ing and shipping, and more productive fishing. Military sonar systems are important for national defense.
This intentional and unintentional introduction of sound in the ocean associated with activities beneficial to humans must be balanced against known deleterious effects on marine mammals. Strandings of beaked whales in certain environments are clearly associated with the use of mid-range tactical military sonar. There are documented behavioral responses of beluga whales to icebreakers 50 km away. Gray whales and killer whales have shown multi-year abandonment of critical habitats in response to anthropogenic noise. Although there are many documented, clearly discernable responses of marine mammals to anthropogenic sound, reactions are typically subtle, consisting of shorter surfacings, shorter dives, fewer blows per surfacing, longer intervals between blows, ceasing or increasing vocalizations, shortening or lengthening duration of vocalizations, and changing frequency or intensity of vocalizations. Although some of these changes become statistically significant in given exposures, it remains unknown when and how these changes translate into biologically significant effects at either the individual or the population level.
The basic goal of marine mammal conservation is to prevent human activities from threatening marine mammal populations. The threat from commercial whaling was obvious, but it is harder to estimate the population consequences of activities that have less immediately dramatic outcomes, such as those with indirect or small but persistent effects. The life histories and habitat of marine mammals compounds these problems. Marine mammals are long lived and slow to mature. Many species have long periods of dependency. They are highly social and show behavioral plasticity, with complex development of behavior. Furthermore, many of these behaviors occur underwater where they are difficult to document. This makes it particularly difficult to estimate the effects that a short term exposure may have as it ripples through the lifetime of an individual, or as effects on different individuals ripple through the population. Even extreme effects, including death, are not necessarily observed.
The status of any population is the consequence of the accumulation of many effects; resulting in marginal changes in survival and reproduction over time. In addition, the end result is often so far removed in time from the proximate causal events that they cannot simply be traced post hoc. The existence of several comparable populations with different status and different exposure can be used to reduce the number of candidate primary