Methods

On the Cayenne Peninsula, nest counts were taken every morning during the nesting season (early April to mid-August), together with nightly patrols (1900–0600 hours) to count nesting females. Based on this intensive monitoring effort, we assumed that no nesting attempt was overlooked on these beaches.

At Kourou, monitoring was limited to weekly nest counts for most of the nesting season. Daily nest counts were made during the peak nesting season only (May–June).

Monitoring at west oceanic nesting sites (Pointe Isère, Farez, Irakompapi, and Organabo) consisted of a daily census of nesting crawls (early May to mid-August). For the year 2002, no distinction was made between nesting and false crawls at Irakompapi, because the latter was considered exceptional on this wide beach. For all narrower nesting sites, differentiation between nesting and false crawls was made.

At Awala-Yalimapo, a daily count of nest crawls was performed (early April to mid-August) from 2001 to 2003. Nesting females were counted 1 night per week. In 2004, only the daily count of nest crawls was performed.

Because some oceanic nesting sites were not monitored over the entire nesting season, the total number of nests was estimated by applying the statistical model proposed by Gratiot et al. (2006):

• Where A = amplitude of the sinus function (highest number of nests estimated for the nesting season),

• tp = duration of the nesting season (in days),

• T = mean position of the nesting season in the year (in days, starting from 1 January),

• B = residual number of nests laid outside of the main nesting season (in number of nests). This corresponds to the period of the year during which nesting is incidental (less than 5 nests per week).

In this model, the adjustment of the nesting season to a sinus function provides an estimate of the annual number of nests, with an error lower than 15% when the monitoring effort is distributed all over the nesting season and exceeds 30 days (Gratiot et al. 2006).

Remote sensing data (SPOT satellite images) by using 50 satellite images, covering 1986–2004, were previously analyzed by Gardel and Gratiot (2004) to study the migration of mudbanks along the coast. The image data set principally contained some SPOT images with a resolution of 15 m. Depending on several parameters, including tide level during image acquisition and turbidity of coastal water, images were reanalyzed to estimate the dynamic of sandy beaches and their potential suitability for nesting.

Results

Because information for some remote sites is scarce (i.e., absence of continuous monitoring in Irakompapi nesting site) and because the counting of nesting crawls sometimes underestimates the real nesting effort, we present here the minimum numbers of leatherback turtle nests laid in French Guiana: 23,107 for 2001, 12,229 for 2002, 13,480 for 2003, and 11,012 for 2004 (Table 1). Over this time period, the beach of Awala-Yalimapo did not suffer major geomorphic change and has hosted a regular proportion of the leatherback turtle nests recorded in French Guiana (42% ± 2%). However, great fluctuations appeared in the number of nests recorded in each of the other nesting sites, giving highly variable individual contribution to the annual nest estimate for French Guiana (Table 1).

Table 1. Estimated number of leatherback turtle nests laid in French Guiana 2001-2004.^a

View Fullscreen

The French Guiana coastline is characterized by alternating zones of mudbanks and interbanks. During the mid-1980s, 6 coastline areas were welded by mudbanks (black lines labeled from 1 to 6 in the eastward direction; Fig. 1). The preserved beaches were located in the west (Awala-Yalimapo, Pointe Isère, Irakompapi areas) (SPOT image of 1986). In the early 1990s, mudbanks nos. 1 and 2 migrated westward and mud progressively covered the beach of Pointe Isère and Irakompapi. During this time period, the beach at Awala-Yalimapo was the only major sandy beach for the entire French Guiana coastline and thus the only suitable nesting site. A narrow beach was forming at the same time in Cayenne Peninsula (SPOT images of 1991 and 1992). Four years later, mudbank no. 5 left the Cayenne Peninsula, and the beach of Kourou received a large amount of sand, recovering suitable conditions for the nesting (SPOT images of 1995 and 1996).

Since the mid-1990s, major geomorphic events occurred around the mouth of the Mana River (Fig. 2). In 1997, a significant amount of mud migrated from mudbank no. 1 to accumulate on the beach of Awala-Yalimapo. As a consequence, this beach lost part of its attractiveness for marine turtles (SPOT image of 1997 and Landsat Image of 1998). At the same time, 2 large sandy beaches were formed between Pointe Isère and Irakompapi areas, offering new nesting possibilities. Since 2000, Awala-Yalimapo Beach has suffered from the arrival of mud, coupled with significant erosion of its sandy areas. At the same time, a crucial geomorphic change occurred at the mouth of the Mana Estuary. The spit-like feature that was developing along the right side of the Mana River became an isolated distal tip (Pointe Isère). As a result, the mud supplied from mudbank no. 2 accumulated in the former Mana River mouth (Fig. 2). Under the new hydrodynamic conditions, the sand supplied by the Mana River accreted predominantly along the isolated distal tip (Pointe Isère), progressively reducing the suitability of the Awala-Yalimapo Beach for nesting.

View Figure

• Download as Powerpoint
• Download Figure

In the Cayenne Peninsula region, the 2001–2004 time period was characterized by the migration of a small mudbank, probably part of mudbank no. 6. The interaction of the mudbank with beaches induced alternate phases of small-scale sand drift but was not associated with any phase of generalized siltation and preserved the nesting grounds.

Discussion

During the first years of the leatherback turtle survey in French Guiana, the main nesting areas were located on western remote oceanic beaches (Pritchard 1973). These nesting spots moved westward, to reach the east side of the Maroni River. The number of nests along the west beaches for the 1977–1979 period highlights the relative importance of each individual nesting site (Table 2). From year to year, the beaches bordering the east side of the Maroni Estuary (Les Hattes, Awara) welcomed an increasing percentage of nesting females (22%, 27%, and 52% in 1977, 1978, 1979, respectively), so that, from 1980 to 2001, the monitoring effort mostly focused on the Maroni Estuary area.

Table 2. Estimated number of leatherback turtle nests laid in western French Guiana nesting beaches (from Fretey and Lescure 1979).

View Fullscreen

Over the last 30 years, the beach of Awala-Yalimapo has represented the only permanent nesting beach in French Guiana. This nesting ground has been considered a representative estimator of the turtle nesting effort for the region, assuming that it was hosting more than 90% of the leatherback turtle nests laid in French Guiana (Girondot and Fretey 1996). This assumption was used until the late 1990s (Chevalier and Girondot 1998a; Chevalier et al. 1999).

Data gathered over the last decades led to the assessment of the leatherback turtle population size. Fretey and Lescure (1979) suggested an estimate of 13,966–19,596 nesting females. The exceptionally high nesting season observed in French Guiana in 1988 led former estimations to be reevaluated (Fretey and Girondot 1989). An estimated number of 5100–9700 nesting females was finally published for the Suriname/French Guiana region (Spotila et al. 1996).

Based on the same data set, population trends were estimated. An increasing number of nests had been noted from the mid-1970s until the early 1990s (Girondot and Fretey 1996). A sharp decline was then reported (Chevalier and Girondot 1998a), interpreted as a potential decline of the French Guiana leatherback turtle rookery (Chevalier et al. 1998), or even of the entire Guianas nesting aggregation (Chevalier et al. 1999). A shift of the leatherback rookery to other nesting beaches was foreseen (Chevalier and Girondot 1998a, 1998b) but did not seem relevant to explain the apparent decline of nests recorded in Awala-Yalimapo as an aerial survey did not reveal any other major nesting grounds (Chevalier et al. 1998).

Our analysis of remote sensing data highlights the fact that major geomorphic events, like beach creation or erosion, occur periodically along the coast of French Guiana. These coastal evolutions have undoubtedly influenced the relative importance of the Awala-Yalimapo nesting beach over the last decades. Indeed, a comparison of nest number trends recorded in Awala-Yalimapo, with the sandy shoreline evolution in French Guiana, reveals that the highest numbers of nests recorded in Awala-Yalimapo (late 1980s to early 1990s) coincide with the period during which no other beaches were available. Moreover, the observed nest decline in Awala-Yalimapo (late 1990s) corresponds with the creation of several other beaches along the coastline of French Guiana.

In light of the remote sensing data and the evolution of sandy beaches along the coast, the statement asserting that the number of nests recorded in Awala-Yalimapo was, for a long period of time, close to the total number of leatherback nests laid in French Guiana must be qualified.

Recent nest estimates show that, for the 2001–2004 period, Awala-Yalimapo still hosted a significant and remarkably stable percentage of the total number of leatherback nests recorded in French Guiana. But, together with remote sensing data, it also highlights the geographic evolution of the leatherback nesting effort in the region over the 1994–2004 period of time; whereas, Awala-Yalimapo was most likely the only beach available during the early-1990s, other beaches have progressively appeared, offering new nesting possibilities in the region. Depending on beach accretion or erosion, the relative importance of any single beach fluctuates considerably at a regional scale. Thus, it is unlikely that a single nesting site hosted a constant proportion of leatherback nesting attempts over multiple decades.

Most recent works on trend analysis in French Guiana and Suriname aimed at estimating potential numbers of leatherback nests that could have been laid on unmonitored nesting areas. When doing so, nest number trends seem to have been stable or have even shown a slight increase (Girondot et al. 2007). Nevertheless, because of the lack of reliable data for some nesting areas, several assumptions have been used, and a real understanding of leatherback nesting trends will need several years of regional beach monitoring to be clarified.

As previously proposed (e.g., Chevalier and Girondot 1998a; Chevalier et al. 1999), we recommend that each single beach be monitored during the leatherback nesting season in the Guianas. Such monitoring effort could be significantly reduced by distributing the scientific coverage throughout the nesting season (Gratiot et al. 2006). Monitoring 20 to 30 days per beach would be sufficient to assess nesting dynamics and would provide valuable information to update the regional conservation status of this critically endangered species.