Practical considerations regarding estimating Ne when generations overlap

Robin S Waples
DOI: https://doi.org/10.1101/2024.06.04.597384
2024-06-06
Abstract:Researchers studying species in nature often find it challenging to apply methods based on simplistic models of reality. Here I consider how some real-world complications influence demographic estimates of effective population size (Ne) when generations overlap. The most widely-used model (by Hill) expresses Ne as a function of variance in lifetime reproductive output (LRO) of the N1 members of a newborn cohort. The model by Hill assumes stable age structure and constant population size, in which case mean LRO =kbar = 2. In real-world applications, researchers often ask whether unbiased estimates can be obtained under the following conditions: (1) When kbar is not 2 for empirical data; (2) If cohorts are defined at a later age than newborns; (3) If survival to age at sexual maturity (alpha) is not random; (4) When some or all null parents (those with LRO=0) are not sampled. Using analytical methods and computer simulations, I show that: (1) Because variance in offspring number is positively correlated with the mean, Ne^ will be biased using raw data when kbar is not 2, but this bias can be overcome by rescaling var(LRO) to its expected value when kbar = 2. (2) The cohort can be defined at any age <=alpha, provided that (a) LRO data cover the full lifespan (e.g., production of newborns by newborns, or production of adults by adults), and (b) survival to age alpha is random. (3) If juvenile survival is family-correlated, defining cohorts at age alpha avoids upward bias in Ne^ that occurs if newborn cohorts are used. (4) Missing some or all null parents has no effect on Ne^, provided that data are rescaled to kbar=2.
Evolutionary Biology
What problem does this paper attempt to address?