Content uploaded by Simon James Pittman
Author content
All content in this area was uploaded by Simon James Pittman on Oct 19, 2017
Content may be subject to copyright.
ORIGINAL ARTICLE
Long-term spatial dynamics in vegetated seascapes:
fragmentation and habitat loss in a human-impacted
subtropical lagoon
Rolando O. Santos
1
, Diego Lirman
1
& Simon J. Pittman
2,3
1 Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, FL, USA
2 Biogeography Branch, National Centers for Coastal Ocean Science, U.S. National Oceanic & Atmospheric Administration, Silver Spring, MD,
USA
3 Centre for Marine & Coastal Policy Research, Marine Institute, Plymouth University, Plymouth, UK
Keywords
Anthropogenic disturbances; GIS; habitat
fragmentation; historical aerial photography;
landscape ecology; seagrass; spatial pattern
metrics.
Correspondence
Rolando O. Santos, Rosenstiel School of
Marine and Atmospheric Science, University
of Miami, 4600 Rickenbacker Causeway,
Miami, FL 33149, USA.
E-mail: rsantos@rsmas.miami.edu
Accepted: 18 December 2014
doi: 10.1111/maec.12259
Abstract
Vegetated coastal seascapes exhibit dynamic spatial patterning, some of which
is directly linked to human coastal activities. Human activities (e.g. coastal
development) have modified freshwater flow to marine environments, resulting
in significant changes to submerged aquatic vegetation (SAV) communities.
Yet, very little is known about the spatially complex process of SAV habitat
loss and fragmentation that affects ecosystem function. Using habitat mapping
from aerial photography spanning 71 years (1938–2009) for Biscayne Bay
(Florida, USA), we quantify both SAV habitat loss and fragmentation using a
novel fragmentation index. To understand the influence of water management
practices on SAV seascapes, habitat loss and fragmentation were compared
between nearshore and offshore locations, as well as locations adjacent to and
distant from canals that transport freshwater into the marine environment.
Habitat loss and fragmentation were significantly higher along the shoreline
compared with offshore seascapes. Nearshore habitats experienced a net loss of
3.31% of the total SAV mapped (2.57 km
2
) over the time series. While areas
adjacent to canals had significantly higher SAV cover, they still experienced
wide fluctuations in cover and fragmentation over time. All sites exhibited
higher fragmentation in 2009 compared with 1938, with four sites exhibiting
high fragmentation levels between the 1990s and 2000s. We demonstrate that
freshwater inputs into coastal bays modify the amount of SAV and the frag-
mentation dynamics of SAV habitats. Spatial changes are greater close to shore
and canals, indicating that these coastal developments have transformative
impacts on vegetated habitats, with undetermined consequences for the provi-
sioning of ecosystem goods and services.
Introduction
Submerged aquatic vegetation (SAV) assemblages com-
posed of seagrasses and macroalgae are keystone compo-
nents of coastal benthic ecosystems, where they provide
important ecological, societal and economic services
(Orth et al. 2006; Barbier et al. 2011). SAV assemblages
are known to significantly contribute to carbon sequestra-
tion (Duarte et al. 2005), provide food and shelter to
economically and ecologically important species (Gilland-
ers 2007) and are essential habitat for endangered marine
species such as the green turtle (Chelonia mydas), West
Indian manatee (Trichechus manatus) and dugongs (Du-
gon dugon) (Hemminga & Duarte 2000). SAV habitats
also influence water quality by the uptake of nutrients
and the deposition and binding of sediments (Koch et al.
2007a), and facilitate organic, inorganic and trophic
transfers to adjacent habitats such as salt marshes,
Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH 1
Marine Ecology. ISSN 0173-9565
mangrovestands and coral reefs (Davis et al. 2009;
Hyndes et al. 2014).Thus, SAV habitats are essential for
the resilience of marine and estuarine ecosystems, and the
growth, development and survival of juvenile and adult
populations of key marine taxa (e.g. herbivorous fishes,
apex predators; Gillanders et al. 2003; Mumby et al.
2004; Olds et al. 2012). In this study, we used a unique
data set of aerial images from 1938 to 2009 to evaluate
historical patterns of change in SAV communities of
Biscayne Bay, Florida, USA, and determine whether water
management actions have influenced the extent (cover)
and spatial pattern (fragmentation) of SAV seascapes.
Like other coastal ecosystems (i.e. coral reefs, man-
groves), SAV communities have experienced significant
global declines in the recent past as a consequence of
anthropogenic disturbances (Waycott et al. 2009). Sea-
grass habitats have disappeared worldwide at a rate of
110 km
2
year
1
between 1980 and 2006 (Waycott et al.
2009), with an estimated 14% of seagrass species experi-
encing an elevated risk of extinction (Short et al. 2011).
Declines in SAV abundance worldwide are mostly caused
by water quality degradation, especially due to nutrient
loading and sediment run-off (Fourqurean & Robblee
1999; Duarte 2002; Santos et al. 2011). Other distur-
bances such as thermal pollution, dredging, vessel
grounding and boat propeller scarring have also been
associated with significant SAV losses (Orth et al. 2006).
Concepts and analytical techniques developed in terres-
trial landscape ecology provide a framework that can be
readily applied to assess broad-scale SAV patterns and
disturbance impacts (Wedding et al. 2011). A benthic
seascape, analogous to a terrestrial landscape, is applied
here to describe a spatially heterogeneous area of the sea
floor composed of various discrete habitat patches (Grob-
er-Dunsmore et al. 2009; Pittman et al. 2011). The degra-
dation and transformation of SAV seascapes may be
characterized by two main aspects: habitat loss and frag-
mentation. The former is the reduction in the amount or
the proportion of habitat occupied by SAV within the
seascape, while the latter refers to the breaking apart of
large, continuous patches into smaller units. Both habitat
loss and fragmentation change the spatial arrangement of
the seascape. A growing body of evidence, mostly from
terrestrial landscape ecology, has demonstrated distinct
ecological impacts of habitat loss and fragmentation
(Fahrig 2003; McGarigal et al. 2005; Liao et al. 2013).
The terrestrial ecology literature suggests that the effects
of habitat fragmentation are generally much weaker than
the effects of habitat loss (Fahrig 2003); however, critical
thresholds for organisms and ecosystem function will vary
depending on the system (Andr
en 1994; Pardini et al.
2010). For example, negative effects of habitat loss and
fragmentation on forest patches are dependent on the
dispersal strategies, extent intra-specific competition and
growth rates of organisms (With & Crist 1995; Bonte
et al. 2010; Liao et al. 2013), which in turn have the
potential to change the species composition, successional
stages and local extinction rates of plant species (Lau-
rance et al. 2006; P€
utz et al. 2011). While planning for
conservation, these two processes of habitat degradation
should be quantified separately as they can be managed
independently with different restoration approaches (e.g.
conserving large areas versus many small areas), and their
effects on populations and biodiversity may differ in
magnitude and direction (Ewers & Didham 2006; Smith
et al. 2009).
Habitat loss and fragmentation studies in marine eco-
systems are relatively rare (Robbins & Bell 1994; Bell
et al. 2007; Bostr€
om et al. 2011). Both of these types of
degradation can have either independent or interactive
effects on the resilience and persistence of SAV and influ-
ence faunal connectivity among habitats, species diversity
and ecological interactions (e.g. competition, predation,
foraging behavior; Irlandi & Crawford 1997; Hovel & Lip-
cius 2001; Fahrig 2003; Bostr€
om et al. 2011). Until
recently, historical mapping and monitoring studies have
concentrated on changes in the areal extent of SAV (i.e.
habitat loss or recovery), largely ignoring spatial configu-
ration and fragmentation and how these may influence
seagrass ecology (Cunha & Santos 2009; Santos et al.
2011; Cuttriss et al. 2013). Here, using a historical record
of >70 years (1938–2009), we quantified, independently,
habitat loss and fragmentation to provide a better under-
standing of how SAV seascapes respond to anthropogenic
disturbances such as the modification of freshwater
deliveries.
In Biscayne Bay, Florida, SAV seascapes are influenced
by water-management practices that regulate freshwater
discharges into littoral areas. Over the last 50 years, the
hydrology of the South Florida watershed has been modi-
fied by the construction of a massive water management
system that has altered the quantity, quality and delivery
method of fresh water into the coastal bays (Browder &
Ogden 1999). The modifications to the watershed hydrol-
ogy have resulted in significantly lower total freshwater
delivery, a reduction in the proportion of fresh water
delivered through overland sheet flow and groundwater
sources, and a switch from historical diffuse deliveries to
pulsed, point-sources of discharge through man-made
canals (Wang et al. 2003; Lirman et al. 2008; Stalker et al.
2009). These hydrological modifications have affected the
abundance and composition of Biscayne Bay’s nearshore
SAV communities. For example, previous research has
shown that the abundance and plant species composition
of SAV are directly related to salinity patterns, with areas
of low and highly variable salinity (i.e. adjacent to canals
2Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH
Spatial-temporal dynamics in SAV seascapes Santos, Lirman & Pittman
that discharge fresh water) exhibiting lower SAV
abundance and high variability in per cent cover (Lirman
et al. 2008). Applying a landscape ecology approach, San-
tos et al. (2011), observed distinct SAV seascape struc-
tures related to salinity regimes using data from only
1 year and one season. As a significant expansion of this
previous research, a temporal change-analysis of SAV sea-
scape characteristics was conducted over a 70-year period
to: (i) develop novel methods to assess the spatio-tempo-
ral trends of SAV habitat fragmentation independently of
habitat loss; and (ii) use SAV seascape dynamics in Bis-
cayne Bay as a case study to understand how patterns of
habitat loss and fragmentation relate to watershed man-
agement. We hypothesized: (H
1
) a significant regional
decrease in SAV habitat cover and increase in SAV sea-
scape fragmentation over the 71-year study period; (H
2
)
seascapes in close proximity to shore and canals experi-
ence greater loss of SAV and fragmentation than more
distant seascapes; and (H
3
) seascapes in close proximity
to canals experience more dynamic patterns of SAV habi-
tat cover and fragmentation than more distant seascapes.
Material and Methods
Study site
The study area was the western shoreline of Biscayne Bay,
Florida (Fig. 1), a shallow subtropical lagoon adjacent to
the city of Miami (population 2,500,000) and downstream
of the Florida Everglades. The natural hydrology of the
Biscayne Bay watershed was modified with the construc-
tion of the Central and Southern Florida Project (CS&F)
water-drainage system completed in the 1960s (Browder &
Ogden 1999). The Comprehensive Everglades Restoration
Plan (CERP) has been designed, in part, to recover the
natural and historical hydrology of the Everglades and
coastal lagoons of South Florida. CERP has specific goals
to restore the amount of fresh water reaching Florida and
Biscayne Bay, as well as to modify the way the fresh water
is delivered (Light & Dineen 1994; McIvor et al. 1994;
Browder & Ogden 1999). The nearshore SAV seascapes are
of special concern within the restoration framework as
these are the areas presently exhibiting the widest fluctua-
tions in salinity and where the impacts of restoration pro-
jects would be concentrated.
Study design
Six sites were selected along the western shore of central
Biscayne Bay (Fig. 1). The study sites were divided into
two types: (i) ‘distant’ from freshwater canals (n =3 sites,
mean distance to canal =2.77 0.94 km), and (ii) ‘adja-
cent’ to freshwater canals (n =3 sites, mean distance to
canal =0.54 0.10 km). The canals adjacent to survey
sites were: Snapper Creek, Black Point and Mowry
Canals. These canals have high discharge rates (average
flow of 4 m
3
s
1
; SFWMD-DBHYDRO (http://my.sfwmd.
gov/dbhydroplsq/show_dbkey_info.main_menu); Browder
et al. 2010), and were constructed >50 years ago, result-
ing in long-term patterns of discharge. There are other
sources of fresh water (e.g. natural creeks, groundwater
seepages) for Biscayne Bay; however, the freshwater
contributions of these sources have been significantly
reduced over the years (Caccia & Boyer 2005; Stalker
et al. 2009). Freshwater pulses from man-made structure
haven been linked to changes in the salinity and nutrients
regimes of the bay (Caccia & Boyer 2005).
Following Santos et al. (2011), sites adjacent to canals
were located in close proximity to canals with the largest
average discharge rates. The sites distant from canals were
randomly selected along the shoreline, but were all
located >1km
2
from a canal. Within sites, the habitats
were further divided into shoreline (<200 m from shore)
and offshore (>200 m) based on previous work that
identified the shoreline habitats as areas with significantly
lower and more variable salinity (Santos et al. 2011; Lir-
man et al. 2014). For the historical analyses, nine repre-
sentative periods, 5–10 years apart, were selected from
aerial photographs collected from 1938 to 2009 (Support-
ing Information Appendix S1). The selection of specific
years for analysis was based on the availability and quality
of the aerial imagery.
Benthic habitat mapping
Image processing
The SAV seascape maps for each year were created using
high-resolution, digital black-and-white aerial photo-
graphs obtained from digital archives held by federal and
state agencies’ digital archives (Appendix S1). All imagery
was first processed to standardize the resolution, optical
properties and area of sampling, and then geo-rectified
using the United States Geological Survey topographic
map as a spatial reference. The resolution of all aerial
photographs was re-sampled to 1-m pixel size, and a his-
togram equalization and convolution filtering technique
was applied to control for the contrast and textural opti-
cal variability among years and sites. The majority of the
aerial photographs had 1-m pixel size, but the most
recent years (1991–2009) had 0.35-m pixel size (Appendix
S1); therefore, the most recent years were re-sampled to
1 m to standardize the mapping procedure. The 1-m res-
olution still provided adequate pixel size to delineate even
the smaller patches observed. Like the filtering technique,
the re-sampling of the aerial photographs helped in the
smoothing of the image and therefore in the reduction of
Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH 3
Santos, Lirman & Pittman Spatial-temporal dynamics in SAV seascapes
noise and salt-pepper effects. Finally, a 500-m radial buf-
fer was used to extract and standardize the area mapped
for each site. A radius of 500 m was used as this distance
includes the extent of the nearshore habitats where the
influence of the CERP canals and projects will be concen-
trated (Lirman et al. 2008; Lirman et al. 2014).
Mapping procedure
Submerged aquatic vegetation seascape maps were created
by hand-digitizing and delineating individual SAV
patches. The digitization procedure was standardized by
setting all photographs to a 50% contrast level and a
1:2500 scale, with a minimum mapping unit of 20 m
2
(the
size of the digitization cursor). Seagrass patches were man-
ually digitized because the optical properties (i.e. bright-
ness, tone, texture) varied significantly within and among
photographs. The contrast level used to analyse the photo-
graphs clearly highlighted patches with dense SAV cover
(>50% cover), which facilitated the digitization process
and reduced misclassification. All photograph interpreters
(n =3) were trained to follow a set of digitization rules to
limit variability among observers/interpreters. In addition,
all preliminary maps were subjected to quality assessment
by the lead interpreter using the digitization rules and by
comparing preliminary maps to a computer-automated
classification. The availability of ground-truth points was
scarce as the majority of the images were taken before
2005 when the nearshore seagrass monitoring program
was initiated in Biscayne Bay (Lirman et al. 2008). Prior
spatial accuracy using ground validation surveys showed
that this mapping procedure produced classified benthic
maps with a spatial accuracy of 60–80% (Santos et al.
2011). Similar studies have used archived aerial photo-
graphs to monitor seascape changes in nearshore areas
with high thematic accuracy (Sheppard et al. 1995; Zhari-
kov et al. 2005). The total area of seagrasses mapped
within the six sites was 2.57 km
2
, with a mean area per
site of 0.43 km
2
(0.16 SD).
Fig. 1. Study region: Biscayne Bay, Miami,
Florida. Six sites were selected: three adjacent
to freshwater canals (in black) and three
distant from canals (in grey).
4Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH
Spatial-temporal dynamics in SAV seascapes Santos, Lirman & Pittman
Spatial-pattern analysis
Spatial-pattern metrics
Spatial-pattern metrics provide quantitative information
that measure and describe the spatial patterning of sea-
scapes and can be grouped broadly into two categories:
those that quantify the composition of the patch mosaic
and those that quantify the configuration or spatial
arrangement of seascape elements (McGarigal & Cush-
man 2002; Wedding et al. 2011). Metrics quantifying
seascape composition are used to document habitat
losses and gains, whereas metrics of configuration are
used to assess habitat fragmentation. Spatial-pattern met-
rics were applied to benthic habitat maps with the soft-
ware FRAGSTATS v. 4 (McGarigal et al. 2012) to
quantify the spatial composition (percentage cover) and
configuration (fragmentation) of SAV seascapes
(Table 1). As recommended by Sleeman et al. (2005)
and McGarigal et al. (2005), patch density (PD), land-
scape division (LD), area-weighted mean perimeter to
area ratio (AWMPAR) and mean radius of gyration
(GYRATE_MN) were selected here to quantify the spa-
tial configuration of SAV patches and measure the rate
of SAV seascape fragmentation over time. These metrics
quantify four distinct characteristics of spatial pattern:
habitat size, compactness, habitat subdivision and habitat
geometry (McGarigal et al. 2005; Table 1) and are robust
across spatial scales (i.e. grain and/or extent size; Wu
et al. 2002; Wu 2004), total areal coverage and aggrega-
tion of the target habitat (Neel et al. 2004; Sleeman
et al. 2005; Cushman et al. 2008), and have been utilized
previously to assess the effects of SAV seascape fragmen-
tation on tropical and temperate fish and invertebrate
species (Hovel & Lipcius 2001; Salita et al. 2003; Bos-
tr€
om et al. 2011).
Habitat loss and fragmentation patterns
Spatio-temporal patterns in metric values were explored
and analysed using principal component and vector
analyses followed by linear regression models.
An index of area change (G; Frederiksen et al. 2004a)
was applied to quantify the proportion of the total area
of SAV either lost or gained between two consecutive
sampling periods. G was calculated as follows:
G¼ðarea lost þarea gainedÞ=ðarea lost þarea gained
þarea unchangedÞ(1)
The index, which ranges from 0 (no change) to 1 (a
complete change in SAV area with either 100% lost or
gained), is used to quantify changes in seascape composi-
tion independent of measurements of spatial configura-
tion.
The configuration metrics used to quantify fragmenta-
tion were analysed using two approaches: (i) a multivari-
ate principal component analysis (PCA) followed by a
vector analysis (VA) to examine the contribution of each
metric to fragmentation patterns, as well as the magni-
tude of fragmentation between consecutive sampling peri-
ods; and (ii) a simple metric of fragmentation, the
‘fragmentation index’ as described below:
Multivariate analyses: The multivariate data (i.e. values
for the four seascape metrics) calculated for each site and
time interval were used in a PCA ordination. The co-
ordinates of each site within the PC1 and PC2 plane were
used to measure the direction and length of the vector
connecting sites between consecutive sampling periods.
The resulting vectors were used to measure the magni-
tude (i.e. length of vector) and direction (i.e. movement
towards fragmentation or expansion/clustering, referred
to hereafter as defragmentation) of change in SAV sea-
scape state between intervals (Appendix S2). The vector
length was standardized by the number of years elapsed
between images. Lastly, a hierarchical cluster analysis
(CA) was used to discern robust groupings of sites/years
that shared similar values of the metrics used to evaluate
spatial and temporal patterns of SAV fragmentation. The
CA was performed using the Euclidean dissimilarity
matrix of the PCA scores of each site/year. The PCA and
CA were performed in PRIMER v. 6.
Fragmentation index: As the four spatial-pattern metrics
used quantify different aspects of spatial properties (habi-
tat size, compactness, habitat subdivision and habitat
geometry), these metrics were integrated into a single
fragmentation index for simplicity. By developing a single
fragmentation index, a simpler temporal analysis can be
used to assess the trajectory of SAV seascapes. Similar
approaches have been used previously to assess the effects
of fragmentation on species diversity, probability of
occurrence, and abundance of terrestrial and marine spe-
cies independent of habitat loss (McGarigal & McComb
1995; Trzcinski et al. 1999; Kaufman 2011). The metrics
used here were collapsed into the following fragmentation
index (FragIndex):
FragIndex ¼4pðPD LD AWMPAR 1=Gyrate MNÞ
(2)
All metrics were standardized to produce a FragIndex
ranging from 0 (low fragmentation) to 1 (high fragmen-
tation).
Statistical Analyses
Statistical analyses were performed using JMP v. 10. A
Shapiro–Wilk test on the dependent variables was used to
Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH 5
Santos, Lirman & Pittman Spatial-temporal dynamics in SAV seascapes
test for normality and check for other assumptions of para-
metric analyses. Variables were Box–Cox transformed
when the normality test failed. One-way analyses of vari-
ance (ANOVA) were used to test for differences in the
index of relative change, fragmentation index and changes
in SAV composition between areas distant and adjacent to
freshwater canals (testing for H
1
and H
3
). Analyses of
covariance (ANCOVA) were used to evaluate differences in
loss of SAV and SAV seascape fragmentation between areas
distant and adjacent to freshwater canals (testing for H
2
).
Results
H
1
–SAV cover and SAV seascape fragmentation
A regional decrease in SAV cover and transition to a
more fragmented SAV seascape was detected (PCA results
below). A net amount of 0.085 km
2
of SAV was lost
across all sites over the whole study period (1938–2009),
representing a loss of 3.31% of the total SAV mapped
(2.57 km
2
) (Table 2). Areas adjacent to freshwater canals
showed higher average net loss than areas distant from
canals. Two of the three sites adjacent to canals, Black
Point Canal (BP) and Snapper Creek (SC), had net losses
of 11.4% and 11.5%, respectively. Unexpectedly, the high-
est net loss (17.8%) was recorded for Black Point Lagoon
(BL), one of the sites distant from canals. Two of the
three distant sites, Chicken Key (CK) and Turkey Point
(TP), presented net gains of 1.2% and 9.9%, respectively
(Table 2). No statistics were performed to test for the sig-
nificance of this pattern due to the limited number of
sites examined.
The PCA based on the spatial-pattern metrics selected
to describe the fragmentation of SAV habitats (e.g. PD,
LD, AWMPAR and GYRATE_MN) showed that 91% of
the variation in SAV spatial configuration was explained
by the first two principal components (Table 3 and
Appendix S3). The first PC explains 71% of the variation,
and separates sites based on the level of fragmentation,
with higher values representing higher fragmentation
(Table 3). This first axis was positively correlated with
PD, LD and AWMPAR, which indicate, respectively, an
Table 1. Spatial pattern metrics used to quantify composition and configuration of submerged aquatic vegetation seascape patterns in
Biscayne Bay. ZLAND was used to assess habitat loss, and the remaining four metrics were used to create the fragmentation index
4p(PD*LD*AWMPAR*1/Gyrate_MN.
Metric Acronym Category Aspect Description Formula
Percentage of the
total landscape
PLAND Composition Area/density Percentage of the total landscape
made up of the corresponding class
PLAND ¼Pn
j¼1aij
Að100Þ
Patch density PD Configuration Area/density Number of patches of a certain class
divided by the total landscape area
n patches per ha
PD ¼ni
A
Mean radius of
gyration
GYRATE_MN Configuration Area/density Measure of patch extensiveness
Larger patches will have higher
GYRATE values
GYRATE MN ¼Pn
j¼1Pz
r¼1hijr
z
ni
Landscape division LD Configuration Interspersion/
contagion
Probability that two randomly chosen
pixels in the landscape are not
situated in the same patch
LD ¼1P
n
i¼0ðai
AÞ2
Area-weighted
mean perimeter-
area ratio
AWMPAR Configuration Shape Measure of patch-shape complexity AWMPAR ¼Pi¼NP
i¼1
pi
ai:ai
Pi¼NP
i¼1ai
Table 2. Patterns of change in submerged aquatic vegetation area in Western Biscayne Bay over the >70-year record evaluated here (1938–
2009). Per cent net change calculated as the proportion of net area change across all years with the total areal extent sampled at each site (site
area). Per cent net positive and negative change highlighted in bold gray and black, respectively.
Site abbreviation Name Canal treatment Site area (km
2
) Net area changes (km
2
) %Net change
BL Black Point Lagoon Distant 0.41 –0.073 –17.78
CK Chicken Key Distant 0.34 0.004 1.17
TP Turkey Point Distant 0.71 0.07 9.93
BP Black Point Canal Adjacent 0.39 –0.044 –11.42
CP Convoy Point Adjacent 0.33 0.003 0.94
SC Snapper Creek Adjacent 0.39 –0.045 –11.55
Total 2.57 –0.085 –3.31
6Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH
Spatial-temporal dynamics in SAV seascapes Santos, Lirman & Pittman
increase in the number of patches, the probability that
two randomly selected locations within the seascape are
not situated in the same contiguous habitat patch and
the shape complexity of the patches. GYRATE_MN was
negatively associated with PC1 as an increase in this met-
ric represents an increase in continuity or structural con-
nectedness of the patches. Using a cluster analysis and
the scores on the first PC, the PCA biplot was divided
into three regions representing low, medium and high
fragmentation (Fig. 2 and Appendix S3). All sites exhib-
ited higher fragmentation in 2009 compared with 1938,
with four sites exhibiting highest fragmentation levels
between the 1990s–2000s (Fig. 2).
H
2
–Temporal trends in SAV cover and fragmentation
Submerged aquatic vegetation cover decreased signifi-
cantly over time, while fragmentation increased signifi-
cantly over time (Fig. 3). Contrary to what we
hypothesized, the rates of change (i.e. slopes of the fitted
lines) of fragmentation and SAV cover were not statisti-
cally different between the canal adjacent and distant
treatments (ANCOVA, P >0.05; Fig. 3). However, when
the metrics were separated into shoreline (0–200 m) and
offshore buffers (200–500 m), there was a significantly
higher rate of habitat loss and fragmentation in shoreline
habitats compared with offshore habitats (Fig. 4). Within
the shoreline buffer, sites adjacent to canals had higher
rates of habitat loss and fragmentation; however, the dif-
ference between the canal adjacent and distant treatments
was not statistically significant.
H
3
–SAV loss and fragmentation dynamics
Looking only at the initial and final values spanning a
>70-year interval masks the highly dynamic changes
that have taken place between decades that show that
the SAV seascape along Biscayne Bay’s western shore-
line is indeed influenced by the presence of freshwater
canals and the historical transformation of the bay’s
salinity regimes. While areas adjacent to canals had, on
average, significantly higher SAV cover (one-way ANO-
VA, P <0.05), the temporal changes in the amount of
SAV were significantly more dynamic in these areas, as
illustrated by the higher relative change index in sites
adjacent to canals (one-way ANOVA, P <0.05; Fig. 5a).
This is further illustrated by the significantly higher
proportion of SAV habitat that was lost and gained in
areas adjacent to canals (one-way ANOVA, P <0.05;
Fig. 5b).
The vector analysis measured the magnitude of
fragmentation (black vectors pointing right within the
PCA biplot) and defragmentation (gray vectors pointing
left) between sampling intervals adjusted by the number
of years between images (Fig. 3). The magnitude of
fragmentation and defragmentation (black and gray
vectors) was not significantly different (Fig. 6), thus
explaining the limited net change in SAV cover
recorded over the >70-year record (Table 2). However,
the magnitude of both vectors was significantly higher
(two-way ANOVA, P <0.05) in areas adjacent to
canals (Fig. 6), indicating more spatially dynamic con-
ditions within these areas.
Table 3. Principal components analysis (PCA) eigenvalues and
variable loading. Four variables that measure different spatial
components of habitat fragmentation were considered in the PCA.
Variable loadings in axis 1 were used to quantify spatial arrangement
and fragmentation temporal dynamics.
Eigenvalues and variable loadings Axis 1 Axis 2
Eigenvalues 2.87 0.79
Percentage 71.70 19.60
Cumulative percentage 71.70 91.30
Principal component analysis
variable loadings
Patch density (PD) 0.57 0.152
Mean radius of gyration
(GYRATE_MN) 0.46 0.64
Area-weighted mean perimeter
to area ratio (AWMPAR)
0.54 0.138
Landscape division (LD) 0.42 0.74
Fig. 2. Classification of fragmentation seascape state within each site. Classification was based on the principal component analysis and overlay
cluster analysis (see Fig. 6 for details). Grey =low fragmentation; Dark Grey =medium fragmentation; Black =high fragmentation. Black dotted
line divides sites distant (up) and adjacent (down) from a freshwater canal.
Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH 7
Santos, Lirman & Pittman Spatial-temporal dynamics in SAV seascapes
Discussion
Our analysis of the structure of nearshore SAV habitats
of Biscayne Bay, Florida, USA, over a 71-year time period
has shown declines in SAV cover and increases in sea-
scape fragmentation for this highly modified coastal
lagoon located adjacent to the city of Miami. However,
the net loss in SAV cover was relatively low (approxi-
mately 3%) across the study period (1938–2009). This is
in clear contrast to recent reports of significant declines
in worldwide seagrass abundance (Duarte 2002; Waycott
et al. 2009), but in agreement with examples of seagrass
populations that have been stable or increasing over time
(Frederiksen et al. 2004b; Hernandez-Cruz et al. 2006;
Lyons et al. 2010, 2013). By contrast, the spatial configu-
ration of the SAV seascapes shifted significantly from
continuous (i.e. seascape dominated by few large patches)
to fragmented seascapes (i.e. many small patches,
Fig. 3. Submerged aquatic vegetation (SAV)
cover and Fragmentation Index over time
measured within the 500 m from shore
spatial extent. Solid and dashed lines
(standard error =shaded area) represent
temporal trend in habitat loss and
fragmentation within areas adjacent and
distant from a freshwater canal. Based on the
ANCOVA results, the rate of habitat loss and
fragmentation appears to be equal in both
canal treatments; however, lines within SAV
habitat cover were not coincident (i.e.,
parallel with different intercept –different
initial and final values).
Fig. 4. Submerged aquatic vegetation cover
and Fragmentation Index over time measured
within the shoreline habitats (0–200 m –left
panel) and offshore habitats (200–500 m –
right panel). Solid and dashed lines (standard
error =shaded area) represent temporal
trend in habitat loss and fragmentation
within areas adjacent and distant from a
freshwater canal. Based on the ANCOVA
results, the rate of habitat loss and
fragmentation appears to be equal in both
canal treatments but significant within the
nearshore area; however, no significant
trends were observed within the offshore
areas.
8Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH
Spatial-temporal dynamics in SAV seascapes Santos, Lirman & Pittman
perforated SAV meadows) over the same time period.
These conflicting patterns (i.e. small change in SAV cover
but significant fragmentation) highlight the importance
of incorporating both composition and configuration
metrics into comprehensive assessments of SAV habitats.
The low taxonomic resolution of the mapping approach
used in this historical analysis may mask changes in spe-
cies and community composition, such as replacements
of euhaline (e.g. Thalassia testudinum,Halimeda spp.) for
more ephemeral mesohaline taxa (Halodule wrighttii,
Laurencia sp.) (Collado-Vides et al. 2011). Such species
replacements (i.e. from slower-growing to fast-growing
SAV species) may add to the highly dynamic nature of a
stressed SAV community and may be partly responsible
for the patterns documented here.
Historically, few large-scale mapping and monitoring
studies have considered the significance of SAV seascape
fragmentation dynamics and how this spatial transforma-
tion may influence the resilience of SAV populations (Bell
et al. 2007; Cunha & Santos 2009; Montefalcone et al.
2010; Cuttriss et al. 2013). Fragmentation of SAV sea-
scapes is likely to have a greater effect on bed persistence
than changes in cover (Sleeman et al. 2005). Within-bed
SAV cover can be highly dynamic as significant changes
in plant biomass are often recorded between seasons (Lir-
man et al. 2014). By contrast, bed expansion through the
colonization of new propagules or the extension of exist-
ing clones through rhizome elongation can be a slow pro-
cess. Evidence of this is the common lack of recovery of
propeller scars within T. testudinum beds in Florida
where rhizomes are not able to bridge the denuded gap
and so the scars persist for years or decades. Seagrasses,
the main ecosystem engineers within SAV patches in
Biscayne Bay, rely heavily on rhizome extension for
recovery after disturbance and bed expansion (Duarte &
Sand-Jensen 1990; Jensen & Bell 2001; Kendrick et al.
2005; Sintes et al. 2005). Therefore, the fragmentation of
SAV habitats, which creates gaps among patches and
increases the amount of edges, influences seagrass ecosys-
tem function directly (Duarte & Sand-Jensen 1990; Kend-
ab
Fig. 5. Submerged aquatic vegetation (SAV)
seascape patterns calculated for the period
1938–2009. (a) Index of area change; and (b)
percentage of SAV area gained and lost.
Metrics were statistically different between
canal treatments based on a one-way
ANOVA (P <0.05).
Fig. 6. Vector lengths calculated from the coordinates of principal
component analysis biplot quantifying the magnitude of
fragmentation (black) and defragmentation (grey) by canal treatment.
Dots represent the mean values and bars/whiskers the standard error.
No significant differences in vector length were found between
fragmentation/defragmentation, but significant differences between
canal treatments were observed (two-way ANOVA, P <0.05).
Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH 9
Santos, Lirman & Pittman Spatial-temporal dynamics in SAV seascapes
rick et al. 2005) and can make plant populations more
vulnerable to local extinction (Liao et al. 2013). The high
mortality associated with small SAV patches could be
linked to lower anchoring capabilities and higher erosion
influenced by the higher edge : area ratio of fragmented
habitats, leading also to higher susceptibility to physical
disturbances (Vidondo et al. 1997; Kendrick et al. 2005;
Sintes et al. 2005; Duarte et al. 2007). By contrast, con-
tinuous SAV seascapes composed of larger patches can be
more stable and resilient by stabilizing sediments, reduc-
ing erosion and re-suspension, and enhancing resource
accumulation and allocation (Duarte & Sand-Jensen
1990; Fonseca & Bell 1998; Sintes et al. 2005).
In this study, for the first time, the long-term spatial
dynamics of SAV habitat loss and fragmentation were
quantified in relation to the disturbance associated with
freshwater discharges into a coastal bay. Net habitat loss
from 1938 to 2009 was more common within the sites
adjacent to freshwater canals (two out of three sites). By
contrast, two out of three sites classified as distant from
canals showed net gains. The rate of habitat loss and frag-
mentation was higher within shoreline habitats that are
close to the point of freshwater discharge. By exploring
the trajectories of individual sites over >70 years, we can
begin to understand how composition and configuration
may be responding differently to the environment and
providing seemingly contradictory results (e.g. loss of
SAV cover with a concurrent decrease in fragmentation;
Fig. 7). There are different alternative scenarios of SAV
habitat loss and SAV seascape fragmentation associated
with different seascape transformation types (e.g. perfora-
tion, dissection, subdivision, shrinkage and attrition) and
trade-offs between disturbance and succession processes
(Forman 1995; McGarigal et al. 2005). Sites can exhibit
an increase in cover without a reduction in fragmentation
if large SAV patches increase in above-ground biomass
and steadily expand over time on a localized portion of
the seascape, but fail to expand into denuded areas (illus-
trated by a downward progression along the SAV cover
vertical axis in Fig. 7). Conversely, sites can exhibit
reduced cover without a change in fragmentation if
remaining patches only ‘thin out’ in unfavorable environ-
mental conditions. Within sites, resources need to be
allocated to the maintenance of above-ground biomass
(which influences SAV cover) and below-ground biomass
that results in rhizome extension and thus bed expansion.
Under favorable conditions, SAV patches can increase in
biomass/cover as well as expand into suitable habitat and
fill out gaps among patches. When conditions are not
consistently favorable, trade-offs may lead to conflicting
and dynamic patterns such as documented here. Under
extreme or persistent unfavorable conditions, this may
lead to both habitat loss and fragmentation (illustrated by
the black dashed line in Fig. 7). Under this scenario, the
SAV seascape is subject to habitat perforation, subdivi-
sion, and shrinkage and attrition processes that lead to a
gradual breakdown and formation of discrete fragments
(Forman 1995). The site south of BP was the only loca-
tion that showed both significant habitat loss and frag-
mentation. The pulsed discharge of fresh water from this
canal creates a highly variable salinity environment that
can experience low-salinity conditions within a few hours
and remains <5 psu for several days (Wang et al. 2003;
Lirman et al. 2008). These pulsed freshwater disturbances
have been associated with the reduction in SAV cover
and dominance, and abundance of species with high
turn-over rates (Zieman et al. 1999; Lirman et al. 2008;
Collado-Vides et al. 2011), thus contributing to both the
loss and fragmentation of SAV habitats. Depression of
productivity and short-shoot mortality of T. testudinum,
the dominant seagrass species of Biscayne Bay, have been
previously associated with low-salinity events, and high
and frequent salinity fluctuations (Lirman & Cropper
2003; Herbert & Fourqurean 2008).
Sites that showed only habitat loss were subject to
localized removal of biomass and/or shrinkage and attri-
tion of existing patches (illustrated by dotted and dash-
dot arrows in Fig. 7). Sites that showed only fragmenta-
tion were subject to a balance among habitat subdivision,
shrinkage and the formation of small patches. The CP site
located close to Mowry Canal showed the highest rate of
fragmentation, but not declines in cover. Similar patterns
were observed at TP, a site classified as distant from
canals. Although fragmentation patterns at TP could not
be associated with freshwater pulses, there are other local-
ized disturbances in this area that could have influenced
such spatial processes. For example, the Turkey Point
Nuclear Generating Station has been operating in this
area since 1967 (Dolan 2012) and studies have observed
significant declines in the abundance and density of sea-
grass up to 1 km from the heated effluent of this power
plant (Zieman & Wood 1975; Dolan 2012). In addition
to thermal stress, hypersalinity conditions, which are
known to be detrimental to seagrass (Koch et al. 2007b;
Herbert & Fourqurean 2008), could be associated with
the fragmentation observed at this site (Florida Power &
Light Company 2009; Hughes et al. 2009; Dolan 2012).
Other studies that used spatial pattern metrics in addi-
tion to areal extent identified an increase in the fragmen-
tation state of SAV seascapes and an increase or neutral
trend in SAV cover (Cunha et al. 2005; Cuttriss et al.
2013). Hernandez-Cruz et al. (2006) observed an expan-
sion of seagrass cover within an embayment, but the
nearshore portion of the study area influenced by efflu-
ent discharges revealed increases in patchiness and
fragmentation. Montefalcone et al. (2010) noticed that
10 Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH
Spatial-temporal dynamics in SAV seascapes Santos, Lirman & Pittman
the abundance patterns of seagrasses in the Mediterra-
nean Sea were not correlated with coastal development,
but that fragmentation measures were indeed influenced
by human-induced disturbances on seagrass meadows.
A confounding environmental factor that may have
played a role in the observed SAV historical patterns is
the increase in nutrient availability associated with fresh-
water discharges. Habitats of Biscayne Bay in proximity
to canals have been shown to have high N availability
(Caccia & Boyer 2005). This was confirmed by studies
that showed high N content within the tissue of macroal-
gae and seagrasses within Western Biscayne Bay (Collado-
Vides et al. 2011; Lirman et al. 2014). While not tested, it
is possible that increased nutrients result in faster growth
and higher SAV cover in closer proximity to canals (Her-
bert & Fourqurean 2008), supporting our observation of
higher SAV cover near canals. The increased above-
ground productivity, however, may not necessarily trans-
late into high resilience to the negative salinity impacts
and may not prevent seascape fragmentation. In addition,
increased sediment organic content linked with elevated
productivity and turnover rates has the potential to cre-
ate sulfate-reducing conditions, which have been impli-
cated in seagrass die-offs in neighboring Florida Bay
(Fourqurean & Robblee 1999; Koch et al. 2007b). Thus,
the temporally and spatially dynamic influence of salinity
and nutrients on SAV may have been responsible for the
patterns of composition and configuration reported here
and may explain why both cover and fragmentation were
higher in some of the areas adjacent to canals. Clearly,
controlled experiments that isolate the effects of low
salinity and high nutrient concentrations on above- and
below-ground seagrass biomass and algal productivity are
needed to provide further insights into the influence of
water management practices on SAV seascapes.
In conclusion, the SAV seascape in Biscayne Bay has
been highly dynamic in time and space during the extent
of 71 years included in this study, with changes in SAV
composition and configuration being greatest in the habi-
tats closest to shore and freshwater canals. More impor-
tantly, our results highlight the importance of quantifying
habitat loss and habitat fragmentation independently to
tease apart patterns related to the amount and removal of
SAV habitat versus those related to the spatial arrange-
ment and configuration of SAV patches. These results
illustrate the importance of incorporating measurements
of SAV seascape spatial characteristics into existing moni-
toring and restoration programs to have a complete indi-
cation and accurate projection of coastal habitat resilience
and recovery from anthropogenic disturbances. In addi-
tion, seascape spatial characteristics are especially relevant
when evaluating the cascading effects that changes in
Fig. 7. Conceptual submerged aquatic vegetation (SAV) seascape transformation illustrating habitat loss/gain (vertical axis) and fragmentation
(horizontal axis). SAV patches illustrated in black. Different arrows depict the most common process of SAV seascape transformation across time
in Biscayne Bay.
Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH 11
Santos, Lirman & Pittman Spatial-temporal dynamics in SAV seascapes
SAV seascapes can have on associated macrofauna. Stud-
ies in terrestrial ecology have indicated that thresholds
exist beyond which abrupt decline in habitat suitability
occurs and where fragmentation effects become signifi-
cant (Fahrig 2003); however, there is still a limited
understanding about how marine nektonic populations
response to fragmentation relatively with the amount of
habitat within the seascape. Therefore, a seascape
approach is essential to comprehend how marine habi-
tats’ spatial properties influence the growth, survivorship
and ecological interactions of marine species, and hence
the quality of nursery functions and fisheries productivity
of nearshore environments. By combining remote sensing
techniques with landscape ecology and conventional mar-
ine ecology, our study provides a quantitative framework
by which change in seascape spatial patterning can be
monitored and measured to support the implementation
of adaptive management strategies.
Acknowledgements
We are indebted to our volunteers S. Denka and A. Ros-
inski for their help in the mapping process. This research
was conducted under permit BISC-2011-SCI-0028. Fund-
ing was provided by National Oceanic & Atmospheric
Administration Educational Partnership Program and
Living Marine Resources Cooperative Science Center, the
Army Corps of Engineers, the US Department of the
Interior’s Critical Ecosystem Studies Initiative and the
RECOVER Monitoring and Assessment Program (MAP).
We thank the anonymous reviewers of this study as well
as Dr Serafy for their insightful suggestions.
References
Andren H. (1994) Effects of habitat fragmentation on birds
and mammals in landscapes with different proportions of
suitable habitat: a review. Oikos,71, 355–366.
Barbier E.B., Hacker S.D., Kennedy C., Koch E.W., Stier A.C.,
Silliman B.R. (2011) The value of estuarine and coastal
ecosystem services. Ecological Monographs,81, 169–193.
Bell S.S., Fonseca M.S., Stafford N.B. (2007) Seagrass
ecology: new contributions from a landscape perspective.
In: Larkum A.W.D., Orth R.J., Duarte C.M. (Eds),
Seagrasses: Biology, Ecology and Conservation. Springer,
Netherlands: 625–645.
Bonte D., Hovestadt T., Poethke H.J. (2010) Evolution of dispersal
polymorphism and local adaptation of dispersal distance in
spatially structured landscapes. Oikos,119, 1921–1925.
Bostr€
om C., Pittman S.J., Simenstad C., Kneib R.T. (2011)
Seascape ecology of coastal biogenic habitats: advances,
gaps, and challenges. Marine Ecology Progress Series,427,
191–217.
Browder J.A., Ogden J.C. (1999) The natural South Florida
system II: pre-drainage ecology. Urban Ecosystems,3, 245–
277.
Browder J.A., Liehr G.A., Johnson D., Buck E., Jackson T.L.
(2010) Epifaunal communities of mainland nearshore South
Biscayne Bay. 2010 Annual Report to CERP-Recover
Monitoring and Assessment Plan, SE MAP Project 3.2.35
and 3.2.4.7.
Caccia V.G., Boyer J.N. (2005) Spatial patterning of water
quality in Biscayne Bay, Florida as a function of land use
and water management. Marine Pollution Bulletin,50, 1416–
1429.
Collado-Vides L., Mazzei V., Thyberg T., Lirman D. (2011)
Spatio-temporal patterns and nutrient status of macroalgae
in a strongly managed region of Biscayne Bay, Florida, USA.
Botanica Marina,54, 377–390.
Cunha A.H., Santos R.P. (2009) The use of fractal geometry to
determine the impact of inlet migration on the dynamics of
a seagrass landscape. Estuarine Coastal & Shelf Science,84,
584–590.
Cunha A.H., Santos R.P., Gaspar A.P., Bairros M.F. (2005)
Seagrass landscape-scale changes in response to disturbance
created by the dynamics of barrier-islands: a case study
from Ria Formosa (Southern Portugal). Estuarine Coastal &
Shelf Science,64, 636–644.
Cushman S.A., McGarigal K., Neel M.C. (2008) Parsimony in
landscape metrics: strength, universality, and consistency.
Ecological Indicators,8, 691–703.
Cuttriss A.K., Prince J.B., Castley J.G. (2013) Seagrass
communities in southern Moreton Bay, Australia: coverage
and fragmentation trends between 1987 and 2005. Aquatic
Botany,108,41–47.
Davis S.E. III, Lirman D., Wozniak J.R. (2009) Nitrogen and
phosphorus exchange among tropical coastal ecosystems. In:
Nagelkerken I. (Ed.), Ecological Connectivity Among Tropical
Coastal Ecosystems. Springer, Netherlands: 625–645.
Dolan T.E. (2012) A Case Study of Turkey Point Nuclear
Generating Station: Perception and Power in Environmental
Assessment. Open Access Theses, University of Miami.
Duarte C.M. (2002) The future of seagrass meadows.
Environmental Conservation,29, 192–206.
Duarte C.M., Sand-Jensen K. (1990) Seagrass colonization:
patch formation and patch growth in Cymodocea nodosa.
Marine Ecology Progress Series,65, 193–200.
Duarte C.M., Middelburg J.J., Caraco N. (2005) Major role of
marine vegetation on the oceanic carbon cycle.
Biogeosciences,2,1–8.
Duarte C.M., Fourqurean J.W., Krause-Jensen D., Olesen B.
(2007) Dynamics of seagrass stability and change. In:
Larkum A.W.D, Orth R.J., Duarte C.M. (Eds), Seagrasses:
Biology, Ecology and Conservation. Springer, Netherlands:
271–294.
Ewers R.M., Didham R.K. (2006) Confounding factors in the
detection of species responses to habitat fragmentation.
Biological Reviews,81, 117–142.
12 Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH
Spatial-temporal dynamics in SAV seascapes Santos, Lirman & Pittman
Fahrig L. (2003) Effects of habitat fragmentation on
biodiversity. Annual Review of Ecology, Evolution &
Systematics,34, 487–515.
Florida Power and Light Company (2009) Site Certification
Application, Turkey Point Units 6 & 7. June 2009. Available
at: http://www.nrc.gov/reactors/new-reactors/col/turkey-
point.html (accessed 20 December 2013).
Fonseca M.S., Bell S.S. (1998) Influence of physical setting on
seagrass landscapes near Beaufort, North Carolina, USA.
Marine Ecology Progress Series,171, 109–121.
Forman R.T.T. (1995) Land Mosaics: The Ecology of Landscapes
and Regions. Cambridge University Press, Cambridge, UK:
656.
Fourqurean J.W., Robblee M.B. (1999) Florida Bay: a history
of recent ecological changes. Estuaries,22, 345–357.
Frederiksen M., Krause-Jensen D., Holmer M., Laursen S.
(2004a) Spatial and temporal variation in eelgrass (Zostera
marina) landscapes: influence of physical setting. Aquatic
Botany,78, 147–165.
Frederiksen M., Krause-Jensen D., Holmer M., Laursen S.
(2004b) Long-term changes in area distribution of eelgrass
(Zostera marina) in Danish coastal waters. Aquatic Botany,
78, 167–181.
Gillanders B.M. (2007) Seagrass, fish, and fisheries. In: Larkum
A.W.D, Orth R.J., Duarte C.M. (Eds), Seagrasses: Biology,
Ecology and Conservation. Springer, Netherlands: 503–536.
Gillanders B.M., Able K.W., Brown J.A., Eggleston D.B.,
Sheridan P.F. (2003) Evidence of connectivity between
juvenile and adult habitats for mobile marine fauna: an
important component of nurseries. Marine Ecology Progress
Series,247, 281–295.
Grober-Dunsmore R., Pittman S.J., Caldow C., Kendall M.S.,
Frazer T.K. (2009) A landscape ecology approach for the
study of ecological connectivity across tropical marine
seascapes. In: Nagelkerken I. (Ed.), Ecological Connectivity
Among Tropical Coastal Ecosystems. Springer, Netherlands:
493–530.
Hemminga M., Duarte C.M. (2000) Seagrass Ecology.
Cambridge University Press, Cambridge, UK: 312.
Herbert D.A., Fourqurean J.W. (2008) Ecosystem structure
and function still altered two decades after short-term
fertilization of a seagrass meadow. Ecosystems,11, 688–700.
Hernandez-Cruz L.R., Purkis S.J., Riegl B.M. (2006)
Documenting decadal spatial changes in seagrass and
Acropora palmatacover by aerial photography analysis in
Vieques, Puerto Rico: 1937–2000. Bulletin of Marine Science,
79, 401–414.
Hovel K.A., Lipcius R.N. (2001) Habitat fragmentation and
extinction thresholds on fractal landscapes. Ecology,82,
1814–1829.
Hughes J.D., Langevin C.D., Brakefield-Goswami L. (2009)
Effects of hypersaline cooling canals on aquifer salinization.
Hydrogeology Journal,18,25–38.
Hyndes G.A., Nagelkerken I., McLeodd R., Connolly R.M.,
Lavery S., Vanderklift M.A. (2014) Mechanisms and
ecological role of carbon transfer within coastal seascapes.
Biological Reviews,89, 232–254.
Irlandi E.A., Crawford M.K. (1997) Habitat linkages: the effect
of intertidal saltmarshes and adjacent subtidal habitats on
abundance, movement, and growth of an estuarine fish.
Oecologia,110, 222–230.
Jensen S., Bell S. (2001) Seagrass growth and patch dynamics
cross-scale morphological plasticity. Plant Ecology,155, 201–
217.
Kaufman K.A. (2011) Seagrass patch dynamics in areas of
historical loss in Tampa Bay, FL, USA. Graduate Theses and
Dissertations, University of South Florida. http://
scholarcommons.usf.edu/etd/3178/.
Kendrick G.A., Duarte C.M., Marb
a N. (2005) Clonality in
seagrass, emergent properties and seagrass landscapes.
Marine Ecology Progress Series,290, 291–296.
Koch E.W., Ackerman J.D., Verduin J., van Keulen M. (2007a)
Fluid dynamics in seagrass ecology –from molecules to
ecosystems. In: Larkum A.W.D, Orth R.J., Duarte C.M.
(Eds), Seagrasses: Biology, Ecology and Conservation.
Springer, Netherlands: 625–645.
Koch M.S., Schopmeyer S.A., Kyhn-Hansen C., Madden C.J.,
Peters J.S. (2007b) Tropical seagrass species tolerance to
hypersalinity stress. Aquatic Botany,86,14–24.
Laurance W.F., Nascimento H.E., Laurance S.G., Andrade
A.C., Fearnside P.M., Ribeiro J.E., Capretz R.L. (2006) Rain
forest fragmentation and the proliferation of successional
trees. Ecology,87, 469–482.
Liao J., Li Z., Hiebeler D.E., El-Bana M., Deckmyn G., Nijs I.
(2013) Modeling plant population size and extinction
thresholds from habitat loss and habitat fragmentation:
effects of neighboring competition and dispersal strategy.
Ecological Modelling,268,9–17.
Light S.S., Dineen J.W. (1994) Water control in the
Everglades: a historical perspective. In: Davis S.M., Ogden
J.C. (Eds), Everglades: The Ecosystem and Its Restoration. St.
Lucie Press, Delray Beach: 47–84.
Lirman D., Cropper W.P. (2003) The influence of salinity on
seagrass growth, survivorship, and distribution within
Biscayne Bay, Florid: field, experimental, and modeling
studies. Estuaries,26, 131–141.
Lirman D., Deangelo G., Serafy J., Hazra A., Hazra D., Herlan
J., Luo J., Bellmund S., Wang J., Clausing R. (2008)
Seasonal changes in the abundance and distribution of
submerged aquatic vegetation in a highly managed coastal
lagoon. Hydrobiologia,596, 105–120.
Lirman D., Thyberg T., Santos R., Schopmeyer S., Drury C.,
Collado-Vides L., Bellmund S., Serafy J. (2014) SAV
communities of Western Biscayne Bay, Miami, Florida,
USA: human and natural drivers of seagrass and macroalgae
abundance and distribution along a continuous shoreline.
Estuarine and Coasts,37, 1243–1255.
Lyons M.B., Roelfsema C.M., Phinn S.R. (2010) Long term
monitoring of seagrass distribution in Moreton Bay,
Australia, from 1972–2010 using Landsat MSS, TM, ETM+.
Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH 13
Santos, Lirman & Pittman Spatial-temporal dynamics in SAV seascapes
Paper presented at the geoscience and remote sensing
symposium (IGARSS), Honolulu, HI, 25–30 July 2010.
Lyons M.B., Roelfsema C.M., Phinn S.R. (2013) Towards
understanding temporal and spatial dynamics of seagrass
landscapes using time-series remote sensing. Estuarine
Coastal & Shelf Science,120,42–53.
McGarigal K., Cushman S. (2002) Comparative evaluation of
experimental approaches to the study of habitat
fragmentation effects. Ecological Applications,12, 335–345.
McGarigal K., McComb W.C. (1995) Relationships between
landscapes structure and breeding birds in the Oregon coast
range. Ecological Monographs,65, 235–260.
McGarigal K., Cushman S., Regan C. (2005) Quantifying
terrestrial habitat loss and fragmentation: a protocol. http://
www.umass.edu/landeco/teaching/landscape_ecology/labs/
fragprotocol.pdf.
McGarigal K., Cushman S., Ene E. (2012) FRAGSTATS v4:
Spatial pattern analysis program for categorical and
continuous maps. University of Massachusetts, Amherst.
http://www.umass.edu/landeco/research/fragstats/
fragstats.html.
McIvor C.C., Ley J.A., Bjork R.D. (1994) Changes in the
freshwater inflow from the Everglades to Florida Bay
including effects on biota and biotic process: a review. In:
Davis S.M., Ogden J.C. (Eds), Everglades: The Ecosystem and
Its Restoration. St. Lucie Press, Delray Beach: 117–146.
Montefalcone M., Parravicini V., Vacchi M., Albertelli G.,
Ferrari M., Morri C., Bianchi C.N. (2010) Human influence
on seagrass habitat fragmentation in NW Mediterranean
Sea. Estuarine Coastal & Shelf Science,86, 292–298.
Mumby P.J., Edwards A.J., Arias-Gonzalez J.E., Lindeman
K.C., Blackwell P.G., Gall A., Gorzynska M.I., Harborne
A.R., Pescod C.L., Renken H., Wabnitz C.C., Llewellyn G.
(2004) Mangroves enhance the biomss of coral reef fish
communities in the Caribbean. Nature,427, 533–536.
Neel M.C., McGarigal K., Cushman S.A. (2004) Behaviour of
class-level landscape metrics across gradients of class
aggregation and area. Landscape Ecology,19, 435–455.
Olds A.D., Connolly R.M., Pitt K.A., Maxwell P.S. (2012)
Habitat connectivity improves reserve performance.
Conservation Letters,5,56–63.
Orth R.J., Carruthers T.J.B., Dennison W.C., Duarte C.M.,
Fourqurean J.W., Heck K.L., Hughes A.R., Kendrick G.A.,
Kenworthy W.J., Olyarnik S., Short F.T., Waycott M.,
Williams S.L. (2006) A global crisis for seagrass ecosystems.
BioScience,56, 987–995.
Pardini R., Bueno A.D.A., Gardner T.A., Prado P.I., Metzger
J.P. (2010) Beyond the fragmentation threshold hypothesis:
regime shifts in biodiversity across fragmented landscapes.
PLoS One,5, e13666. doi:10.1371/journal.pone.0013666.
Pittman S., Kneib R.T., Simenstad C.A. (2011) Practicing
coastal seascape ecology. Marine Ecology Progress Series,427,
187–190.
P€
utz S., Groeneveld J., Alves L.F., Metzger J.P., Huth A. (2011)
Fragmentation drives tropical forest fragments to early
successional states: a modelling study for Brazilian Atlantic
forests. Ecological Modelling,222, 1986–1997.
Robbins B.D., Bell S.S. (1994) Seagrass landscapes: a terrestrial
approach to the marine subtidal environment. Trends in
Ecology & Evolution,9, 301–304.
Salita J.T., Ekau W., Saint-Paul U. (2003) Field evidence on
the influence of seagrass landscapes on fish abundance in
Bolinao, Northern Philippines. Marine Ecology Progress
Series,247, 183–195.
Santos R.O., Lirman D., Serafy J.E. (2011) Quantifying
freshwater-induced fragmentation of submerged aquatic
vegetation communities using a multi-scale landscape
ecology approach. Marine Ecology Progress Series,427, 233–
246.
Sheppard C.R.C., Matheson K., Bythell J.C., Murphy P., Myers
C.B., Blake B. (1995) Habitat mapping in the Caribbean for
management and conservation: use and assessment of aerial
photography. Aquatic Conservation: Marine and Freshwater
Ecosystems,5, 277–298.
Short F.T., Polidoro B., Livingstone S.R., Carpenter K.E.,
Bandeira S., Bujang J.S., Calumpong H.P., Carruthers T.J.B.,
Coles R.G., Dennison W.C., Erftemeijer P.L.A., Fortes M.D.,
Freeman A.S., Jagtap T.G., Kamal A.H.M., Kendrick G.A.,
Kenworthy W.J., La Nafie Y.A., Nasution I.M., Orth R.J.,
Prathep A., Sanciangco J.C., van Tussenbroek B., Vergara
S.G., Waycott M., Zieman J.C. (2011) Extinction risk
assessment of the world’s seagrass species. Biological
Conservation,144, 1961–1971.
Sintes T., Marba N., Duarte C., Kendrick G. (2005) Nonlinear
process in seagrass colonization explained by simple clonal
growth rules. Oikos,108, 165–175.
Sleeman J.C., Kendrick G.A., Boggs G.S., Hegge B.J. (2005)
Measuring fragmentation of seagrass landscape: which
indices are most appropriate for detecting change? Marine
and Freshwater Research,56, 851–864.
Smith A.C., Koper N., Francis C.M., Fahrig L. (2009)
Confronting collinearity: comparing methods for
disentangling the effects of habitat loss and fragmentation.
Landscape Ecology,24, 1271–1285.
Stalker J.C., Price R.M., Swart P.K. (2009) Determining spatial
and temporal inputs of freshwater including submarine
groundwater discharge, to a subtropical estuary using
geochemical tracers, Biscayne Bay, South Florida. Estuaries
and Coasts,32, 694–708.
Trzcinski M.K., Fahrig L., Merriam G. (1999) Independent
effects of forest cover and fragmentation on the distribution
of forest breeding birds. Ecological Applications,9, 586–593.
Vidondo B., Duarte C.M., Middelboe A.L., Stefansen K.,
L€
utzen T., Nielsen S.L. (1997) Dynamics of a landscape
mosaic: size and age distribution, growth and demography
of seagrass Cymodocea nodosa patches. Marine Ecology
Progress Series,158, 131–138.
Wang J.D., Luo J., Ault J. (2003) Flows, salinity, and some
implications on larval transport in South Biscayne Bay,
Florida. Bulletin of Marine Science,72, 695–723.
14 Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH
Spatial-temporal dynamics in SAV seascapes Santos, Lirman & Pittman
Waycott M., Duarte C.M., Carruthers T.J.B., Orth R.J.,
Dennison W.C., Olyarnik S., Callandie A., Fourqurean J.W.,
Heck K.L., Hughes A.R., Kendrick G.A., Kenworthy W.J.,
Short F.T., Susan L.W. (2009) Accelerating loss of seagrasses
across the globe threatens coastal ecosystems. Proceedings of
the National Academy of Sciences of the United States of
America,106, 12377–12381.
Wedding L.M., Lepczyk C.A., Pittman S.J., Friedlander A.M.,
Jorgensen S. (2011) Quantifying seascape structure:
extending terrestrial spatial pattern metrics to the marine
realm. Marine Ecology Progress Series,427, 219–232.
With K.A., Crist T.O. (1995) Critical thresholds in species’
responses to landscape structure. Ecology,76, 2446–2459.
Wu J. (2004) Effects of changing scale on landscape pattern
analysis: scaling relations. Landscape Ecology,19,
125–138.
Wu J., Shen W., Sun W., Tueller P.T. (2002) Empirical
patterns of the effects of changing scale on landscape
metrics. Landscape Ecology,17, 761–782.
Zharikov Y., Skilleter G.A., Loneragan N.R., Tarranto T.,
Cameron B.E. (2005) Mapping and characterizing
subtropical estuarine landscapes using aerial photography
and GIS for potential application in wildlife conservation
and management. Biological Conservation,125,87–100.
Zieman J.C., Wood E.J.F. (1975) Effects of thermal pollution
on tropical-type estuaries, with emphasis on Biscayne Bay,
Florida. In: Johannes R.E., Wood E.J.F. (Eds), Tropical
Marine Pollution 12. Elsevier Oceanography Series,
Amsterdam: 75–98.
Zieman J.C., Fourqurean J.W., Iverson R.L. (1999) Seagrass
dieoff in Florida Bay: long-term trends in abundance and
growth of turtle grass, Thalassia testudinum.Estuaries,22,
460–470.
Supporting Information
Additional Supporting Information may be found in the
online version of this article:
Appendix S1. Description of the aerial photographs.
Appendix S2. PCA and Vector Analysis Illustration.
Appendix S3. Site Classification Based on PCA.
Marine Ecology (2015) 1–15 ª2015 Blackwell Verlag GmbH 15
Santos, Lirman & Pittman Spatial-temporal dynamics in SAV seascapes
