No abstract

Pesticide contamination resulting from agricultural runoff depends on the time period between application and rainfall. In Western Cape orchard areas, the last pesticide application of the growing season in summer takes place at the end of February. Pesticides, total phosphates and total suspended solids (TSS) were measured in the Lourens River at the beginning of April 1999 prior to the first rainfall of the rainy season and in the middle of April during high discharge following the first rainfall of 9.6 mm/d. Pre-runoff samples indicated only contamination with total endosulfan (α, β, sulphate) at levels up to 0.06 µg/l. Runoff during the first rainfall event resulted in an increase in the levels of endosulfan, chlorpyrifos and azinphos-methyl, to 0.16, <0.01 and 0.38 µg/l, respectively, in water samples and 245, 344, and 244 µg/kg in suspended sediments. In terms of chemical load the single rainfall event caused a loss of 15.1 g/h endosulfan, 1.8 g/h chlorpyrifos and 20.5 g/h azinphos-methyl. The second rainfall event caused no measurable increase in pesticide levels, although the amount of rain was even higher (14.4 mm/d). Levels of both total phosphate and TSS were also increased during the first runoff event. Transient contamination levels exceeded the target water quality range proposed by the South African Department of Water Affairs and Forestry (DWAF). The Lourens River site downstream of the farming area is identified as a site where potentially toxic conditions might arise.

No abstract

Drift from pesticide spray application can result in contamination of nontarget environments such as surface waters. Azinphos-methyl (AZI) and endosulfan (END) deposition in containers of water was studied in fruit orchards in the Western Cape, South Africa. Additionally, attention was given to the contamination in farm streams, as well as to the resulting contamination of the subsequent main channel (Lourens River) approx. 25 km downstream of the tributary stream inlets. Spray deposit decreased with increasing distance downwind and ranged from 4.7 mg m(-2) within the target area to 0.2 mg m(-2) at 15 m downwind (AZI). Measured in-stream concentrations of both pesticides compared well with theoretical values calculated from deposition data for the respective distances. Furthermore, they were in the range of values predicted by an exposure assessment based on 95th-percentile values for basic drift deposition (German Federal Biological Research Centre for Agriculture and Forestry [BBA] and USEPA). Pesticide deposition in the tributaries was followed by a measurable increase of contamination in the Lourens River. Mortality of midges (Chironomus spp.) exposed for 24 h to samples obtained from the AZI trials decreased with decreasing concentrations (estimated LC50 from field samples = 10 microg L(-1) AZI; lethal distance: LD50 = 13 m). Mortality in the tributary samples averaged 11% (0.5-1.7 microg L(-1) AZI), while no mortality was discernible in the Lourens River samples (0.041 microg L(-1)). The sublethal endpoint failure to form tubes from the glass beads provided was significantly increased at all sites in comparison with the control (analysis of variance [ANOVA], Fisher's protected least significant difference [PLSD], p < 0.01).

We evaluated the potential interaction of pesticide effect and predatory fish on behavior and mortality of a stream mayfly. Experiments in laboratory stream microcosms compared the activity, drift, and mortality of Baetis mayfly nymphs in the absence of fish with that in the presence of Cape galaxias (Galaxias zebratus), both species inhabiting the same stream environments in the Western Cape of South Africa. These two predator treatments were combined with exposure either to no pesticide or to 0.2 microg/L of the organophosphate insecticide azinphos-methyl (AZP) or 0.2 microg/L of the pyrethroid insecticide fenvalerate (FV). Such pesticide levels are known from transient spraydrift- or runoff-related pesticide input into running waters. Each of the six combinations was replicated six times as 30-min trials during the day and effects were analyzed using 2 x 2 factorial analysis of variance (ANOVA). A single exposure to either fish or pesticide significantly increased the absolute activity of mayflies, measured as number of animals visible on top of stones, and the absolute mayfly drift in the fish treatment and in the FV treatment but did not increase the mortality above 0.8%. The combination of predator presence and sublethal pesticide exposure resulted in a significant increase in the mortality rate, to about 9% in the AZP x fish and 25% in the FV x fish treatment, although the activity and drift rates were not increased compared with the single-stressor treatments. Two-by-two factorial ANOVA and the comparison of expected and measured responses indicated that the mortality resulted from a synergistic interaction of the two stressors. The observed mortality was without exception caused by predation of the fish on drifting mayflies. The relative drift rate in the FV x fish treatment was decreased, again due to a synergistic interaction, which suggests an active drift avoidance reaction of the mayflies exposed to the combined pesticide x fish treatment, in contrast with those exposed just to FV. We conclude that the drift response of mayflies to transient sublethal pesticide exposure results in a synergistically increased adverse effect in the presence of predatory fish.

No abstract

No abstract

No abstract