Assessment of Landscape Characterization and Classification Methods

ii Executive Summary The absence of a hierarchical approach to classification that integrates landscape and riverine ecological theory hampers scientific and regulatory endeavors to define natural resource management strategies protective of watershed dynamics. It is my intent to review existing aquatic and landscape classification methodologies and measures of ecosystem condition and to develop a synthesis paper that first describes the components and uses of existing methods, and that evaluates the strengths and weaknesses of existing approaches. Finally, a suggested template for " riverscape " classification based on the assessment of existing systems and a suggested case study applying the classification template will be presented. Additional basic research supporting hierarchical classification will be outlined. Classification systems often focus solely on ecosystem structure, ignoring ecosystem processes and patch dynamics responsible for establishing system heterogeneity and driving resource distribution. Structural classification systems have stemmed from the classic work conducted by Bailey (1976) and Omernik (1987). Rohm et al. (1987) and Whittier et al. (1988) found that the approaches used by Bailey (1976) and Omernik (1987) could serve as a geographic framework for classifying streams. However, ecoregions alone were insufficient to classify fish assemblages, predict resource pattern, and diagnose biotic impairment (Hughes et al. 1994, Boulton 1999, Meixler and Bain 1999). Classification systems based solely on ecosystem structure omit ecological concepts important in predicting resource distribution and pattern such as disturbance and recovery processes classification systems focus on the physical characteristics (structure) of landscapes without attempting to explicitly classify and map types and rates of critical ecosystem processes across multiple scales. As such, existing stream classification systems are often arbitrary, developed by creating classes out of a continuum of channel form. Owing to the discontinuous, patchy nature of stream networks, classification approaches omitting a process-based hierarchical framework are limited in their utility. To accurately partition ecosystem variability, infer system structure and function across landscapes, predict watershed response to disturbance, diagnose lotic system impairment, and design effective recovery strategies, a classification approach embracing the fundamental tenets of hierarchical patch dynamics (HPD) theory is needed – an approach emphasizing the nested, discontinuous hierarchies of patch mosaics and the centrality of pattern, process, and scale. The discontinuum view central to the HPD theory recognizes general trends and trend reversals in habitat characteristics along the longitudinal profile of the stream network, and serves as a framework to understand the ecological importance of the unique habitat patterns created by …