Mapping microphytobenthos in the intertidal zone of Northern ...

Mapping microphytobenthos in the intertidal zone of Northern
France using high spectral resolution field and airborne data
ABSTRACT
V. Carrère
Laboratoire Biogéochimie et Environnement du Littoral (LABEL)
UMR CNRS 8013 ELICO
Université du Littoral Côte d’Opale
MREN 32 av. Foch F 62930 Wimereux France
Email : carrere@mren2.univ-littoral.fr
Three simple techniques for estimating microphytobenthos Chlorophyll a (Chl a) concentration in intertidal flats
from high spectral resolution field spectra were compared and then applied to CASI and ROSIS images in order to
map spatial distribution of the biofilm. These techniques rely on relationships between different ways to quantify
Chl a absorption around 673 nm from field reflectance spectra and surface Chl a concentration derived from field
samples. The best approach should provide an estimate that is independent of sediment physical properties (grain
size, moisture content …) and type as well as illumination conditions in order to be easily applicable. In a first step,
high spectral resolution field reflectance spectra were acquired over tidal flats from the French shore of the Eastern
English Channel, which differed in sediment types and dynamics. The sites were studied at low tide and at various
time of year in order to include microphytobenthos seasonal blooms and illumination variations. Results on field
reflectance spectra show that all three methods are rather robust (R 2 > 0.8) which could imply little influence of the
measurement conditions on the reflectance at the considered wavelength range (570 – 720 nm) and a great stability
of the instrument used in the field. These techniques were then applied to CASI and ROSIS data acquired over one
of the site. Validation was performed by comparing the Chl a concentrations derived from the images to specific
field sites. Best results were obtained for scaled band area but Chl a was either overestimated (CASI) or
underestimated (ROSIS). These discrepancies, probably related to scaling effects and airborne data quality, point to
the need for further investigations on the influence of sampling strategy, Chl a estimation and sediment optical
properties with respect to spectral signatures.
Keywords: Chlorophyll a, reflectance data, microphytobenthos, CASI, ROSIS, intertidal flats.
1. INTRODUCTION
Information on microphytobenthos Chlorophyll a (Chl a) surface concentration on spatial scales ranging from
meters to kilometers can provide valuable information on biomass ([1], [2], [3], [4], [5]) and primary productivity
([6], [7], [8], [9]) on intertidal mudflats. The high level of primary productivity of benthic microalgae ([4], [7],
[10], [11], [12], [13]) represents a major energy source available for higher trophic levels. Therefore, a precise
evaluation is required in order to get a good knowledge of temporal and spatial variations of microphytobenthos
dynamics. The short term dynamics has been recently documented by Guarini et al. [13] and Blanchard et al. [14]
about temperate intertidal mudflats. Guarini et al. [15] have shown in the intertidal area of the Bay of Marennes-
Oléron (France) that the spatial distribution of microphytobenthic biomass exhibits a patchy pattern with maximum
values in the center of aggregates, the size of which changes with seasons (larger in summer during the productive
period than in winter).
Some studies have also tried to establish relationships between microphytobenthos Chl a surface concentrations and
prediction of critical erosion shear stresses ([|16], [17], [18], [19], [20]). Information on sediment surface stability
parameters on spatial scales ranging from meters to kilometers will provide valuable information on the dynamics
of the seabed morphology. It has been shown that activity of benthic microorganisms can increase the threshold of
erosion considerably by forming a network of extracellular polymeric substances (EPS) and by smoothing the
sediment surfaces ([18], [19], [21], [22], [23]).
_________________________________
Presented at the 3rd EARSeL Workshop on Imaging Spectroscopy, Herrsching, 13-16 May 2003
395

At present, mapping sediments or biomass in intertidal areas by conventional methods involves extensive sampling
programs that are often difficult in practice and expensive in time and manpower ([15], [24]). No matter how
extensive such programs are, the accuracy of the resultant maps is limited by the need to extrapolate from sample
sites to the whole area, usually by linking similar sites in a series of contours. Remote sensing may offer a more
efficient method of mapping sediment distribution if sediment types or features are unique in their reflection of
light and infrared radiation, since a much less extensive sampling program with more accurate extrapolation to the
whole would be involved. The potential of remote sensing for mapping and monitoring intertidal areas has been
realized and successfully applied by workers using both aerial photography and satellite imagery ([25], [26]).
Detailed mapping of intertidal seaweeds has been carried out in Canada by Zacharias et al. [27] and in France by
Bajjouk et al. [28] using Compact Airborne Spectrographic Imager (CASI). Remote sensing at higher spectral and
spatial resolutions using instrument such as CASI offers the prospect of extremely detailed land cover mapping and
the potential to model erosion and accretion rates of intertidal surfaces. Developments in hyperspectral remote
sensing have opened up the possibility of quantifying individual photosynthetic pigments within vegetation.
However, most published papers present empirical relationships between pigment concentration and various indices
([29], for example). Very few rely on the understanding of the physics of the reflectance signal to quantitatively
estimate pigment concentrations.
The objective of this paper is to present a comparison between three simple, physically based, techniques developed
to quantitatively estimate Chl a surface concentration in the intertidal zone from high spectral resolution field
spectrometer reflectance data. To test the robustness of these approaches, regression equations developed were used
to predict the concentrations in newly acquired field data sets.. These algorithms were then applied to airborne
CASI and ROSIS data. Finally, validation was performed, when possible, by using reference sites to compare Chl a
estimation from the images to field samples concentrations.
2. METHODOLOGY
2.1. Study sites
Selected study sites for field spectral measurements are located along the French coast of the Eastern English
Channel (Figure 1). To allow a wide range of concentrations and variable species, several campaigns took place at
various times of year. Measurements were performed at low tide, in 4 different test sites, reflecting different
conditions. From North to South:
- Site 1, Wimereux: this site is typical of an hydrodynamically exposed sandy beach habitat that can be
encountered on this shore. Measurements were performed on an area located in the upper intertidal zone, rather flat
and without sharp topographical features such as ripple marks, high pinnacles or deep surge channels. The substrate
was homogeneous medium size sand (200-250 µm, modal size), typical of the surrounding sandy habitat. Due to
the highly dynamic environment, microphytobenthos is resuspended and surface concentrations at low tide are low.
- Site 2, Baie d’Authie: here, several sites were measured in this macrotidal estuary where sandy to mixed
sandy flats prevail. The substrate composition varied from fine to medium grain sand mainly, with some coarse and
very fine grain mixed sandy sites. Microphytobenthos Chl a surface concentration varies as a function of sediment
grain size, estuary dynamics and seasonal blooms.
- Site 3, Baie de Somme: the Baie de Somme is the second ranked estuarine system, after the Seine
estuary, and the largest sandy-muddy (72 km 2 ) intertidal area on the French coasts of the Eastern English Channel.
This site is similar to the Baie d’Authie but the estuary being larger, sediment spatial variation occurs at a larger
scale, meaning that the measurements performed covered less diverse sediment types. The substrate grain size
varied mainly between 125 and 250 µm where most measurements were performed.
- Site 4, Baie de Seine: this site is very different from the previous ones as mudflats dominate where
measurements were performed. Particle size < 63 µm corresponded to more than 50% of the sediment composition.
396

Figure 1. Location of the test sites used to develop the general relationships between field spectral reflectance and
Chl a concentration.
2.2. Spectral measurements and field sampling
Reflectance spectra covering the 350 – 2500 nm spectral range with a spectral resolution of 1 (Visible) to 3 nm
(Infrared) were acquired for each station with a portable field spectrometer (ASD FieldSpec FR). Two optics were
selected (8° and 18°) depending on surface characteristics, covering a field of view of roughly 154 and 784 cm 2 at 1
m height. For each study sites, several stations were measured (between 10 and 22) to cover a wide range of
concentrations and sediment types. Ten spectra were acquired for each station and mean reflectance and standard
deviation were calculated for each set. Fewer measurements were acquired in Wimereux where, as indicated by
spectra visualization, there was very little evidence of microphytobenthos most of the time. The resulting mean
reflectance is the spectral signature of the station. This spectral signature reflects surface composition (sediment
mineralogical composition, pigments related to microphytobenthos species, moisture content, etc.) through the
presence of specific, well located, absorption features and surface physical properties (sediment grain size, etc.)
through the general spectral shape related to wavelength dependent diffusion for example.
Sampling and spectra acquisition occurred in spring 2001 and 2002 and summer 2001. An additional data set,
acquired on May 30th 2002 in Baie de Somme, was used for algorithm validation. All data sets were acquired on
days when low tide was close to solar noon in order to minimize directional effects on reflectance and optimize
surface exposure conditions.
For each station, sediment samples were randomly collected in the instrument field of view (FOV) using a 1.55 cm
diameter syringe or a 3.6 cm diameter corer, with 4 and 6 samples per measurement site for the 8° and 18° optics
respectively when using syringes and 3 samples for corers and the 18° FOV. The cores were pushed into the
sediment to a depth of 1 cm, where most of the cells are concentrated ([30], [31], [32], [33], [34]), carefully
removed and then stored in a cool box, brought back to the laboratory and stored in the dark at -20°C. In the
laboratory, sections of sediment were placed in acetone and pigments were extracted for 4 hours in the dark at 4°C
([35]). After extraction, samples were centrifuged at 4000 rpm for 15 min. Chlorophyll a concentrations (Chl a,
mg) in the supernatant were determined by spectrophotometry following Lorenzen [36]:
Chl a = V [(11.64 (O.D.663 – O.D.750) – 2.16(O.D.645 – O.D.750) + 0.1(O.D.630 – O.D.750)], (1)
Where V: extraction volume and O.D.λ : optical density at wavelength λ (nm)
397

Chl a concentrations in the supernatant have then been converted in terms of Chl a per surface unit (Chl a, mg.m -2 )
taking into account the surface of the sampling units.
2.3. Reflectance spectra analysis
In the field reflectance spectra, Chl a absorption is located around 673 nm which is a different wavelength position
than what is published for pure pigment (665 nm, [37], [38]). This is probably due to the fact that this wavelength
position is for pure pigment in a solvent and not in a leaving cell. Some tests we performed in the laboratory with
monospecific cultures also show an absorption located around 673 nm. The importance of the absorption is directly
related to pigment concentration and was estimated using three different and simple approaches (Figure 2).
2.3.1. Single band ratio (Figure 2a)
This approach is a method that has been classically used in spectroscopy and remote sensing in general to measure
amounts of atmospheric or surface components. It is based on differential absorption concept ([39], [40]) which
consists of a simple ratio between reflectance at maximum absorption (Rb) and reflectance outside the absorption,
referred to as the continuum (Rc). Here we used a ratio between reflectance at 673 nm (absorption) and reflectance
at 720 nm (continuum).
2.3.2. Normalized ratio (or scaled band depth) (Figure 2b, 2c)
Since the objective is to develop a simple method that has to be independent from measurement conditions, one
needs to perform some kind of normalization in order to remove the influence of other parameters on the spectral
signature and concentrate on the chlorophyll absorption itself. For example, it is well known that sediment grain
size influences the diffusing part of light, therefore changing the general shape of the spectrum through a
modification of the spectral dependency of the diffusion ([41], [42]). Spectral shape is also influenced by sediment
optical properties such as refracting index related to sediment composition. Sediment moisture content will also
modify the spectral signature (enhanced liquid water absorptions, level change of the general reflectance).
Removing the continuum of the spectrum allows isolating the spectral feature from other effects. This approach was
proposed by Clark and Roush [43] who determined the depth of a spectral absorption feature and related it to frost
grain size and by Clark [44] and Clark et al. [45] and references therein for applications to rocks and minerals. The
scaled band depth Db is calculated as the difference between the continuum reflectance Rc and the reflectance
spectrum Rb in the deepest part of the absorption band, normalized by the continuum reflectance:
Db = (Rc – Rb) / Rc. (2)
While this band depth method is a valuable approach to the problem, its reliance on a single band causes the
accuracy of the results to suffer in the presence of noise in the reflectance spectrum. Noise-induced changes in the
spectrum affect the depth of the absorption feature and may give erroneous chlorophyll concentration estimates.
2.3.3. Integral after continuum removal (scaled band area) (Figure 2d)
To minimize any effects of instrumental noise on chlorophyll retrievals, we also tested the use of the scaled area of
the absorption feature Ab rather than simply the scaled absorption band depth. This approach was used by Nolin and
Dozier [46] to estimate snow grain size from hyperspectral data. Scaled band area is a dimensionless quantity and is
calculated by integrating the scaled absorption band depth over the wavelengths of the absorption feature:
Ab = ∫λ (Rc – Rb) / Rc. (3)
The basic assumption is that the noise is randomly distributed Gaussian noise (“white noise”). While the exact
distribution of sensor noise is not known, a normal distribution is a reasonable assumption and has been used by
others in examining sensor noise characteristics ([47]). By integrating over the absorption feature, the fluctuations
caused by noise should average out and produce an estimate closer to the true value. Integration was restricted to
the [650 – 720 nm] wavelength range in order to exclude influence of other absorbing pigments such as Chlorophyll
c (absorption at 590 and 635 nm).
398

For each method, regression laws where calculated between absorption estimated from spectra and mean Chl a
concentration as determined from the samples. Abnormal values that could be related to sampling strategy were
eliminated before calculating the mean Chl a concentration and the goodness of fit in each case.
Figure 2. Estimation of chlorophyll a absorption: a – simple band ratio; b – principle of continuum removal. Dashed lines
represent mean surface reflectance (10 measurements) plus or minus standard deviation; c – normalized ratio (after continuum
removal); d – scaled band area after continuum removal.
2.4. Airborne data
Airborne data were acquired over the Baie d’Authie site. This site was selected because it is characteristic of the
estuaries of the French shore of the Eastern English Channel. Baie d’Authie is a macrotidal estuary showing a great
variety of sediment types and areas where the biofilm is present at low tide.
399

2.4.1. CASI data
CASI overflights occurred in September 2000, under poor weather conditions (overcast skies), preventing
concurrent atmospheric measurements. Field reflectance spectra were nonetheless acquired over four reference
sites. Empirical atmospheric corrections were performed using the empirical line technique and two of the reference
sites. Three flight lines were necessary to cover the entire Bay but we tested the algorithms on a subscene
corresponding to the area were the reference sites were located and the biofilm was present.
The CASI instrument was flown in a “MERIS like” configuration with an additional band centered at 673.9 nm to
map Chl a. Table 1 summarizes the main characteristics of the CASI data.
Table 1. Characteristics of CASI data
2.4.2. ROSIS data
Band # Central wavelength (nm) Band width (nm)
1 411.9 14.0
2 443.3 14.2
3 489.8 14.2
4 510.3 14.2
5 560.1 12.4
6 619.7 14.4
7 665.4 10.6
8 673.9 8.8
9 681.9 8.8
10 704.5 12.6
11 733.2 10.6
12 760.8 6.8
13 776.1 18.4
14 865.2 24.0
Spatial resolution: 2 m
ROSIS data were acquired in the framework of the 2001 HySens Campaign, in August 2001. The weather
conditions were good, allowing for concurrent atmospheric and reflectance measurements. Aerosol optical depth
derived from atmospheric measurements was used by DLR in their atmospheric correction procedure. Data were
also georeferenced using aircraft and ground GPS information. Field reference sites were spatially located using
GPS.
ROSIS is an airborne imaging spectrometer, making use of a two-dimensional CCD array for imaging
simultaneously 115 spectral bands of 512 picture elements perpendicular to the flight direction. The spatial
resolution was 2 m, the spectral range 416.5 – 872.5 nm with a spectral sampling interval of 4 nm.
The data appeared fairly noisy so additional preprocessing was necessary. MNF transform was performed to
remove most of the noise, the first 5 channels were removed and a reference site (dry sand) was used to refine
atmospheric corrections.
3. RESULTS
3.1. Field reflectance spectra
Results for the regression of simple ratio, normalized ratio and scaled band area versus field samples Chl a
concentration are presented in Figure 3. The dependency of absorption depth (simple ratio) or normalized
absorption depth (normalized ratio) to Chl a concentration can be approximated by an exponential function fit
(Beer-Lambert-like law) (Figure 4a and b). The functional relationship between scaled band area and Chl a
concentration was assumed a linear function (Figure 4c).
400

Figure 3. Regression laws to estimate chlorophyll a
concentration: a – simple band ratio; b – normalized
ratio; c – scaled band area.
These preliminary results are rather satisfactory as the R 2 > 0.8 for all three approaches which seem to give
equivalent results. This would indicate very little impact of measurement conditions on the estimation of the
absorption due to Chl a. This might point to the fact that there is little variation in sediment grain size with respect
to the wavelength range considered or little change in water content or that our sampling was not fully
representative of the observed conditions. A more detailed analysis of the influence of sampling strategy and
sediment optical properties with respect to spectral signatures is under progress.
Another point is that the instrument we used is rather stable, with very little noise that is obviously already removed
by averaging 10 spectra at a time (very small standard deviation, see Figure 3b).
401

Figure 4. Inversion of data from the Bay of Somme May
2002 campaign: a – simple band ratio; b – normalized
ratio; c – scaled band area.
Each technique was then validated using a newly acquired data set in Baie de Somme (May 30, 2002). Chl a
concentration was estimated by inverting the resulting functions and compared to the concentrations measured from
the samples. As shown on Figure 4, there is a strong correlation (R 2 > 0.9) between estimated and measured Chl a
concentration, no matter which algorithm is used, for the range of Chl a concentrations encountered that day (50 –
200 mg.m -2 ). More measurements under different conditions are still needed to crosscheck these results.
3.2. Airborne data
In a first step, field reflectance spectra were convolved to airborne sensor bandpasses. New regressions were
derived for the three approaches. Then a ratio image was produced as well as, after continuum removal, normalized
402

atio and scaled band area images. Chl a maps were derived by inverting the appropriate regressions. Validation
was performed, when possible, by using reference sites to compare Chl a estimation from the images to field
samples concentrations.
3.2.1. CASI data
As shown on Figure 5, the results are rather satisfactory but not as good as for field spectra. This can probably be
related to the lower spectral resolution of CASI data. This assumption is consistent with a better fit for scaled band
area which in fact is equivalent to degrading spectral resolution.
403
Figure 5. Regression laws to estimate chlorophyll a
concentration after convolution to CASI bandpasses :
a – simple band ratio; b – normalized ratio; c – scaled
band area.

Resulting Chl a maps (Figure 6) appear qualitatively correct but the retrieved concentration range seems to show an
overestimation of Chl a based on our knowledge of the area. Unfortunately, it was impossible to validate these
concentrations because field samples collected at the time of the overflight were not properly frozen, preventing an
accurate estimate of Chl a concentration.
Figure 6. Chlorophyll a surface concentration maps derived from CASI data: a – “false color composite” (red: 865 nm; green:
673 nm; blue: 489 nm); b – simple band ratio; c – normalized ratio; d – scaled band area.
3.2.2. ROSIS
As shown in Figure 7, regressions show a better fit than for CASI data. This is probably due to the fact that ROSIS
spectral resolution is closer to field spectra resolution. All three approaches appear again equally robust (R 2 = 0.86).
When applied to ROSIS images (Figure 8), as for CASI maps, Chl a spatial distribution is qualitatively correct but
the concentration range is narrower.
Validation through comparison with field samples (Table 2) shows a strong underestimation of Chl a concentration.
This discrepancy could be explained in part by the noise in the data, assumption which seems to be confirmed by
the fact that the Scaled Band Area gives better results than the other approaches. Instrument calibration can also be
404

a potential factor. More certainly, it can also be attributed to the scaling factor between field spectral measurements
and pixel size. It is well known that the biofilm is not spatially homogeneous. Therefore, field sampling and field
spectral measurements are not fully representative of the spatial distribution of the microphytobenthos. Moreover,
we are estimating Chl a concentration from the field samples up to a 1 cm depth which also leads to an
overestimation of Chl a with respect to what the field or the airborne instrument measures.
405
Figure 7. Regression laws to estimate chlorophyll a
concentration after convolution to ROSIS bandpasses: a –
simple band ratio; b – normalized ratio; c – scaled band
area.

Figure 8. Chlorophyll a surface concentration maps derived from ROSIS data: a – “false color composite” (red: 752.5 nm;
green: 672.5 nm; blue: 632.5 nm); b – simple band ratio; c – normalized ratio; d – scaled band area
Table 2 . Comparison between Chl a concentrations (mg.m -2 ) derived from field samples and from ROSIS data –
NR: Normalized ratio; SBA: Scaled Band Area.
SITES SAMPLES RATIO NR SBA
I 474.39 12.90 10.23 27.44
J 199.05 13.99 13.4 23.53
K 360.45 8.38 6.92 22.37
L 91.76 10.56 6.78 19.98
M 517.88 47.36 48.37 69.04
N 204.72 11.20 10.17 25.23
O 98.86 24.55 20.86 39.80
P 5.12 42.43 35.70 53.60
Q 108.14 52.62 49.90 71.07
R 149.04 11.34 13.42 31.25
S 17.93 6.09 7.37 15.53
T 3.39 2.19 1.12 4.72
U 12.28 4.73 7.23 14.38
406

4. CONCLUSIONS
Preliminary results regarding the simple approaches developed here indicate that all three methods are rather robust
(R 2 > 0.8) when applied to field spectra. There seems to be very little difference between the three, indicating either
that there is very little influence of measurement conditions in the spectral range considered or that the field
sampling technique averages out this impact. In addition, the instrument used in the field is fairly stable which
explains why integrating over the absorption feature does not dramatically improve the estimation.
When applied to airborne CASI and ROSIS data, all three techniques appear also equally robust when looking at
the new regression laws derived from field spectra convolved to instrument bandpasses. Better results were
obtained for ROSIS data which was to be expected as ROSIS has a higher spectral resolution than CASI and
contiguous bands.
Inversion of the resulting functions on the images produced Chl a concentration maps that appeared qualitatively
correct. Unfortunately, CASI maps could not be validated through comparison with field samples but appeared
overestimated based on our knowledge of the area. On the other hand, ROSIS Chl a maps could be validated and
show an underestimation of Chl a concentration when compared to reference sites.
These discrepancies could mainly be explained by the scaling factor between field spectral measurements and
airborne instrument pixel size. It is well known that the spatial distribution of the biofilm is generally fairly patchy;
therefore field sampling and spectral measurements might not be representative of a 2 x 2 m pixel. In addition, both
data sets present different characteristics for the same spatial resolution. CASI data have a coarser spectral
resolution than field spectra and were acquired under poor weather conditions, preventing for an accurate
atmospheric correction. ROSIS data have a higher spectral resolution and contiguous bands but are fairly noisy
which partly explains why the scaled band area approach gives better results as it “smoothes” the noise. In both
cases, instrument calibration might also be an important factor.
These preliminary results clearly show that further studies are required to better understand the influence of the
sampling strategy (particularly estimating Chl a concentration within the first centimeter of sediment) and sediment
optical properties on the spectral signature at various scales. Field measurements have to be extended to other sites
with different sediment characteristics. Sampling strategies have to be compared and validated in order to
extrapolate to airborne instrument pixel size. We are also investigating the possibility of using this approach to
identify and quantify other pigments, leading to an estimation of biodiversity as pigments can be used as
fingerprints for different groups of microalgae.
ACKNOWLEDGEMENTS
This work was supported by the French “Programme National d’Environnement Côtier”. CASI data were acquired
thanks to special funding from Université des Sciences et Techniques de Lille. ROSIS data were acquired as part of
the 2001 HySens campaign. The author would particularly like to thank the DLR HySens Team for providing
atmospherically and geometrically corrected data and for their great flexibility in acquiring data under such strong
constraints. Samuel Degézelle and Nicolas Spilmont helped with the Chlorophyll a extraction and Aurélie Robin
characterized the Bay d’Authie sediment samples.
REFERENCES
[1] BLANCHARD, G.F., GUARINI, J.M., RICHARD, P., GROS, P., AND MORNET, F., 1996: Quantifying the short-term
temperature effect of light-saturated photosynthesis of intertidal microphytobenthos. Mar. Ecol. Prog. Ser.
134:309-313.
[2] BLANCHARD, G.F., GUARINI, J.M., GROS, P., AND RICHARD, P., 1997: Seasonal effect on the relationship
between the photosynthetic capacity of intertidal microphytobenthos and temperature. J. Phycol. 33:723-728.
[3] JONGE DE, V.N., 1980: Fluctuations in the organic carbon to chlorophyll a ratio for estuarine benthic diatom
populations. Mar Ecol Prog Ser 2:345-353.
[4] MCINTYRE, H.L., GEIDER, R.J., AND MILLER, D.C., 1996: Microphytobenthos: the ecological role of the ‘secret
garden’ of unvegetated, shallow-water marine habitats. Distribution, abundance and primary production. Estuaries
19:186-201.
407

[5] CARIOU-LE GALL, V., AND BLANCHARD, G.F., 1995: Monthly HPLC measurements of pigment concentration
from an intertidal muddy sediment of Marennes-Oléron Bay, France. Mar Ecol Prog Ser 121:171-179.
[6] PINCKNEY, J.L., AND ZINGMARK, R.G., 1991: Effects of tidal stage and sun angles on intertidal benthic
microalgal productivity. Mar Ecol Prog. Ser. 76:81-89.
[7] PINCKNEY, J.L., AND ZINGMARK, R.G., 1993: Modeling the annual production of intertidal benthic microalgae
in estuarine ecosystems. J Phycol 29:396-407.
[8] BLANCHARD, G.F., 1994 : Caractéristiques photosynthétiques du microphytobenthos d’une vasière intertidale.
C.R.A.Sc. Paris (Life Sciences) 317:633-673.
[9] BLANCHARD, G.F., AND CARIOU–LE GALL, V., 1994: Photosynthetic characteristics of microphytobenthos in
Marennes- Oléron bay, France : preliminary results. J. Exp. Mar. Biol. Ecol. 182:1-14.
[10] ADMIRAAL, W., 1984: The ecology of estuarine sediment-inhabiting diatoms. Progress in Phycological
Research 3:269-322.
[11] COLIJN, F., AND DE JONGE, V.N., 1984: Primary production of microphytobenthos in the Ems-Dollard Estuary.
Mar Ecol Prog Ser. 4:9-21.
[12] MILLER, D.C., GEIDER, R.J., AND MCINTYRE, H.L., 1996: Microphytobenthos: the ecological role of the
“secret garden“ of unvegetated, shallow-water marine habitats: II Role in sediment stability and shallow water food
webs. Estuaries 19:202-212.
[13] GUARINI, J.M., BLANCHARD, G.F., GROS, P., GOULEAU, D., AND BACHER, C., 2000: Dynamic model of the
short term variability of microphytobenthic biomass on temperate intertidal mudflats. Mar Ecol Prog Ser 195:291-
303.
[14] BLANCHARD, G.F., GUARIN,I J.M., BACHER, C., AND HUET, V., 1998 : Contrôle de la dynamique à court terme
du microphytobenthos intertidal par le cycle exondation-submersion. C.R.A.Sc. Paris (Life Sciences) 321:501-508.
[15] GUARINI, J.M., BLANCHARD, G., BACHER, C., GROS, P., RIERA, P., RICHARD, P., GOULEAU, D., GALOIS, R.,
PROU, J., AND SAURIAU, P.G., 1998: Dynamics of spatial patterns of microphytobenthic biomass : inferences from a
geostatistical analysis of two comprehensive surveys in Marennes-Oléron Bay (France). Mar Ecol Prog Ser
166:131-141.
[16] RIETHMÜLLER, R., HAKVOORT, J.H.M., HEINEKE, M., HEYMANN, K., KÜHL, H., AND WITTE, G., 1998:
Relating erosion shear stress to tidal flat surface color. In: Black K.S., Paterson D.M., Cramp A. (Eds): Sedimentary
Processes in the Intertidal Zone, 139:283-293.
[17 RIETHMÜLLER, R., HEINEKE, M., KÜHL, H., AND KEUKER-RÜDIGER, R., 2000: Chlorophyll a concentration as
an index of sediment stabilization by microphytobenthos? Continental Shelf Research 20:1351-1372.
[18] UNDERWOOD, G.J.C., AND PATERSON, D.M., 1993: Seasonal changes in diatom biomass, sediment stability
and biogenic stabilization in the Severn Estuary. J. Mar. Biol. Ass. UK 73:871-887.
[19] MADSEN, K.N., NILSON, P., AND SUNDBÄCK, K., 1993: The influence of benthic micro algae on the stability of
a subtidal sediment. J. Exp. Mar. Biol. Ecol. 170:159-177.
[20] HAKVOORT, J.H.M., HEINEKE, M., HEYMANN, K., KÜHL, H., RIETHMÜLLER, R., AND WITTE, G., 1998: A
basis for mapping the erodability of tidal flats by optical remote sensing. Mar. Freshwater Res. 49:867-873.
[21] DADE, W.B., DAVIS, J.D., NICHOLS, P.D., NOWELL, A.R.M., THISTLE, D., TREXLER, M.B., AND WHITE, D.C.,
1990: Effects of bacterial exopolymer adhesion on the entrainment of sand. Geomicrobiology Journal 8:1-16.
[22] PATERSON, D.M., YALLOP, M.L., AND GEORGE, C., 1994: Spatial variability in the sediment erodability on the
Island of Texel. In: W.Krummbein, D.M. Paterson and L.J. Stal (Eds): Biostabilization of sediments, pp.107-120.
[23] AMOS, C.L., BRYLINSKY, T.F., SUTHERLAND, T.F., O’BRIEN, D., LEE, S., AND CRAMP, A. 1998: The stability
of a mudflat in the Humber estuary, South Yorkshire, UK. In: K.S., Paterson D.M., Cramp A. (Eds): Sedimentary
Processes in the Intertidal Zone, Black, 139:25-43.
[24] SEURONT, L., AND SPILMONT, N., 2002: Self-organized criticality in intertidal microphytobenthos patch
patterns. Physica A
[25] FOLVING, S., 1984: The Danish Wadden Sea: Thematic mapping by means of remote sensing. Folia
Geographica Danica 15:1-56.
[26] DOERFFER, R., MURPHY, D., 1989: Factor analysis and classification of remotely sensed data for monitoring
tidal flats. Helgolander Meeresunters 43:275-293.
[27] ZACHARIAS, M., OLAF, N., AND BORSTAD, G., 1992: An assessment and classification of a multispectral
bandset for the remote sensing of intertidal seaweeds. Can. J. Remote Sens. 18:263-274.
[28] BAJJOUK, T., GUILLAUMONT, B., AND POPULUS, J., 1996: Application of airborne imaging spectrometry
system data to intertidal seaweed classification and mapping. Hydrobiologia 326/327:463-471.
[29] BLACKBURN, G.A., 1998: Quantifying Chlorophylls and Carotenoids at Leaf and Canopy Scales: An
Evaluation of Some Hyperspectral Approaches. Remote Sens. Environ. 66:273-235.
[30] CADÉE, G.C., AND HEGEMAN, J., 1974: Primary production of the benthic microflora living on tidal flats in the
Dutch Wadden Sea. Neth. J. Sea Res. 8:260-291.
408

[31] BAILLIE, P.W., 1987: Diatom size distribution and community stratification in estuarine intertidal sediments.
Estuar. Coast. Shelf Sci. 25:193-209.
[32] ADMIRAAL, W., ARKEL, M.A., VAN, BARETTA, J.W., COLIJN, F., EBENHÖH, W., JONGE, V.N., DE KOP, A.,
RUARDIJ, P., AND , H.G.J., 1988: The construction of the benthic submodel. In: Baretta, J. & Ruardij, P. (Eds.):
Tidal flat estuaries. Simulation and analysis of the Ems estuary. Ecological studies 71, Springer-Verlag, Heidelberg,
p. 105-152.
[33] DELGADO, M., 1989: Abundance and distribution of microphytobenthos in the bays of Ebro Delta (Spain).
Estuar. Coast. Shelf Sci. 29:183-194.
[34] JONGE DE, V.N., AND COLIJN, F., 1994: Dynamics of microphytobenthos biomass in the Ems estuary. Mar
Ecol Prog Ser 104:185-196.
[35] BRUNET, C., 1994 : Analyse des pigments photosynthétiques par HPLC : communautés phytoplanctoniques et
productivité primaire en Manche orientale. Ph. D. Thesis. Univ. Paris VI 367 p.
[36] LORENZEN, C.J., 1967: Determination of chlorophyll and phaeopigments : spectrometric equations. Limnol
Oceanogr 12:343-346.
[37] JEFFREY, S.W., AND HUMPHREY, G.F., 1975: New spectrophotometric equations for determining chlorophylls
a, b, c1 and c2 in higher plants algae and natural phytoplankton. Biochem. Physiol. Pflanz 167:191-194.
[38] JEFFREY, S.W., MATOURA, R.F.C., AND WRIGHT, S.W., 1977: Phytoplankton pigments in Oceanography.
SCOR, UNESCO Publishing, 661 p.
[39] FOWLE, F.E., 1912: The spectroscopic determination of aqueous vapor. Astrophys. J. 35:149-153.
[40] FOWLE, F.E., 1913: The determination of aqueous water vapor above Mount Wilson. Astrophys. J. 37:359-
372.
[41] HUNT, G.R., AND SALISBURY, J.W., 1970: Visible and near-infrared spectra of minerals and rocks: I Silicate
minerals. Modern Geology 1:283-300.
[42] LEU, D.J., 1977: Visible and near-infrared reflectance of beach sands : A study on the spectral
reflectance/grain size relationship. Remote Sens. Environ. 6:169-182.
[43] CLARK, R.N., AND ROUSH, T., 1984: Reflectance spectroscopy: quantitative analysis techniques for remote
sensing applications. J.Geoph.Res. 89:6329-6340.
[44] CLARK, R.N., 1999: Chapter 1: Spectroscopy of rocks and minerals and principles of spectroscopy. In: A.N.
Rencz (Ed:) Manual of Remote Sensing, John Wiley and Sons, New York, p. 3-58.
[45] CLARK, R.N., SWAIZE, G.A., LIVO, K.E., KOKALY, R.F., SUTLEY, S.J., DALTON, J.B., MCDOUGAL, R.R., AND
GENT, C.A., 2002: Imaging spectroscopy: Earth and planetary remote sensing with the USGS Tetracorder and
Expert Systems. J. Geoph. Res., in Press.
[46] NOLIN, A.W., AND DOZIER, J., 2000: A hyperspectral method for remotely sensing the grain size of snow.
Remote Sens Environ 74:207-216.
[47] SABOL, D.E., JR., ADAMS, J.B., AND SMITH, M.O., 1992: Quantitative subpixel spectral detection of targets in
multispectral images. J. Geoph. Res. 97:2659-2672.
409