Immediately after the tide and before the measurements
were taken, we removed one leg (fourth right walking
leg taken at the joint) from each participant in the
nesting group (i.e. the female, the attached male and all
satellites). The legs were frozen immediately in dry ice or
liquid nitrogen and transported from Seahorse Key to the
University of Florida where they were held at 80C for
later genomic DNA extraction. We allowed each crab to
recover in the wading pool for at least 1 h prior to release
(horseshoe crabs have a rapid clotting response and many
crabs have legs missing naturally). We saw most nesting
again on the following day (no mortality occurred).
We evaluated paternity for each clutch using the
methods described in Brockmann et al. (1994) (conducted
by C.N. and W.P.). Briefly, for adults, we ground a
1-cm3 sample of muscle tissue from each leg in a mortar
and pestle precooled with liquid nitrogen. For 16–24
larvae from each clutch, we ground the entire larva in a
1.5-ml microcentrifuge tube with a tube-shaped metal
pestle. We extracted DNA from these ground tissues using
standard phenol: chloroform procedures. We used each
genomic DNA as a template in a polymerase chain
reaction (PCR) to amplify a single microsatellite locus
developed previously. PCR products were resolved on
ethidium-stained, 7% polyacrylamide gels. We determined
paternity by comparing electrophoretic bands
of each larva with the mother, attached male and all
satellites that participated in the nesting.
Paternity analysis was possible with a single locus for
two reasons. First, the locus is extremely polymorphic,
with 22 alleles observed in 33 adults from our previous
study (Brockmann et al. 1994). Although we did not make
a precise assessment of the allelic frequencies in the population
of this study, inspection of the gels indicated a
genetic diversity similar to the previous study. That study
documented that all alleles had relatively low allelic frequencies,
with the most common allele having a frequency
of 0.14. So, even in the worst-case scenario where
the female, the attached male and one satellite male had
the six most common alleles, there was still a 52% chance
that any randomly drawn sperm from the population
would be recognized. The mean and median allele frequency
was 0.05, so the power to recognize extragroup
fertilizations was approximately 0.7, considerably greater
than this worst-case example. In our previous study,
fertilization by sperm with an allele not present in the
putative parents was only 3% (11 of 324). Even if all of
these were due to sperm left in the sand by other males
(some could be due to mutation, misgenotyping and mishandling
of samples), the rate is sufficiently low not to
affect the general trends and conclusions of either our
840 ANIMAL BEHAVIOUR, 60, 6