Aquatic Toxicology 78 (2006) 303–314 Effects of the phenylurea herbicide diuron on natural riverine microbial communities in an experimental study St´ ephane Pesce, C´ eline Fajon ∗ , Corinne Bardot, Fr´ ed´ erique Bonnemoy, Christophe Portelli, Jacques Bohatier Laboratoire de Biologie des Protistes, UMR CNRS 6023, Universit´ e Blaise Pascal, 63177 Aubi` ere Cedex, France Received 16 January 2006; received in revised form 16 March 2006; accepted 19 March 2006 Abstract The effects of the phenylurea herbicide diuron (10 gl -1 ) on natural riverine microbial communities were investigated using a three-week laboratory microcosm study. During the first six days, a latency period was observed both in the algal and the bacterial communities despite favorable abiotic conditions and independently of diuron exposure. From the second week, an intense algal bloom (chlorophyll a concentrations and cell abundances) was observed in the uncontaminated microcosms but not in the treated microcosms. The bloom stimulated the bacterial community and led to an increase in heterotrophic bacterial production ([ 3 H]thymidine incorporation), activity (CTC reduction) and cell abundance. In parallel, shifts in bacterial community composition were recorded by polymerase chain reaction (PCR)–temporal temperature gradient gel electrophoresis (TTGE) analysis, whereas no major variation was detected using the fluorescent in situ hybridization (FISH) method. In the treated microcosms, the diuron acted not by damaging the initial communities but by inhibiting the algal bloom and indirectly maintaining constant bacterial conditions throughout the experiment. These inhibitory effects, which were recorded in terms of abundance, activity and diversity, suggest that exposure to diuron can decrease the recovery capacities of microbial communities and delay the resumption of an efficient microbial food web despite favorable environmental conditions. © 2006 Elsevier B.V. All rights reserved. Keywords: Community analysis; Bacteria; Microalgae; Microcosm; Diuron; Lotic system 1. Introduction Pollution of aquatic ecosystems by pesticide contamination is a major environmental concern. Numerous studies have been conducted to determine the occurrence of these xenobiotics in European streams and rivers (e.g. Gerecke et al., 2002; Carabias- Mart´ ınez et al., 2003; Cerejeira et al., 2003). In France, the Jauron River (Puy de Dˆ ome), a first- and second-order stream located 25 km SE of Clermont-Ferrand, is monitored by a pes- ticide pollution watch net. This river which is characterized by meso-eutrophic conditions in spring and summer (Fajon et al., unpublished data) is frequently polluted by numerous chemical compounds, especially the phenylurea herbicide diuron (3-(3,4- dichlorophenyl)-1,1-dimethylurea). This molecule was detected 20 times between 2000 and 2003 (monthly sampling) in six sam- ∗ Corresponding author. Tel.: +33 473407465; fax: +33 473407670. E-mail address: celine.fajon@univ-bpclermont.fr (C. Fajon). pling sites located on the Jauron river with a range from 0.05 to 4.21 gl -1 (Phyt’eauvergne, 2004). In French rivers diuron is detected in 38% of surface waters and its concentration ranges from 0.05 to 20.3 gl -1 (IFEN, 2000). This substituted urea herbicide used for total control of weeds and mosses inhibits photosynthesis by preventing oxygen production (Wessels and Van der Veen, 1956) and blocks electron transfer at photosystem II of photosynthetic microorganisms and plants. The intensive use of this xenobiotic leads to surface water contamination by soil leaching and overland flow (Garmouma et al., 1997; Louchart et al., 2000; Barra Caracciolo et al., 2005) in agricul- tural (Thurman et al., 2000; Mitchell et al., 2005) and urban areas (Revitt et al., 2002; Blanchoud et al., 2004). The toxic properties of diuron mean that this contamination poses signif- icant toxicological risk to resident aquatic organisms. Because of their physiological characteristics, planktonic and periphytic microalgae represent potential primary targets for herbicides in lotic ecosystems. Previous studies based on single-species tox- icity tests have reported a wide variation in algal sensitivity to 0166-445X/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.aquatox.2006.03.006