Here is my original question:
Dear EvolDir folks,
I have a sample from a single cohort of juvenile frogs and am wondering if
I can apply the linkage disequilibrium method (as implemented in the
NeEstimator program) to this sample to get an estimate of the effective
number of the parents ( *N*b) that produced the juvenile cohort. Can
anyone verify or refute this? I have microsatellite genotype data for the
juveniles as well as the adults, so I will have an *N*b estimate from the
temporal method as well. I am hoping to use the LD method to get an
additional estimate of *N*b.
Any of your comments on this issue will be greatly appreciated!
Thank you,
Ivan
-----------
And here are the answers I received:
Yes, you certainly can get an estimate of parental Nb from your samples of
juvenile frogs (you can also get another estimate from your sample of
adults). You can get such an estimate from NeEstimator, but you might want
to consider as an alternative a program (LDNe) we wrote and have described
in a paper in press at Mol. Ecol. Notes. LDNe has a couple of advantages
over other methods based on linkage disequilibrium:
* it implements the bias correction described in Waples 2006 (Con.
Genetics 7:167-184.)
* It provides separate estimates based on different user-specified
criteria for excluding rare alleles. this facilitates evaluation of
potential biases from use of microsat data.
it is easy to use and reads standard input files (FSTAT, GenePop). you
can access the program, user's manual, and related files at
http://fish.washington.edu/xfer/LDNE/
you might also consider using OneSamp (recently described in an online
early paper by Tallmon et al. in Mol. Ecol. Notes). this program uses
approximate Bayesian computation methods to derive an estimate of effective
size based on a single sample.
for your temporal estimate, if you have sampled the adults non-lethally
you have Plan I samples and you should make sure you use an estimator that
accounts for the correlation in allele frequencies between parents and
offspring. this requires an estimate of N, the total number of adults your
initial sample was (presumably randomly) sampled from.
best wishes, and let me know if you have any questions.
Robin Waples
Regarding your question posted on EvolDir:
First of all I suspect you already have the references on this method?
Here are some - there may be more:
Bartely D, Bagley M, Gall G, Bentley B (1992). Use of linkage
disequilibrium data to estimate effective size of hatchery and natural fish
populations. Conservation Biology, 6, 365-375.
England PR., Cornuet J-M, Berthier P, Tallmon DA, Luikart G (2006). Estimating
effective population size from linkage disequilibrium: severe bias in small
samples. Conservation Genetics, 7, 303-308.
Hill WG (1981). Estimation of effective population size from data on
linkage disequilibrium. Genetical Research, Cambridge, 38, 209-216.
Waples RS (1991). Genetic methods for estimating the effective size of
cetacean populations. Report of the International Whaling Commission
(special issue 13), 279-300.
Waples RS (2005). Genetic estimates of contemporary effective population
size: To what time periods do the estimates apply? Molecular Ecology, 14,
3335-3352.
Weir BS, Hill WG (1980). Effect of mating structure on variation in
linkage disequilibrium. Genetics, 95, 477-488.
I've used Ne estimator to estimate the number of breeders from a single
cohort of amphibian progeny as you outline. With my dataset (40 tadpoles per
population and 9 microsatellite markers), I definitely received estimates
within a plausible range, when applying this method. Not knowing the actual
number of breeders in the various populations, I had no way of testing the
reliability however. I can only say that in general the results appeared in
the range we expected, and that large populations yielded a larger
Ne-estimate than small populations.
Nonetheless, I think you should be very careful to use any of the Ne
methods isolated, and the results should always be interpreted with caution.
To employ several methods in combination and use a large number of
polymorphic markers would of course be preferable.
I would be very keen to see the answers you get, so please post them, if
you receive anything useful. Also, I would appreciate to learn your opinion
and experiences with the method once/if you chose to apply it to your data.
Cheers,
Morten
Morten E. Allentoft
M.Sc., Ph.D. candidate
School of Biological Sciences
University of Canterbury
Private Bag 4800
Christchurch
New Zealand
-----------
Ivan C. Phillipsen
Department of Zoology
Oregon State University
Corvallis, OR 97331-2914
philliiv@science.oregonstate.edu