Biology Reference
In-Depth Information
authors have suggested that if sex estimation is achieved with DNA, then sex can be treated
as a known (
Hummel and Herrmann, 1991; Saunders, 2000; Stone, 2000; Gibbon et al., 2009
).
Sex estimation from DNA ideally should work well because the X and Y chromosomes have
distinct sequences (
Saunders, 2000; Stone, 2000
). However, there exist potential problems
with sexing from DNA alone that are discussed further by Cabana and colleagues in this
volume (Chapter 16). These authors point out the problem of false negatives. The absence
of regions specific to the Y chromosome does not indicate a female necessarily, because the
Y chromosome is already very small and genetic material is more easily degraded, especially
in bioarchaeological samples.
Stone (2000)
explains that there are two methods for sexing from DNA: repetitive
sequences (which are chromosome specific) and single copy genes found on both the
X and Y chromosomes (
Stone, 2000
). When using degraded ancient DNA, the repeat
sequences are easier to amplify. The presence of a male product must indicate the subject
is male, but the absence does not necessarily mean that the subject is female (
Stone, 2000
).
In a comparison of molecular sex estimation to morphological estimation in subadults, there
does not appear to be enough DNA left in the subadult skeleton to work successfully
(depending on the age of the individual), though the molecular method may ultimately be
more accurate when DNA is present (
Saunders, 2000
). Several problems with determining
sex from DNA are contamination, DNA degradation, and the great time and financial
investment (
Buikstra and Ubelaker, 1994; Saunders, 2000
) given that chemical/molecular
approaches to sexing in subadults can be greatly biased taphonomically by burial environ-
ment (
Stone, 2000
). For a discussion of the problems of contamination and DNA degradation
in genetic analysis, see Cabana et al. (Chapter 16), this volume.
Quantitative Sexing and Demography
In demography research, the estimation of sex involves calculating the probability that
the sex is male or female.
Konigsberg and Frankenberg
include a comprehensive introduc-
tion to methods of calculating sex ratios in demography later in this volume (Chapter 11),
but I will provide a brief overview. In order to calculate the probability that a skeleton
belongs to one sex or another, there must be a sample of individuals of known sex from
a similar population as the unknown to develop a sex ratio for that sample (number of males
to number of females). This information is prior information,
11
which is then used in the
calculation of the probability that the unknown is male. You calculate the probability that
the individual is male, because the probability that it is female is one minus the probability
of it being male. Developing a sex ratio for the sample in question is first accomplished by
using a reliable osteological assessment of sex, such as the Phenice method described earlier,
or with molecular methods. Ideally, multiple variables will indicate the sex of individuals in
this sample, so that there is little ambiguity. This approach is very different from traditional
sex estimation and assessment, which is described throughout this chapter. Instead of esti-
mating sex, the question becomes, “What is the probability of sex given the observation of
a certain feature?”
Konigsberg and Frankenberg
thoroughly explain this theoretical
11
This approach is Bayesian in nature. Refer to
Konigsberg and Frankenberg
(Chapter 11) and Uhl (Chapter 3)
for more information on Bayesian statistical approaches.
