Turtle distribution

Hawksbill distribution in our study area closely follows the contour of hard-bottom areas (sparse-hard bottom and coral reef habitats).  Our detailed turtle distribution maps (see the Maps page) highlight the importance of hard-bottom areas for hawksbills.  We believe that more marine benthic maps and inventories are needed to assist in detecting and monitoring other hawksbill aggregations in the region.  

Turtles captured

To date, we have made a total of 886 turtle captures, corresponding to 849 hawksbills and 37 green turtles.  Of the hawksbill captures, 127 corresponded to previously tagged individuals.  Hawksbills captured ranged from 18- 69 cm of straight carapace length (SCL), but the majority of turtles have between 25 and 35 cm SCL.

At present, we can not explain the near-absence of large juveniles and adult-sized turtles. However, given the fact that size-class structure has remained constant for the past 6 years, it is probable that: 1) hawksbills are transient in the area and as they grow they leave, and/or 2) these turtles have low survival rates, and numerous small recruits keep reaching the area. 

Site Fidelity

The time interval between hawksbill captures ranged from 23-1519 days (mean = 404).  Distances moved from a previous capture ranged from 13 to 10,503 meters (mean 598).  Most displacements, however, were of less than 1000 m, suggesting a high site fidelity for most individuals.  Except for two individuals, all recaptured turtles have been found at the same site where they had been first tagged, despite the proximity of other known hawksbill sites. Similar observations of site fidelity in juvenile hawksbills have been reported by Limpus (1992) and van Dam and Diez (1998) in Australia and Puerto Rico, respectively. Thus, is seems crucial for hawksbill conservation that such sites are protected, since many depend on them for a number of years. 

However, we have also found that hawksbill site fidelity does not always occur.  We found two individuals that moved over > 9,000 m from Playa Norte and near-by Muelle Oeste to Bahía de las Aguilas in the south.  This suggests that at least some of these young turtles are capable of undertaking greater movements than what we had previously believed within or outside the study area.  Also, the fact that both turtles moved south could indicate that the northwestern sites are transient habitats for these turtles as they recruit into the area.  Furthermore, for the first time in 2003, we received news of international recapture of some of our tagged turtles.  In February 2003, a Miskito fisher capture one of our tagged hawksbills near Cayo Gorda, approximately 125 km NW of Barra Kruta in the coast of Honduras (reported by Josiah Townsend, from the Florida Museum of Natural History, Gainesville).  Another recapture was made in July of 2003 near La Guajira, Colombia (reported by Carlos Pinzón, of the Santa Marta Turtle Foundation).  Both individuals were greater than 35cm SCL, suggesting that perhaps when our turtles reach a certain size, they take off and migrate to other areas.  Currently we lack the data to support this hypothesis.  Only more recapture data, on a long term basis, can settle this question.   

Growth

Growth rates calculated for captures over an interval greater than a year ranged from about 2 to 10 cm / year (mean = 5).   No significant differences were found between size-classes in growth rates, however, this could be due to the under-representation of larger juveniles in our sample. However, we detected significantly different growth rates according to site. Turtles at Lanza Zó grew at a significantly greater rate than turtles at all other sites (mean = 7 cm/year).

Abundance trends

The overall mean of observed hawksbills per unit of effort (OPUE) for 2000-2002 was 4.4 animals per hour (range 1-9.5).  When plotting the OPUEs by site, certain sites seem to have consistently higher OPUEs than others. However, this difference was not statistically significant when we analyzed data for 2000-2002.  Only the years of 2000-2002 were selected for hypothesis testing because it was only after 2000 that the length of our unit of survey effort was standardized to one hour.  None of the observed OPUE differences could be attributed to a year effect in this data set either.