I am not an expert, but I would guess that not all plant species are equal. Some types of plants are annuals and reproduce by producing thousands of seeds per year and have huge population sizes and are pollinated by the wind. Other types of plants such as some trees reproduce MUCH more slowly and are less promiscuous. Some plants rarely reproduce by seeds and pollination at all, and instead spread by clones.

I also would guess that is depends on whether you are looking a rather local population diversity, or some type of global diversity. Also, has the plant been carried around the world by human activity in the past 3,000 years or less, or was it spread around the world by natural forces over the past 100,000 years or more.

Even within some relatively highly related plants such as palm trees or apple trees, there a huge differences in human influence on individual species or even strains within a species.

Brian Thomas Foley Thank you for the clarification,,,,,,,,

Number of loci required will depend on context. e.g. How polymorphic the markers are, the scale of genetic differentiation you are looking at, what sort of signal you are actually trying to derive, etc.

Given some prior information, you can simulate data and do a power analysis to examine the capacity of to accurately diagnose using simulated data (i.e. we simulate two populations and could look at assignment accuracy w/ various numbers of msat loci, assuming some a priori level of polymorphism per marker). There are also many studies along these lines which have been published- try a Google Scholar search with keyworks "microsatellite power analysis" + whatever process you are interested in (e.g. 'hybridization", or "parentage", etc)

Thank you Tyler Chafin for your insight. It is helpful.