Abstract
Effective population size (N(e)) is a quantity of central importance in evolutionary biology and population genetics, but often notoriously challenging to estimate. Analyses of N(e) are further complicated by the many interpretations of the concept and the alternative approaches to quantify N(e) utilising different properties of the data. Each method is also informative over different time scales, suggesting that a combination of approaches should allow piecing together the entire continuum of N(e), spanning from the recent to more distant past. To test this in practice, we inferred the N(e) continuum for 45 populations of nine-spined sticklebacks (Pungitius pungitius) using whole-genome data with both LD- and coalescent-based methods. Our results show that marine populations exhibit the highest N(e) values in contemporary, recent, and historical times, followed by coastal and freshwater populations. The results also demonstrate the impact of both recent and historical gene flow on N(e) estimates and show that simple summary statistics are informative in comprehending the events in the very recent past and aid in more accurate estimation of NeC , the contemporary N(e), as well as in reconstruction and interpretation of recent demographic histories. Although our sample size for each large population is limited, we found that GONE can provide reasonable N(e) estimates. However, due to challenges in detecting subtle genetic drift in large populations, these estimates may represent the lower bound of N(e). Finally, we show that combining GONE and CurrentNe2, both sensitive to population structure, with MSMC2 provides a meaningful interpretation of N(e) dynamics over time.