SHOREBIRD COMMUNITIES ON BRITISH ESTUARIES - FACTORS AFFECTING COMMUNITY COMPOSITION

1. Environmental gradients across 109 British estuaries were examined in relation to the communities of shorebirds in winter. Ordination and classification techniques were employed on the 5-year means of peak-w inter high-tide counts for the 13 most common species, converted to de nsities, for the two periods, 1969-75 (first period) and 1981-85 (seco nd period), in order to quantify changes in community composition betw een the two periods. Multiple regression within GLIM was then used to relate the DCA ordination scores to physical, climatic, geographic and water chemistry variables. 2. In an ordination analysis of the combin ed data for the two periods, axis 1 had a left-right gradient of speci es favouring coastal lagoons (avocet) to those favouring sandy estuari es (bar-tailed godwit, sanderling, knot). Dunlin and grey plover (favo uring muddy estuaries) occupied similar positions on the ordination as did turnstone and purple sandpiper (favouring rocky estuaries). 3. Tu rnstone, purple sandpiper, redshank and oystercatcher underwent large changes in numbers between the two time periods, which affected their overall contribution to community structure. 4. In terms of changes in community composition between periods, most estuaries have experience d little change, but a small number, e.g. Blyth (Northumberland), Axe, Auchencairn Bay/Rough Firth complex, Irt/Mite/Esk complex, and the Tw eed, underwent large changes. Sites showing most change received more rainfall (Jan - Mar), had higher minimum water temperature, and lower maximum salinity. Community change was not correlated with any aspect of nutrient status. 5. Axis 1 of the sites ordination was positively c orrelated with estuary size, high concentrations of ammonium-N, high b iochemical oxygen demand, and negatively with rainfall, mean air and m inimum water temperature, maximum salinity and mean percentage of diss olved oxygen. Axis 1 therefore represents a left-right gradient of inc reasing estuary size containing higher quantities of organic matter, h igher oxygen demand and lower oxygen availability. Multiple regression analysis within GLIM showed the effect of these nutrient-status indic ator variables operated to some extent independently of estuary size. 6. Hierarchical classification of the 109 estuaries, based on their wa der communities, revealed four groups. Group 1 represented species fav ouring muddy estuaries; group 4 ones with a dominance of species favou ring sandy shores; groups 2 and 3 represented estuaries intermediate b etween these types. Stepwise discriminant analysis indicated that nort hing, tidal range and total estuary area were significant discriminati ng variables between the four groups. Estuaries in group 4 tended to h ave a greater area of intertidal feeding habitat and wider entrances, corresponding to large estuaries. 7. The implications of these finding s to estuary conservation and development proposals, are discussed.