No abstract

The East River (Dong Jiang), a major tributary of the Pearl River (Zhu Jiang, the second largest river in China by discharge), is situated in southern China, which has the highest rates of urbanization and development on Earth. The East River also provides 80% of Hong Kong's water supply. However, there have been no ecological studies to examine the combined impacts of changes in land use and water quality degradation on this river ecosystem. We tested the hypothesis that land-use disturbance and water quality degradation would significantly reduce benthic biodiversity in the East River by investigating macroinvertebrate community composition and relating it to data on water quality and catchment condition. The percentage of total impervious area within each catchment (%TIA-an indicator of land-use disturbance) was negatively related to a composite water quality index-the ERWQI-we developed for the East River. Modeling by partial least squares projection to latent structures (PLS) showed that family richness and relative abundance index (RAI) of macroinvertebrates were strongly influenced by both %TIA and ERWQI. Multi-response permutation procedure (MRPP) tests showed highly significant differences in family richness composition and RAI of macroinvertebrates among sites in the upper, middle, and lower course of the East River. MRPP also revealed differences in the family richness composition of nighttime drift samples between upper and middle site groups. Abundance (individuals m −3 ) and total family richness of drifting macroinvertebrates at each site were positively related to %TIA (range: 1.0-8.5%), while drift biomass was negatively related to dissolved oxygen and positively related to total suspended solids. Thus, human disturbances associated with land-use changes (increasing %TIA) and nutrient Electronic supplementary material The online version of this article (

Semistatic bioassays of endosulfan with native and introduced Australian fish indicated that they are sensitive to this insecticide at low concentrations. The introduced European carp (Cyprinus carpio) was the most sensitive species, with a measured 96‐h LC50 of 0.1 μg/L, whereas the native eastern rainbow fish (Melanotaenia duboulayi) and silver perch (Bidyanus bidyanus) were the least sensitive, with measured 96‐h LC50 values of 2.4μg/L. The measured 96‐h LC50 values of endosulfan for bony bream (Nematolosa erebi), golden perch (Macquaria ambigua), and rainbow trout (Oncorhynchus mykiss) were 0.2, 0.5, and 1.6 μ/L, respectively. Toxicities of endosulfan in Sydney mains water and Mehi River water were similar when rainbow fish, golden perch, and silver perch were used as test animals. The initial measured concentration was about 42% of the nominal concentration, whereas final measured concentration after 24 h in semistatic tests declined to 5.6% of the nominal concentration. The toxicity of endosulfan was similar in Australian native and introduced fish and in overseas fish, but varied markedly with test method. The 96‐h LC50 values for one species, the rainbow fish, varied from 0.5μg/L when reported as measured concentrations in a flow‐through test to 11.4 μg/L as nominal values in a static test. These differences span much of the variation found in 96‐h LC50 values for all fish species described in the world literature.

No abstract