Multilevel structure-function relationship in impaired stream ecosystems . From theory to management applications

Increasing urbanization across the world has led to increased research on ecology in urban settings in the last decade. Urban ecological studies have investigated both impacts of urban development on native ecosystems and the dynamics of urban environments themselves as ecosystems. In both areas of research, streams of urban areas have an important part to play because their position in the landscape makes these ecosystems particularly vulnerable to impacts associated with landcover change. Urban stream ecosystems are affected by multiple stressors and their effects are synthetized in the “urban stream syndrome” theorized in Walsh et al. 2005. Consistent symptoms of the urban stream include flashier hydrograph, elevated concentrations of nutrients and contaminants, altered channel morphology and stability and reduced biodiversity, with increased tolerant species. These ecological effects are often accompanied by other symptoms not observed in all urban areas, such as reduced baseflow and increased suspended solids. In impaired freshwater ecosystems, it is known that ecological integrity can be subdivided into two components, structural and functional integrity. Structural indicators of ecosystem health may be defined as the qualitative and quantitative composition of biological communities. Fish and macroinvertebrate assemblages have been the main focus for assessing structural integrity, although a variety of alternative targets such as benthic algal communities, protozoans, and macrophytes have also been used. Macroinvertebrate assemblages play a central ecological role in many stream ecosystems and are among the most ubiquitous and diverse organisms in fresh waters. Macroinvertebrates are easily recognizable and classifiable and some taxa are representative of every different habitat and condition (sensibility or tollerance to pollution and environmental changes) and so it is easy to aggregate results of macrobenthos analyses into synthetic indeces (such as STAR_ICMi). Function indicators instead, that have a much shorter history, are complementary to structural indicators and refer to the autoecology of biological communities and ecological attributes within the ecosystem in which they are located. In Water Framework Directive, develops by European Union to advance more comprehensive water legislation, the river basins with above mentioned characteristics are defined heavily modified water bodies (HMWBs). HMWBs have unique water quality characteristics that, in most cases, are comparatively different from normal stream conditions upstream of the discharge or at regional reference sites. Reference sites are commonly used in bioassessment studies to identify undisturbed or pristine conditions and hence management targets. The increase of urban development often results in the absence of reference sites in HMWBs and this leads in difficulties to define a target condition for restoring urban stream sites. The WFD requires that all waters achieve good ecological status and only slightly deviate from natural reference conditions, which has become the main objective of most restoration projects in Europe. The ecological status is quantified in many European member states using multi-metric indices, and good ecological status corresponds to a specific score value. However, there is little information on the limiting effects of large-scale pressures on the biological metrics. As suggested by numerous research works, the scale to approach river investigations can be considered from the microhabitat level to basin scale. A river may be analysed across a variety of levels, which can be ordered into a hierarchy, with different degree of sensitivity and recovery time. Impacts of human activity are becoming increasingly unacceptable to a global community that focuses on environmental sustainability. Therefore, whole catchment approach management have been developed to preserve stream ecosystems or restore damaged ecosystems, and mitigate against further damage. The individuation of which factors set limits to biological community development and their respective values is of great interest for river managers and river restoration campaigns. In urban streams is usually hard to assess causal relationships among specific stressors and responses of biological communities using the most common statistical tools. Using macroinvertebrate assemblages as biological indicators in micro- and mesohabitat level works, applied statistics may be viewed as an elaboration of the linear regression model and associated estimation methods of least square. In whole basin analyses, data variability is high and classic statistical approach may even become uninformative. Moreover, the effects of many stressors (local and global) may influence simultaneously the response of biological community leading to a decrease of statistical model fit. In this perspective, alternative statistical approaches are necessary. In 1978 Koenker and Bassett theorized the quantile regression in econometric sciences, a robust alternatives to the least squares estimator for the linear model. Other authors introduced this kind of regression in ecology declaring that quantile regression allows the various stressors to be considered as “constraints” to the distribution of biological communities, without compromising the model causal relationship. Aim of this work is to assess the overall pressure of human activities in river basins of Lombardy piedmont and floodplain area and to relate changes in the biological communities as a result of habitat loss and changes in both hydromorphological and physico-chemical properties. In this area, many rivers have a “channelized” nature with straight section, clear of river bank tree and uniform bed morphology. Flow regulation and modification have also been widespread. The quantity and timing of water availability have been altered for irrigation and industrial purposes, through the construction of dams and reservoir for water supply. Changes in water quality are also common, in particular in lowland areas where urbanization and agriculture are more strong. For these reasons, the work is focusing on different scale (microhabitat, site, river reach and basin levels) to have a better resolution and understanding of existing dynamics among structural and functional indicators and pressures in impaired environments. These areas undergo different stresses (habitat loss, changes in physico-chemical properties and changes in flow) that affect the integrity of the ecosystems. Assessing the condition of ecosystems is a prerequisite to reduce the induced anthropogenic pressure. Decision-making in river restoration programs can also be helped by multilevel kind of information.