Compartmentalization in environmental science and the perversion of multiple thresholds

Nature and living organisms are separated into compartments. The self-assem bly of phospholipid micelles was as fundamental to the emergence of life an d evolution as the formation of DNA precursors and their self-replication. Also, modern science owes much of its success to the study of single compar tments, the dissection of complex structures and event chains into smaller study objects which can be manipulated with a set of more and more sophisti cated equipment. However, in environmental science, these insights are obta ined at a price: firstly, it is difficult to recognize, let alone to take i nto account what is lost during fragmentation and dissection; and secondly, artificial compartments such as scientific disciplines become self-sustain ing, leading to new and unnecessary boundaries, subtly framing scientific c ulture and impeding progress in holistic understanding. The long-standing b ut fruitless quest to define dose-effect relationships and thresholds for s ingle toxic agents in our environment is a central part of the problem. Deb ating single-agent toxicity in splendid isolation is deeply flawed in view of a modern world where people are exposed to low levels of a multitude of genotoxic and non-genotoxic agents. Its potential danger lies in the unwarr anted postulation of separate thresholds for agents with similar action. A unifying concept involving toxicology and radiation biology is needed for a full mechanistic assessment of environmental health risks. The threat of s ynergism may be less than expected, but this may also hold for the safety m argin commonly thought to be a consequence of linear no-threshold dose-effe ct relationship assumptions. (C) 2000 Elsevier Science B.V. All rights rese rved.