Time‐scales
• mitochondrial data can give little or no information
on recent events (perhaps within 1000 years –
depends on mutation rates),
• nuclear markers can give information on more
recent events, but probably not within the last few
generations,
• Estimates of effective population size must be within
b d fid li it
20
very broad confidence limits,
• Sample sizes for robust inference of gene flow,
• Cost and technical expertise,
• The “CSI factor” and unrealistic expectat