92 Marine Mammals, Extinctions of Thus, for example, recreational boating activities continue to crowd needed habitat for manatees in Florida despite decades of documentation that manatees not tolerate powerboat activities, and fishing interests continue to set nets in vaquita habitat despite widespread recognition that incidental take is driving the population to extinction Species in our first category have reasonable probabilities of survival Northern elephant seals escaped the period of vulnerability associated with small population size, and other taxa seem well on their way Some taxa or populations probably will not persist Western gray whales and northern right whales, particularly the eastern North Pacific population, will survive the new century only with the most rigorous imaginable protection We are less optimistic about species in our second category, because their ultimate survival depends on conscious economic restraint by human cultures, and a possible reevaluation of values regarding the survival of marine mammals and other species in habitats also used or coveted by people Under this premise we anticipate that the vaquita will be the next marine mammal extinction at the species level Excessive subsistence harvests, anthropogenic noise, contaminants, oil spills, and depletion of genetic diversity are issues that have at least occasionally been invoked as risk factors for extinction of marine mammals We find that there are relatively few taxa or populations clearly falling toward extinction as the direct result of any one of these factors Subsistence harvest by native peoples is without question a serious risk factor for Cook Inlet beluga whales, and may have been a crucial precursor to the extinction of Steller’s sea cow However, the western Arctic bowhead whale population has been growing steadily for years despite regular annual subsistence harvest Thus, subsistence harvests are manageable risk factors and need not be regarded universally as unacceptable practice None of the other listed factors are alone causing widespread extinction risk, although there are isolated examples for each Of greater concern here is the problem of significant effects arising from the interaction of multiple factors The best-known cases involve mass mortalities that result from disease outbreaks Such outbreaks often result from immune suppression, which in turn may result from contaminants or from natural disturbances such as the toxic byproducts of certain phytoplankton Interacting factors often are a problem near human population centers, and are difficult, if not impossible, to manage Thus, in our view, the danger of many risk factors discussed here is not the direct effect of a single factor, but the synergistic effects of multiple factors that may be less damaging when separated from one another The inexorable risks posed to marine mammals by human population expansion are further problematic because of inherent uncertainties in population data for many species, confounding efforts to identify and track the status of taxa jeopardized by contact with human populations Geographic Regions of Greatest Concern with Regard to Anthropogenic Extinctions of Marine Mammals As suggested above, marine mammal taxa and populations constrained to life near human population centers are in general most vulnerable to anthropogenic extinction Extinction risks will be greatest where human cultures have the fewest economic options when confronted with the need for restraint in order to solve conservation problems Such circumstances are most likely in ‘‘developing’’ countries at lower and middle latitudes, where large concentrations of people face ongoing economic shortfalls Over the next century we anticipate the greatest extinction risks for coastal marine or aquatic marine mammals in southern and southeastern Asia, eastern Europe, central America, and central Africa Marine mammals in the southern hemisphere generally should be at lower risk of extinction over the next century than those in the north Most major human population centers and most cases of coastal habitat degradation occur in the northern hemisphere However, many southern populations remain small because of past excessive harvests Thus, distantly located southern taxa and populations will be removed from the risk of extinction only to the extent that all forms of harvest are regulated with extreme caution and conservatism Moreover, small distant populations are at risk of incorrect conclusions about number of individuals, trend in numbers, or demographic characteristics because of the statistical limitations associated with survey and demographic data that can be obtained only infrequently and at high cost General Approaches Toward Minimizing the Rate of Anthropogenic Extinctions of Marine Mammals Excessive directed harvests have caused more cases of jeopardized marine mammal taxa and populations than any other single factor Thus the most direct and straightforward approach to the control of extinction risk for marine mammals is a precautionary approach to the concept of marine mammal harvest on a global scale Fortunately, this is the approach currently adopted by many governments, and by international regulatory cooperatives such as IWC, for some species In this context we offer three points of caution First, the protections provided by international treaties and conventions such as the IWC, and by individual governments, not necessarily extend to all marine mammals There are high-profile protective protocols, with active ongoing oversight, for the larger cetaceans, some of the pinnipeds, sea otters, and polar bears Many small cetaceans, some pinnipeds, and some sirenians are not actively or explicitly protected at the national or international level Second, many of the populations subject to active protection are small in number, and as a consequence will be subject to risks associated with stochastic events, both natural and anthropogenic, for decades if not centuries Thus, some extinctions are possible because of unequal conservation effort and because the most aggressive protection cannot eliminate all risk factors Third, the crucial process of detecting trends in small distant populations is so costly that errors are likely in determining which populations are most seriously jeopardized Thus, despite the best of human intentions, protective effort may be misdirected or inappropriately withheld The group of risk factors associated with human population growth will almost certainly cause some extinctions of coastal marine mammals in the current century Here the outcome is in our view more certain, and the methods of prevention more intractable, than in the simpler cases of small