In order to study the effects of environmental pollution on the genetic diversity of natural populations, two field populations of the cladoceran Daphnia longispina were sampled: one in a reference site (uncontaminated) and the other in a site historically stressed with acid mine drainage (AMD). Five hypotheses were formulated: (1) the stressed population presents a higher tolerance to lethal levels of copper than the reference population, (2) differences in tolerance to lethal levels, observed between the two populations, are due to the loss of sensitive lineages, (3) differences in tolerance to lethal levels of copper between the two populations are due to the appearance of new genotypes, (4) the acquisition of tolerance to lethal levels of copper involved changes in life-history patterns and fitness costs under optimal conditions, and (5) historical contamination by AMD resulted in tolerance differences to sublethal levels between populations, within categories similarly tolerant to lethality, specifically, lineages with similar tolerance to lethal levels from both populations show differences in tolerance to sublethal levels, the stressed population being more tolerant to sublethal levels of contamination than the reference population. Over 125 acclimated cloned lineages, from each population, were exposed to different copper concentrations for 24 hours. At the end of each assay, mortality and feeding inhibition were monitored. Life-cycle traits under optimal conditions were also monitored (time to first brood, number of neonates per brood, inter-brood time, body length, and ingested algae). At lethal levels of copper, significant differences were found in the frequency of sensitive lineages between the two populations. The stressed population did not include the most sensitive lineages, though the most tolerant ones were also present in the reference population. Thus, the hypothesis of presence of new genotypes in the stressed population resulting in an overall increase tolerance of the population at lethal levels was rejected. Changes in life-history patterns were detected, though they were not fully consistent with predictions of life-history theory. Furthermore, these changes in life-history patterns did not involve fitness costs. The fifth hypothesis was rejected since, at sublethal levels of copper, no significant differences were found between the feeding inhibitions of similarly lethal tolerant lineages of the two populations, with only one exception. Therefore, the present study further confirmed the genetic erosion hypothesis and gave further support to the incorporation of genetic diversity measurements into risk assessments.