Massachusetts

ABSTR4C1’ It is the object of this paper to examine the sediments of the Gulf of Vcnezucla and to relate them to the hydrography. The hydrography is, therefore, described in some detail. New data support the general conclusions of Rcdfield (1955) that the Gulf of Vcnezucla and Calabozo Bay opcratc as two estuarine cells driven by the trade winds and by outflow from Lake Maracaiho. New data also show that water enters the Outer Gulf from the northwest and flows at mid-depths towards the southeast in response to upwelling taking place off the coast of Paraguanh. Sediment parameters that reflect circulation best arc shells of pelagic foraminifera, sand distribution, and silt clay distribution. Of unknown importance is the distribution of resistant fecal pellets and the distribution of carbonate (shell material), The distribution of foraminifera indicated that water does not enter Calabozo Bay over the central part of the sill but rather along both coasts. The silt clay distribution reflects an outflow of water from Calabozo Bay into the Outer Gulf over the sill, a flow that had been first seen in the hydrographic stations. Rainfall data showed the great dependence of water character in the Gulf of Vcnczuela on rainfall and outflow from Lake Maracaibo. Hydrographic stations made in 1958, following a protracted dry period showed differcnccs in salinity of l-3%, between stations occupied in 1954 that wcrc taken following 2 months of heavy rainfall,

IN’IT3ODUCTION RV Atlantis (November 1958 and Febru- The Gulf of Venezuela is a large kid- ary 1960). Eight days were spent collect- ncy-shaped embayment along the north- ing bottom samples, hydrographic data, ern coast of South America between Venc- plankton tows, fish tows, and water sam- zuela and Colombia. It is approximately ples for nutrient and geochemical studies. 200 km long on a line between Aruba and The position of the hydrographic and sedi- the entrance to Lake Maracaibo and 84 ment stations are shown in Fig. 1. The km wide between Espada Point and Para- hydrographic stations are numbered, the guan& Waters from the open Caribbean penetrate the gulf over a broad shallow shelf ( Fig. 1). The Gulf of Venezuela is divided into two parts, Calabozo Bay and the Outer Gulf, by a submarine sill that markedly influences the water movements between the two basins. The basin of Cal- abozo Bay is 9 m deeper than the deepest portion of the sill. The bottom of the Outer Gulf slopes toward the Caribbean Sea somewhat regularly from the coast of Falc6n to the sudden deepening at the shelf edge. Four small islands off the Los Monges group in the northwestern part of the Outer Gulf are within the area of study. Field data presented in this paper (Ta- ble 1) were collected on two cruises of the

l Contribution No. 1395 from the Woods Hole Oceanographic Institution, Woods Hole, Massa- FIG. 1. Hydrographic stations, Gulf of Vent chusctts. zuela. Bathymetric contours in meters. 397 398 JOHN M. ZEIGLER

TABLE 1. Elydrographic stations in the Gulf of Venezuela, November 1958

Station Location Sample Tcmpcr- Oxygen number d;jP’: nture (ml/liter) Latitude Longituclr m cc> 5698 12"14' N 7O"lO' W 33, 0 25.93 36.590 4.17 10 25.92 36.594 4.17 20 25.63 36.598 4.03 30 24.52 36.640 3.28 5699 12"20' N 70"09' W 44 1 26.22 36.515 4.21 10 26.21 36.518 4.31 20 26.19 36.526 4.29 30 25.97 36.549 4.10 40 25.01 36.619 3.78, 5700 12"lO' N 70"20 W 55 1 27.16 36.548 4.31 10 27.15 36.548 4.37 20 25.03 36.598 4.26 30 23.70 36.706 3.69 40 23.12 36.737 3.15 50 23.03 36.739 2.97 5701 ll"60' N 70"20' W 35 1 27.11 36.584 4.27 10 27.08 36.582 4.32 20 25.48 36.589 4.05 30 24.76 36.613 2.95 5702 ll"50' N 70"20 W 42 1 27.24 36.584 4.75 10 27.21 36.583 4.34 20 27.14 36.578 4.27 30 24.91 36.600 2.97 40 23.84 36.690 2.79 5703 ll"50' N 70"25' W 55 1 27.43 36.578 4.27 10 27.32 36.571 4.32 2#0 27.21 36.568 4.26 30 26.42 36.567 4.24 40 23.98 36.685 3.27 50 23.80 36.695 2.84 5704 ll"50' N 70"28' W 56 1 27.43 36.5635 4.26 10 27.23 36.563 4.27 20 27.19 36.566 4.21 30 24.83 36.640 4.16 40 23.95 36.700 3.56 50 23.43 36,.714 2.97 5705 ll"58' N 70"26' W 64 1 27.22 36.574 4.32 10 27.22 36.575 4.37 20 25.93 36.547 4.32 30 24.39 36.644 4.19 40 23.39 36.718 3.26 50 23.27 36.734 3.15 60 23.31 36.701 2.86 5706 12"20' N 70"30' W 71 1 27.72 4.32 10 27.23 36.552 4.32 20 27.23 36.545 4.32 30 25.50 36.598 4.28 40 23.61 36.697 4.05 50 22.48 36.717 3.44 70 21.92 36.746 3.19 5707 12"20' N 70"40' W 78 26.92 36.556 4.37 1; 26.97 36.560 4.35 20 26.83 36.551 4.41 30 24.87 36.610 4.08 40 23.34 36.699 3.51 50 22.96 36.716 3.37 70 22.77 36.714 3.29 HYDROGRAPHY AND SEDIMENTS OF TIIE GULF OF VENEZUELA 3,99

TABLE 1. Continued

Station Location Water Sample Tcmpcr- number ature Latitude Longitude yz$ “c”“Bm (Cl 5708 12”OO’ N 70”40 W 65 1 27.20 36.550 4.32 10 27.20 36.601 4.28 20 27.04 36.544 4.31 30 24.74 36.655 4.20 40 23.65 36.687 3.91 50 23.05 36.786 3.16 60 22.81 36.735 3.19 5709 11”51’ N 70”40’ W 56 1 27.32 36.588 4.21 10 27.21 36.593 4.30 20 27.19 36.578 4.28 30 25.09 36.593 4.03 40 24.15 36.652 3.11 55 23.84 36.688 2.95 5710 ll”50 N 70”50’ W 51 1 27.86 36.352 4.41 10 27.62 36.363 4.50 20 27.08 36.547 4.40 30 25.02 36.~87 4.14 40 24.64 36.636 3.58 50 24.54 36.618 3.24 5711 12”35’ N 70’50’ W 78 1 27.22 36.481 4.24 15 27.20 36.494 4.29 30 26.97 36.582 4.27 45 24.37 36.652 3.6-i’ 60 23.12 36.729 3.43 75 23.12 36.728 3.40 5712 12”35’ N 71”OO’ W 75 1 27.12 36.558 4.27 15 27.13 36.554 4.30 30 26.95 36.557 4.30 45 26.43 36.620 4.50 60 23.61 368.709 3.62 70 23.40 36.711 3.29 5713 12”20’ N 71”OO’ W 43 1 27.32 36.458 4.32 10 27.32 36.460 4.59 20 25.23 36.601 4.62 24.19 36m.666 3.73 2 24.11 36.670 3.69 5714 12”OO’ N 71”OO’ W 20 1 27.25 36.566 4.32 5 27.26 36.574 4.27 12 27.28 36.562 4.30 5715 ll”50 N 71005’ W 12 1 28.29 32.912 4.43 10 28.09 36.280 4.32 5716 ll”40 N 71”20’ W 25 1 28.60 32.932 4.50 10 28.47 32.9 10 4.48 20 28.34 33.227 3.81 5717 11”20’ N 71”50’ W 16 1 27.85 27.012 4.51 12 28.34 31.990 3.75 5718 ll”20’ N 71”40’ W 24 1 28.03 30.684 4.47 10 28.19 30.843 4.57 20 28.26 33.449 4.18 5719 ll”30’ N 71”40’ W 24 1 28.19 32.410 4.21 10 28.19 32.410 4.21 22 28.33 33.639 3.70 5719A ll”30’ N 71”34’ W 24 1 28.21 32.883 4.54 10 28.22 32.897 4.50 22 28.25 34.026, 3.97 400 JOHN M. ZEIGLER

TABLE 1. Continued

Station Location water Sample Tempcr- Oxygen number ature (ml/liter) Latitude Longitude w m d(%tp (Cl 5720 ll"30' N 71"30' w 23 1 28.03 28.6rll 4.62, ,5 28.04 28.667 4.64 10 28.06 28.662 4.64 5721 ll"O6' N 71"20' W 9 1 28.34 34.743 4.48 7 28.14 34.876 4.64 5722 ll"20' N 71"20' W 17 1 28.18 33.857 4.46 6 28.20 33.842 4.48 13 28.16 34.054 4.27 5723 ll"20' N 71"OO' w 17 1 28.34 34.456 4.27 7 28.35 34.453 4.28 14 28.32 34.440 4.287 5724 ll"40 N 71"00' w 17 1 28.14 33.374 4.24 5 28.11 33.382 4.2'1 13 28.08 34.352 3.92 5725 ll"20' N 70"40' w 2,l 1 26.70 36.627 4.09 8 26.60 36.6286, 4.08 5726 ll"30' N 70"40' w 30 1 26.95 36.595 4.24 14 26.97 36.599 4.27 28 26.95 36.601 4.23 5727 ll"40' N 70"40' w 36 3 27.02 36.605 4.23 13 27.01 36.603 4.21 23 27.00 36.599, 4.16 33 26.70 36.599 3.62 5728 ll"40' N 70"26' W 43 1 26.96 36.599 4.27 10 27.00 36.599 4.28 20 27.02 36.593 4.27 30 24.21 36.663 3.89 40 24.01 36.684 3.22 5729 ll"30' N 70"30' W 30 1 26.62 368.607 4.21 10 26.63 36.609 4.20 20 26.56 36.605 4.09 28 26.29 36.603 3.78 5730 ll"30' N 70"20' W 30 3 24.87 36.648 3.54 16 24.87 36m.640 3.58 29 24.76 36.638 3.36 5731 ll"40 N 70"20 W 42 1 26.82 36.606 4.23 14 26.83 36.610 4.24 27 23.85 36.695 2.79 40 23.73 36.705 3.22 sediment stations are not. All hydro- made in nonwindy seasons the existing graphic stations, including some made in limited hydrographic descriptions will be 1954 by the Atlantis, were occupied dur- incomplete. ing the windy season. Previous writings, ACKNOWLEDGMENTS summarized by Murphy ( 1936)) show that this part of South America has a wet and The assistance of the officers and crew a dry season and that the strength of the of the RV Atlantis and of the author’s col- trade winds that are a primary driving leagues is gratefully acknowledged. The force of surface water is also seasonal. The study would not have been possible with- outflow from Lake Maracaibo also fluctu- out the encouragement and cooperation ates from year to year ( Carter 1955). of the Republic of Venezuela through Therefore, until hydrographic surveys are its representative Dr. Ramon Perez Mena HYDROGRAPHY AND SEDIMENTS OF THE GULF OF VENEZUELA 401 of the Ministry of Mines and Hydrocar- NE bons, I want to thank the Creole Pctro- leum Corporation for its hospitality and interest, particularly to Dr. Parke Dickey and Mr. L. A. Earlston Doe. Dr. Dickey was also kind enough to read the manuscript. Dr. A. C. Redfield gave generously of his time to reading and interpreting this manuscript. His assistance is gratefully acknowledged. The writer wishes to thank Mrs. Barbara Gill and Mr. Carlyle Hayes for laboratory analyses and Professor K. 0. Emery for his valuable suggestions regard- ing the manuscript. The cruises and subsequent work were financed by the Geophysics Branch of the Office of Naval Research under Contract Nonr 2196( 00)-NRO 83004.

IIYDROGRAPIIY A description of the hydrography of the Gulf of Venezuela, based on a survey made in December 1954, has been pub-

lished by Redfield (1955). Our observa- FIG. 2. Cross section A-A’, Gulf of Venezuela, tions made in November 1958 generally showing vertical distributions of tcmpcrature, sa- confirm his conclusions but are somewhat linity, and oxygen. more detailed. The circulation of the gulf appears to be The Maracaibo outflow appears as a influenced by the wind, the introduction tongue of low-salinity water that projects of freshwater from Lake Maracaibo, and toward the center of Calabozo Bay, the submarine sill. Both s~u~ys were flanked on either side by water of greater made during the period of strong north- salinity. The distribution of the isohalines east trade winds. The outflow of fresh- suggests a westward indraft of saline wa- water from the lake was much greater in ter along the coasts, with a compensating 1954 than in 1958 and resulted in lower escape of water across the central part of salinities and some differences in the pat- the sill. This distribution is somewhat dif- tern of circulation, ferent from that observed by Redfield in The circulation of the Gulf of Venezuela 1954. In 1954, the volume: of water de- consists of two estuarine cells delimited by rived from Lake Maracaibo was much the submarine sill that separates Calabozo greater, as evidenced by the lower salini- Bay from the Outer Gulf. Freshwater from ties of the bay water. The outflow ap- Lake Maracaibo dilutes the water of Cala- pears to have spread northwestward along bozo Bay, particularly in the upper layers. the coast and to have suppressed the in- Over the sill and on its eastward slope, flow along the Guajira coast, thus distort- where the surface waters of Calabozo Bay ing the salinity pattern without changing mix with those of the Outer Gulf, there is its fundamental character. a marked change in salinity (Figs. 2 and The hydrography of the Outer Gulf ap- 3). The distribution of salinity at the sur- pears to be dominated by upwelling that face ( Fig. 3) indicates the position of this occurs off the western coast of the penin- transition zone. sula of Paraguana as the result of the 402 JOHN M. ZEIGLER

72” 710 700 guanj coast but rise to the surface in the offing of the Gulf of Coro. The horizontal extent of the upward . . . .*. water movements illustrated in this section is shown in Fig. 6, in which the contours show the depths of specified isolines. The contours for the 24C isotherm and the 36.7& isohalinc (Fig. 6A,B) suggest that the indraft of dee,p water moves in a southeasterly direction from offshore. Such water does not upwell to the sea . . surface. The contours for the 26C isother- mal surface and the 36.6%0 salinity surface break the surface off the Gulf of Coro ( Fig. 61),E). Water of this salinity extends westward into shallow water along the coast of Falcbn and northwesterly where it appears to overflow less saline water lying at a greater depth (Fig. 6E). The temperature of the upwelled water ......

. . 3.:. ,.L

TEMPERATURE

Frc. 3. Distribution of temperature and salinity 1 m below the sdrfacc, Gulf of Vcnezucla. northeast trade winds (Fig. 4). The net- work of stations occupied in 1958 permits the description of this phenomenon in some detail. Fig. 5 shows the distribution of salinity, temperature, and oxygen along a section that approaches the peninsula from the northwest, then turns to the south and west parallel to the coast. These isolines all slope upward in a similar pattern. The isolines for 36X$& salinity, 26C, and 4 ml/liter of oxygen all rise from a depth of about 50 m at the continental slope to 20 m as the coast is approached. FIG. 4. Cross section B-B’ showing upwelling Thev remain at this depth along the Para- near the coast of Paragud. HYDIlOGRAPIlY AND SEDIMENTS OF TIIE GULF OF VENEZUELA 403 appears to have increased on reaching the surface, and in conscquencc water of 26C 0 is more limited in its extent at the surface. The contours for the surface of 4 ml/ 2. liter of oxygen follow a pattern similar to that of 36.6%0 salinity and 26C. The 3 ml/ 4. liter of oxygen surface does not extend seaward across the gulf as do those for 60 60 36.7%0 salinity and 24C. This is explained by the fact that oxygen is not a conscrva- 80 60 tive property, as are temperature and sa- TEMPERATURE, “C, linity, at the depth in question. No water o of such low oxygen content was observed offshore. The low oxygen content of the 20 upwelling water can be attributed to the decomposition of organisms that sink from cr, 40 the surface, and it is enhanced in accord- 5 ante with the countercurrent principle, $ 60 60 whereby the surface waters enriched by upwelling move more or less in a direction 60 60 opposite to the deeper current (Redfield SALINITY, %o 1955). Water containing less than 3 ml/ 0 liter of oxygen occurs only close to the bottom, where sinking organisms may bc 20 expected to accumulate, and in a zone that extends seaward in the direction from 40 which the temperature and salinity sur- faces indicate the upwelling water to have 60 60 come. In the upper layers of the Outer Gulf, 00 the highest salinities are found in the southeastern quarter surrounding the rc- OXYGEN, me/liter gion off the Gulf of Coro, where upwelled water appears at the surface (Fig. 6). The lowest salinities occur off the coast of Gua- jira and southward along the mixing zone T7--r 11;. 5. Vertical distribution of tcmpcraturc, separating the Outer Gulf from Calabozo salinity, and oxygen showing indraft of water Bay. This distribution suggests a general from the northwest toward tllc Paraguan,i coast. northwesterly movement of the surface waters with a concentration of the major Los Mongcs, as observed during both sur- outflow along the northeast coast of veys. In the Outer Gulf and in the eastern Guajira. part of Calabozo Bay, the differences in The principal difference between the salinity of the surface waters occurred con- water of the Gulf of Venezuela as ob- sistently at all positions sampled and in served in 1958 and as observed in 1954 is general amounted to nearly lsO. In the its greater salinity. Fig. 7 shows the ver- western part of the bay the differences tical distribution of salinity, temperature, were grcatcr, averaging about 3g0. It is and oxygen content at stations in similar impossible to calculate exactly how much positions in the western part of Calabozo more freshwater had diluted the Carib- Bay, the central part of the Outer Gulf, bean water of the gulf in 1954 than in and at the edge of the continental shelf off 1959, but it was approximately 1.8 x 1O1” 404 JOHN M. ZEIGLER

-TEMPERATURE, SALINITY AND OXYGEN ISOTHERMS -BOTTOM CONTOURS IN METERS

FIG 6. Charts showing the depths of specified isolines of temperature, salinity, and oxygen over the Gulf of Venezuela. m3. Of this, about two-thirds of the fresh- tionally high rainfall following 2 years water difference is found in Calabozo Bay of average precipitation. In contrast, the and the remainder in the Outer Gulf. survey of November 1958 followed a pro- Data on the outflow of Lake Maracaibo tracted period when the outflow was small make it clear that the differences in the and frequently negative, that is, when wa- salinity of the gulf are related to variations ter flowed from the gulf into Lake Mara- in rainfall in the Maracaibo Basin, These caibo. When the data for the 12 months data were supplied by the Creole Petro- immediately preceding the surveys are leum Corporation and are based on esti- added together, the annual exchange is as mates, made in accordance with the pro- follows: cedure of Carter ( 1955)) of the monthly Exchange 1954 1958 rainfall corrected for evaporation. The Outflow positive 1.78 x 10” mx 0.18 x 10’” m3 estimated outflows for the years 1952 Outflow negative through 1958 are shown in Fig. 8. The ( inflow ) 0.45 x lOlo rn’ 0.84 x 10” m3 outflows vary greatly from month to Net outflow 1.33 X 10” m” -0.66 X 10’” m” month and frequently arc negative when Very little of the lake water returns to the evaporation exceeds precipitation. This lake during periods of inflow, when the condition occurs especially during winter, inflow is composed of gulf water that is the dry season. The survey of December only slightly diluted. Consequently, the I954 was made after 2 months of excep- positive outflows give the best indication HYDROGRAPHY AND SEDIMENTS OF THE GULF OF VENEZUELA 405

27’C,5 16 37x.s3 44 m///qer of the effect of lake water in diluting the 023 24 25 26 .’ IXI gulf for the two periods. ARUEA CHANNEL ‘T 1 iT I The difference in the volume of outflow 10- ,.A i \ I in 1954 and 1958 is 1.6 x UYO m3. This is about the same as the estimated difference in the water content of the gulf. It is probable that the residence time of water in the gulf is not more than 1 or 2 years. Of this, about two-thirds of the time is spent in Calabozo Bay and one-third in the Outer Gulf.

SEDIMENTS Bottom sediment in the Gulf of Vene- zuela is composed of three principal corn-- ponents: terrigenous silt clay, terrigenous sand, and shells ( carbonate). Tedgenous sand Terrigenous sand is defined as grains of terrigenous minerals that did not pass a 8 40 screen with hole size of 0.062 mm. The $ 5. sand is composed mostly of quartz but includes minor amounts of heavy minerals and feldspars. Five areas where sand is more than a trace can be seen (Fig. 9). The large sand area in the southeast corner of the gulf is probably a relict deposit that washed through the Gulf of Coro before the strait was closed on the east by the tombolo connecting Paraguana to the 20- mainland. Bits of broken barnacles, cal- careous a1 gae, bryozoans, and heavy shelled mollusks are mixed with sand. These animal remains were probably washed from the shallow Gulf of Coro at the same time. Sand in the central part of the gulf, off Espada Point, was probably eroded from sandy outcrops on the sea floor, One of the samples contained many oolitcs and Frc. 7. Comparison of tcmpcraturc, salinity, bits of massive limestone. Limestones and oxygen at stations occupied in 1054 and in have been examined in formations exposed 1958. on shore by the author and reported by Renz ( 1956)) and since neither oolites nor sand were found between the offshore Sand near the Los Monges Islands prob- sandy area and the coast, it is assumed ably was derived from those islands or that some of the coastal formations from outcrops now below the sea. Rocks cropped out on the sea floor and were of the islands, as reported by Carmona eroded when sea level was lower and wa- and Bellizzia ( 1952, p, 6, S), are igneous ter turbulence greater. Present depth of or metamorphic hornblendites and aplites water in this sandy area is 60-65 m. that contain quartz stringers. How fast the 406 JOHN M. ZEIGLER

+ 10

52 53 54 55 56 57 58 YEAR Frr,. 8,, Estimated outflow from Lake Maracaibo into Calabozo Bay 1952-59 and dates o f 1 graphic surveys. present islands are being eroded or how the sediment in Calabozo Bay and extends much quartz is being supplied to the sea outward into the Outer Gulf (Fig. 10). is not known. It is surprising that no Most of the fine material is produced by striking increase could be seen in the erosion along the Guajira or Falcon coasts amounts or kinds of heavy minerals pres- or is carried in suspension in the outflow ent in the sediments near the islands. Pos- from Lake Maracaibo. Nearshore circula- sibly the recent rate of erosion does not tion under the influence of trade winds supply large amounts of material to the drives the muddy coastal water into Cala- sea. bozo Bay, which is a natural topographic Sand eroded from the Guajira and Fal- collecting basin. Most of the fine material con coasts is being moved westward near settles out before the water enters the cir- the coast by waves driven before the trade culation of the Outer Gulf. However, the winds. The result is that sand is accumu- muddy water that escapes from Calabozo lating along the western end of Calabozo Bay does so across the center of the ridge, Bay. Fine sand, which is present in small and a tongue of silt and clay, outlined by amounts all over the Gulf of Venezuela, is the 60% contour, concides closely with the probably carried from the land by wind. zone of escape. That the less saline water escaping from Calabozo Bay is taking this Silt clay route is also indicated by Fig. 6D, which The term silt clay includes all materials shows the 26C isotherm greatly depressed smaller than 0.062 mm and not soluble in by the outflowing water over a tonguc- acid. A blanket of fine material dominates shaped area. HYDROGRAPIIY AND SEDIMENTS OF TIIE GULF OF VENEZUELA 407

WEIGHT PER CENT TERRIGENOUS SAND CONTOUR INTERVAL .0, t IO PER CENT

72O 71” 70” 72’ 710 I FIG. 9. Distribution of tcrrigcnous sand, Gulf of Venczueln. FIG. 10. Distribution of terrigenous silt clay, Gulf of Vcnczuela.

Carbonate (shells) inantly pelagic, that is, broken and whole Broken and whole shells form a major shells of globigerina and swimming mol- sedimentary constituent in the outer part lusks plus other foraminifera. The more of the Gulf of Venezuela. Inasmuch as all central and eastern shelly areas are char- of the carbonate particles readily visible acterized by mixtures of heavy shelled, under low magnification were shell frag- badly worn (often only hinge lines re- ments except for the one sample that con- maining) mollusk shells, and foraminifera, tained oolites and broken limcstoncs, it both pelagic and agglutinated. The inner was concluded that a map of the carbon- edge of the carbonate deposits is domi- ate content of the sediment would repre- nated by pelagic foraminifera and thin sent a reasonable approximation to shell fragile shell fragments. Parker and Curray content. Samples were analyzed for ap- ( 1956) and Curray (1960) found shell de- proximate carbonate content by acidifica- posits in deeper waters in the Gulf of tion and back-titration ( Jackson 1958). Mexico that were composed, at least in The carbonate content is generally low part, of the shells of nearshorc organisms. over Calabozo Bay and most of the south- Among other things, these authors used ern part of the Gulf of Venezuela, except the occurrence of the shallow-water forms for two areas off the Gulf of Coro; the as evidence that some of the shell deposits carbonate content increases abruptly to were, in reality, relict shorelines, The rel- the north, and one can describe the sedi- atively narrow transition zone in which the ments of the northern part of the Gulf of shell content increases rapidly from 20 to Venezuela as muddy or sandy shell dc- 40% in the Gulf of Venezuda suggests posits. that this zone may also rcprescnt a relict The shelly areas can be distinguished shoreline similar to those in the Gulf of somewhat in terms of dominant character- Mexico; however, R. II. Parker (personal istics; for example, the area around Los communication ) failed to find distinct Monges contains many algal bank-type or beach or lagunal deposits (based on shells) reef-type carbonate remains such as bryo- in 17 samples selected from the zone of zoans or small corals. Sediments along the increasing shell content, although he did deeper outer edge of the Gulf of Venczu- find shallow-water shells in some of the ela, and particularly near Aruba, are dom- samples. Inasmuch as the entire Gulf of 408 JOHN M 1. :ZEIGLER

washing in dilute hydrochloric acid, al- WEIGHT PER CENT though they bleach to a gray color. Inves- FECAL PELLETS tigations of the mud, which was kept frozen, failed to reveal organisms that could be responsible for the pellets (How-

. . . . ard Sanders, personal communication ) ; however, it is known that some polychaete 12” worms produce pellets similar to those we found ( Moore 1955). Inasmuch as these pellets survived drying and screening, it is assumed that their resistant nature is their unusual characteristic, because fresh pellets would have been destroyed during analysis. If this assumption is correct, one might speculate on the meaning of this well-defined distribution in terms of a 72” 71” 70” process such as glauconitization, which would cause fragile pellets to become J?K. 11. Distrihtion of resistant fecal pellets,

hydrography indicates that the bottom 71” 70’ currents are probably weak. Perhaps the PELAGIC FORAMINIFERA pellets represent the range of some as yet unidentified animal, or they may indicate an environment that is favorable to glau- conitc formation. The close relationship between glauconite formation and organic matter has been pointed out by Takahashi I and Yagi (1929) and by Takahashi (1955), who also believe that an oxygen-low cnvi- ronment is conducive to glauconite formation. Perhaps the fecal pellet distribution thercforc represents the distribution of reducing conditions. Pyritized tests of foram- inifcra were collected from core samples in Calabozo Bay. Pyrite was present from near the surface of the cores to 3 m below 72’ 71” 700 the surface, indicating the prescncc of a reducing environment at least below the FIG. 12. Numbers of pelagic foraminifera per bottom-water interface. Our hydrographic gram of dry scdimcnt, Gulf of Vcnezucla. studies revealed low concentrations of ox- Foraminifera ygen in Calabozo Bay. Redfield reported even lower values at a time when the out- Shells of dead benthonic and pelagic flow of lake water was greater. It is pos- foraminifera constitute a visible and im- sible that during the summer wet season, portant part of the sediment of the Gulf when the outflow of the lake is cvcn of Venezuela. The various species from greater, organic production is higher, and these samples were identified by Bermu- winds are lighter, the water over the bot- dcz ( 1960). During routine examination tom may become anoxic. A period of calm of sediment constituents, it became appar- coupled with an increase in organic detri- ent that benthonic foraminifera occurred tus from lake discharge would probably all over the Gulf of Venezuela, including use up the already small amounts of oxy- Calabozo Bay, but pelagic species were gen present. Production of organic matter entirely missing from the greater part of in Lake Maracaibo is among the highest Calabozo Bay ( Fig. 12). in the world, and the outflow from the Pelagic foraminifera drift with the cur- lake through Tablazo Bay is directly into rents, and their preferred habitat seems to Calabozo Bay, the main locus of the pel- be the open ocean seaward of the conti- lets. The physical setting, therefore, leads nental shelves, although they probably one to believe that oxygen-poor waters are also live and reproduce to some extent the norm and anoxic bottom water is not over the more seaward parts of the broad itipossible. The present study cannot es- shelves. In a sense, therefore, their shells tablish validity for the idea that partially represent microscopic drift bottles and glauconitized fecal pellets represent oxy- under some conditions might be useful to gen-low environments, but it does present track oceanic currents that invade near- supporting evidence. Coincidentally, Van shore waters or open bays such as the Gulf Andel and Postma (1954) reported the of Venezuela. This hypothesis was tested largest concentrations of fecal pellets in by counting the number of shells of pe- the Gulf of Paria near Trinidad to be in lagic species in dried sediment samples the Guiria Trough, a basin within the from the Gulf of Venezuela ( Fig. 12). In gulf near the Venezuelan coast, No hydro- spite of the numerous errors that can af- graphic data for the trough were given. fect the count, such as errors in sampling, 410 JOHN M. ZEIGLER

graphic features of the Gulf of Venezuela. Thus, terrigenous sand predominates along the coasts of Calabozo Bay where wave action provides its source and wind-driven currents its distribution, while terrigenous silt clay underlies the central part of the basin. In contrast, shell fragments make up the major part of the bottom deposits in the deeper waters of the Outer Gulf. Pelagic species of foraminifera are limited almost wholly to the Outer Gulf and pene- tratc the warmer, less saline, and more tur- bid waters of Calabozo Bay only in small numbers, being carried by wind-driven currents along the coasts. In some areas where hydrographic data are lacking, the distribution of sediments, and particularly FIG. 13. Circulation in the Gulf of Venczucla. of the pelagic foraminifera, may bc used Heavy arrows represent movement of clccper wa- to supplement details of the circulation. ter, light arrows upper water layers. In Fig. 13, the general character of the circulation of the Gulf of Venezuela, as splitting, or counting, the overall distribu- deduced from the hydrography and sedi- tion is thought to be significant. Not only ment distribution, is shown diagrammat- are foraminifera scarce in Calabozo Bay ically. (entirely absent from 23 stations in the Water draining from Lake Maracaibo central part of the bay), but those few across Calabozo Bay must remain more in that are present are confined to two bands the basin than close to the coast, even of water adjacent to the two coasts. They though the tongue of fresher water coming are present in small numbers over the sub- out of the lake will be shifted more to the marine sill, but their numbers increase one west and north in times of strong trade or two orders of magnitude in a short dis- winds. Inflow of saline water into the bay tance seaward, while they disappear alto- is confined to the two coasts, while dis- gether in going in the opposite direction charge from the bay is across the central to the central basin of Calabozo Bay. The part of the sill. The primary inflow of writer interprets this to mean that very water into the Gulf of Venezuela is in the little open gulf water enters Calabozo Bay deeper layers from northwest to southeast. directly across the sill, but rather that This water, in turn, upwells along the water from the open gulf enters the bay coast of ParaguanA. Wind drift in the along both coasts and water from Cala- shallow layers is an important part of the bozo Bay moves seaward over the sill. mechanism, but this drift does not pene- This circulation must be reasonably per- trate far into Calabozo Bay, turning in- mancnt, because the shell counts represent stead toward the coasts or toward the open samples of the bottom that required years northwest part of the gulf. Circulation in to deposit. Counts from cores taken in Calabozo Bay is weak, and water spends Calabozo Bay show that the scarcity of twice as much time there as in the Outer pelagic foraminifera persists some meters Gulf, this number being based on compari- below the present sea bottom at the places son of hydrographic stations taken in 1954 where cores were taken. and 1958. SUMMARY REFERENCES I~EILMUL)E’L, I’. J. 1900. Foraminifcras plancto- The distribution of sediments has been nicos clcl Golfo de Venczucla. Man. Sot. Ci- shown to be related to the general hydro- enc. Nat. La Salle, 20( 55). IIYDROGRAPIIY AND SEDIMENTS OF TIIE GULF OF VENEZUELA 411

CAIWONA, C. L., AND A. BELLIZZIA. 1952. Re- shelf, northwest Gulf of Mexico. Bull. Am. conocimicnto Gcol6gico dcl Flanco nor-ori- Assoc. Petrol. Geologists, 410: 2428-2439: cntnl de la Sicrre dc Pcrijk e Isln do Toas, REDFIELD, A. C. 1955. The hyclrography of the Estado Zulia y dcl Archipi&go dc Los Mon- Gulf of Venezuela. Papers Marinc Biol. jes, depcndcncia Federal. 18 p. Occnnog. Deep-Sea Rcs., 3: (Suppl. ) 115- CAIITM, D. B. 1955. The water balance of the 133. Lake Maracaibo Basin during 1946-1953. RENA, 0. 1956. Crctaccous in wcstcrn Venczu- Drcxel Inst. Tcchnol., Lab. Climatol., Publ. ela ancl the Guajira ( Colombia ). ( Unpub- Climatol., 3 : 209-227. lishcd paper prescntcd at the 20th Intern. CURRAY, J. R. 1960. Scdimcnts and history of Gcol. Congr. Mexico City, 1956. ) Holocene transgression, continental shelf, northwest Gulf of Mexico, p. 221-266. In TAKAHASIII, JUN-ICIII. 1955. Synopsis of glau- F. P. Shepard and others, [eds.], Rcccnt scd- conitization, p. 503-512. In P. Trask, [cd.], Am. Assoc. Pe- imcnts, northwest Gulf of Mexico. Am. As- Rcccnt marine sediments. soc. Petrol. Geologists, Tulsa, Oklahoma. trol. Geologists, Tulsa, Oklahoma. -. 1961. Late Quatcrnary sea lcvcl: a AND T. YAGI. 1929. Peculiar mucl- cliscussion. Bull. Gcol. Sot. Am., 72: 1707- g&ins and their relation to the origin of 1712. glauconite. Econ. Gcol., 241: 838-852,. A~OORE, D. G. 1955. Rate of deposition shown VAN ANDEL, T., AND H. POSTMA. 1954. Rcccnt by relative abundance of foraminifera. Bull. scdimcnts of the Gulf of Paris. Vcrhandcl. Am. Assoc. Petrol. Geologists, 39: 1594- Koninkl. Ned. Akad. Wetenschap. Afdcl. Na- 1600. tuurk. Sect. I, 20. 244 p. MURPIIY, 13. C. 1936. Oceanic birds of South ZEIGLER, J. M. 1959. Scdimcntary cnviron- America, v. 1. Am. Museum Nat. Hist., mcnts on the continental shelf of northern New York. 640 p. South America. Preprints, Intern. Occanog. PARKER, R. II., AND J. R. CUWAY. 1956. Fauna Congr., New York. AAAS, Washington, D.C. and bathymetry of banks on continental p. 670.