Abstract
This paper attempts to explain circumstances under which local may be or may not be best. Natural selection may lead to local adaptation (LA), or it may be constrained by gene flow, founder effects, small population size, genetic drift, and archetype. 'Specialist' species display greater LA than 'generalist' species. Local genotypes are to a certain extent transient, being a consequence of past historical genetic patterns. Two recent meta-analyses found that while local performance exceeded the performance of a randomly chosen nonlocal population in 71% of comparisons, general adaptation across environments was as frequent as LA. Genotypes for restoration are most likely to be effective if they are adapted to current site conditions. As environmental change accelerates, both globally and locally, exceptions to 'local is best' may increase. For these reasons, 'local is best' may be better thought of as a testable hypothesis rather than as a general assumption. While either local or nonlocal plant material may be most effective for restoration practice depending on individual circumstances, local material will continue to be the first choice for restoration practitioners whenever this option is feasible and effective.