Interrelationships of fish and channel environmental conditions with aquatic macrophytes in an Argentine irrigation system

This study examines the relationships between fish, environmental variables and submerged macrophytes within the irrigation system of the lower valley of the Río Colorado in southern Argentina. Using Canonical Correspondence Analysis (CCA), the strongest environmental gradients detected were conductivity and carp ( Cyprinus carpi) biomass per unit area of channel cross-section. These variables were positively associated with each other and also with water turbidity. Sites scoring high on these gradients were mainly drainage channels; those scoring lowest were irrigation channels. The main fish species associated with high carp biomass, high turbidity and high conductivity were carpa ( C. carpio), pejerrey ( Odontesthes bonariensis), madrecita ( Jenynsia lineata lineata) and lisa ( Mugil liza). Dientudo ( Oligosarcus jenynsi) and mojarra ( Astianax eigenmanniorum) were more strongly associated with clearer water, with low carp biomass. In all CCA analyses macrophytes were arranged in similar order along the main conductivity-turbidity-carp biomass gradient. Lowest on this main environmental gradient, and scoring very close to each other, were Potamogeton pectinatus and Chara contraria. Salinity-tolerant species such as Ruppia maritima, Zannichellia palustris and Enteromorpha flexuosa tended to score highest, followed by the surface floating Azolla filiculoides and the filamentous alga Cladophora surera. Within the constraints imposed by conductivity, turbidity was a key predictor of both abundance and distribution of the two dominant plants of the irrigation scheme ( P. pectinatus and C. contraria). Turbidity was strongly predicted by biomass of carp per unit channel cross sectional area, when fine sediment particle content was taken into account. The positive association between carp biomass and water turbidity was both substantial and predictable, and was in turn associated with reduction in submerged plant growth. The biomass of the most widespread nuisance-causing plant species in the channels, Potamogeton pectinatus, could best be predicted ( R = 0.592, P < 0.05) using a multiple regression model utilising four predictor variables: conductivity, nitrate, phosphate and carp biomass. This revised version was published online in July 2006 with corrections to the Cover Date.