Résumé
This study describes the design and application of a stratified sampling strategy of waterbodies to assess and analyze the distribution of cyanobacteria at a regional scale (Ile-de-France, IDF). Ten groups of hydrographical zones were defined within the IDF on the basis of their anthropogenic and geomorphologic characteristics. Sampling effort (
n
=
50) was then randomly allocated according to the number of waterbodies in each group. This sampling strategy was tested in August 2006, using a field probe to estimate total phytoplankton as well as cyanobacteria biomasses. The sampled waterbodies exhibited a wide range of phytoplankton (<1–375
μg
equiv.
Chl
a
L
−1) and cyanobacteria biomasses (<1–278
μg
equiv.
Chl
a
L
−1). 72% of the waterbodies in the IDF were classified as eutrophic (42% hypereutrophic), and 24% of the sites studied were dominated by cyanobacteria. Waterbodies connected to hydrographical networks (
n
=
26) showed significantly higher total (
p
<
0.0001; 3.5 times greater) and cyanobacterial (
p
<
0.001, 3.2 times greater) biomasses than the isolated ones (
n
=
24). No significant overall relationship was found through contingency analysis between waterbody trophic status and global land use categories (urban, periurban, and rural) within their hydrographical zones. However, concerning the waterbodies linked to hydrographical networks, the percentage of land covered by forest appeared as a good indicator of phytoplankton and cyanobacterial biomasses. This observation may be a consequence of lower amounts of nutrients being discharged into waterbodies from highly forested hydrological zone than from urban and/or agricultural areas. Our results illustrate a successful means of selecting representative waterbodies to conduct a regional assessment of cyanobacteria distribution using accessible GIS analyses.