Freshwater Forum, Vol 28 (2010)

Quantifying the relationship beween land cover and biological condition of headwater streams

John Murphy


There has been considerable effort recently to gain insight into the relationships between the characteristics of catchments and the ecological status of streams within rural landscapes. Some such studies have been confounded by the natural changes seen in land use longitudinally along a river, or have been limited by poor replication. As part of the long-running Countryside Survey programme, a random-stratified sample of over 400 replicate headwaters across all British landscape types has been sampled every eight years since 1990. The macroinvertebrate and aquatic plant communities were sampled and the hydromorphological features were surveyed at each site. Adjacent riparian vegetation and habitats were mapped in detail as well as catchment land cover for each stream site. Such a comprehensive dataset allows a robust assessment of the association between catchment and riparian land use and stream biological integrity. Very few significant correlations were found between cover of particular land uses and stream condition. In general there was considerable unexplained variation in the biological data. Only the aquatic plant Mean Trophic Rank score (MTR; an index of nutrient enrichment) showed a consistent association with catchment and riparian land cover. MTR scores were positively associated with forest and natural grassland cover and negatively associated with arable and improved pasture cover. Relationships between biotic measures and land cover were generally not improved by restricting the spatial extent of the analysis from the whole of Britain to a more homogenous landscape type, e.g. easterly lowland; nor was it improved by standardising the observed family richness of macroinvertebrates by a site-specific prediction of reference conditions. This analysis reveals the complexity of relating land cover to stream biology.