RIVER RESEARCH AND APPLICATIONS River Res. Applic. 22: 167–177 (2006) Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/rra.903 TREND, VARIABILITYOR EXTREME EVENT? THE IMPORTANCE OF LONG-TERM PERSPECTIVES IN RIVER ECOLOGY MICHAEL A. REID* and RALPH W. OGDEN Cooperative Research Centre for Freshwater Ecology University of Canberra, ACT 2601, Australia ABSTRACT Temporal variability at extra-annual scales is typically a feature of freshwater ecosystems, particularly in places like Australia where the climate is highly unpredictable and where recent anthropogenic changes have been dramatic. Accounting for this variability when studying freshwater systems is generally not possible if freshwater ecologists rely solely on measurements carried out in the course of studies that rarely exceed a few years in length. Alternative methods can be used to extend temporal perspective; however, we argue these are under-utilised in river ecology. This paper outlines approaches for extending temporal perspective (historical studies, palaeolimnological studies and modelling) and illustrates, using pertinent examples, areas of river ecology where we believe they can be of particular use. We conclude with some suggestions for ways in which cooperation between relevant disciplines can be enhanced. Copyright # 2006 John Wiley & Sons, Ltd. key words: freshwater ecology; palaeoecology; historical studies; ecological modelling; benchmarking INTRODUCTION In 1986, Weatherhead wrote a landmark paper questioning the frequent citing of so-called unusual events in the ecological literature. The central conclusion was that many of the events were only unusual within the context of the short time frame of most ecological studies. Other authors have since made similar assessments of other fields (Smol, 1995; Tilman, 1989), highlighting a key problem for aquatic ecologists trapped by both funding cycles and their own human temporal perspective. These events are not unusual but are an important component of the spatial and temporal habitat template that determines the ecosystem. Some freshwater ecologists appreciate the importance of a long-term perspective (Nicholls, 1997; Smol, 1995; Tilman, 1989; Walker et al., 1995; Weatherhead, 1986), and several palaeolimnologists have written extensively on the value of combining palaeolimnological approaches with ‘neo’ limnology to understand and manage lake eco- systems (Battarbee 1999; Smol 1990a, 1992, 1995, 2002). The value of extended temporal perspective is no less for river systems and is even greater when these systems are characterised by high climatic and hydrologic varia- bility at inter-annual and decadal scales. Compared to more stable systems such as those of North America and Europe (see Puckridge et al. 1998 for a comparison of hydrological variability indices of major river basins), where palaeolimnological study has been most extensive, it is harder to generalize about variable systems on the basis of short-term datasets, which may not capture key ecological cycles or resetting events (Walker and Thoms, 1993). Thus in areas like Australia, where climate and stream flow are highly variable across much of the continent (Puckridge et al. 1998, 2000), long-term variability in river and floodplain ecology and is increasingly seen as a critical factor driving the structure and function of aquatic ecosystems (Kingsford, 2000; Puckridge et al., 1998, 2000; Sheldon et al., 2000; Walker et al., 1995). Despite this, the question of long-term variability is not often addressed in studies of river ecosystems subject to high extra-annual variability, particularly in Australia. This can be partly explained by the fact that long-term instrumental records are largely confined to the physical realm. An example is the Murray River in Australia, where reliable flow records for some parts of the river extend back to the late 19th century (Close, 1990), well Received 13 November 2004 Revised 18 December 2004 Copyright # 2006 John Wiley & Sons, Ltd. Accepted 24 January 2005 *Correspondence to: Michael A. Reid, Cooperative Research Centre for Freshwater Ecology University of Canberra, ACT, 2601, Australia. E-mail: mike.reid@canberra.edu.au