Abstract
The effective size of a population (N
), which determines its level of neutral variability, is a key evolutionary parameter. N
can substantially depart from census sizes of present-day breeding populations (N
) as a result of past demographic changes, variation in life-history traits and selection at linked sites. Using genome-wide data we estimated the long-term coalescent N
for 17 pinniped species represented by 36 population samples (total n = 458 individuals). N
estimates ranged from 8,936 to 91,178, were highly consistent within (sub)species and showed a strong positive correlation with N
([Formula: see text] = 0.59; P = 0.0002). N
/N
ratios were low (mean, 0.31; median, 0.13) and co-varied strongly with demographic history and, to a lesser degree, with species' ecological and life-history variables such as breeding habitat. Residual variation in N
/N
, after controlling for past demographic fluctuations, contained information about recent population size changes during the Anthropocene. Specifically, species of conservation concern typically had positive residuals indicative of a smaller contemporary N
than would be expected from their long-term N
. This study highlights the value of comparative population genomic analyses for gauging the evolutionary processes governing genetic variation in natural populations, and provides a framework for identifying populations deserving closer conservation attention.