Aquatic vascular plants in nitrate-rich calcareous lowland streams : do they respond to phosphorus enrichment and control?

The problems of eutrophication in the lowland rivers of Britain have been exacerbated in the past 13 years by the droughts of 1989-92 and 1995-96. Large growths of filamentous algae have smothered aquatic vascular plants (major primary producers) in lowland calcareous streams and phosphorus enrichment has been blamed for a shift in submerged plant species composition and community structure. This study investigated the impact of phosphorus removal from two sewage treatment works (STWs) situated in the River Wensum catchment area, Norfolk, UK. The mesology of aquatic vascular plants was investigated through a survey at 62 sites spread over four river basins: Wensum, Nar, Bure, Wissey. The landscape is predominantly rural (crops and pasture) not rising above 90 metres (OD). The average background concentrations of phosphorus from diffuse sources were: soluble reactive phosphorus (SRP) 13 microg/L, total dissolved phosphorus 24 ug/L and total phosphorus (TP) 53 microg/L. Before phosphorus stripping (1990-1999), the background contributed about 12% of the TP loads (0.08-0.10 kg/ha/year of TP) and after (2000-2001) it was about 34% for two reasons, the phosphorus stripping plus effects due to higher rainfall (0.13-0.18 kg/ha/year of TP). The level of bioavailable phosphorus from the river-bed sediment was highest at sites impacted by both the effluents and weirs associated with mills. Phosphorus removal effectively reduced the TP load contribution of the two STWs from 42% to 18% of the effluents in the whole catchment. However the predicted levels of SRP at mean flow (140-273 microg/L) and 95% exceedance discharge (415-1039 microg/L) dowstream from the STWs, remained very high. An isozyme technique was developed to identify the eight British species of Callitriche L. (water-starworts). This allowed the chorology of Callitriche in the rivers of Norfolk to be determined. Aquatic vascular plant composition, species Natural Combination of Attributes (NCAs - life history trade-offs) and community structure did not respond to the gradient of nutrients (SRP 11-3600 microg/L, nitrate 6.4-12.9 mg/L, ionised ammonia 16-434 microg/L). Species composition did respond, however, to the biogeographical units and distance of the site from the source (surrogate for species probability of colonisation) and channel depth/substrata (species environmental requirements). Even the most obvious geomorphological factors were only weak predictors of the NCAs. Moreover, the NCAs were linked, although only marginally, to the spatial structure of the river network, a surrogate for species probability of dispersal. Therefore the idea of using the species' attribute approach as currently applied to aquatic vascular plants seems falsely optimistic. The lack of dominance within plant communities suggested that demographic processes as well as inter-species competition might have occurred at the local scale as predicted by the patch dynamics concept. However, the different probabilities of colonisation of the sites as well as the regional species pool may dictate the local processes of plant community structure and composition in a directional river network. This study indicates the need for a unification of the niche and demography theories at multiple spatial and temporal scales. The species' demographic dynamics, the catchment- based habitat heterogeneity and the regional species pool should therefore always be taken into account for conservation and management purposes. Wetlands should be preserved and restored as much as possible to maintain the regional species pool which will ensure the persistence of the riverine flora.