The study consisted of two parts: field work and laboratory work.

Field Work: We obtained a small amount of blood from 36 adult females nesting at Playa Grande, Costa Rica. The blood was stored on specially treated cards called IsoCode® (Schleicher and Schuell, Inc.) For 20 of the 36 females we transported their eggs from the nest to a nearby fenced-in hatchery. Since female Leatherbacks lay more than one nest during the season, we transported multiple nests to the hatchery. In some cases we were able to obtain up to four clutches from a single female.

When the hatchlings emerged from the relocated nests they were collected and a few drops of blood were taken for this genetic study. Then all of the hatchlings were released to the beach.

Laboratory Work: After six months of field work, we returned to the lab at Drexel University with the blood samples. DNA was isolated from the blood spots collected from the 36 female Leatherbacks and from over 1600 hatchlings.

Next we used the Polymerase Chain Reaction (PCR) to amplify several specific sequences of turtle DNA. We selected 5 microsatellite DNA sequences to study as markers for our analysis. Microsatellites are short arrays of DNA made up of simple, repetitive sequences. A typical sequence for a microsatellite is CACACACACACA. The sequence of this microsatellite can also be written as (CA)6 because there are 6 repeats of the simple sequence CA. The 5 microsatellites we studied were developed and used by other sea turtle researchers: Nancy FitzSimmons and Peter Dutton.

We analyzed DNA from the 36 females for each of these microsatellites to build a genotypic profile of the population. Of the 5 microsatellites, 3 (Cc117, Ei8, and Dc99) were chosen for the paternity analysis based on how informative they were. This means that there are different variants (or alleles) of each marker and that many turtles had two different alleles. Alleles can be noted by their size or by arbitrary number or letter designations. For example, one female may have alleles 3 and 5 for marker Cc117, and a second female may have alleles 2 and 4 for the same marker.

Next we compared the alleles for each marker in each hatchling, that is the hatchling genotypes with those of their mothers. The alleles observed in the hatchling that were not inherited from the mother were assumed to come from the father, under the expectations of Mendelian inheritance. Thus, looking at the genotype of a single hatchling, we could infer paternal genotype as illustrated in the example below for the case of single paternity.

When the genotypes of each hatchling within a single clutch or a family (multiple clutches from the same female) are counted, we summarize the genotype distribution as follows:

In this illustration, 25% of the hatchlings have each of the four possible genotypes. This is what is expected if there is no selection for a particular genotype. We then compared the observed distribution of genotypes with what is expected to see if there was significant deviation from expected in any of the clutches.

If more than two paternal alleles are observed, we concluded that there were two fathers or that there was a mutation in the microsatellite DNA sequence of the hatchling with that allele.

Results: Of the 20 families (several clutches within each family) genotyped, 60% demonstrated single paternity (that is the hatchling genotypes showed only 2 paternal alleles). This finding implies that the female mated with only one male prior to the nesting season and stored the sperm from one nesting cycle to the next.

Two of the 20 families displayed multiple paternity (more than 2 paternal alleles were present for at least one locus with relatively equal distributions throughout all clutches of a single female). This implies that a single female mated with two or more males prior to the beginning of the nesting season, and that the sperm was mixed and stored from one nesting event to the next. Mating with more than one male is termed polyandry.

In the remaining 6 families, only one or two hatchlings had an allele that did not match the mother or the paternal alleles of its siblings. It is not possible to determine whether these variations were due to a mutation or to multiple paternity. If they were due to multiple paternity, then the contribution of the second father was very minor. Going back to the illustration of the expected frequency of hatchling genotypes above, we could see that even though there were only two paternal alleles for some of the microsatellite markers in two of these six remaining families, there was significant deviation from the expected distribution frequency. Therefore, these two families may also demonstrate multiple paternity.

We did additional statistical analysis of the inferred paternal genotypes suggesting that it was possible that one male mated with two different females in two cases. This is termed polygyny.

If you would like to see the raw data for the female leatherbacks and all the hatchlings, you can download the PDFs listed below. You will need Acrobat Reader in order to view the PDFs.

Female_Genotypes.pdf
Hatchling_Genotypes.pdf
Nest_Data.pdf

Conclusions: Female leatherback turtles store sperm from one nesting to the next. They do not successfully mate after each clutch is laid.

Multiple paternity occurs with a frequency of at least 10%, but may be as high as 20-40%. Thus female leatherbacks exhibit a polyandrous mating strategy to some extent.

A single male probably mates with more than one female, polygyny. The reasons for this are not known, but we can speculate. There may be fewer males than females in the mating pool, perhaps because there are fewer males altogether. Or male and female behavior may take them to different areas of the ocean, and introduce a location bias when it is mating season.