1 Beaches and coastal dunes constitute the most significant accumulations of subaerially exposed sediment along coasts (Short 1999). Several natural forces are responsible for beach profile changes, and these include waves, tides and currents (Johnson 1919; Cambers 1998). While tidal cycles are important in some coastal processes, tidal range is of much greater geomorphic significance because it controls the vertical distance over which waves and currents are effective along the shoreline (Masselink and Short 1993). Tides generate tidal currents which flow landwards as the tide is rising and seawards as the tide is falling. These tidal currents can move beach sand and are particularly important between islands and at river mouths where flow velocities are high (Cambers 1998). In addition to the tidal cycles that occur over the short-term, a study by Gratiot et al. (2008) has demonstrated that the 18.6 year nodal tidal cycle contributes significantly to regional coastal changes. Currents are important because of their role in transporting sediment, and to a smaller extent, erosion. More recently, Thomas et al. (2013a) showed that beach rotation took place over centurial timescales, with sediment exchange between headlands taking less than one year. Thomas et al. (2013b) also demonstrated that medium and short term beach oscillation occurred between offshore islands and an embayed beach.

2 Nearshore processes shape beach morphology (Hardisty 1990), but there is feedback where morphology influences the processes at work in the nearshore (Komar 1976). Beaches and their adjacent nearshore zones act as buffers to wave energy. As a result, they are sensitive to change over various timescales ranging from a few seconds to several years (Carter 1991). Beaches can both adapt their shape very quickly to changes in wave energy, and also dissipate this energy with minor adjustments of the position of each sand or shingle grain (Pethick 2001).

3 The works of Shepard and LaFond (1940), Shepard (1950, 1973), Bascom (1954), Hayes and Boothroyd (1969), and Norcross et al. (2002) on mid-latitude beaches, have all demonstrated that the beach and near-shore exchange sediment on a winter-summer cycle. They demonstrated that the low, flat swell waves during the summer build up the berm, which causes the beach face to prograde seawards forming a steep profile. During the winter period, high steep storm waves erode the beach face and transport the sediment seawards where it is deposited to form a long-shore bar. Cambers (1998) work on Caribbean beaches has also alluded to the existence of this winter-summer cycle. Subsequent research by Short (1979) and Nordstrom (1980) indicates that the terms “winter” and “summer” are somewhat misleading in that beach response is cyclic, as opposed to seasonal. Their work suggests that there may not be a defined seasonal pattern to beach change, but that the beach will go through cycles of erosion or accretion.

4 The morphology of a beach profile is transitory, and rarely attains a static equilibrium form, due to continual variation in littoral zone processes (King 1972). King (1972) recognized two major types of profiles; the “summer and winter” profiles. Other terms used to describe these two profiles are “storm and normal” by Johnson (1949), and “storm and swell” by Komar (1976).

5 The Institute of Marine Affairs (1986, 1998, 2004), Singh (1997) and Darsan (2005), have all conducted research on beaches in Trinidad, in particular on Cocos Bay and Nariva Swamp. Their research has highlighted the severe coastal erosion at work in this bay, and has also alluded to seasonal changes to the beach morphology. However, their research did not account for the causes of the morphological changes, specifically at shorter timescales.

6 The main aims of this paper are to (a) model the effects of tidal cycles on beach dynamics in an open bay, low-latitude tropical trade wind setting; and (b) address the applicability of the “summer and winter” profiles in this type of environment.

7 Trinidad and Tobago, the most southerly of the Caribbean islands, is situated between 10° 02' to 10° 50' N latitude and 60° 55' to 61° 56' W longitude (Fig. 1). It is located on the continental shelf of South America and immediately adjacent to the outflow of the Orinoco River, and this determines to a great extent the nature and form of its coastal and marine environment. Trinidad and Tobago has a tropical maritime climate with two distinct seasons; a wet and dry season. The prevailing wind system is the north-east trades, and the dominant wind direction is from the north-east in the dry season (December to May), and from the east in the wet season (May to November).

8 The dominant ocean current influence in Trinidad and Tobago is the northern branch of the South Equatorial Current, the Guiana Current. As the Guiana Current approaches Trinidad and Tobago it divides into two streams, with the inner stream passing into the Columbus Channel in a predominantly westerly flow and then into the Gulf of Paria, while the outer passes up along the east coast of Trinidad (Hudson 1988) (Fig. 2). Trinidad’s coasts are influenced by the discharge of the Orinoco River, and the Guiana current that flows along the east coast of South America (Andel 1967). The Orinoco discharge subsequent to the rainy season in Venezuela affects the velocity of the Guiana current due to changes in sediment load and density (Kenny 2008); which in turn may have implications for wave energy and sediment budgets along the east coast of Trinidad.

9 The tidal regime experienced is a function of the tide waves from both the Caribbean Sea and the Atlantic Ocean. The tidal regime is semi-diurnal with periods of approximately 12.5 hours. At high spring tides the maximum range is 1.2 m with some slight variation from north to south. At other times, the tidal range is less than one metre, averaging 50 to 60 cm between high and low tides (Kenny 2008).

10 The Northern hemisphere's winter period with associated swell waves is most often experienced in the Caribbean between the months of October to April, and is usually caused by intense mid-latitude storms in the North Atlantic Ocean (Cambers 1998). During the winter months, swells attack the coast from directions varying from northwest to north-east. On the east coast, deepwater wave attack is primarily from north-east to south-east, with an easterly approach being the most significant.

11 The east coast of Trinidad is rugged in its northern section where the rocks of the Northern Range outcrop. Further south the coastline becomes gentler, where in the central regions the Nariva wetlands are found (Bertrand et al. 1992). The east coast of Trinidad is extremely varied with three stretches of low coast separated by prominent headlands. Manzanilla beach is a barrier beach system found in Cocos Bay, embraced by two prominent headlands at Manzanilla Point and Radix Point (Fig. 2). Cocos Bay has been classified as an open sea beach (Georges 1983).

12 The Cocal area includes the Manzanilla beach in Cocos Bay and the Nariva Swamp. The Cocal sand bar known locally as Manzanilla beach, has a relationship with the Nariva Swamp whereby the swamp is protected from the marine environment by the barrier beach. The beach protects the swamp from salt water intrusion, helping to maintain the correct salinity levels that promote life in this wetland ecosystem (Darsan 2005). The characteristics (including the geology and hydro-geology) of the sand bar are not well known having received little attention (Environmental Management Authority 2001). The Cocal sand bar has a fairly low topography with some sections below sea level (Environmental Management Authority 2001).

13 The Manzanilla beach at Cocos Bay located along the Manzanilla/Mayaro Main Road, is about 20 km long, and borders the landward edge of the Nariva Swamp. The barrier beach protects the Nariva Swamp and its distinct wetland ecosystem from the marine environment. The back of the barrier beach has not been driven inland with coastal retreat, and as such, sections of the barrier are becoming progressively narrower.

14 There is also significant erosion along several parts of the Manzanilla beach (Singh 1997; Darsan 2005), particularly near the Nariva River mouth; attributable to fresh water outflow and tidal inflow dynamics. The Nariva River also carries large quantities of particulates and nutrients to Cocos Bay which has implications for marine biota and productivity (Bacon et al. 1979). At several points along Manzanilla beach, the sand bar has been eroded from fresh water outflow and sea water inflow; creating points where salt water is able to directly penetrate and alter salinity in the Nariva Swamp (Environmental Management Authority 2001).

15 Data were collected on a diurnal, lunar (tidal), monthly and seasonal basis to reveal the changes that occur diurnally, from spring to neap tides, monthly and seasonally from dry season (winter period) to wet season (summer period). In order to address the applicability of the “summer and winter” profiles in this tropical environment, data from the dry season and wet season were analyzed against the corresponding winter and summer seasons experienced in mid-latitude beaches.

16 Data for the dry season (winter period) from 15th–23rd December 2005 were collected initially over one week to obtain diurnal changes. During this week, 3 stations (north, central, and south) were monitored in the morning and afternoon, corresponding to low and high tide cycles. Data collection included field measurements, sediment sample collections, and observations. After this diurnal data collection phase, data were collected over a lunar cycle to get an appreciation for the effects of spring and neap tides on 31st December 2005 and 6th January 2006 respectively at all 9 sites along the beach. Sites were selected based on accessibility and an attempt to get data coverage across the bay. Sites 8 and 9 had to be relocated and re-named 8B and 9B respectively after revetment works commenced in the southern section of the bay (Fig. 3). Data collection for these subsequent phases was done whenever possible at low tide, to maximize beach profile lengths.

17 Data for the wet season (summer period) from 11th– 18th June 2006 was also collected on a diurnal basis in order to compare the results with the data from the winter period for the same 3 sites. These 3 sites were also monitored for an entire month to understand how the beach profiles responds to lunar changes; from Full Moon on 11th June 2006 to Full Moon on 11th July 2006. Thereafter, all 9 sites were surveyed for a spring and neap tide on 25th June and 2nd July 2006 respectively to monitor these tidal effects on the coastline, and also to compare them with the surveys from the winter period.

18 Data was subsequently collected over a 14 month period from 9th August 2006–1st September 2007 on a monthly basis to investigate whether the beach actually goes through a yearly cycle, by being able to compare data with that of the previous year. Since data already existed for winter and summer periods in the previous year, these existing data could be cross-referenced with this new 14 month data to investigate any trends or cycles. This fieldwork was conducted once a month on every Full Moon corresponding to a spring tide.

40 Waves, tides and currents are responsible for beach profile changes (Johnson 1919; Cambers 1998; Saleh Salem et al. 2011). Masselink and Hughes 2003 noted the importance of not only tidal cycles, but also tidal range. Tidal range is of greater geomorphic significance because it controls the vertical distance over which waves and currents are effective along the shoreline (Masselink and Short 1993). Tides also generate tidal currents which flow landwards as the tide is rising and seawards as the tide is falling. Tidal type affects the intensity of tidal currents, as the velocity of water movement will be greater in semi-diurnal regimes than for mixed or diurnal types because of the shorter interval between high and low tide (Woodroffe 2002). At Manzanilla beach, the semi-diurnal tidal regime does play an influential role in the beach dynamics. The changes attributed to tidal cycles and seasonal changes are modeled below.

46 It should be acknowledged that results presented are based on observations of just under 2 years of data collection from December 2005 to September 2007. While a longer dataset would undoubtedly make the following conclusions more robust, the analysis has provided a window of understanding into the beach dynamics at Cocos Bay, Manzanilla. The beach changes which occur tidally and seasonally at Cocos Bay can be modeled. Diurnally, erosion occurs with rising tide and increased wave energy, whereas accretion takes place with falling tide. The general trend observed for the spring/neap cycle was erosion during spring tide conditions and accretion in neap tide conditions. The spring/neap tide changes are also controlled by seasonal changes. In the winter period (dry season), the erosion that occurred after spring tide was not repaired in neap tide; therefore the profile recorded negative change. In the summer period (wet season), the erosion after spring tide conditions was repaired in neap tide conditions due to the presence of and shoreward migration of bars, which had a relationship with lower wave energies.

47 The different pattern in the spring/neap cycle observed in the summer was due to the presence of bars and their relative positions on the profile. Three possible scenarios were modeled to show the beach changes in summer due to the formation and migration of bars:

1. Shoreward bar migration and beachface erosion due to trough’s location on the beachface.
2. Shoreward bar migration and beachface accretion as the bar becomes welded to the beachface.
3. Beachface erosion whose sediment fuels bar formation in the surf zone.

48 In terms of beach volume changes, the diurnal cycle exhibited the least changes, with average diurnal beach volume changes ranging between 0.03 m³/m and 0.42 m³/m of sediment. The beach volume changes during spring tide conditions (-2.87 m³/m) outweighed that occurring during neap tide conditions (0.17 m³/m). The average volume change across sites during the spring/neap cycle in winter was erosion of 3.53 m³/m, whereas it was only 0.42 m³/m in the summer period. The volumetric changes in the spring/ neap cycle during the winter period outweighed that during the summer due to stronger wave energy in the winter period.

49 This beach does not have a winter-summer cycle as exists in mid-latitude beaches. All sites do not respond similarly across seasons from the winter period to the summer period. While some sites eroded from summer to winter, others accreted. It is also apparent that the beach generally contained more sediment volume in the summer months during lower wave energies as compared to the winter months. However, the winter-summer exchange of sediment across sites was not confirmed. Manzanilla’s beach response may therefore be described as dynamic with some seasonal control.