WaterSubject

Few habitat survey, classification or assessment methodologies have been developed specifically for urban or heavily engineered rivers, although there are some notable exceptions. This lack of urban-focused methodologies is unfortunate because more general river surveys are often of inappropriate style and resolution to discriminate between reaches of urban river, tending instead to group them into a single homogenous category of ‘bad’ or ‘poor’ habitat quality. As a result, urban rivers have become undervalued. However, the EC Water Framework Directive defines a category of ‘modified water body’ that includes urban rivers and requires that reference conditions should be developed to support river restoration aims. This has produced a major opportunity for research to focus on survey, classification and decision support systems to aid in the assessment, management and rehabilitation of urban rivers.

An urban river survey (URS) has been proposed by some of the present authors, which is a modification of the Environment Agency’s river habitat survey (RHS). Semiquantitative indices and classifications based upon URS data have also been developed. Here we present a refinement of the previously reported classifications and we develop a simple decision and scenario modelling system for the initial exploration of urban river habitat and rehabilitation potential that is particularly appropriate for application to entire urban catchments within a Geographic Information System. Therefore, the aim of this paper is to describe a suite of methods that are specifically designed to classify urban river stretches according to their habitat characteristics using URS data and to support identification of those river stretches that might be candidates for rehabilitation. The methods combine to form a decision support system that can be applied at the catchment scale to identify and prioritise river stretches for rehabilitation through a change in their engineering/management and to carry out preliminary scenario modelling of the potential impact of adopting such management changes.

It is important to stress that the system is concerned entirely with the quality, diversity and complexity of physical habitat within and between urban river stretches. It does not have a specific target ecology, but aims at an increase in habitat quality, diversity and complexity with the assumption that this will provide a context for a more diverse ecology. Furthermore, because the focus is entirely upon physical habitat, the outputs from the system need to be compared with information concerning potential constraints on (i) the actual ecological success of physical habitat change (e.g. constraints attributable to water and sediment quality) or (ii) the feasibility of any proposed changes (e.g. flood defence and land use limitations).

The methods described in this paper were all developed using information gathered within the upper River Tame catchment, West Midlands, UK. Previous work on the URS and on the development of urban river classifications was based on two surveys of approximately fifty 500 m stretches of river, but the refined classifications and decision support system presented here are based on analysis of a much larger data set comprising URS surveys of one hundred and six 500 m stretches that was undertaken in 2003 as part of an EU Life Demonstration Project entitled ‘The Sustainable Management of Urban Rivers and their Floodplains’ (SMURF). While full details of the URS and the indices derived from it can be obtained from previously published work, a brief summary is presented here as a context for the revised classifications and decision support system that have been developed within the SMURF project.

The Tame river network was subdivided into stretches of river of a single ‘engineering type’ or particular combination of (i) cross-sectional profile (seminatural, restored, cleaned, enlarged, two-stage, resectioned), (ii) planform pattern (seminatural, straight, artificially meandered/sinuous, recovered) and (iii) level of reinforcement (none, bed only, one bank, bed and one bank, both banks, full). URS surveys were applied to 500 m stretches of a single engineering type.

Because of similarities between the URS and RHS, details of the URS survey are not presented here. However, it is important to note that the URS gathers a large amount of information that is a significant resource for exploring characteristics of particular urban river stretches. This paper focuses on classifications of urban river stretches and how these can be used to underpin a simple decision support system that forms a first level in assessing the degree to which stretches may benefit from modified management.

The classifications are developed from synthetic indices of URS observations. The indices each describe well-defined components of the urban river environment: their materials, physical habitat and vegetation properties. Each of the indices combines several, often categorical, observations into a single, integrative, semiquantitative index. Each of the indices was selected because it described an important habitat attribute. Each index covers a similar numerical range and represents a gradient in magnitude or intensity of the attribute that it describes.