Combining surveys of river habitats and river birds to appraise riverine hydromorphology

Abstract

1. Biological monitoring is vital to river conservation. Aside from providing census data, regular monitoring may detect population trends that reflect the degradation or remediation of riverine environments. Birds are major candidates for this purpose because of their connections to riverine food webs and river habitat features. However, much information on factors affecting river bird distributions is qualitative. 2. River bird populations have been surveyed annually in the U.K. since 1974. The value of this monitoring could be increased if links between population trends and changes in riverine landscapes or channel hydromorphology were better understood. We modelled the relationships between bird species' distributions recorded by the British Trust for Ornithology's Waterways Breeding Bird Survey (WBBS) and hydromorphology recorded by the Environment Agency's River Habitat Survey (RHS). 3. Regression models were built for 20 bird species associated with the river channel and neighbouring floodplains. This provided one of the first illustrations in ecology of generalized estimating equations (GEEs) for extending generalized linear models to instances where individual observations are not statistically independent. GEEs could offer a range of benefits to freshwater biologists, as they are a simple approach to analysing hierarchical, correlated data while maintaining the ease of interpretation with a range of link functions and data types. 4. Upland, fast water species and reedbed specialists showed the closest links with RHS, followed by lowland, slow water species, where correlations to RHS varied widely in magnitude. The distributions of floodplain wading birds were poorly modelled by RHS, suggesting that floodplain land uses as quantified by RHS were poor predictors of distributions. Marked preferences between upland and lowland channels were evident for most species. Specific features were also important, including hydraulics, bank/channel vegetation, depositional features and anthropogenic structures. 5. Monitoring the abundance and distribution of a range of bird species, with a diversity of life histories, could be a valuable tool for assessing trends in riverine landscapes and whole river basins. RHS captures useful information about channel hydromorphology and vegetation, but needs supplementing with additional environmental information to describe (i) the wider riverine landscape (e.g. more detailed floodplain variables, proximity of other waterbodies) and (ii) water chemistry. Although further work is required, this study suggests that WBBS and RHS add mutual value in appraising the river environment. We advocate the further development of birds as indicators in riverine

Linking hydrologic, physical and chemical habitat environments for the potential assessment of fish community rehabilitation in a developing city

“…Generally, previous assessment approaches focus on specific species (e.g., endangered, threatened or native species) Bain and Meixler, 2008), bioindicators (Hughes, 2005;Feld and Hering, 2007;Vaughn et al, 2007;Hughes et al, 2010), recruitment index (Armstrong and Hightower, 2002;Fox, 2004) or detection of impaired habitat and/ or processes (Bellmore et al, 2012). Successful implementation of these methods depends critically on connecting the underrepresented taxa with the mechanisms responsible for their reduction/ elimination but often requires substantial scientific expertise.…”

Characterizing physical habitats in rivers using map-derived drivers of fluvial geomorphic processes

“…These approaches can broadly characterize eco-regions for biological communities, but there is potential for further characterization of river types. The inclusion of geomorphic drivers on channel forms and conditions could offer greater potential for better understanding of the primary geomorphological and ecological processes shaping fluvial systems (Chessman et al, 2006;Vaughan et al, 2007;Palmer, 2009).…”

Accounting for weather and time-of-day parameters when analysing count data from monitoring programs

“…In these cases, imperfect detection must be accounted for if detectability covaries with the dimension to be compared. Here we focused on the use of count data to infer on abundance variations across spatial variations of an ecological factor, which is a common output of large scale monitoring programs Posthuma and De Zwart 2006;Vaughan et al 2007). As many ecological factors such as habitat characteristics are measured very locally, these studies often need a measure of relative abundance on a very local scale (e.g.…”