Biology Reference
In-Depth Information
one of the goals is also to reduce genetic diversity/adaptive potential.
I did not have a means to estimate the
N
c
of the parents that produced the
sampled worms from each household especially since the parents are
likely from different breeding years (i.e. different
N
t
's;
Figure 8.1
A). As
a surrogate, I tested for a correlation between the “best” estimate-
N
e
's and
the infection intensities recorded for each house (
Table 8.2
), which
implicitly assumes large census populations beget large populations.
There was a significant correlation between infection intensity and “best”
estimate-
N
e
(
r
11). Thus, in these samples
N
e
may be
a good tracker of
N
c
. It would be encouraging if this result holds in future
studies because that means
N
e
estimates might be useful to monitor not
only adaptive potential, but also intensity data following an
Ascaris
treatment program. More data are certainly needed, but it is interesting to
speculate that these correlations suggest that the
Ascaris
subpopulations
do not exist at high densities (e.g. mean intensity per person was ~2.5 in
Jiri
62
) where an asymptotic relationship between
N
e
and
N
c
would be
relevant (
Figure 8.2
). In comparison, an asymptotic relationship may be
more pertinent in parasites that have high infection intensities per host
(hundreds to thousands) such as several trichostrongylid nematodes of
livestock. Interestingly, among nematodes, the latter group is largely
where drug resistance has been reported.
4,5
The overall metapopulation
N
e
(
N
eT
) is also of interest in relation to
dynamics that occur among subpopulations (e.g. equal subpopulation
contributions to the migrant pool versus extinction/recolonization
dynamics). My goal in this section is to compare an estimate of
N
eT
using
Wright's island model
71
to an estimate of
N
eT
from the single-sample
estimators. I caution the combining of samples across subpopulations
(and across years as in this data set) and the subsequent use of these
single-sample estimators has not been quantitatively tested as a means to
estimate
N
eT
. Thus, the following should be treated as a thought exercise
rather than definitive conclusions. I used the entire data set of 1094 worms
and obtained an LD-
N
e
estimate of 1062 (95% CI: 975
0.61,
p
0.047,
n
¼
¼
¼
e
1161, at the 0.02
cutoff) and SA-
N
e
estimate of 1645 (95% CI: 1502
1789). The harmonic
mean of these two estimates yields a “best” estimate of
N
eT
¼
e
1291. In
Wright's island model,
71
N
eT
is a function of subpopulation
N
e
and genetic
differentiation (
F
ST
) such that
N
eT
z
nN
e
1
F
ST
;
(8.2)
where
n
is the number of subpopulations and each subpopulation has
the same
N
e
. This model assumes that subpopulations contribute equally
to the migrant pool. As can be seen in Eq.
(8.2)
, as genetic differentiation
increases among subpopulations,
N
eT
can exceed the sum of the sub-
population effective sizes.
49,72
This is because while each subpopulation
