For this study, two beaches were patrolled: Ya:lima:po-Les Hattes (located on the mouths of Mana and Maroni Rivers) and Apo'tïlï-Pointe Isère (not influenced by the fluvial races; cf. map). The Ya:lima:po beach was patrolled from the night of 20 April 1987 to the night of 7 August 1987 (Fretey and Girondot, 1988). Volunters worked two hours and were relayed. For the 28,000 turtles seen, the time of observation and the stage of the nesting process were entered in a computer. Then, knowing the length of each nesting stage (Fretey 1981), estimated times of arrival were computed. The Apo'tïlï beach was patrolled from the night of 6 June 1988 to the night of 22 June 1988 and the same procedures were followed.

On Ya:lima:po beach there is a relationship between high tide and the time of leatherback's arrival on the beach. This phenomenon is visualized with the linear regression of the number of turtles per hour as a function of the date. However, for the seven lunar cycles, these regressions have slope values higher than the slope obtained with the times of high tide. The difference between the time of maximun number of animals arriving and the time of high tide is null when the high tide is in the evening, and from 3 (Vigie) to 5 (Bois Tombé 2b, or 'BT2b') hours when tbe high tide is in the morning (Figure 1).

We explain these observations by two antagonist mechanisms. On one hand, the carrier effect of the rising tide facilitates the arrival of turtles. On the other hand, the fluvial currents prevent the leatherbacks from arriving on Ya:lima:po beach. The Apo'tïlï beach is only subject to tidal currents. The northwest direction of the longshore Guiana current (Prost 1986) carries the fluvial currents far from Apo'tïlï beach. So, the turtles can arrive all the night on this beach with a slight peak at rising tides.

The hourly difference between high tide and the time of arrival can be interpretated by the same mechanisms. The fluvial currents are inversed earlier in the morning for spring-tides than in the evening for neap-tides. As referred to the high tide, the leatherbacks will arrive on the beach sooner in the morning than in the evening.

For morning (AM) high tides, we explain the difference between time of arrival in Vigie and in BT2b by a specific effect of these mechanisms on different parts of Ya:lima:po beach. We propose two explanations for this:

- The fluvial current is higher in Vigie than in BT2b (Girondot. personal observation)

- A sand bank appears in front of Vigie during low tides and many turbulences are provoked during neap-tides.

Figure 2 presents the percentage of arrivals per night on different parts of Ya:lima:po beach. The maximum of frequentation in Vigie was seen for full and new moons (spring-tides). This arrival in Vigie appears to be facilited when there are spring-tides. This effect is not obselved for other parts of this beach. These observations suggest that access on Vigie is submitted to high constraints that prevent the latherback's arrivals.

The objectives for this study are:

- To understand how the different currents in the estuary of Maroni and Mana River interact to influence arrival time of leatherback turtles,

- To study simultaneously the distribution of arrival time on all the estuary's beaches.

LITERATURE CITED

Fretey, J. and M. Girondot. 1988. Nidification de la Tortue Luth sur le littoral de Guyane française pendant la saison 1987. Ann. Soc. Sci. Nat. Char.-Mar. 7(6):729-737, maps, figs., tabls.

Prost, M.T. 1986. Observations sur I'évolution morphosédimentaire du littoral guyanais. Intl. Symp. Sea Level Changes and Quaternary Shoreline, Sao Paulo, Brasil, July l986. 16p, maps, figs.