1. It is frequently asserted that the adaptation of metal tolerance involves a cost, i.e. tolerant individuals have a lowered fitness on uncontaminated soil. To date, however, evidence for a cost has been highly circumstantial. One well-known example of metal tolerance, copper tolerance in Mimulus guttatus, is used in this paper to search for a cost.
2. The absence of highly tolerant plants in the field, despite the ability to select, rapidly and artificially, for extreme levels of tolerance, indicates that strong selection against extreme tolerance occurs in this species.
3. Negative genetic correlations between plants selected for increased degree of tolerance and other fitness traits were sought to test the trade-off hypothesis that energy/resources needed to manifest tolerance are diverted away from other essential traits.
4. The case for using replicated selection lines when using this methodology to demonstrate a cost is put forward. It is argued that replication is required to show unambiguously that any correlations found result from the pleiotropic effects of the gene(s) that govern the adaptation and not the consequence of linkage disequilibrium.
5. No clear evidence to support the trade-off hypothesis was found.