Biology Reference
In-Depth Information
used here would reflect uneven sex ratios and variation in reproductive
success of the previous breeding generation. In an extensive review by
Frankham,
69
0.42) among
species for which variation in reproductive success and uneven sex ratios
were taken into account to obtain demographic estimates of
N
e
. Thus, the
Ascaris
value falls just on the edge for what is known from single gener-
ation
N
e
/
N
c
estimates of other species.
The following may be a bit of an extrapolation because of the restrictive
assumptions of the island model,
74
but I think it is a useful exercise in
what genetic data and a
N
e
/
N
c
ratio might be able to tell us. Under the
assumptions of Wright's island model
47,74
genetic differentiation is
a function of subpopulation
N
e
and migration rate (
m
) where
the mean
N
e
/
N
c
ratio was 0.35 (95% CI: 0.28
e
1
4N
e
m
þ
1
$
F
ST
z
(8.3)
As discussed above, the island model might approximate the
Ascaris
population dynamics in Jiri. Thus, it seemed reasonable to estimate the
effective number of migrants per generation (
N
e
m
) from Eq.
(8.3)
. Using
a
F
ST
of 0.023
41
,
N
e
m
10.61. If the
N
eT
/
N
cT
ratio of 0.28 also represents the
ratio within subpopulations, then that means about 38 census worms/per
generation are migrants into the foci of transmission around households.
This does not mean all 38 census worms become adults or even infect
a person. It would be more appropriate to say a minimum of 10 migrant
census worms infect people (necessarily adult worm infections because
N
e
m
represents individuals that contribute to the gene pool), but up to 38
census worms infecting a household were acquired from another trans-
mission focus per worm generation. A key point here is “per worm
generation.”
Ascaris
adult worms live about 1 year in their host.
11
Thus,
one might conclude generation time is 1 year and, therefore, 10
¼
38
migrant worms per year cause infections. However, as noted above, the
long-lived egg stages of
Ascaris
will increase generation time. Thus, these
10
e
38 migrant worms will be spread out likely over several years.
Above I have focused on using single-sample estimators to estimate
the
N
e
of the parents that generated the infections in the sampled
households. One can also estimate long-term or coalescent
N
e
that reflects
the historical evolutionary dynamics of a population. Such an estimate
may provide a historical baseline for what the parasite's
N
e
was like
prior to the implementation of a control program. Waples
49
provides a
summary about estimating long-term
N
e
. Here, I illustrate estimation of
long-term
N
e
with the Jiri
Ascaris
data while also highlighting some of
the caveats discussed by Waples.
49
Long-term
N
e
requires an estimate
of
e
4
N
e
u
, which means an estimate of
u
is also needed. Importantly,
an accurate estimate of
N
e
via an estimate of
q ¼
will be dependent on
q
a reliable estimate of
u
;a10
change in
u
leads a 10
change in the
N
e
