The effective size (N_e) of a population (and especially of a
small or endangered one) is an important quantity, affecting both the
inbreeding in the population and the rate of loss of genetic variability. 
Often one can observe the actual number of individuals in a population
(N), and use that information to infer N_e.  It is desirable to know
the ratio of N_e to N.  Many authors have attempted to estimate the ratio
N_e/N for particular populations and species, however the interpretation
of such quantities is difficult when population sizes fluctuate and
generations overlap.  A potentially more informative quantity, lambda,
is the ratio of the effective number of reproducing adults in a
single breeding season to the census number of reproducing adults in that
breeding season. Especially in species with high fecundity and high 
juvenile
mortality, it is difficult to estimate lambda or the effective size of a
population by demographic methods (you simply cannot count all the
reproductively successful offspring born to each parent!).  In such
situations, estimating the variance effective population size from the
observed change in allele frequencies over time is a possibility. I will
present a Markov Chain Monte Carlo method to estimate lambda by maximum 
likelihood from sample allele frequencies drawn over time from a 
population.