The Rocky Mountains in the USA and Canada encompass the interior cordillera of western North America, from the southern Yukon to northern New Mexico. Annual weather patterns are cold in winter and mild in summer. Precipitation has high seasonal and interannual variation and may dier by an order of magnitude between geographically close locales, depending on slope, aspect and local climatic and orographic conditions. The region's hydrology is characterized by the accumulation of winter snow, spring snowmelt and autumnal base¯ows. During the 2±3-month `spring runo' period, rivers frequently discharge 470% of their annual water budget and have instantaneous discharges 10±100 times mean low ¯ow.
Complex weather patterns characterized by high spatial and temporal variability make predictions of future conditions tenuous. However, general patterns are identi®able; northern and western portions of the region are dominated by maritime weather patterns from the North Paci®c, central areas and eastern slopes are dominated by continental air masses and southern portions receive seasonally variable atmospheric circulation from the Paci®c and the Gulf of Mexico. Signi®cant interannual variations occur in these general patterns, possibly related to ENSO (El NinÄ o±Southern Oscillation) forcing.
Changes in precipitation and temperature regimes or patterns have signi®cant potential eects on the distribution and abundance of plants and animals. For example, elevation of the timber-line is principally a function of temperature. Palaeolimnological investigations have shown signi®cant shifts in phyto-and zoo-plankton populations as alpine lakes shift between being above or below the timber-line. Likewise, streamside vegetation has a signi®cant eect on stream ecosystem structure and function. Changes in stream temperature regimes result in signi®cant changes in community composition as a consequence of bioenergetic factors. Stenothermic species could be extirpated as appropriate thermal criteria disappear. Warming temperatures may geographically isolate cole water stream ®shes in increasingly con®ned headwaters. The heat budgets of large lakes may be aected resulting in a change of state between dimictic and warm monomictic character. Uncertainties associated with prediction are increased by the planting of ®sh in historically ®shless, high mountain lakes and the introduction of non-native species of ®shes and invertebrates into often previously simple food-webs of large valley bottom lakes and streams. Many of the streams and rivers suer from the anthropogenic eects of abstraction and regulation. Likewise, many of the large lakes receive nutrient loads from a growing human population.
We concluded that: (1) regional climate models are required to resolve adequately the complexities of the high gradient landscapes; (2) extensive wilderness preserves and national park lands, so prevalent in the Rocky Mountain Region, provide sensitive areas for dierentiation of anthropogenic eects from climate eects; and (3) future research should encompass both short-term intensive studies and long-term monitoring studies developed within comprehensive experimental arrays of streams and lakes speci®cally designed to address the issue of anthropogenic versus climatic eects.