Ecosystems (2010) 13: 338-351
DOI: 10.1007/s 10021-010-932 2-9
ECOSYSTEMS!
) 2010 The A uthor(s). This article is published w ith open access at Springerlink.com
Spatial Synchrony in Intertidal
Benthic Algal Biomass in Temperate
Coastal and Estuarine Ecosystems
Daphne van der Wal,* A nnette Wielemaker-van den Dool,
and Peter M. J. Herman
Netherlands Institute of Ecology (NIOO-KNAW), Centre for Estuarine and Marine Ecology, P.O. Box 140, 4400 AC Yerseke,
Netherlands
A
bstr a c t
M icrophytobenthos plays a vital role in estuarine
and coastal carbon cycling and food webs. Yet, the
role of exogenous factors, and thus the effects of
climate change, in regulating m icrophytobenthic
biomass is poorly understood. We aim ed to unravel
the m echanism s structuring m icrophytobenthic
biomass b oth w ithin and across ecosystems. The
spatiotem poral distribution of the biom ass of
intertidal benthic algae (dom inated by diatoms)
was estim ated w ith an unprecedented spatial ex ­
ten t from tim e-series of Norm alized Differential
V egetation Index (NDVI) derived from a 6-year
period of daily A qua MODIS 250-m images of se­
ven tem perate, m ostly turbid, estuarine and coastal
ecosystems. These NDVI tim e-series w ere related to
m eteorological and environm ental conditions.
Intertidal benthic algal biomass varied seasonally in
all ecosystems, in parallel w ith m eteorology and
w ater quality. Seasonal variation was m ore p ro ­
nou n ced in m u d th a n in sand. In teran n u al varia­
tion in biom ass was small, b u t synchronized yearto-year biomass fluctuations occurred in a n u m b er
of disjointed ecosystems. Air tem perature ex ­
plained in teran n u al fluctuations in biom ass in a
n u m b er of sites, b u t the synchrony was m ainly
driven by the w ind/w ave climate: high w ind
velocities reduced m icrophytobenthic biomass, ei­
th er thro u g h increased resuspension or reduced
em ersion duration. Spatial variation in biomass was
largely explained by em ersion duration and m ud
content, both w ithin and across ecosystems. The
results im ply th at effects on m icrophytobenthic
standing stock can be anticipated w h en the position
in the tidal fram e is altered, for exam ple due to sea
level rise. Increased storm iness will also result in a
large-scale decrease of biomass.
I n t r o d u c t io n
an n u al productivity of benthic microalgae alone is
estim ated to contribute about 500 m illion tons of
carbon globally (Cahoon 1999). The m ajor p art of
the benthic m icroalgae inhabit intertidal sedim ents
in estuarine and coastal ecosystems (Heip and
others 1995; C ahoon 1999). They constitute a p ri­
m ary carbon source for food webs (Miller and
others 1996; H erm an and others 1999), and in flu ­
ence sedim ent dynam ics by form ing a sedim ent
stabilizing biofilm (Holland and others 1974).
K ey w ord s: m icrophytobenthos; NDVI; rem ote
sensing; inundation; spatial scales; synchrony.
B enthic algae are a vital link in carbon cycling in
shallow w aters (Gattuso and others 2006); the
R eceived 28 A ugust 2009; accepted 1 F ebruary 2010;
published online 5 M arch 2010
A u th o r C o n trib u tio n s: D. v an der W al conceived an d designed the
study, com piled th e data, p erform ed th e analyses an d w ro te th e article. A.
W ielem aker-van den Dool processed th e satellite im agery to derive NDVI
data. P. M. J. H erm an c o n trib u ted to study design, (statistical) analysis
an d in te rp reta tio n an d edited th e article.
*Corresponding author; e-mail: d.vanderw al@ nioo.knaw .nl
338
Synchrony in Intertidal B enthic Algal Biomass
Intertidal sedim ents in estuarine and coastal
ecosystems experience fluctuations in tem perature,
irradiance, desiccation, nutrients, grain-size, and
activity and grazing by benthic fauna, resulting in
variability in m icrophytobenthic biomass an d spe­
cies com position at different tem poral scales (Ad­
m iraal and Peletier 1980; U nderw ood 1994;
B lanchard an d others 1997). Intertidal zones also
show large spatial heterogeneity for m any of these
factors (for exam ple, irradiance as a function of
em ersion duration in turbid w aters, an d sedim ent
grain-size), and so a large spatial variation in m i­
crophytobenthic biomass is also expected (Pinck­
ney and Zingm ark 1993; M acIntyre an d others
1996; U nderw ood and K rom kam p 1999; Guarini
and others 2008).
Given th eir global im portance, estuarine and
coastal w ater bodies are a m ajor focus of concern
regarding th e p otential impacts of anthropogenic
clim ate change (Harley an d others 2006). Observed
and projected climate trends include changes in (air
and sea) tem perature, sea level, tidal range, p re ­
cipitation, river discharge an d turbidity, and storm
frequency and intensity (Bates an d others 2008),
all of w hich m ay potentially affect intertidal m i­
crophytobenthic biomass an d prim ary production.
Thus, th ere is a need to unravel the factors
structuring m icrophytobenthic biomass, so th at the
effects of clim ate change can be better predicted.
We hypothesize th at the key to unraveling these
factors is th e extent of synchrony in m icrophyto­
b enthic biom ass across ecosystems. Spatial syn­
chrony in populations, th at is, the coincident
changes (tem poral coherence) in abundance of
geographically disjointed populations, m ay arise
from th ree prim ary m echanism s: dispersal am ong
populations, trophic interactions w ith populations
of o th er species th a t are them selves spatially syn­
chronous (such as benthic m acrofauna) or mobile,
and congruent dependence of population dynam ics
on a synchronous exogenous (random ) factor
(such as tem perature) (Liebhold and others 2004).
Thus, w h e n m icrophytobenthic biom ass fluctuates
in th e same w ay in different ecosystems, it will be
governed largely by factors th a t are effective at
such a large geographical scale, w hereas any local
departure will point to the influence of locally
operating factors (Beukem a an d Essink 1986).
Advances in rem ote sensing allow a synoptic
estim ation of intertidal m icrophytobenthic biomass
over large spatial scales (for example, Combe and
others 2005; Van der W al an d others 2008), b u t to
date very few studies have taken a m ulti-tem poral
basin-w ide or inter-ecosystenr approach. In con­
trast, in th e m arine (Behrenfeld and others 2006)
339
and terrestrial (Pettorelli and others 2005) realm ,
satellite rem ote sensing is w idely used to assess
changes in th e biom ass of photo-autotrophs over
vast areas. A com m on m easure for photo-autotrophs is th e Norm alized Differential Vegetation
Index, NDVI, com puted from reflectance in th e red
(Rr ) and near-infrared (RNœ) part of the electro­
m agnetic spectrum , th at is: NDVI = (RNm — Rr )/
(Rnir + R r )- P hoto-autotrophs absorb m ost of the
incom ing light, notably in th e red. Thus, NDVI
gives higher values w ith increasing biomass, cover
or h ealth (Tucker 1979). In the absence of rnacrophytes, NDVI is a good proxy for chlorophyll-«?
(Chl-fl) and m icrophytobenthic biomass in in te r­
tidal sedim ents (Krom kam p an d others 2006; Van
der Wal and others 2008). Retrieval of NDVI from
m oderate resolution satellite sensors, such as Aqua
MODIS (Huete an d others 2002), enables m icro­
phytobenthic biomass to be m onitored sim ulta­
neously in m ultiple ecosystems. Such inform ation
can also facilitate the incorporation of benthic algae
into m odels of carbon and nitrogen cycling in
coastal an d estuarine areas.
This article addresses the spatiotem poral varia­
tion of intertidal benthic algae, derived from a 6year daily tinre-series of Aqua MODIS 250-rn
imagery. The study is aim ed at ( 1 ) determ ining the
extent of synchrony in m icrophytobenthic biomass
across ecosystems and the coupling of tem poral
variation in m icrophytobenthos and exogenous
factors (m eteorology and w ater quality) and (2)
determ ining the extent of spatial structuring of
m icrophytobenthos by em ersion duration and
sedim ent type, both w ithin an d across tem perate
estuarine and coastal ecosystems.
M
ethods
Study Sites
Seven estuarine and coastal w ater bodies in the
N etherlands and U nited Kingdom w ere selected
(Figure 1; Table 1), all w ith extensive intertidal
areas and all experiencing a sem i-diurnal tidal re ­
gime. The D utch sites include th e W esterschelde
(site WES) and th e Ems-Dollard (EMS), two
eutrophic, turbid, m acrotidal estuaries, and the
Oosterschelde (OOS), an estuary th at is relatively
clear and saline due to construction of dams in the
upper estuary and a storm surge barrier at the
m outh, as w ell as the w estern W adden Sea (site
WAD), a eutrophic turbid m esotidal basin bordered
by a series of barrier islands. The three exposed and
turbid sites in east England (Devlin an d others
2008) include th e m acrotidal open coast m udflats
340
D. van der Wal and others
Time-Series of NDVI from MODIS 250
Satellite Images
EMS,E
HUM
WAD
WAS
United
Kingdom
North Sea
N etherlands
THA
600000
600000
F igure 1. Study sites: The W esterschelde (WES), Oosterschelde (OOS), W adden Sea (W AD), and E m s-Dollard (EMS) in the N etherlands, and the Tham es (THA),
W ash (W AS), and Hum ber (HUM) in the U nited K ing­
dom. Light gray areas indicate intertidal areas above
M ean Low W ater Spring (MLWS). C oordinate system is
in UTM 31/W G S84 (m).
and sandflats of the Thames estuary (THA), the
W ash (WAS), a m acrotidal em baym ent, and Spurn
Bight at th e m o u th of the m acrotidal H um ber
estuary (HUM). The seven sites are not n u trien tlim ited for m icrophytobenthos (Nienhuis 1992;
Aldridge an d Trim m er 2009).
The sedim ents are dom inated by benthic rnicroalgae, in particular diatom s (Admiraal and Peletier
1980; Heip and others 1995). B enthic m acroalgae
(for example, Enteromorpha sp. an d Ulva sp.) occur
in all sites, m ainly attached to suitable substrate
such as em pty shells or shellfish beds. N ienhuis
(1992) presented a carbon budget of prim ary p ro ­
ducers and estim ated th at benthic m acroalgae
contribute less th a n 10% of the benthic algae in site
OOS, less th a n 5% in site WAD and even less in
sites WES an d EMS. M acroalgal cover in the British
sites is com parably low (for exam ple, Aldridge and
Trim m er 2009).
T able 1.
A total of 2,179 daily com posite images (Septem ber
2002-2008) from Aqua MODIS (EOS PM) w ere
distributed by th e Land Processes Distributed Active
Archive C enter (LP DAAC), located at the USGS
Earth Resources O bservation and Science (EROS)
C enter (http://w w w .lpdaac.usgs.gov). The Aqua
MODIS Surface Reflectance Daily L2G Global 250-rn
SIN grid product (MYD09GQ.005) contains surface
reflectance w ith 250-rn resolution in a red ban d R r
(620-670 nrn, central w avelength 645 nrn) and a
near-infrared band R mR (841-876 nrn, central
w avelength 859 nm ). Equatorial crossing tim e is
13:30 for Aqua, b u t the daily products contain the
best observations over a 24-h period (in practice
acquired betw een 9:35 and 14:05 UTC) w ith respect
to solar zenith angle, view angle and cloud cover.
The data are atm ospherically corrected using aerosol
and w ater vapor inform ation derived from MODIS
data and taking into account th e directional p ro p ­
erties of th e observed surface, and corrected for
adjacency effects. Images (reprojected to UTM/
WGS84) w ere used in conjunction w ith the associ­
ated MOD09GA quality inform ation products. Pixels
w ith highest quality reflectance data (for example,
solar zenith angle < 85°) w ere selected using the
500-rn resolution quality assessm ent (QC_500), and
pixels w ith clouds/cirrus, cloud shadow, pixels
adjacent to clouds, and pixels w ith fire, snow or ice
w ere flagged using the 1-krn resolution quality
assessm ent (state_l 1cm). For pixels th a t w ere not
flagged, NDVI was calculated in a Geographical
Inform ation System (ESRI ArcGIS), provided R r was
greater th a n 0% an d R^or was greater th a n 2% .
A lthough w e used only high-quality surface
reflectance data, the NDVI will still contain some
Site Characteristics
Site
W ES
OOS
WAD
EM S
THA
W AS
H UM
Longitude (°E)
Latitude (°N)
Intertidal area (kin2)
Spring tidal range (nr)
M ud (%), intertidal only
K d (n U 1)
Secchi depth (nr)
Salinity
DIN (mg/1)
P (mg/1)
Si (mg/1)
3° 50'
51°20'
70
4.4
13
3.15
0.44
23.4
1.562
0.092
1.114
4°00'
51°35'
50
3.1
5
0.90
1.60
31.2
0.379
0.035
0 .3 7 4
5°40'
53°20'
990
2.0
17
2.27
0.69
27.7
0.383
0 .0 2 9
0.665
TOO'
53°25'
180
2.6
40
4.17
0.56
21.9
1.366
0.071
1.768
1°00'
51°35'
120
3.9
0°15'
52°55'
170
4.9
0°05'
53°35'
30
4.2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
See ‘ ‘Materials and methods ' ' for sourees. Water quality variables are averaged over September 2002-2008.
Synchrony in Intertidal B enthic Algal Biomass
noise. U nderestim ation of NDVI can be due to
residual cloud cover an d residual atm ospheric
contam ination; bo th effects increase w ith optical
p ath length (Huete and others 2002). O verestim a­
tion of NDVI at large solar or view angles due to
anisotropic effects such as shadow -casting and
m ultiple scattering, as reported for m any vegeta­
tion types (Van Leeuw en an d others 1999; Pettorelli and others 2005) can be neglected in our study,
as th e surface of an intertidal flat can be assum ed
Lam bertian.
Supratidal areas w ere excluded using a land/
w ater m ask w ith a 500-rn buffer. Seagrass (Zostera
sp.) m eadow s (notably in sites EMS, WAD, and
THA) and p erennial saltm arshes (for example,
Spartina sp.) w ere excluded based on vegetation
m aps and aerial photographs. Low -density stands
of annuals (for exam ple, Salicornia sp.) w ere not
excluded, and no distinction betw een benthic
m acroalgae an d benthic m icroalgae could be m ade.
NDVI values (with NDVI > 0 to exclude w ater)
w ere bin n ed per m o n th for each pixel and subse­
quently averaged per site.
Ground-Tru thing
G round-truthing the m onthly averaged NDVI of 250rn satellite image pixels at the appropriate spatiotem poral scale requires an extensive in situ data set. In
situ Chl-fl data from the period 2002-2005 (Ri­
jksw aterstaat 2006) w ere available for groun d-tru­
thing at site WES only. Every 1-3 m onths, the top
1 cm of the sedim ent was sampled (4 cm2 surface) at
approxim ately 90 well distributed (but fixed) in ter­
tidal stations (n = 2439 samples in total), and Ch\-a
content (in pg/g) of the samples was determ ined
using HPLC techniques. For each campaign, satellitederived NDVI depended significantly on in situ Ch\-a
(for exam ple R2 = 0.47, SE = 0.0297, F 1 A 5 = 39.512,
P < 0.0001, for M ay 2005). M onthly site-averaged
NDVI also depended significantly on m onthly
site-averaged Chl-a (NDVI = 0.00450 x Chl-a +
0.0285, R 2 = 0.52, SE = 0.0149, F L 2 2 = 24.143, P <
0.0001 ). M oreover, m onthly site-averaged NDVI and
Chl-fl exhibited a similar seasonal variation, w ith a
m axim um in June. Thus, spatial patterns and tem ­
poral (interannual and seasonal) trends in Chl-a
could be significantly represented by NDVI.
Time-Series of Meteorological and Water
Quality Variables
M eteorological tinre-series included m on th ly aver­
ages of m ean and m axim um hourly w ind veloci­
ties, m axim um w ind gust, m ean, m inim um and
341
m axim um daily air tem perature, global radiation,
sun duration and surface air pressure an d days of
air frost (Royal N etherlands Institute of M eteorol­
ogy, h ttp ://w w w .k n rn i.n l) for sites WES an d OOS
(w eather station Vlissingen) and sites WAD and
EMS (station De Kooij, Den Helder), an d m o n th ly
averages of m ean, m inim um and m axim um daily
air tem perature, sun duration and days of air frost
(UK M et Office, http://w w w .rnetoffice.gov) for site
THA (Met Office district East Anglia) and sites WAS
and HUM (Met Office district East and N ortheast
England).
M onthly averages of significant w ave height Hs
in the N orth Sea (Rijkswaterstaat W aterbase,
http ://w w w .w aterb ase.n l) w ere used for sites WES,
OOS, and THA (North Sea station Euro Platform)
and sites WAD, EMS, WAS, and HUM (North Sea
station K13).
In situ w ater quality tinre-series (Rijkswaterstaat
W aterbase, http ://w w w .w aterb ase.n l) of th e D utch
w aters included Secchi depth, light extinction
coefficient (Kd), nu trien ts [dissolved inorganic
nitrogen (DIN), orthophosphate (P) and silicate
(Si)], and salinity in the surface w ater at approxi­
m ately four fixed m onitoring stations per site,
averaged per m onth, per site.
Emersion Duration and Sedim ent Type
Em ersion duration was approxim ated for each
pixel by the nu m b er of unclouded images in w hich
the pixel is exposed (that is, has an NDVI > 0) as a
percentage of th e total nu m b er of unclouded im a­
ges. R ecent elevation data (available for all sites
except WAS) w ere com bined w ith inform ation on
tidal levels to qualitatively validate th e retrieval of
em ersion duration from th e satellite images.
S edim ent data w ere available for the D utch sites
(for exam ple, R ijksw aterstaat Sedim entatlas, h ttp ://
w w w .w addenzee.nl), com prising surface samples
collected in the period 1990-2008 in sites WAD and
EMS (7502 samples in total), site WES (1,418
samples) and site OOS (1025 samples). M ud (par­
ticles < 63 pm) percentages w ere determ ined from
laser particle sizer analysis. M ud percentages at the
sample points w ere interpolated in a GIS by an
inverse distance pow er 2 algorithm , and a distinc­
tion was m ade betw een m uddy (that is, >20%
m ud) and sandy ( < 2 0 % m ud) sedim ents. For the
British sites, a broad distinction betw een m u d and
sand was taken from recent Landranger O rdnance
Survey m aps (scale 1:50000) an d used in a quali­
tative w ay only.
342
D. van der Wal and others
Data Analysis
0.20
ANO VA w as carried ou t to test w h e th e r m onthly
averaged NDVI depended on th e categorical predic­
tors Year, M onth, Site and the first-order interaction
term s (each year starting in Septem ber to guarantee
a com plete ANO VA design). As ANO VA requires the
residuals of th e factor levels to be independent,
m odel residuals of the m on th ly tinre-series w ere
checked for autocorrelation at each site; no signifi­
cant (tem poral) autocorrelation was found.
Norm al (that is, average) m on th ly values of
NDVI, m eteorological and w ater quality variables
w ere calculated over th e 6-year period. Tinre-series
of m o n th ly anom alies of all variables w ere defined
as th e departure front these norm al m on th ly val­
ues. The im pact of m eteorology and w ater quality
on NDVI was tested by Pearson product m om ent
correlations. Synchrony was dem onstrated by cor­
relations of (anom alies of) NDVI betw een sites.
The im pact of sedim ent type and em ersion
duration on NDVI was quantified using ANO VA
(effects site, sedim ent type (sand/nrud), em ersion
(in 5 classes), and sedim ent x em ersion, on NDVI
averaged per pixel over the full period).
A 5% significance level was applied to all statis­
tical tests.
0.15
□
0.10
0.05
WES
OOS
0.00
2002
2003
2004
2005
2006
2007
2008
2009
0.20
0.15
D 0.10
0.05
WAD
EMS
0.00
2002
2003
2004
2005
2006
2007
2008
2009
2003
2004
2005
2006
2007
2008
2009
0 .2 0 r
0.15 -
S
0.10
z
-
0.05 0.00
—
2002
Year
Figure 2. Tim e-series of m on th ly m ea n NDVI for the
sev en sites.
R esults
Seasonal and Year-to-Year Variation
in M icrophytobenthos
NDVI tinre-series show a striking seasonal cycle
(Figure 2). This seasonal variation is significant for
all sites (one-w ay ANOVA per site). NDVI is lowest
in December. NDVI peaks in June, b u t in sites
WAD, OOS, and HUM m axim um NDVI occurs in
Septem ber and in site EMS in April. The blooms
also differ in m agnitude in each site (Figure 2).
ANOVA confirms differences in NDVI betw een
m onth (A 1 .3 3 0 = 59.51, P < 0.001), betw een sites
(A, 3 3 0 = 100.88, P < 0.001), and betw een m onth x
sites (F66 ,3 3 0 = 3.17, P < 0.001); but not betw een
years (A.îîo = 0.99, P = 0.424) and not betw een
m onth x year
(F5 5 .H 0 = 1-34,
P = 0.065)
and
year x sites (Fi 0 i i 0 = 0.91, P = 0.600). A post hoc
Tukey HSD test reveals that NDVI is lowest in sites
WAD, THA, and WES, interm ediate in EMS and WAS
and highest in sites HUM and OOS.
Synchrony in M icrophytobenthos
The strongest correlations in m onthly NDVI are
found betw een sites WAD an d EMS (r = 0.79,
n = 72, P < 0.001), b u t cross-correlations are also
significant b etw een o ther sites, confirm ing seasonal
synchrony. The exception is NDVI in site THA,
significantly correlating w ith NDVI in site OOS and
WES only.
The m onthly anomalies of NDVI also show sig­
nificant cross-correlation betw een all D utch sites,
dem onstrating in teran n u al synchrony, except b e­
tw een WAD an d OOS (Table 2). The strongest
correlations are betw een WAD an d EMS, and b e­
tw een OOS and WES; both sets of sites are situated
closest to each other. There is no cross-correlation
betw een the D utch an d British sites (except for a
negative correlation of THA w ith WAD and EMS)
and there is no significant cross-correlation b e­
tw een th e British sites (Table 2).
Correlations of th e anom alies are generally
stronger in th e w in ter half-year (O ctober-M arch)
th a n in the sum m er half-year (April-Septenrber).
For example, r = 0.74, n = 36, P < 0.001 for the
correlation betw een NDVI anom alies in site WAD
and EMS in w inter. M oreover, correlation betw een
NDVI anom alies in site WAS and HUM is signifi­
cant in w inter ( r = 0 . 5 0 , « = 36, P = 0.002),
pointing to year-to-year synchrony betw een the
Synchrony in Intertidal B enthic Algal Biomass
T able 2.
343
Year-to-Year Synchrony in NDVI: Correlations of M onthly NDVI Anom alies B etw een Sites
Site
W ES
WES (W esterschelde)
OOS (O osterschelde)
WAD (W adden Sea)
EMS (Ems-Dollard)
THA (Thames)
WAS (Wash)
HUM (Humber)
0.47*
0.32*
0.31*
- 0 .1 9
0.03
0.03
W AD
OOS
0.47*
EM S
0.32*
0.22
0.22
0.30*
-0 .0 7
0 .0 4
0.06
0.58*
-0 .3 1 *
0.13
0.02
0.31*
0.30*
0.58*
- 0 .2 5 *
- 0 .0 7
0.01
THA
W AS
H UM
- 0 .1 9
- 0 .0 7
- 0 .3 1 *
- 0 .2 5 *
0.03
0.0 4
0.13
- 0 .0 7
-0 .1 4
0.03
0.06
0.02
0.01
- 0 .1 1
0.12
- 0 .1 4
- 0 .1 1
0.12
Significant correlations (P < 0.05) are marked (*).
tw o m ost n o rth ern British sites. In contrast, corre­
lations betw een sites WAD an d THA and betw een
sites EMS an d THA are n eith er significant in the
w in ter half-year nor sum m er half-year. In the
sum m er half-year, th e strongest correlations of
NDVI anom alies are found betw een sites WAD and
EMS (r = 0.47, n = 36, P = 0.004) and betw een
sites WES and OOS (r = 0.42, n = 36, P = 0.010).
Im pact of Meteorology and W ater Quality
on M icrophytobenthos
Obviously, tinre-series of m onthly NDVI are sig­
nificantly correlated w ith variables w ith a seasonal
com ponent, including a positive correlation w ith
tem perature, global radiation, sun hours and
salinity and Secchi depth and a negative correlation
w ith w ind velocity, significant w ave height, days of
air frost, Kd, an d some nutrients. NDVI has the
strongest correlation w ith tem perature (for exam ­
ple, r = 0.79, n = 72, P < 0.001 in site WES) and
global radiation (for example, r = 0.73, n = 72,
P < 0.001 in site WES).
In contrast, m onthly anomalies of both NDVI and
m eteorological an d w ater quality variables are
generally n o t w ell correlated (Tables 3, 4), sug­
gesting th at year-to-year variation in benthic algae
is n o t w ell explained by these exogenous factors.
However, anom alies of NDVI and w ind velocity
(and w ave height) are negatively correlated in a
n u m b er of sites, particularly in the upper intertidal
zone; this effect is strongest in site EMS. Most
D utch sites also show a significant positive corre­
lation betw een anom alies of NDVI and surface air
pressure. Anom alies of NDVI and Secchi depth are
positively correlated for site WAD. In site WES,
tem perature best explains year-to-year variation in
NDVI (but m axim um w ind gust is also a significant
factor). In th e British sites WAS and HUM, (m ini­
m um ) tem perature and days of air frost m ost sig­
T able 3. Im pact of Exogenous Factors on Year-to-Year Variation in NDVI in the D utch Sites: Correlation of
M onthly Anom alies of NDVI an d Anom alies of M eteorology and W ater Quality
V ariable
W ind velocity, m ean (nr/s)
W ind velocity, m ax (nr/s)
W ind velocity, m ax gust (nr/s)
Global radiation (J/cnr2)
Sun (hours)
Temperature, m ean (°C)
Temperature, nrin (°C)
Temperature, m ax (°C)
Air frost (days)
Surface air pressure, m ean (hPa)
K d (n U 1)
Secchi depth (nr)
Significant w a v e h eigh t N orth Sea (nr)
W ES
W AD
OOS
EM S
A ll
H igh
A ll
H igh
A ll
H igh
A ll
H igh
-0 .1 0
- 0 .1 4
- 0 .1 4
-0 .0 6
- 0 .1 1
0.22
0.24*
0.22
- 0 .0 2
0.06
- 0 .0 3
0.06
-0 .0 9
-0 .1 8
- 0 .2 3
-0 .2 4 *
0.16
0 .1 4
0.33*
0.30*
0.36*
- 0 .0 3
0.17
-0 .0 9
0 .1 4
- 0 .2 3
- 0 .2 0
- 0 .2 3
- 0 .2 5 *
- 0 .0 7
- 0 .0 1
0.05
0.08
0.07
0.07
0.22
- 0 .0 8
0.12
- 0 .1 9
- 0 .2 3 *
- 0 .2 7 *
- 0 .2 7 *
0.01
0.06
0.09
0.10
0.12
0.07
0.23*
- 0 .1 0
0.09
- 0 .2 1
- 0 .1 6
- 0 .1 5
- 0 .1 6
- 0 .2 0
- 0 .1 5
0.07
0.10
0.05
- 0 .1 3
0.06
- 0 .1 6
0.28*
- 0 .1 1
- 0 .2 1
- 0 .2 2
- 0 .2 3
- 0 .0 3
-0 .0 0
0.12
0.13
0.13
-0 .1 4
0.25*
0.03
0.26*
-0 .1 8
- 0 .2 6 *
- 0 .3 0 *
- 0 .3 3 *
0.09
0.09
0.12
0.11
0.16
- 0 .0 5
0.34*
0.02
0.03
- 0 .2 3 *
-0 3 2 -0 3 4 —0 3 7 0.05
0.11
0.08
0.05
0.13
- 0 .0 8
0 .3 4 ;
- 0 .0 2
- 0 .0 2
—0 .3 5 ;
Significant correlations (P < 0.05) are marked (*). Distinction between all areas, and upper intertidal only (high: emersion duration > 50%).
344
D. van der Wal and others
T able 4. Im p a c t o f E x o g e n o u s F a cto rs o n Y e a r -to -Y e a r V a r ia tio n in N D V I in th e B r itish Sites: C o r r e la tio n o f
M o n t h ly A n o m a lie s o f N D V I a n d A n o m a lie s o f M e te o r o lo g y
THA
V ariable
Sun (hours)
Temperature, m ean (°C)
Temperature, m in (°C)
Temperature, m ax (°C)
Air frost (days)
Significant w a v e h eigh t N orth Sea (m)
HUM
W AS
A ll
H igh
A ll
H igh
A ll
H igh
0.09
0.11
0.09
0.11
-0 .1 0
0.03
- 0 .0 3
0.11
0.07
0 .1 4
- 0 .0 1
0.11
- 0 .1 2
0.24*
0.31*
0.15
-0 .3 0 *
- 0 .1 4
-0 .0 6
0.37*
0.41*
0.28*
-0 .3 1 *
-0 .2 5 *
-0 .2 3
0.13
0.22
0.03
- 0 .2 6 *
0.05
- 0 .1 5
0.20
0.25*
0.12
-0 .2 4 *
- 0 .0 3
Significant correlations (P < 0.05) are marked (*). Distinction between all areas, and upper intertidal only (high: emersion duration > 50%).
nificantly explain in teran n u al variation in benthic
algae, although site WAS is also affected by
anom alies in w ind/w ave climate (Table 4).
The effect of w ind/w ave climate occurs
th ro u g h o u t th e year (that is, for all m onths), b u t is
strongest in th e w in ter half-year. For example,
correlation of anom alies of m ean w ind velocity and
NDVI is only significant in th e w inter half-year for
the highest shores in site WAD (r = —0.35, n = 36,
P = 0.038). The effect of tem perature is also sig­
nificant in b oth w in ter and sum m er, except for the
effect of m in im um tem perature in site HUM and
days of air frost in site WAS and HUM, w hich are
only significant in th e w inter half-year.
We did n o t obtain higher correlation coefficients
w h e n applying a tim e lag of one season betw een
anom alies of NDVI and exogenous factors. For
10km
r^
■* OOS
"' -
Im pact of Emersion Duration
and Sedim ent Type
on M icrophytobenthos
Highest values for NDVI are generally found in the
upper intertidal flats (Figures 3, 4). Extrem ely high
values of NDVI in the eastern part of site OOS are
associated w ith benthic m acroalgae and m ussel bed
cultures (Figure 3). NDVI increases w ith em ersion
20 km
------
v
t l
A iv
example, no significant correlation was found b e­
tw een average tem perature anom alies in w inter
and NDVI anom alies in the subsequent spring: a
h otter or colder w inter did no t lead to a change in
m icrophytobenthic biom ass in spring. However,
the available nu m b er of years is (too) lim ited for
such a tim e-lag analysis.
F igure 3. Spatial
variation in long-term
m ean NDVI.
-* ■
\
• -:/
S'
i
W-
EMS
/y
i
WAD ^
°
WES
.y* - -,
NDVI
<0.05
0.05-0.10
■ 10 .1 0 -0 .1 5
■ 0.15 - 0.2°
■ >0.20
WAS
mM
5 km
HUM
W - 5 km
2 km
Synchrony in Intertidal B enthic Algal Biomass
345
F igure 4. Spatial
variation in em ersion
percentage.
10 km
Emersion
<20
□
■
20-40
40-60
60-80
5 km
0.20
□ W ES
S OOS
■ WAD
a THA
□ WAS
□ HUM
EMS
0.15
Q 0.10
0.05
0.00
m
<20
IA
20-40
40-60
60-80
>80
Emersion (%)
F igure 5. R elationship b e tw een long-term m ean (±SE )
NDVI and em ersion duration for the sev en sites.
duration in all sites (Figure 5), and for a given
em ersion duration class NDVI increases w ith m ud
co ntent at m ost sites (Figure 6).
An ANOVA m odel was constructed to disentangle
the effects of em ersion duration and sedim ent type
on NDVI. M ean NDVI depends on site (^ 6 ,4 0 3 6 4 =
883, P < 0.001), an d on em ersion (^ 4 ,4 0 3 6 4 = 9315,
P < 0.001), sedim ent type (FÍA 0 í 6 4 = 2101, P <
0.001) an d th eir interaction (_F4,4 0 3 5 4 = 120, P <
0.001). The m odel explains 62% of the variation
(.R2 = 0.62, F 15 ,4 0 3 6 4 = 4389, P < 0.001).
NDVI is n o t only higher in m ud th a n in sand, bu t
the peak bloom is also m ore pronounced in m uddy
sedim ents, w ith a few exceptions (such as site
WAS) (Figure 7). Given similar sedim ent, NDVI
increases w ith em ersion duration, an d seasonal
variation is in some sites m ore pronounced in the
upper intertidal zone, b u t the effect is m inor w h en
com pared to th e effect of sedim ent type (Figure 7).
There is no clear evidence for a tem poral shift in
bloom w ith increasing em ersion duration or m ud
content.
Ecosystems w ith short m ean em ersion duration
(such as sites THA and WAD) generally have a low
long-term m ean NDVI, w hereas sites th at are ex ­
posed longer (such as HUM) have higher long-term
m ean NDVI values (Figure 8A). The relationship
betw een em ersion duration and NDVI is significant
in sand (Figure 8C), b u t no t in m u d (Figure 8B).
Note th at a large variation in m ud content m ay be
present w ithin the sedim ent category 'm u d '. M ud
content in site EMS for exam ple is m uch larger
th a n in the o ther D utch sites (Table 1). Regression
equations of NDVI and em ersion duration are not
significantly different for m uddy and sandy sedi­
m ents (tested w ith a hom ogeneity of slopes m odel,
and subsequent test for offset differences), except
w h e n a paired test is carried out (slope higher in
m ud th a n in sand). Corrected for em ersion d u ra­
tion, sedim ent type and their interaction using
ANOVA on a per pixel basis, residual NDVI is p o ­
sitive for sites OOS and EMS (that is, NDVI is higher
th a n can be expected based on sedim ent type and
em ersion duration) and negative for sites WES and
THA.
D is c u s s io n
Satellite-Derived NDVI as a Proxy
for M icrophytobenthic Biomass
NDVI is used as a proxy for benthic Chl-a and in
particular for intertidal m icrophytobenthic biomass.
G round-truthing in th e W esterschelde show ed a
346
D. van der Wal and others
0.20
OOS
0.15
0.15
O 0.10
Q 0.10
0.05
0.05
0.00
0
F igure 6. Long-term
m ean (±SE ) NDVI as a
fu n ction of m ud content
and em ersion for the
D utch sites.
0.20
WES
20
40
60
Mud (% )
80
100
0.00
20
40
60
80
100
Mud (%)
0.20
0.20
WAD
Emersion (%)
< 20 %
EMS
0.15
0.15
Q 0.10
^
0.10
z
- a - 20-40%
0.05
0.05
60-80%
0.00
0
20
40
60
80
100
0.00
40-60%
>80%
20
0.20
0.20
WES
0.15
0.15
S 0.10
~z.
0.05
0.05
0.00
0.00
0.20
EM:
0.15
S
0.05
0.05
0.00
0.00
0.20
THA
0.15
a
WAD
0.15
0.10
0.20
WAS
0.15
0.10
0.05
0.05
0.00
ó
V V-% % % < ' f
a 0.10
0.00
Si
Q 0.1 0
cf%<ƒ"
Month
Month
—♦ — Mud, high
—■—Sand, high
60
80
100
Mud (%)
Mud (%)
0.20
40
— Mud, low
—Sand, low
F igure 7. Seasonal variation in norm al m on th ly (±SE )
NDVI depending o n sed im en t type (m ud/sand) and
em ersion (low: < 5 0 % em ersion, high: >50% em ersion).
significant coefficient of determ ination of about
50% betw een NDVI an d Chl-rz, both in space and in
time. Correlations betw een in situ Chl-rz and NDVI
derived from in situ or airborne sensors reported in
the literature are com parably low (for example,
M urphy an d others 2005; Van der Wal an d others
2008) or slightly higher (for example, Krom kam p
and others 2006). A lthough there m ay be m ore
effective spectral indices to retrieve Chl-rz (for
example, M urphy and others 2005; Serôdio and
others 2009), perform ance of NDVI is found to be
consistent w ith in and across (tem perate) estuaries
(Krom kam p and others 2006).
The m oderate correlation betw een NDVI and
Chl-fl in th e sedim ent can be partly attributed to a
m ism atch in spatial support. In our study, each
ground sample represents 4 cm 2 only, w hereas
each satellite pixel takes into account spatial vari­
ation in benthic algal biomass w ith in an area of 250
by 250 nr. M oreover, a nu m b er of factors m ay
confound the satellite signal. At the subpixel scale,
nrixels w ith surface w ater will inevitably reduce
NDVI. Low -density Salicornia stands will increase
the NDVI in sum m er in a lim ited nu m b er of high
shore places. In addition, shellfish beds an d m ussel
cultivation plots m ay yield an increased NDVI lo ­
cally; for exam ple, m ussel beds cover less th a n 1%
of th e intertidal area in site WAD (Beukem a and
Delcker 2007). B enthic algae consist of both nricrophytobenthos and nracroalgae. However, even
in the site w ith the highest abundance of nracroalgae, th at is site OOS, nracroalgae contribute less
th a n 10% an d nricrophytobenthos m ore th a n 90%
to the an n u a l prim ary production of benthic algae
(Nienhuis 1992). Thus, bare sedim ent and nricrophytobenthic bionrass w ith varying bionrass dom i­
nate all sites an d NDVI is generally proportional to
m icrophytobenthic bionrass.
The satellite sensor can only detect Chl-rz at the
(near-(surface of th e sedim ent, an d therefore
Synchrony in Intertidal B enthic Algal Biomass
0 .1 5
0 .1 5
(a )
^OOS
0 .1 0
>
o
0 .1 0
* E M S /* H U M
.^A A/AS
*W ES
•n/VAD
0 .0 5 ■ THA
y = 0.00113X+ 0.0402
R2 = 0.56, P=0.0518
n=7, S E = 0.0148
-----------1
---------1-----------1-----------1
0 .0 0
20
40
60
Emersion i %)
80
>
o
2
0 .1 5
(k )
hum
TOOS
*
EMSI ? I
__
*
V ^ -^ W A S
W A D ^ ^ ^ M WES
Mud:
y = 0.000841 x + 0.0596
R2 = 0.44, P =0.1030
n=7, S E =0.0194
0 .0 0 ---------- 1----------- 1-----------1-----------1
40
EMS
WAD.
2
# THA
0 .0 5
20
60
Emersion (
underestim ates the total standing stock of m icro­
p hytobenthos as sam pled using cores. Light is m ore
effectively absorbed in m uddy sedim ents th a n in
sand (Kühl and Jorgensen 1994), w hereas sedi­
m en t m ixing is orders of m agnitude higher in sand
(M iddelburg and others 2000), Chl-rz concentration
generally has a steep vertical gradient in m ud and is
relatively constant w ith depth in sand. A special
feature of epipelic diatom s is their ability to m igrate
w ith in th e (mostly m uddy) sedim ent w ith the tidal
cycle (time after em ersion) and daylight (solar ze­
n ith angle) (Pinckney and Zingm ark 1991; Jesus
and others 2009). Vertical m igration of epipelic
diatom s to th e surface of the sedim ent causes an
increase in NDVI u n til about 1 h after exposure,
followed by a stabilization of NDVI during daytim e
em ersion (Palmer an d R ound 1967; Paterson and
others 1998). The satellite will capture this m ore or
less stable state in NDVI w ith regard to daylight, bu t
n o t necessarily w ith regard to em ersion time. Note
th at tidal phasing is site-dependent (Serôdio and
Catarino 1999). For exam ple, low w ater springs
occur consistently at daw n and dusk in the Thames
and Erns estuaries.
Seasonal Variation in M icrophytobenthos
We found a significant seasonal p attern in the
dynam ics of b enthic algae in all ecosystems studied.
Others have observed th at m icrophytobenthic
biomass can vary seasonally as a function of tem ­
p erature, irradiance, n u trie n t concentrations, and
resuspension, resulting in an increase in spring
an d /o r au tu m n , an d a decrease in sum m er gen er­
ally attributed to an increase in grazing by benthic
fauna (Admiraal and Peletier 1980; Colijn an d Dijlcema 1981; U nderw ood 1994; M acIntyre and
others 1996; B lanchard and others 1997), although
in some (tem perate) ecosystems no seasonal vari­
ation is found (for example, Brotas and others
1995; Thornton an d others 2002; M éléder and
others 2007). Santos an d others (1997) attributed
th e contradictive findings in seasonality to the fact
th at m any in situ studies are based on sampling
F igure 8. Long-term
m ean (±SE ) NDVI as a
fu n ction of m ean (±SE)
em ersion percentage, for
A entire sites, B m uddy
sedim ents only, and C
sandy sedim ents only.
(C)
0 .1 0
>
o
80
WES
Sand:
0 .0 5
y = 0.001308X+ 0.0302
R2 = 0.79, P =0.00762
n = 7, S E= 0.0113
0 .0 0
20
40
60
347
80
Emersion %)
schedules of once or twice a m onth, despite a large
daily and weelcly/fortnightly variation in m icro­
phytobenthos. Our m ethod, binning daily NDVI
values per m onth, allow ed an unequivocal study of
seasonal variations.
In the open coast intertidal flats of the Thames
(seasonal), variations in NDVI w ere small, sug­
gesting th at local factors could be m ore im portant
in regulating m icrophytobenthic biomass. Lack of
seasonality in m icrophytobenthic biom ass in the
nearby Colne estuary was attributed to severe
grazing pressure th ro u g h o u t the year (Thornton
and others 2002).
Spatial differences in the onset of th e m icro­
phytobenthic bloom have b een attributed to dif­
ferences in em ersion duration, an d hence, in turbid
waters, irradiance (Admiraal an d others 1984), b u t
this effect was no t found at the ecosystem scale in
our study. Spatial differences in the m agnitude and
tim ing of seasonality have also been attributed to
species com position of the diatom assemblages
(Admiraal and others 1984; U nderw ood 1994; Sah an an d others 2007). Sandy sedim ents, generally
found in m ore exposed areas, contain propo rtio n ­
ally m ore epipsam m ic algae (attached to the sub­
strate), w hereas m uddy sedim ents, generally
dom inating areas sheltered from current and w ave
action, contain m ore (mobile) epipelic algae
(Round 1971; Paterson an d H agerthey 2001), al­
th o u g h such assemblage differences also occur in
similar sedim ents (U nderw ood 1994). Epipelic
assemblages show higher seasonal variability in
diatom diversity and biom ass th a n epipsam m ic
assemblages (U nderw ood 1994; Hamels an d others
1998; Sahan an d others 2007). O ur study con­
firm ed th at seasonal variation for a given site was
largest in m ud, and seasonal variation was lim ited
in exposed sites such as those in the Thames and
the W ash. Note th at NDVI m ay be sensitive to the
background properties of th e sedim ent surface
(Huete and others 1985; Combe and others 2005,
b u t see M urphy an d others 2005), such as m ud
content, w hich m ay itself show seasonal variation,
w ith highest values typically found in early au-
348
D. van der Wal and others
tu rn n (Van der W al and H erm an 2007; Van der Wal
and others 2008). This effect was no t obvious from
our results. M ud content can also vary over the
years, lim iting the applicability of a single sedim ent
survey as used in our study, although broad p a t­
terns in sedim ent characteristics persist for years
(Van der W al an d H erm an 2007).
Year-to-Year Variation
in M icrophytobenthos
In terannual changes in benthic algal biomass w ere
small relative to their seasonal dynamics. Part of the
year-to-year variation operated on a large spatial
scale, in parallel in different ecosystems in the
Netherlands, dem onstrating regional synchrony. This
synchrony was m ainly driven by the w ind/w ave
climate: w inds/w aves had a negative effect on b en ­
thic algal biomass, w hich was m ost pronounced in
th e upper intertidal zone. The influence of w ind/
w ave climate on benthic algae biomass can thus be
substantial even in tide-dom inated estuaries such as
th e Oosterschelde and W esterschelde. Two m echa­
nisms m ay be responsible for the negative correlation
of w ind velocity and NDVI. First, increased in u n d a ­
tion due to w ave set-up w ould limit the effective
photoperiod (length of tim e available for photosyn­
thesis). Second, increased w ind velocities m ay induce
resuspension. Resuspension of benthic algae by tu r­
bulence and shear stress generated by waves and
currents and subsequent transport by currents can be
substantial in w ave-dom inated estuaries and lagoons
(Demers and others 1987; De Jonge and Van Beusekorn 1995), b u t also in very shallow w aters in tidedom inated estuaries (Baillie and W elsh 1980; Lucas
2003). Field observations show th at this can drasti­
cally reduce m icrophytobenthic biomass, generally
followed by a recovery period of one or m ore weeks
(C olijnandD ijkem a 1981; Adm iraal and others 1984;
Underw ood 1994; Blanchard and others 1997; De
B rouw er and others 2000).
A part from w ind/w ave climate, tem perature
anom alies in th e 6-year period studied w ere a sig­
nificant driver for anom alies in NDVI. Low tem p er­
atures (and days of air frost) accounted for regional
synchrony in algal biom ass in th e w inter half-year in
the tw o m ost n o rth ern (coldest) British sites. This is
in accordance w ith m ost previous studies th a t show a
positive effect of tem perature on algal grow th up to a
m axim um tem perature (Blanchard and others
1997), although other studies dem onstrate little ef­
fect of tem p erature (A dm iraalandP eletier 1980) o ra
m ore com plex regulation by tem perature, for
exam ple by influencing grazing pressure (Thompson
and others 2004).
In th e Thames, anom alies in m eteorology and
NDVI w ere n o t correlated, in accordance w ith the
lim ited seasonality in benthic algal biom ass found.
Spatial Variation in M icrophytobenthos
M icrophytobenthic biomass varies across spatial
scales. At very small spatial scales (centim eters),
patchiness of m icrophytobenthos is generally
attributed to interspecies interactions, notably
grazing (Saburova and others 1995; M urphy and
others 2008). Variability on interm ediate scales
(centim eters to tens of m eters) has been associ­
ated w ith th e interplay of sedim ent, m icrotopography, an d diatom s (Plante and others 1986;
Saburova and others 1995; B lanchard an d others
2000). The lim ited nu m b er of in situ studies car­
ried ou t at the w hole ecosystem scale dem o n ­
strated the presence of patterns in m icrophyto­
benthic biom ass at these large spatial scales (tens
of m eters to kilometers) (G uarini and others
1998). In a study in the Tagus estuary in Portugal,
Brotas and others (1995) show ed th at approxi­
m ately 62% of th e variability of m icrophytoben­
thic biomass at this scale could be explained by
sedim ent type an d elevation.
O ur study highlighted the large scale patterns in
m icrophytobenthic biom ass and show ed th at the
position in the tidal fram e (em ersion duration) and
sedim ent type largely explained their spatial vari­
ability w ithin ecosystems. Our study also dem o n ­
strated, for th e first time, th at these factors explain
spatial variability across ecosystems (thus, at a scale
of about 100-1000 km). A similar am o u n t of vari­
ance was explained w ith in and across the ecosys­
tems, and explained variance was com parable to
th at reported by Brotas and others (1995) for the
Tagus estuary. At still larger spatial scales, latitude
effects (for exam ple, tem perature gradients an d day
length) m ay becom e m ore im portant.
The increase in NDVI w ith increasing em ersion
duration m ay be associated w ith low er w ater con­
tents for sedim ents on th e higher shores, facilitat­
ing algal grow th (U nderw ood and Paterson 1993).
However, it also suggests th at light is a lim iting
factor for benthic algal grow th in all sites. Less light
a tten u atio n in the Oosterschelde could explain the
larger NDVI values w h e n corrected for em ersion
duration and sedim ent type, across the entire tidal
fram e (com bined w ith the higher proportion of
benthic nracroalgae). Note th at in such clear w a ­
ters, positive NDVI values m ay occur even if there
is overlying w ater, m aking the estim ates of benthic
algae biomass and th e em ersion m aps unreliable
here.
Synchrony in Intertidal B enthic Algal Biomass
Our study also confirm ed the results of in situ
studies, reporting low er biomasses of m icrophyto­
benthos in sand th a n in m u d (M acIntyre an d o th ­
ers 1996). This can be partly explained by a
difference in m icrophytobenthic species com posi­
tion and th eir life traits. In addition, sands tend to
be low er in n u trien ts an d resuspension losses and
grazing pressure m ay be higher (U nderw ood and
Paterson 1993; De Jong and De Jonge 1995;
U nderw ood and K rom kam p 1999; H erm an and
others 2001), all contributing to a low er m icro­
ph y to b en th ic standing stock.
C o n c l u s io n s
The satellite-derived tinre-series of NDVI have
provided inform ation on the spatiotem poral distri­
b u tio n of b enthic algae w ith an unrivalled spatial
extent. The results dem onstrate an in tera n n u al
regional synchrony (superim posed on a seasonal
cycle) in intertidal m icrophytobenthic biomass
across a n u m b er of ecosystems, m ainly driven by
th e w in d /w av e clim ate and to some extent by
tem perature, and reveal th at the long-term aver­
aged biom ass of benthic algae can be w ell explained
by em ersion percentage an d sedim ent type, both
w ith in and across ecosystems. The results im ply
th at effects on m icrophytobenthic biom ass and
prim ary production can be anticipated w h en the
position of th e m udflats and sandflats in the tidal
fram e is altered, for exam ple due to sea level rise. In
addition, increased storm iness will result in a largescale decrease of benthic algae biomass, either
th ro u g h resuspension or through a reduction of the
effective em ersion duration.
ACKNOWLEDGMENTS
This study is carried out as part of FOKUZ. W e grate­
fully acknowledge Dick de Jong (Rijkswaterstaat
Dienst Zeeland) for providing us w ith in situ data of
th e W esterschelde and Oosterschelde. F urther sedi­
m en t data of th e Oosterschelde w ere provided by the
M onitor Taskgroup of the N etherlands Institute of
Ecology. This is publication 4685 of the Netherlands
Institute of Ecology (NIOO-KNAW).
OPEN ACCESS
This article is distributed under the term s of the Cre­
ative Commons A ttribution Noncomm ercial License
w hich perm its any noncom m ercial use, distribution,
and reproduction in any m edium , provided the ori­
ginal author(s) and source are credited.
349
REFERENCES
A dm iraal W, Peletier H. 1980. Influence of seasonal variations of
tem p eratu re a n d light on the grow th rate of cultures and
n a tu ra l populations of intertidal diatom s. M ar Ecol Prog Ser
2:35-43.
A dm iraal W, Peletier H, B rouw er T. 1984. The seasonal succes­
sion patterns of diatom species on an intertidal m udflat: a n
experim ental analysis. Oikos 42:30M 0.
Aldridge JN, Trim m er M. 2009. M odelling the distribution and
g row th of 'problem ' green seaw eed in the M edw ay estuary,
UK. Hydrobiologia 629:107-22.
Baillie PW, W elsh BL. 1980. The effect of tidal resuspension on
th e distribution of intertidal epipelic algae in a n estuary.
E stuarine Coast M ar Sei 10:165-80.
Bates BC, Kundzew ics ZW, W u S, Palutikof JP (Eds). 2008.
Clim ate change a n d W ater. Technical Paper IPCC. Geneva:
IPCC Secretariat.
B ehrenfeld MJ, O 'M alley RT, Siegel DA, M cClain ChR, Sarm i­
ento JL, Feldm an GC, M illignan AJ, Falkowski PG, Letelier
RM, Boss ES. 2006. C lim ate-driven trends in contem porary
ocean productivity. N ature 444:752-5.
B eukem a JJ, Essink K. 1986. C om m on patterns in the flu ctu a ­
tions of m acrozoobenthis species living at different places on
tidal flats in th e W adden Sea. Hydrobiologia 142:199-207.
B eukem a JJ, D ekker R. 2007. Variability in a n n u a l recruitm ent
success as a determ inant of long-term a n d large-scale va ria ­
tio n in a n n u a l p roduction of intertidal W adden Sea m ussels
(Mytilus edulis). Helgol M ar Res 61:71-86.
B lanchard GF, G uarini J-M , Gros P. R ichard P. 1997. Seasonal
effect on the relationship betw een the photosynthetic capacity
of intertidal m icrophytobenthos a n d tem perature. J Phycol
33:723-8.
B lanchard GF, Paterson DM, Stal LJ, R ichard P. Galois R, H uet V,
Kelly J, H oneyw ill C. de B rouw er J, Dyer K, Christie M, Seguignes M. 2000. The effect of geom orphological structures on
poten tial biostabilisation by m icrophytobenthos o n intertidal
m udflats. C ont Shelf Res 20:1243-56.
Brotas V, Cabrita T, Portugal A, Serôdio J, Catarino F. 1995.
Spatio-tem poral distribution of th e m icrophytobenthic b io ­
m ass in intertidal flats of Tagus E stuary (Portugal). H ydrobi­
ologia 300-301:93-104.
C ahoon LB. 1999. The role of benthic m icroalgae in neritic
ecosystem s. O ceanogr M ar Biol A n n u Rev 37:47-86.
Colijn F, Dijkem a KS. 1981. Species com position of benthic
diatom s and distribution of chlorophyll a on a n intertidal flat
in the D utch W adden Sea. M ar Ecol Prog Ser 4:9-21.
Combe J-P, L auneau P. Carrère V, D espan D, M éléder V, Barillé
L, Sotin C. 2005. M apping m icrophytobenthos biom ass by
n o n -lin ear inversion of visible-infrared hyperspectral images.
R em ote Sens E nviron 98:371-87.
De B rouw er JFC, Bjelic S, De Deckere BMGT, Stal LJ. 2000.
Interplay betw een biology a n d sedim entology in a m udflat
(Biezelingse Ham, W esterschelde, The N etherlands). Cont
Shelf Res 20:1159-77.
De Jong DJ, De Jonge VN. 1995. Dynam ics a n d distribution of
m icrophytobenthic chlorophyll-a in th e W estern Scheldt
estuary (SW N etherlands). Hydrobiologia 311:21-30.
De Jonge VN, V an B eusekom JEB. 1995. W ind a n d tide induced
resuspension of sedim ent a n d m icrophytobenthos in th e Ems
estuary. Lim nol O ceanogr 40:766-78.
350
D. v an der W al and others
Dem ers S, T herriault J-C, B ourget B, B ah A. 1987. Resuspension
in th e shallow sublittoral zone of a m acrotidal estuarine
environm ent: w in d influence. Lim nol O ceanogr 32:327-39.
Devlin MJ, Barry J, Mills DK, G ow en RJ, Foden J, Sivyer D, Tett
P. 2008. Relationships b e tw ee n suspended particulate m a te ­
rial, light a tte n u a tio n a n d Secchi depth in UK w aters. E stuar
Coast Shelf Sei 79:429-39.
G attuso J-P, Gentili B, D uarte CM, Kleypas JA, M iddelburg JJ,
A ntoine D. 2006. Light availability in th e coastal ocean: im ­
pact on the distribution of benthic p h otosynthetic organisms
a n d their distribution to prim ary production. Biogeosciences
3:489-513.
G uarini J-M , B lanchard GF, B acher C, Gros P, Riera P, R ichard P,
G ouleau D, Galois R, Prou J, Sauriau PG. 1998. Dynam ics of
spatial patterns of m icrophytobenthic biomass: inferences
from a geostatistical analysis of tw o com prehensive surveys in
M arennes-O léron Bay (France). M ar Ecol Prog Ser 166:13141.
G uarini J-M , Sari N, M oritz C. 2008. M odelling th e dynam ics of
th e m icroalgal biom ass in sem i-enclosed shallow -w ater eco­
systems. Ecol M odel 211:267-78.
Hamels I, Sabbe K, M uylaart K, B arranguet C, Lucas C, H erm an
P, V yverm an W. 1998. O rganisation of m icrobenthic com ­
m unities in intertidal estuarine flats, a case study from the
M olenplaat (W esterschelde estuary, the N etherlands). E ur J
Protistol 34:308-20.
Harley CDG, Hughes AR, H ultgren KM, M iner BH, Sorte CJB,
T hornber CS, Rodriguez LF, T om anek L, W illiam s SL. 2006.
The impacts of clim ate change in coastal m arine systems. Ecol
Lett 9:228M 1.
Heip CHR, Goosen NK, H erm an PMJ, K rom kam p J, M iddelburg
JJ, Soetaert K. 1995. Production a n d consum ption of biolog­
ical particles in tem perate tidal estuaries. Oceanogr M ar Biol
A n n u Rev 33:1-149.
H erm an PMJ, M iddelburg JJ, V an de Koppel J, Heip CHR. 1999.
Ecology of estuarine m acrobenthos. Adv Ecol Res 29:195-240.
H erm an PMJ, M iddelburg JJ, Heip CHR. 2001. B enthic c o m m u ­
nity structure a n d sedim ent processes on a n intertidal flat: r e ­
sults from the BCOFLAT project. C ont Shelf Res 21:2055-71.
H olland AF, Zingm ark RG, D ean JM . 1974. Q uantitative evi­
dence concerning th e stabilization of sedim ents by m arine
b enthic diatom s. M ar Biol 27:191-649.
H uete A, D idnan K, M iura T, Rodriguez BP, Gao X, Ferreira LG.
2002. O verview of th e radiom etric a n d biophysical p e rfo r­
m ance of the MODIS vegetation indices. R em ote Sens E nvi­
ro n 83:195-213.
H uete AR, Jackson RD, Post DF. 1985. Spectral response of a
p lan t canopy w ith different soil backgrounds. R em ote Sens
E nviron 17:37-54.
Jesus B, Brotas V, Ribeiro L, M endes CR, C artaxana P, Paterson
DM. 2009. A daptations of m icrophytobenthos assem blages to
sedim ent type a n d tidal position. C ont Shelf Res 29:1624-34.
K rom kam p JC, M orris BP, Forster RM, Honeyw ill C, H agerthey
S, Paterson DM. 2006. R elationship of intertidal surface sedi­
m en t chlorophyll concentration to hyperspectral reflectance
a n d chlorophyll fluorescence. Estuaries Coasts 29:183-96.
K ühl M, Jorgensen BB. 1994. The light field of m icrobenthic
com m unities: radiance distribution a n d m iscroscale optics of
sandy coastal sedim ents. Lim nol Oceanogr 39:1368-98.
Liebhold A, Koenig WD, B jornstad ON. 2004. Spatial synchrony
in population dynam ics. A n n u Rev Ecol Bvol Syst 35:467-90.
Lucas C. 2003. Observations of resuspended diatom s in the
tu rb id tidal edge. J Sea Res 50:301-8.
M acIntyre HL, Geider RJ, M iller DC. 1996. M icrophytobenthos:
th e ecological role of th e 'secret garden' of u nvegetated
shallow w a ter m arine habitats. I. D istribution, abundance and
prim ary production. Estuaries 19:186-201.
M éléder V, Rincé Y, Barillé L, G audin P, Rosa P. 2007. Spatio­
tem poral changes in m icrophytobenthos assem blages in a
m acrotidal tidal flat (Borgneuf Bay, France). J Phycol
43:1177-90.
M iddelburg JJ, B arranguet C, Boschker HTS, H erm an PMJ,
M oens T, Heip CHR. 2000. The fate of intertidal m icrophyto­
b e nthos carbon: a n in-situ 13C labelling study. Lim nol Oce­
anogr 45:1224-34.
M iller DC, Geider RJ, M acIntyre HL. 1996. M icrophytobenthos:
th e ecological role of the "secret garden" of unvegetated,
shallow -w ater m arine habitats n. Role in sedim ent stability
a n d shallow -w ater food w ebs. Estuaries 19:202-12.
M urphy RJ, T olhurst TJ, C hapm an MG, U nderw ood AJ. 2005.
E stim ation of surface chlorophyll-a on a n em ersed m udflat
using field spectrom etry: accuracy of ratios a n d derivativebased approaches. In t J R em ote Sens 26:1835-59.
M urphy RJ, T olhurst TJ, C hapm an MG, U nderw ood AJ. 2008.
Spatial v ariation of chlorophyll on estuarine m udflats d e te r­
m ined by field-based rem ote sensing. M ar Ecol Prog Ser
365:45-55.
N ienhuis PH. 1992. E utrophication, w a ter m an agem ent a n d the
functioning of D utch estuaries a n d coastal lagoons. Estuaries
15:538-48.
Palm er JD, R ound FB. 1967. Persistent, vertical m igration
rhythm s in benthic m icroflora. VI. The tidal a n d diurnal n a t­
ure of the rh y th m in the diatom Hantzschia virgata. Biol Bull
132:44-55.
Paterson DM, H agerthey SB. 2001. M icrophytobenthos in c o n ­
trasting coastal ecosystems: biology a n d dynam ics. In: Reise K,
Ed. Ecological com parisons of sedim entary shores. Heidelberg:
Springer, p 105-21.
Paterson DM, W iltshire KH, Miles A, B lackburn J, Davidson I,
Yates MG, M cG rortym S, E astw ood JA. 1998. M icrobiological
m ediation of spectral reflectance from intertidal cohesive
sedim ents. Lim nol O ceanogr 43:1207-21.
Pettorelli N, Vik JO, M ysterud A, Gaillard J-M , T ucker CJ,
Stenseth NC. 2005. Using the satellite-derived NDVI to assess
ecological responses to e n vironm ental change. Trends Ecol
Bvol 20:503-10.
Pinckney J, Zingm ark RG. 1991. Effects of tidal stage a n d sun
angles o n intertidal benthic m icroalgal productivity. M ar Ecol
Prog Ser 76:81-9.
Pinckney J, Zingm ark RG. 1993. Biomass a n d p roduction of
b enthic m icroalgae com m unities in estuarine habitats. E stu­
aries 16:887-97.
Plante R, P lante-C uny M-R, Reys J-P. 1986. Photosynthetic
pigm ents of sandy sedim ents o n th e n o rth M ed iterranean
coast: th eir spatial distribution a n d its effect on sam pling
strategies. M ar Ecol Prog Ser 34:133M 1.
Rijksw aterstaat. 2006. M onitoring v a n de effecten v a n de v e r­
ruim ing 4 8 7 4 3 ' (MOVE). R eport RIKZ/2007.003.
R ound FB. 1971. B enthic m arine diatom s. Oceanogr M ar Biol
A n n u Rev 9:83-139.
Saburova MA, Polikarpov IG, B urkovsky TV. 1995. Spatial
structure of a n intertidal sandflat m icrophytobenthic com ­
Synchrony in Intertidal B enthic Algal Biomass
m u n ity as related to different spatial scales. M ar Ecol Prog Ser
129:229-39.
Sahan E, Sabbe K, Creach V, H ernandez-R aquet G, V yverm an
W, Stal LJ, M uyzer G. 2007. C om m unity structure a n d sea­
sonal dynam ics in diatom biofilm s a n d associated grazers in
intertidal m udflats. A quat M icrob Ecol 47:253-66.
351
Tucker CJ. 1979. Red a n d photographic infrared linear com bi­
nations for m onitoring vegetation. R em ote Sens E nviron
8:127-50.
U nderw ood GJC. 1994. Seasonal a n d spatial v ariation in epipelic
diatom assem blages in th e Severn estuary. D iatom Res 9 :45172.
Santos PJP, Castel J, Souzasantos LP. 1997. Spatial distribution
a n d dynam ics of m icrophytobenthos biom ass in the Gironde
estuary (France). O ceanol Acta 20:549-56.
U nderw ood GJC, Paterson DM. 1993. Seasonal changes in d ia­
to m biomass, sedim ent stability a n d biogenic stabilization in
th e Severn estuary. J M ar Biol Assoc UK 73:871-87.
Serôdio J, Catarino F. 1999. Fortnightly light a n d tem perature
variability in estuarine intertidal sedim ents a n d implications
for m icrophytobenthos prim ary productivity. A quat Ecol
33:235-41.
U nderw ood GJC, K rom kam p J. 1999. Prim ary p roduction by
ph y to p lan k to n a n d m icrophytobenthos in estuaries. Adv Ecol
Res 29:93-153.
Serôdio J, C artaxana P, Coelho S, Vieira S. 2009. Effects of
chlorophyll fluorescence o n th e estim ation of m icrophyto­
b e nthos biom ass using spectral reflectance indices. R em ote
Sens E nviron 113:1760-8.
Thom pson RC, N orton TA, Haw kins SJ. 2004. Physical stress and
biological control regulate the producer-consum er balance in
intertidal biofilms. Ecology 85:1372-82.
T h ornton DCO, Dong LF, U nderw ood GJC, Nedw ell DB. 2002.
Factors affecting m icrophytobenthic biomass, species com po­
sition a n d p roduction in th e Colne E stuary (UK). A quat
M icrob Ecol 27:285-300.
Van der W al D, H erm an PMJ. 2007. Regression-based synergy of
optical, shortw ave infrared a n d m icrow ave rem ote sensing for
m onitoring the grain-size of intertidal sedim ents. R em ote Sens
E nviron 111:89-106.
Van der W al D, H erm an PMJ, Forster RM, Ysebaert T, Rossi F,
K naeps B, Plancke YMG, Ides SJ. 2008. D istribution and
dynam ics of intertidal m acrobenthos predicted from rem ote
sensing: response to m icrophytobenthos a n d environm ent.
M ar Ecol Prog Ser 367:57-72.
Van L eeuw en JD, H uete AR, Laing TW. 1999. MODIS vegetation
index com positing approach: a prototype w ith AVHRR data.
R em ote Sens E nviron 69:264-80.