RESUMO
The structure of intertidal benthic diatoms assemblages in the Tagus estuary was investigated during a 2-year survey, carried out in six stations with different sediment texture. Nonparametric multivariate analyses were used to characterize spatial and temporal patterns of the assemblages and to link them to the measured environmental variables. In addition, diversity and other features related to community physiognomy, such as size-class or life-form distributions, were used to describe the diatom assemblages. A total of 183 diatom taxa were identified during cell counts and their biovolume was determined. Differences between stations (analysis of similarity (ANOSIM), R = 0.932) were more evident than temporal patterns (R = 0.308) and mud content alone was the environmental variable most correlated to the biotic data (BEST, ρ = 0.863). Mudflat stations were typically colonized by low diversity diatom assemblages (H' ~ 1.9), mainly composed of medium-sized motile epipelic species (250-1,000 µm(3) ), that showed species-specific seasonal blooms (e.g., Navicula gregaria Donkin). Sandy stations had more complex and diverse diatom assemblages (H' ~ 3.2). They were mostly composed by a large set of minute epipsammic species (<250 µm(3) ) that, generally, did not show temporal patterns. The structure of intertidal diatom assemblages was largely defined by the interplay between epipelon and epipsammon, and its diversity was explained within the framework of the Intermediate Disturbance Hypothesis. However, the spatial distribution of epipelic and epipsammic life-forms showed that the definition of both functional groups should not be over-simplified.
RESUMO
The aim of this work is to assess the use of (SPOT) multispectral visible infrared remote sensing to study microphytobentos assemblages in a shellfish ecosystem (Bay of Bourgneuf, France). SPOT satellite images (acquired at low tide in spring or autumn between 1986 and 1998) were calibrated using in situ radiometric data, and the normalised vegetation index (NDVI) obtained from these images showed microphytobenthos on bay mudflats. Proliferation was mainly along a north-south strip, essentially localised around the +2 m isobath and covering a surface area of 19 to 25% of the total mudflat area studied (420 to 550 ha). Three factors seem to be responsible for the spatial structure of the assemblages: bathymetry, nutrient input from the Falleron River and its channel, and the location of oyster-farming areas. Although spatial and spectral resolutions of multispectral remote sensing data have certain limitations, this approach opens up a new field of application for hyperspectral remote sensing, particularly for synoptic mapping of biomass distribution.
