3 Conceptual history of landscape-scale riverine research

A series of landmark papers has developed the conceptual framework for landscape-scale riverine research. Hynes (1975) described the complex interactions linking terrestrial and aquatic ecosystems. Vannote et al. (1980*) provided a coherent framework from which to predict and understand how interactions between terrestrial and aquatic ecosystems control instream processes and how these interactions change in a predictable way from forested headwater streams to mainstem rivers. Frissell et al. (1986) described how climate, geology, and land form provide a series of nested controls through which instream habitats are shaped and maintained, and to which biota respond. Ward (1989) described the four-dimensional nature of river ecosystems and the importance to aquatic biota and processes of maintaining those connections across landscapes and through time. Tonn (1990*) offered a conceptual framework outlining how continental, regional, water body type, and local environmental processes and features (filters) govern fish assemblage patterns across all those spatial extents. Schlosser (1991) emphasized the importance of spatial and temporal heterogeneity and connectivity to river ecosystems and their biota. And, Hawkins et al. (1993) promoted the concept of hierarchical drivers of stream habitat features. Poff (1997) added a framework for using explicit multi-scale landscape constraints to predict and understand distribution, abundance, and local assemblage composition. Ward (1998) described the pathways by which landscape disturbances organize and constrain instream habitats. The next conceptual leap came from Fausch et al. (2002*), who promoted the need for a continuous view of rivers and encouraged researchers and managers not to view the river as a series of disjoint reaches but to consider “the entire spatially heterogeneous scene of the river environment, the riverscape, unfolding through time.” Burcher et al. (2007) applied the concept of an ecological cascade or series of impacts to understand how land-cover change can alter biological responses. Building on earlier ideas in which processes of higher scales form boundary conditions for lower scales (Angermeier and Winston, 1999), they proposed the land cover cascade as a conceptual framework for organizing the diverse relationships between landscape-scale phenomena and instream biological responses.

These theoretical frameworks have supported a large body of research to understand and to quantify interactions between landscape conditions over large spatial extents and instream responses. A series of review papers and research compendia organizes and tracks growth in the field of landscape riverine research.

A special issue of Freshwater Biology (1997) first began to assemble the ideas and findings of this new perspective on river systems. Johnson and Gage (1997*) reviewed the state of the art at the time. They concluded that the advent of GIS technologies and the need for wide-ranging policies and management programs had pushed landscape-scale research to the forefront of riverine ecology. They synthesized the landscape metrics in common use for landscape analysis and the statistical tools available for linking landscape patterns to instream conditions. Johnson and Gage concluded that these new spatial and statistical tools have enabled ecologists to study patterns and relationships over larger and more diverse extents than previously possible. Other papers in the special issue compared local versus landscape perspectives on Michigan trout streams (Wiley et al., 1997), examined the role of catchment land use patterns on stream integrity (Allan and Johnson, 1997*), land use on macroinvertebrate assemblages in New Zealand (Townsend et al., 1997), landscape condition on water chemistry (Johnson and Gage, 1997*), and multi-scale impacts on species traits of macroinvertebrate assemblages (Richards et al., 1997).

A second set of papers was assembled in Freshwater Biology in response to the First International Symposium on Riverine Landscapes. In that special issue, Wiens (2002*) promoted the direct application of concepts from the field of landscape ecology, which had traditionally focused on terrestrial landscapes, to aquatic systems. Papers in the special issue described linkages between landscapes and both hydrogeomorphic processes (Poole, 2002) and biological responses (Malmqvist, 2002), detailed advances in remote sensing and GIS technologies (Mertes, 2002) and applied landscape concepts to riverine management (Poudevigne et al., 2002).

A comprehensive and detailed review of research linking land use to instream conditions was recently provided by Allan (2004a*). He outlined pathways by which land use at the catchment and riparian scales affects instream habitats (Table 2) and synthesized over 100 research papers correlating land use with metrics of stream health. Allan found that the most common landscape composition metric in studies linking land use and land cover to aquatic response has been the percentage of land cover types in either the catchment or the riparian area. Across multiple projects, he was able to conclude that a high proportion of forest cover in a catchment is normally associated with positive stream conditions. Conversely, agricultural or urban areas in the catchment have been documented to have a negative influence on downstream river conditions (Allan, 2004a*). Allan (2004b) also raised several key challenges to linking landscapes and the ecological status of rivers.

Durance et al. (2006*) reviewed 658 papers dealing with the management of fish in rivers and found that only 27 explicitly considered the impact of scale. They were able to summarize some catchment-scale pathways by which landscapes affect fish abundance and distribution such as climate, geology, vegetation, and past land use (Table 3) (Durance et al., 2006*).

In 2006, a large volume of research exploring and quantifying landscape influences on stream habitats and biological assemblages was assembled by Hughes et al. (2006b). This collection of 30 research papers focused on research linking catchment-scale landscape condition to instream habitat and fish responses. The editors concluded that there were four key challenges in studying river systems in a landscape context: determining appropriate units for measuring and interpreting the riverscape, understanding the mechanisms by which land use alters river habitats and biota, measuring and understanding how spatial factors interactively affect aquatic habitats and biota, and collecting and interpreting appropriate landscape and riverine data. Major knowledge gaps requiring additional research included improving river-landscape classification, determining the appropriate spatial and temporal scales at which data should be captured, improving predictive models where data are limited, and improving our measures of connectivity among river networks and their landscapes.

Most recently, Johnson and Host (2010*) reviewed the landscape-aquatic literature from 1986 to 2008, revealing a 20-fold increase in the number of publications over that time. They also found that as the scales of the studies increased so did the importance of landscape versus site-scale predictors. Because of the use of varied terms and extents for ‘site’ and ‘landscape’, they recommended that editors demand explicit descriptions of both. In the same issue, Poole (2010) pointed out that stream hydrogeomorphology is also at the roots of riverine landscape ecology by providing mechanistic foundations for how fluvial landscapes shape stream ecosystems.