Content uploaded by Simon James Pittman
Author content
All content in this area was uploaded by Simon James Pittman
Content may be subject to copyright.
Chapter 14
A Landscape Ecology Approach for the Study
of Ecological Connectivity Across Tropical
Marine Seascapes
Rikki Grober-Dunsmore, Simon J. Pittman, Chris Caldow,
Matthew S. Kendall and Thomas K. Frazer
Abstract Connectivity across the seascape is expected to have profound conse-
quences for the behavior, growth, survival, and spatial distribution of marine species.
A landscape ecology approach offers great utility for studying ecological connectiv-
ity in tropical marine seascapes. Landscape ecology provides a well developed con-
ceptual and operational framework for addressing complex multi-scale questions
regarding the influence of spatial patterning on ecological processes. Landscape
ecology can provide quantitative and spatially explicit information at scales relevant
to resource management decision making. It will allow us to begin asking key ques-
tions such as ‘how much habitat to protect?’, ‘What type of habitat to protect?’, and
‘Which seascape patterns provide optimal, suboptimal, or dysfunctional connectiv-
ity for mobile marine organisms?’. While landscape ecology is increasingly being
applied to tropical marine seascapes, few studies have dealt explicitly with the issue
of connectivity. Herein, we examine the application of landscape ecology to better
understand ecological connectivity in tropical marine ecosystems by: (1) review-
ing landscape ecology concepts, (2) discussing the landscape ecology methods and
tools available for evaluating connectivity, (3) examining data needs and obstacles,
(4) reviewing lessons learned from terrestrial landscape ecology and from coral reef
ecology studies, and (5) discussing the implications of ecological connectivity for
resource management. Several recent studies conducted in coral reef ecosystems
demonstrate the powerful utility of landscape ecology approaches for improving
our understanding of ecological connectivity and applying results to make more
informed decisions for conservation planning.
Keywords Seascape ecology ·Landscape ecology ·Connectivity ·Spatial scale ·
Pattern metrics ·Fish
R. Grober-Dunsmore (B)
Institute of Applied Sciences, Private Bag, Laucala Campus,
University of South Pacific, Suva, Fiji Islands
e-mail: rikkidunsmore@gmail.com; dunsmore l@usp.ac.fj
493
I. Nagelkerken (ed.), Ecological Connectivity among Tropical Coastal Ecosystems,
DOI 10.1007/978-90-481-2406-0 14, C
Springer Science+Business Media B.V. 2009
494 R. Grober-Dunsmore et al.
14.1 Conceptual Framework
Tropical marine ecosystems often exist as dynamic and spatially heterogeneous
seascapes in which different habitat types (e.g., coral reef, seagrass, open water,
mangrove, sand) are connected to one another by a variety of biological, physical,
and chemical processes (Fig. 14.1). Water movements, including tides and currents,
facilitate the exchange of nutrients, chemical pollutants, pathogens, sediments, and
organisms among components of the seascape. The active movement of organisms
also connects habitat patches across the seascape (Sale 2002, Gillanders et al. 2003).
For example, many tropical marine species exhibit complex life histories that utilize
resources from spatially and compositionally discrete habitat patches (Parrish 1989,
Pittman and McAlpine 2003). Highly mobile species can connect patches through
daily foraging movements, including tidal and diel migrations, as well as, broader
scale excursions for spawning and seasonal migrations (Zeller 1998, Kramer and
Chapman 1999; see Chapters 4 and 8). Furthermore, many species of fish and crus-
taceans exhibit distinct shifts in habitat through ontogeny (Dahlgren and Eggleston
2000, Nagelkerken and van der Velde 2002). The ability of an organism to suc-
cessfully navigate among several (often critical) ‘ontogenetic stepping stones’ or to
move successfully to spawning locations will likely be influenced by both the com-
position of the seascape (i.e., the patch type and the abundance and richness of patch
types) and the spatial configuration or spatial arrangement of patches (e.g., distance
Benthic habitat map classes
Colonized pavement
Linear reef
Patch reef
Sand
Seagrasses
Muddy sand
0 0.5 km
67°4′0″ W
17°58′0″ N
67°2′0″ W
Fig. 14.1 IKONOS image with marine portions classified into six benthic habitat types for the La
Parguera coast of Puerto Rico
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 495
to suitable patches, juxtaposition of complementary resources). The composition
and configuration of the seascape encompasses many quantifiable structural fea-
tures that are likely to influence ecological connectivity, with some configurations
providing better connectivity for a species (or assemblage) than others (Mumby
2006, Grober-Dunsmore et al. 2007, Pittman et al. 2007b).
Improving our understanding of ecological connectivity in tropical marine
ecosystems is one of the most pressing needs of resource managers and decision-
makers today. For example, optimally-connected seascapes for specific species can
be identified and mapped providing valuable spatially explicit information in sup-
port of resource management activities such as the design of Marine Protected Areas
(MPAs). In addition, such information can also contribute to the design of optimally-
connected habitat-restoration projects. At present, we have little knowledge of the
behavior of tropical marine organisms at spatial and temporal scales relevant to
their key life-cycle movements. Consequently, we remain largely ignorant of the
spatial and temporal patterns of ecological connectivity that are likely to exist in
marine environments. This greatly inhibits our ability to understand the influence of
seascape patterning on connectivity.
Several important research questions emerge from this conceptual framework
which can and must be asked of marine environments: (1) ‘How does the spatial
patterning of the seascape influence connectivity?’, (2) ‘What are the factors that
inhibit or facilitate exchange or flows of materials and energy among spatial ele-
ments of the seascape?’, (3) ‘How does loss of habitat or change in habitat con-
figuration alter connectivity and thus change the functioning of the seascape?’, and
(4) ‘What and where are the optimally connected seascapes’?
Recent technological advances in remote sensing, acoustic telemetry, Geograph-
ical Information Systems (GIS), and spatial statistics now allow us to capture, man-
age, and analyze the data needed for connectivity studies in a spatially explicit way
and at appropriately broad spatial scales (Crooks and Sanjayan 2006). Integrating
spatial technologies with the discipline of landscape ecology provides both the oper-
ational and conceptual frameworks necessary to tackle these complex ecological
problems at multiple spatial scales. Landscape ecology is a discipline that deals
with environmental complexity including spatial heterogeneity and the importance
of scale (Wu 2006) and has demonstrated great utility in the examination of ecolog-
ical connectivity in terrestrial environments (With et al. 1997, Crooks and Sanjayan
2006). An extensive suite of concepts, terminology, and analytical tools for under-
standing the linkages between spatial patterning of land surfaces and ecological
processes have recently been developed (Turner 2005). Landscape ecologists have
shown that a spatially explicit and quantitative examination of fluxes in heteroge-
neous systems (e.g., Turner 1989) is the key to improving our understanding of how
the physical structure and temporal dynamics of spatial mosaics influence ecological
connectivity (Crooks and Sanjayan 2006).
Many tropical marine organisms, particularly fish and crustaceans, exhibit a
strong linkage with benthic structure. For this reason, landscape ecology has
been strongly and increasingly advocated as an ecologically meaningful approach
for examining species-environment relationships in a wide range of structured
496 R. Grober-Dunsmore et al.
shallow-water marine habitat types (Robbins and Bell 1994, Irlandi et al. 1995,
Pittman et al. 2004, Grober-Dunsmore 2005, Grober-Dunsmore et al. 2008). We
argue here that the application of a landscape ecology approach to the highly hetero-
geneous structures that typify coral reef ecosystems will assist in understanding the
interactions between movement behavior and the spatial patterning of the seascape.
Ultimately, this should lead to more ecologically-meaningful decision making in
resource management.
This chapter examines the value of applying a landscape ecology approach when
attempting to examine ecological connectivity in tropical marine ecosystems by:
(1) presenting a landscape ecology conceptual framework for understanding con-
nectivity, (2) reviewing existing landscape ecology methods and tools for evaluat-
ing connectivity, (3) discussing data needs and limitations, (4) reviewing lessons
learned from terrestrial landscape and coral reef ecology studies, and (5) discussing
the many implications for resource management. The focus here is on highly mobile
species, with particular emphasis on marine fish, but we also draw on examples
from terrestrial systems to highlight some similarities and differences in apply-
ing such an approach in marine systems. We do not address marine connectivity
from a metapopulation (Hanski 1998) or genetic perspective (Cowen et al. 2006),
although these approaches can also be spatially explicit and overlap in techniques
and terminology does sometimes occur. Furthermore, we will not deal with the many
approaches for studying larval connectivity. Instead, emphasis is placed on connec-
tivity associated with active movement of individuals across the benthic seascape.
These techniques, however, may also be applicable to studies of nutrient fluxes or
other exchanges of materials across the seascape.
14.1.1 Definitions and Concepts
14.1.1.1 Some Commonly Used Landscape Ecology and GIS Terms
Various concepts, terminology, structural relationships, and analytical techniques
used in terrestrial landscape ecology are also appropriate for studying ecological
patterns and processes in shallow-water benthic seascapes (Carleton Ray 1991,
Robbins and Bell 1994, Table 14.1). In landscape ecology terminology, patches are
the basic spatial element in the landscape, and have been defined simply as a rela-
tively homogeneous nonlinear area that differs from its surroundings (Forman and
Godron 1986). In coral reef ecosystems, a wide variety of patch types have been
classified (e.g., linear reef, patch reef, seagrass, sand, or some more biologically spe-
cific class such as gorgonian-dominated hard-bottom; Mumby and Harborne 1999,
Kendall et al. 2002). Patch types differ from habitat types, in that habitat types do not
infer any structural boundaries, although these terms are often used interchangeably.
The structural boundaries of patches are referred to as edges or ecotones, each of
which can be represented as a sharp boundary or a gradual transition from one struc-
tural type or community to another. A patch or an aggregation of patches can form
a corridor, which is a linear feature that differs from its surroundings and connects
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 497
Table 14.1 Definitions for major concepts of landscape ecology, with examples of application to
coral reef ecosystems (adapted from Forman 1995)
Concept Definition Coral reef example
Matrix The dominant element in a landscape Sand or seagrass
Patches Patches are the basic spatial element
in the landscape
Reef patch
Mosaic A combination of different patch types
that are usually interspersed
amongst one another
Patches of seagrass, reef, sand,
mangrove
Seascape A heterogeneous marine area that can
exist at a wide range of scales and
may be described as a mosaic
pattern or spatial gradient
The home range of a fish is an
ecologically meaningful seascape
Seascape
structure
The composition and spatial
arrangement of patches, but may
also include bathymetric complexity
or structure in the water column
The distribution, diversity, and spatial
geometry of structure at relevant
spatial scales
Patch context The position of a patch relative to
surrounding seascape elements
A patch can be surrounded by seagrass
or sand habitat
Heterogeneity The uneven, non-random distribution
of objects
Distribution of habitat patches that
comprise a reef
Seascape
connectivity
The degree to which the seascape
facilitates or impedes movement
among resource patches
Cross-shelf movement of grunts
through ontogeny
Structural
connectivity
Physical linkages within a seascape A map of a reef area portrays
structural connectivity
Potential
connectivity
Measure of connectivity that
incorporates indirect and limited
information on mobility of
organism
Extrapolating vagility of all jacks
based on information on one species
Actual
connectivity
Measure of connectivity that quantifies
the movement of individuals
through a habitat or landscape
Spatial information from acoustic
tracking of fish, conchs, or lobsters
Functional
connectivity
How the structure of the seascape
interacts with the properties of the
organisms, disturbances, or
materials to influence how they
move
How the spatial arrangement of
seagrass beds influences the
movement of grunts
Stepping stone
connectivity
A row of small patches (stepping
stones) can connect an otherwise
disconnected set of patches
Seagrass patch connecting reef
patches in sand matrix
Spatial scale/
Temporal
scale
A measure of the resolution or extent
perceived or considered
Depends on question asked and may
be selected using species home
range or other ecological processes
Extent The size of the study area or the
duration of time under consideration
Theareaofinterest
Grain The finest level of spatial or temporal
resolution possible within a given
data set
The smallest unit or minimum
mapping unit (e.g., a 1 m2patch
reef)
498 R. Grober-Dunsmore et al.
patches (Forman and Godron 1986). More broadly, an aggregation of patches may
form a mosaic or habitat mosaic. The most abundant and well-connected compo-
nent of the landscape or seascape is sometimes referred to as the matrix (Forman
and Godron 1986). These elements of landscapes are usually arbitrarily defined and
typically determined by the observer and depend upon the perspective, scale, and
question of interest (Wiens et al. 1993).
The seascape can also be considered as a spatial unit or sampling unit (i.e.,
seascape unit) within a GIS, within which seascape structure can be quantified
and characterized into two categories of structure: (i) composition, and (ii) con-
figuration. Essentially, landscape composition, also referred to as marine landscape
composition (Grober-Dunsmore et al. 2004, Pittman et al. 2004) and subsequently
seascape composition (sensu Pittman et al. 2007b), encompasses the variety and
abundance of patch types, whereas landscape or seascape configuration (also
referred to as spatial arrangement) is the physical distribution of patches in space
(Dunning et al. 1992, Pittman et al. 2004).
GIS are routinely used in quantitative landscape ecology. A GIS is an organized
collection of specific computer hardware, software, geographic data, and personnel
designed to efficiently capture, store, update, manipulate, analyze, and display all
forms of geographically referenced information. The two major types of internal
data organization used in GIS are raster (grid) and vector. Raster systems superim-
pose a regular grid over the area of interest and associate each cell or pixel with
one or more data records (Malczewski 1999). Vector systems are based primarily
on coordinate geometry and take advantage of the convenient division of spatial
data into point, line, and polygon types. GIS are frequently used to investigate ques-
tions regarding ecological connectivity including the application of algorithms or
methods for quantifying spatial pattern from habitat maps.
14.1.1.2 What is Connectivity?
The term connectivity appears in diverse contexts in the ecological science literature,
sometimes with much ambiguity, in both terrestrial and marine research. To land-
scape ecologists, it often refers to the interactive pathways that link organisms and
ecological processes with landscape elements (Crooks and Sanjayan 2006). Changes
in composition and configuration are capable of altering the physical connectiv-
ity of landscapes (see Section 14.2). Each species’ unique biological and behav-
ioral characteristics interact with the physical landscape structure to determine the
functional connectivity of a particular landscape. To understand this complexity,
connectivity is often described and quantified in three ways: (1) structural, (2) poten-
tial, and (3) actual connectivity (Calabrese and Fagan 2004, Fagan and Calabrese
2006). Structural connectivity usually refers to spatial characteristics of the phys-
ical structure of the environment. It is the type of connectivity that one envisions
when examining a map, and is typically measured by quantifying the configura-
tion of a landscape with limited reference to the movement of organisms, materials,
or energy (Crooks and Sanjayan 2006; Fig. 14.2). Potential connectivity considers
some limited, albeit, indirect information on the dispersal or movement ability of
the organism or process of interest (Fagan and Calabrese 2006). Actual connectivity
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 499
seagrass
seagrass
seagrass
reef
reef
reef
reef reef
(a) High
connectivity
(c) Low
connectivity
(b) Medium
connectivity
Optimal foraging,
refuge, and ontogenetic
transitioning between
patch types
Optimal foraging only
for adults, and with high
mortality for
transitioning juveniles
Sub-optimal for both
juveniles and adults due
to energetic constraints
and predation risk
reef
Fig. 14.2 Schematic representation of different seascape structure for a hypothetical species that
requires multiple patch types in close proximity. (a) Highest or optimal connectivity occurs where
all three essential resources exist in close proximity and even juveniles are able to easily traverse
between patches. Seascapes (b)and(c) represent seascapes with sub-optimal configuration for our
species with only adults able to traverse between seagrass and coral reefs due to greater distances
of travel required over unsuitable low structure patches (i.e., sand) to reach essential resources.
Dispersal ability and movement patterns of organisms will influence the degree to which these
seascapes are connected
quantifies the movement of individuals through a habitat or seascape (e.g., acous-
tic tracking), thereby providing a direct measure of the potential linkages that may
exist among habitat patches or seascape elements. The latter two types of connec-
tivity are synonymous with functional connectivity, i.e., measures which examine
how the structure of the landscape interacts with ecological processes such as dis-
turbances or the movement of organisms and other materials across the seascape
(Wiens 2006).
14.1.1.3 The Importance of Spatial Scale
The relationship between scale and pattern is considered one of the most important
issues in ecology (Levin 1992, Schneider 2001). Our perception and measurement
of pattern is determined by scale selection, and species respond to pattern individual-
istically at a range of spatial scales (Wiens and Milne 1989; Fig. 14.3). In landscape
ecology, the concept of scale has two subcomponents: grain and extent (Forman and
Godron 1986, Turner 1989). Spatial grain refers to the size of the sample unit area.
Spatial extent is the overall area encompassed by an investigation. Therefore, extent
500 R. Grober-Dunsmore et al.
m2km2
10 m2km2× 103
km2× 102
km2× 10
goby
damselfish
butterflyfish
reef crab
conch
parrotfish
lobster
experimental
jack
ecologist
MPA
manager
fishery
manager
Fig. 14.3 Diagrammatic representation of the scaling windows or domains of various organisms
occupying tropical marine ecosystems, and of the humans who manage or conduct research on
these resources across spatial scales (adapted from Wiens et al. 2002)
and grain define the upper and lower limits of resolution of a study and constrain
any inferences about the scale-dependency of ecological phenomena (Wiens 1989).
Spatial resolution is an alternative term for spatial grain that is usually indicated by
pixel or cell size or the minimum mapping unit (MMU). The thematic resolution,
or the level of habitat classification, can also differ with various maps. Maps delin-
eate habitat classes or patch types, but the level or detail of information can vary.
Often a hierarchical habitat classification scheme is created to define and delineate
habitat maps. A hierarchical scheme allows users to collapse or expand the level
of detail depending upon their specific needs. For instance, at the lowest level of
thematic resolution, a map may indicate a patch as soft bottom. At a finer resolu-
tion, the same patch would be classified as seagrass, and at an even finer thematic
resolution, the same patch may have information on the species, relative height and
density of seagrasses. Varying spatial resolution and thematic resolution of data will
have important consequences for studies of the influence of patterns in the seascape
(Kendall and Miller 2008).
14.1.2 The Emergence of Landscape Ecology in Tropical
Marine Ecology
The importance of seascape composition and configuration has long been recog-
nized as an important structuring mechanism for coral reef fishes. Early studies
highlighted the importance of ecological interactions between adjacent habitat types
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 501
in a mosaic of patch types (i.e., seagrasses, coral reefs, mangroves) (Gladfelter et al.
1980, Ogden and Gladfelter 1983, Birkeland 1985, Parrish 1989). More recently, the
importance of the location of various patch types such as mangroves, seagrasses, and
coral reefs to fish species and communities was demonstrated (Nagelkerken et al.
2000a, b, Nagelkerken and van der Velde 2002, Dorenbosch et al. 2005, 2006a).
Furthermore, Dorenbosch et al. (2004a) highlighted the importance of proximity
between adjacent habitat types (seagrasses and coral reefs) on faunal abundance
and diversity. Increasingly, the movements of fishes are being correlated with the
presence of specific habitats (e.g., Verweij et al. 2007).
While these studies highlighted the ecological importance of spatial patterning
of patches and linkages between patches, they are not considered landscape ecol-
ogy studies. Spatial patterning was not explicitly quantified and incorporated as an
explanatory variable in these studies (i.e., surveys were conducted without a quan-
titative spatial context). In addition, they were conducted without the adoption of
a landscape ecology conceptual and operational framework. Several early studies
adopted an island biogeography perspective (MacArthur and Wilson1967) to exam-
ine the importance of patch size and the spatial arrangement of patches (i.e., patch
isolation) using artificial units simulating seagrasses or patch reefs (Molles 1978,
Bohnsack et al. 1994) at relatively fine spatial scales (1−10 m2), though they also
lacked a consideration of the seascape context.
Further recognition of the importance of scale together with spatial technolo-
gies such as GIS and an increased availability of benthic habitat maps allows us
to quantify seascape patterning at multiple spatial scales. Investigators have used
this information to examine the influence of seascape patterning on species dis-
tributions and assemblage richness, biomass, and abundance (Turner et al. 1999,
Kendall et al. 2003, Pittman et al. 2004, 2007a, b, Grober-Dunsmore et al. 2007,
2008; Table 14.2). These studies have successfully used spatial information (fea-
tures and habitat classes) that exist in digital benthic habitat maps as explanatory
variables. Yet, very few marine applications of landscape ecology have targeted the
subject of connectivity directly. Instead, connectivity is increasingly being investi-
gated through examination of the spatial patterns of species abundance, size class,
and movement data (i.e., acoustic telemetry). While these approaches are useful,
such studies often neglect to incorporate and quantify patterns in the seascape struc-
ture. Without spatially explicit information on seascape structure, studies will be
missing a suite of potentially important explanatory variables and results will have
limited application to resource management.
14.2 Operational Framework: Designing a Landscape
Ecology Study
14.2.1 Scale Selection
It is essential to define connectivity from the perspective of the organism, species,
or process of interest (e.g., terrestrial studies: Wiens and Milne 1989, With et al.
1997; marine studies: Pittman and McAlpine 2003, Pittman et al. 2004, 2007b,
502 R. Grober-Dunsmore et al.
Table 14. 2 Examples from coral reef ecosystems of studies that may provide insights into the importance of specific seascape features on connectivity for
mobile marine organisms. Major findings by seascape feature and potential general principles that may be useful for understanding ecological connectivity in
coral reef ecosystems are offered
Landscape feature Major findings Reference
General principle for coral reef
management
COMPOSITION AND ARRANGEMENT
Patch size and
distribution
Patch size and shape influences reef fish
distribution
Grober-Dunsmore et al. (2004), Lugendo
et al. (2007a), Ault and Johnson (1998)
Habitat configuration influences reef fish
distribution, colonization
Contribution of mangrove habitat as a
feeding source may depend on habitat
configuration
Areas with similar configurations may
function similarly
Species diversity differed in isolated and
continuous reef patches
Habitat composition Presence of juvenile grunts related to area
of soft bottom
Kendall et al. (2003), Grober-Dunsmore
et al. (2007, 2008), Pittman et al.
(2004), Chittaro et al. (2004),
Appeldoorn et al. (2003)
Landscape composition influences presence
of reef fishes
Total abundance, species richness, and
species presence correlated to particular
habitats
Specific habitat associations for groups of
reef fishes, though dependence may not
be obligate
Certain species present only where
mangrove/seagrass occurs
Presence of specific
habitats
Assemblage structure differed with
presence of mangrove/seagrass
Nagelkerken et al. (2001, 2002),
Dorenbosch et al. (2006b),
Nagelkerken and van der Velde (2002)
Presence/absence of specific habitats may
influence fish assemblage structure and
fish density
Higher densities of reef fishes on reefs with
seagrass
Densities of nursery species higher where
nursery habitats present
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 503
Table 14. 2 (continued)
Landscape feature Major findings Reference
General principle for coral reef
management
Movement Few movements when reefs separated by
20 m of sand/rubble
Chapman and Kramer (2000), Tewfik and
Bene (2003), Grober-Dunsmore and
Bonito (2009)
Certain features/habitats may serve as
barriers to movement for some species;
spillover may be reduced by
configuration of habitats
Conch density reduced with occurrence of
sand
Movement of reef fishes reduced with
presence of specific seascape features
Edge Reef-associated species almost exclusively
on reef edge
Dorenbosch et al. (2005),
Grober-Dunsmore et al. (2004), Kramer
and Chapman (1999), Jelbart et al.
(2006)
Edges can affect ecological processes
including movement of species; high
area:edge ratio may help retain mobile
marine organisms
Piscivore abundance increased with
perimeter:area ratio
Density predicted to be higher in reserves
with high area:edge ratio
Species richness and density of fishes
affected by edge habitat
Proximity Species richness increased with proximity
to topographically complex areas
Density of nursery species higher adjacent
to seagrass/mangrove
Pittman et al. (2007b), Dorenbosch et al.
(2004a, 2005, 2007), Appeldoorn et al.
(2003), Nagelkerken and Faunce
(2007), Lugendo et al. (2007b)
Proximity to specific habitats may be
important for certain taxa; habitat
proximity may interact with degree of
connectivity; distance between juvenile
and adult habitats may influence
colonization
Similarity in community composition may
be a function of distance between habitats
Lutjanid and haemulid biomass greater near
mangrove/seagrass
Adult densities of nursery species higher on
reefs adjacent to seagrass/mangrove
Density and colonization related to distance
from reef
High species overlap in adjoining habitats
compared to those that were spatially
separated
504 R. Grober-Dunsmore et al.
Table 14. 2 (continued)
Landscape feature Major findings Reference
General principle for coral reef
management
Fragmentation Juvenile blue crab survival significantly
lower in patches separated by expanses
of unvegetated sediment than patches
separated by <1 m of unvegetated
sediment (connected patches)
Hovel and Lipscius (2002) Fragmentation and connectivity influences
population dynamics of blue crabs
Thresholds of
habitat availability
Abrupt decline in species richness and
density (fish and decapods) occurred at
20% seagrass cover
Pittman et al. (2004), Grober-Dunsmore
et al. (2008)
Habitat loss important driver of population
decline
Above 30%, reef fish diversity and density
not controlled by seagrass
At certain threshold, the relative influence
of particular habitats may change
Interactions of
seascape features
Fish assemblages differed between day and
night, likely due to nocturnal migrations
Nagelkerken et al. (2000b), Dorenbosch
et al. (2004b)
Importance of habitats may differ with
proximity/composition of other habitats
Tide-related movements of juvenile
snappers from low-tide shelter habitat to
high-tide shelter habitat (notches)
Movement of fishes between habitats may
be influenced by surrounding habitats
CONNECTIVITY
In mangrove, coral
reef, and seagrass
Biomass of some fishes doubled when adult
habitat connected to mangrove
Mumby et al. (2004), Sheaves (2005),
Mumby (2006), Nagelkerken et al.
(2002), Chittaro et al. (2004), Lugendo
et al. (2006)
Connectedness between mangrove and reef
increased biomass of particular species
Temporal availability of mangroves
influenced by tidal fluctuations
Position of patches relative to tidal flux can
influence connectivity
Reefs with greater connectivity to
mangroves may have increased
immigration and productivity
Algorithms can generate a connectivity
matrix to identify connected corridors of
habitat
Scarus guacamaia largely absent on reefs
isolated from mangroves. Habitat use
suggests movement to offshore reefs with
increased size
Connectivity of reef and mangrove habitat
beneficial to some species
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 505
Table 14. 2 (continued)
Landscape feature Major findings Reference
General principle for coral reef
management
Otolith microchemistry suggests grunts
pass through mangrove nursery before
moving to reef
Different habitats used as feeding grounds
(based on isotopes) by different fish
species
In back-reef High site fidelity to small spatial scales:
<171 m linear distribution range for
haemulids
Verweij and Nagelkerken (2007) Connectivity between back-reef habitats
through fish movement
In reef habitat Individuals of highly vagile species are able
to move among isolated patches in
response to habitat preferences or
resource availability. The continuous
shelter provided by contiguous reef may
allow sedentary species to migrate to
more favorable areas
Ault and Johnson (1998) Patterns in distribution and abundance
established at recruitment are modified
by post-settlement migration. Migration
may differ for isolated reef patches and
connected continuous reefs
In seagrass, reefs,
and saltmarsh
Movement (crabs) increased when seagrass
connected to reefs/marsh
Micheli and Peterson (1999), Verweij
and Nagelkerken (2007)
Predation and foraging movements
influenced by connectivity of matrix
Fishes (snappers) tagged in an embayment
resighted offshore, crossing a 115 m
open sand zone
Direct evidence of connectivity through
movement
506 R. Grober-Dunsmore et al.
Grober-Dunsmore et al. 2007, 2008). The same seascape will likely differ in func-
tional connectivity for different processes, species, and even for different life stages
of the same species. Natural history attributes of specific organisms (e.g., life history
strategy, mobility, dispersal, resource requirements, habitat generalist or special-
ist, behavioral attributes, etc.) must be considered as an important determinant of
the potential response to seascape structure (Pittman et al. 2004, Grober-Dunsmore
et al. 2008), though some generality may be expected to occur among taxa (Stamps
et al. 1987, Sisk et al. 1997, Mitchell et al. 2001). For example, a seagrass patch
boundary may function as a constraint to the scale and direction of movements for a
resident seagrass specialist, but may be relatively inconsequential to a more gener-
alist species (Pittman et al. 2004; Fig. 14.3). In addition, investigations of seascape
connectivity must recognize that for an individual species, connectivity may be an
integrated function of responses to structural characteristics or ecological processes
existing at multiple scales in time and space (Crooks and Sanjayan 2006).
When adopting an organism-based or organism-centered approach, the spatial
resolution of the maps or sample units and the extent of the study should be appro-
priate to the scales at which the organisms responds and utilizes its environment.
For instance, studies focused on understanding seascape connectivity throughout
the daily home range (i.e., routine foraging or territorial movements) would prob-
ably select different scales than studies focused on connectivity throughout the life
cycle (Pittman and McAlpine 2003). Identifying the appropriate scale(s) is essen-
tial for assessing connectivity, though it remains a significant challenge. Without
data at the appropriate scale(s), interpretation of results can be incorrect and mis-
leading due to scale dependencies. Ultimately, scales selected should always be
relevant to the questions being asked (Wiens and Milne 1989, Li and Wu 2004).
To obtain interpretable results, it is highly advisable to conduct pilot studies to
establish a reliable estimate of the relevant spatial and temporal scales. For faunal
studies, the use of acoustic tracking, tagging, and appropriately designed extrac-
tive sampling or visual census surveys can provide data suitable for selecting the
temporal and spatial extents of habitat use (Pittman and McAlpine 2003). Pre-
dictable patterns of behavior, such as migrations (diel, tidal, seasonal, and spawn-
ing), residence times within patches, and home range sizes, provide ecologically
meaningful, organism-based scales for connectivity investigations (e.g., Meyer et al.
2007).
Efforts to define the most influential spatial scales for fish–environment rela-
tionships have demonstrated complex, scale-dependent, and species-specific rela-
tionships (Kendall et al. 2003, Pittman et al. 2004, 2007b, Grober-Dunsmore et al.
2007, 2008; Table 14.2). These exploratory studies have linked fish and crustacean
distributions to variability in seascape structure quantified at multiple spatial scales
from classified benthic habitat maps. Scale issues have importance for the construc-
tion and application of benthic habitat maps. Both the spatial (cell size or minimum
mapping unit) and thematic resolution (level of detail in patch composition) can
affect results and the types of questions that can be asked (Kendall and Miller 2008).
Practical considerations such as data availability and the type of research methods
employed will also affect the scale selected.
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 507
Selecting the grain size and spatial extent is a crucial first consideration in any
investigation of ecological connectivity. In addition, the spatial extent (size of the
study area) and the temporal extent (maximum duration of time for a study or pro-
cess) are important attributes of scale to examine early in the planning stage since
they define the level of detail and set the time and space bounds for the study. These
decisions will impact all stages of an ecological study from budgeting, to data col-
lection and interpretation of results.
14.2.2 Use of Spatially Referenced Faunal Distribution Data
Several types of faunal distribution data can be used to apply landscape ecological
principles to the study of connectivity in coral reef ecosystems. Three approaches
are commonly used in marine ecology and are discussed here in order of increasing
strength of inference for examining seascape connectivity: (1) non-extractive survey
(e.g., visual observation) or extractive sampling (e.g., traps and enclosures, netting),
(2) tag recapture/resighting, and (3) hydro-acoustic telemetry. To be useful these
data types must be spatially explicit, that is they must have geographic coordinates
or other positional information that can be used to understand an organism’s location
with respect to the surrounding seascape.
The type of data selected for study will be largely driven by what aspects of
connectivity are of interest. Marine resource managers are increasingly interested
in understanding: (1) which combinations of habitat types, and more specifically,
which spatial configurations provide functional connectivity or even maximum con-
nectivity for a species or community metric (i.e., species richness)?, (2) the conse-
quences of habitat loss on functional connectivity; and (3) the actual pathways of
movement across the seascape for marine organisms in relation to protected area
boundaries.
It is important to be aware of the tradeoffs when choosing among data sources,
since certain questions can only be addressed with specific types of data and the
availability of information varies, as does the cost of acquiring the information.
Typically, structural connectivity metrics require less data input and are relatively
inexpensive compared to actual connectivity metrics (Fig. 14.4). Metrics of potential
and actual connectivity require considerable information specific to the organism of
study, which often limits the available data for study, when compared to structural
connectivity. Such tradeoffs must be considered when selecting the type of spatially
referenced distribution data for your study.
14.2.2.1 Observational Studies
In coral reef ecosystems, underwater visual counts of fish species (i.e., abundance,
body size, and species composition) (e.g., Brock 1954, Bohnsack and Bannerot
1986) are the most common data collected. Similar data types can also be provided
using extractive sampling via traps or nets, although both passive and active fishing
gears can be highly selective (Recksiek et al. 1991, Rozas and Minello 1998). These
508 R. Grober-Dunsmore et al.
Increasing data requirements
Increasing detail
Structural
Potential
Actual
Increasing sophistication and cost
Increasing data availability
Fig. 14.4 Schematic representation of the tradeoff between information content and data require-
ments among connectivity metrics and approaches (structural, potential, and actual). Both infor-
mation content and data requirements increase going from nearest neighbor to actual movement
rates, as does the level of detail for connectivity increases on the y-axis. Factors such as technologi-
cal sophistication, cost, and availability of information also influence the decision regarding which
approach and metric to use for studying connectivity (modified from Calabrese and Fagan 2004)
techniques are fieldwork-intensive and provide only a snap shot of the fish com-
munity at a given time and place, yet with an appropriate sampling design, driven
by specific questions regarding connectivity, spatial patterns in species abundance
patterns and size distributions can be examined in relation to seascape structure.
This linking of faunal pattern to environmental pattern often serves as an effective
and exploratory precursor to more detailed species–environment studies (Under-
wood et al. 2000). In coral reef ecosystems, several studies document correlative
relationships between underwater visual census data and proximity of adjacent habi-
tat types or the juxtapositioning of seascape elements (Nagelkerken et al. 2002,
Grober-Dunsmore et al. 2004, 2007, 2008, Dorenbosch et al. 2007, Jelbart et al.
2007, Pittman et al. 2007b, Vanderklift et al. 2007; Table 14.2). Slightly stronger
inference can be made regarding connectivity in the case of observational data on
species that undergo habitat-dependent ontogenetic shifts (i.e., different size classes
associated with different habitat types; Nagelkerken et al. 2000a, Christensen et al.
2003, Mumby et al. 2004; Table 14.2). For example, if juveniles of a species only
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 509
occur in seagrasses and adults only occur in coral reefs then it is likely, but not
certain, that juveniles shift from seagrass to coral reefs at some point in time.
Census effort along ecotones or habitat boundaries, if properly timed, in some
cases may provide strong inference of connectivity among seascape elements.
Observing the daily migrations of grunts into adjacent soft-bottom habitats (Ogden
and Ehrlich 1977, Helfman et al. 1982) provides direct evidence of the functional
connectivity of these habitat types, but does not clarify the spatial extent of the con-
nection. Similarly, nets and traps may be deployed along habitat boundaries and
used to infer movement between seascape features (and in the case of nets, direc-
tion of the movement) (Clark et al. 2005). This approach provides some important
information and in general these census techniques have the advantage of being rel-
atively low cost with few technological requirements and have contributed to the
identification of spatial variables that could potentially influence connectivity.
Yet, visual census and extractive sampling alone is rarely sufficient to piece
together the details of connectivity, such as pathways and responses to boundaries
across the seascape. Since the unit of study for these observational methods is at the
species level, rather than the individual level, scientists can only identify species–
habitat associations, which provide only indirect evidence of connectivity. Yet stud-
ies without spatially explicit data can still address important questions on habitat-use
patterns. If information on size and age class has been collected, insights into why
a particular habitat is important can also be inferred. If juveniles are observed in
one habitat and adults in another, one can infer that the two habitat types are con-
nected; however, direct observation of movement is necessary to confirm that these
specific habitat patches are connected. With careful experimental design (by select-
ing varying patches sizes), observational studies can also address questions such
as: ‘How much habitat is needed for a species to occur?’ and ‘How close should
certain patch types or resources be to optimize habitat use?’ These observational
studies do not confirm connectivity, but can provide a valuable initial springboard
for further studies on connectivity and lead to construction of more specific and
testable hypotheses. Furthermore, sufficient information can be gathered to develop
map products and empirical models of connectivity that support decision-making in
resource management (Mumby 2006).
14.2.2.2 Tagging Studies
Mark–recapture/resighting techniques (e.g., subcutaneous dyes, plastic wire, fin
clips) have also been used effectively to examine habitat use in coral reef ecosys-
tems (e.g., Zeller and Russ 1998; Chapter 13). When analyzed using a seascape con-
text, these techniques can provide direct evidence of connectivity (Fig. 14.5), while
remaining relatively inexpensive and requiring materials that are widely available.
Although potentially more powerful than census data in determining connectivity
between landscape elements, the approach still lacks the capability to answer many
key questions regarding functional connectivity. For example, a marked or tagged
fish released in habitat Aand subsequently resighted in habitat Bis obvious evidence
of a connection, yet as with census methods, the timing of movements, exact route
510 R. Grober-Dunsmore et al.
Fig. 14.5 Map of potential connectivity of inshore and offshore reef habitat based on mark–
recapture and visual census studies of Lethrinidae in Vitu Levu, Fiji. Key areas for juveniles, adults,
and spawning aggregations were identified. Small-sized numbers indicate the locations of acoustic
receiving stations distributed inside and outside the MPA (MPA boundaries are depicted by white
lines). Arrows indicate likely areas of movement and connectivity for different life history stages.
Adult fishes move across the forereef to spawn, from the continuous back reef areas. Connectiv-
ity within the continuous back reef habitat occurs for juveniles and adults, however, deep water
channels may serve as a barrier to movement, limiting connectivity among alongshore MPAs
followed, and any intermediary stops along the way are usually not known. Knowing
the trajectory of movement and the response of organisms to particular habitat fea-
tures will be particularly important for considering the flow of energy and materials
among elements of the seascape.
14.2.2.3 Telemetry Studies
The most revealing and sophisticated techniques for addressing connectivity across
the seascape are those that provide spatially continuous, temporally referenced
movement data for individual animals or other mobile components. Hydroacous-
tic telemetry (Holland et al. 1993, Meyer et al. 2000, Starr et al. 2007) and other
tracking techniques, such as close observation of tagged fish (Burke 1995), col-
lect continuous positional data as fish move across the seascape. Tracking can
either be conducted manually (e.g., Beets et al. 2003), in real time with directional
hydrophones (i.e., following the fish by boat with an acoustic receiver), or auto-
mated with an array of fixed receivers (see Chapter 13). Mapping these movements
and their timing provides a wealth of information at spatial and temporal resolution
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 511
Fig. 14.6 Map of actual connectivity of one individual lethrinid inside and outside marine protected
areas in Vitu Levu, Fiji. MPA boundaries are depicted by white lines. Numbers represent acoustic
receiving stations (those not shaded are locations where the fish was not detected). Shaded numbers
indicate locations where the fish was detected. The size of the number is scaled to represent the
number of detections: a larger number indicates a greater number of detections. This fish (tagged
at station seven, indicated in white shading) moved freely across the MPA boundary within the
continuous back reef habitat patch, suggesting that minor modifications in the boundaries of the
MPA may include the daily home range movements of this fish
not available using non-tracking approaches (e.g., Chateau and Wantiez 2007). Con-
nections can be plotted among habitat patches and the timing of transitions can be
identified. Exact pathways of travel and obstacles to dispersal can be identified by
overlaying animal tracks onto benthic maps (Fig. 14.6). The principle disadvantage
of these techniques is that they require an initially high investment in cost, technol-
ogy, and field set-up (Fig. 14.4). Although many telemetry studies have now been
conducted (e.g., Meyer et al. 2007, Starr et al. 2007), few researchers are linking
movement data to spatially explicit information on seascape features (e.g., channels,
patch edges) (Grober-Dunsmore and Bonito 2009). Much could be learned through
reinterpretation of existing tracking data with seascape structure that is typically
represented in benthic habitat maps (Pittman and McAlpine 2003).
14.2.3 Analytical Tools for Examining Seascape Connectivity
Many analytical tools are available to assist our understanding of the structural and
functional connectivity of seascapes and the consequences on species distribution
and behavior. Often adapted from engineering and systems analysis, these methods
512 R. Grober-Dunsmore et al.
have been successfully applied to examine ecological connectivity in both terres-
trial and marine landscapes. Here we describe three types of spatial analytical tools
of particular relevance to seascape analyses: (1) spatial pattern metrics, (2) graph-
theoretic approaches, and (3) computer simulation models. These tools address
three types of connectivity: (1) structural, (2) potential, and (3) actual connectivity
(Calabrese and Fagan 2004). Data requirements and complexity increase from type
1–3 often requiring more site- and species-specific information to address actual
connectivity, but so too does ecological realism and therefore explanatory perfor-
mance (Calabrese and Fagan 2004; Fig. 14.4). In general, structural connectivity
is more easily visualized and measured than functional connectivity; however, it
generally ignores the behavioral response of organisms to the landscape. Most spa-
tial pattern metrics are useful for quantifying seascape structure and can be used to
help explain the influence of seascape composition and configuration (Table 14.2)
including structural connectivity. Structural connectivity metrics can be used to
examine the relationship between seascape structure and species distributions and to
determine whether differences in seascape structure matter. For instance, ‘Does the
proximity or juxtaposition of complementary resources in two discrete patch types
influence species distributions, growth, and movement?’ (e.g., Irlandi and Crawford
1997). However, before such information could be used to design movement corri-
dors or predict dispersal pathways, more information would be required to provide
information on the spatial processes relevant to functional connectivity. It is impor-
tant to realize that seascapes that are structurally connected may not necessarily be
functionally connected for all species.
In contrast, functional connectivity metrics incorporate various levels of move-
ment information; therefore their use broadens the types of questions that can be
answered to include ecological processes. Potential and actual connectivity met-
rics define seascape structure using indirect and direct knowledge of an organism’s
dispersal ability or behavior. Potential connectivity metrics can be parameterized
using estimates of mobility derived from body size or trophic guild (Kramer and
Chapman 1999), or measurements with limited spatial detail, such as mean or max-
imum recapture distances from mark–capture studies. Potential connectivity can be
used to address questions such as, ‘Is there a threshold of habitat below which a
landscape is fragmented?’ (Table 14.2) or ‘How will dispersal pathways be affected
by degradation or removal of certain habitat patches?’ Potential connectivity metrics
are capable of addressing many more resource management questions than struc-
tural metrics and are relatively cost-effective compared to the more data-intensive
actual connectivity metrics.
Metrics of actual connectivity directly link individual movement data to spatially
explicit patterns of landscape structure and are useful for modeling population
dynamics (immigration, colonization, dispersal) in response to landscape features
(Rothley and Rae 2005). They can also predict dispersal pathways, be used to
design networks of reserves, and assess the flexibility in habitat requirements of
certain organisms. Numerous methods exist to provide the most direct estimate of
actual connectivity, though these are generally costly and labor-intensive (Fagan
and Calabrese 2006). Acoustic tracking of the precise movement pathways of indi-
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 513
vidual animals is the most direct measure (Fagan and Calabrese 2006). However,
radiotracking has been used to provide critical long-distance dispersal information
(Gillis and Krebs 2000), mark–recapture can be used to compare dispersal abilities
in different landscapes (Pither and Taylor 1998) and genetic methods to explore the
genetic consequences of connectivity (Andreassen and Ims 2001).
14.2.3.1 Spatial Pattern Metrics
Spatial pattern metrics measure structural connectivity and are usually in the form
of mathematical equations or algorithms designed to quantify the composition and
spatial arrangement of landscapes (Table 14.3). Structural metrics are measured on
maps or GIS images (though they can be calculated from paper habitat maps or hand
delineated polygons from aerial photography). The computer-based approach, how-
ever, provides higher spatial accuracy and greater flexibility in data processing. Soft-
ware packages such as Fragstats v3.3. (McGarigal et al. 2002) offer a wide selection
of structural connectivity metrics such as contagion (the aggregation of patches;
Li and Reynolds 1993), proximity index (the isolation of patches; Gustafson and
Parker 1992), patch cohesion (area-weighted mean perimeter–area ratio divided by
area-weighted mean patch shape index; Schumaker 1996), connectance index (num-
ber of functional joinings between all patches of the same type divided by total
number of possible joinings; McGarigal et al. 2002), and lacunarity (a measure of
the distribution of gap sizes; Plotnick et al. 1993). Several of these metrics quan-
tify similar geometric properties of spatial pattern, and therefore are often collinear
(Riitters et al. 1995). In addition, patch area and patch quality interact with spatial
patterning to determine connectivity, such that a range of metrics and additional
information may also be required to quantify ecologically meaningful structural
connectivity.
Meta-analyses have shown that structural connectivity metrics such as nearest-
neighbor distance or inter-patch distance are often more sensitive to sample size
and less likely to detect a significant effect than functional metrics (Moilanen and
Nieminen 2002; Table 14.3); since they are often applied without any knowledge
of species resource requirements and space-use patterns. In a terrestrial forest sys-
tem, Schumaker (1996) found only weak correlations between nine commonly used
pattern metrics and the results from simulation models of dispersal indicating that
pattern metrics may not always be appropriate for predicting connectivity. In con-
trast, Tischendorf (2001) found strong correlations, but with some highly variable
results using similar comparisons between pattern metrics and simulated dispersal
processes.
Some spatial pattern metrics allow functional information on how species use the
landscape to be taken into consideration. For example, connectance, also referred to
as CONNECT (Fragstats v3.3), can be defined on the number of functional joinings
between patches of a specified patch type. The metric allows the user to input a
threshold distance for a particular species to determine if a pair of patches is con-
nected or not. Then FRAGSTATS computes connectance as a percentage of the
maximum possible connectance given the number of patches. The threshold distance
514 R. Grober-Dunsmore et al.
Table 14.3 A summary of the data-dependent classification framework for connectivity metrics
(from Calabrese and Fagan 2004)
Connectivity
metrics
Type of
connectivity
Habitat-level
data
Species-level
data Method
Nearest neighbor
distance
Structural Nearest neighbor
distance
Patch occupancy Patch-specific field
surveys
Spatial pattern
indices
Structural Spatially explicit None GIS/remote sensing
Scale–area slope Structural None Point or grid based
occurrences
Occurrences
databases,
presence/absence
sampling
Graph theoretic Potential Spatially explicit Dispersal ability GIS/remote sensing
and dispersal
studies
Buffer, radius,
incidence
function
metapopulation
model
Potential Spatially explicit,
including
patch area
Patch occupancy
and dispersal
ability
Multi-year, patch-
specific field
surveys or single
year, patch
occupancy study
with dispersal study
Movement
distance
(emigration,
immigration,
dispersal,
spawning)
Actual Variable,
depends on
method
Movement
pathways or
location specific
dispersal ability
Track movement
pathways,
mark–recapture
studies
can be based on either Euclidean distance or functional distance (McGarigal et al.
2002).
Very few marine examples exist where spatial pattern metrics have been applied
to seascapes (but see Garrabou et al. 1998, Turner et al. 1999, Andrefouet et al.
2003, Pittman et al. 2004, 2007a, b, Grober-Dunsmore et al. 2007, 2008, Kendall
and Miller 2008) (Table 14.2), with no examples of studies that have focused specif-
ically on connectivity. Further studies are needed to determine the ecological rel-
evance of seascape structure as quantified by structural connectivity metrics for
marine species and marine processes. In time, such studies should also provide the
necessary information for evaluating the suitability of pattern metrics to investigate
seascape connectivity.
While exploratory studies that include a wide range of metrics may be fruitful,
marine ecologists should choose pattern metrics judiciously, with some understand-
ing of the approach that will be adopted and the intended purpose for the data. In
addition, several of these metrics quantify similar geometric properties of spatial
pattern and are often collinear (Riitters et al. 1995). However, exploratory statistical
techniques may also be useful in selecting the best predictors amongst similar data
structures or in simplifying complex and collinear multivariate data to a more par-
simonious set of orthogonal variables. Techniques such as Principal Components
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 515
Analysis (PCA) have been used for such a purpose (McGarigal and McComb
1995). Any novel application of these methods (e.g., in marine settings) should
also involve exploration of the behavior of spatial pattern metrics when applied
to data of varying spatial and thematic resolution (Hargis et al. 1998, Saura and
Martinez-Millan 2001). Furthermore, some caution is necessary when interpreting
results from species-metric studies since existing metrics are unlikely to capture all
of the relevant spatial information in marine environments and new metrics specific
to aquatic systems may be required. Ultimately, of course, the metric must be rele-
vant to the questions being asked, and researchers must recognize that some metrics
will not be appropriate for practical applications (Crooks and Sanjayan 2006).
14.2.3.2 Graph Theory
In landscape ecology, graph theoretic approaches, which integrate habitat maps with
information on the movement and behavior of fauna or any other mobile component
of the ecosystem, offer several advantages for connectivity analyses. Graph theoretic
approaches are usually considered a potential connectivity technique (Calabrese and
Fagan 2004), since graphs link structural seascape pattern to estimates of dispersal
ability and thereby offer the ability to go beyond structural connectivity and closer
to an understanding of functional connectivity. A graph includes a set of ‘nodes’
usually indicating the center of a habitat patch, and lines termed ‘edges’ that link
the nodes of two connected patches (reviewed by Urban and Keitt 2001). If the
distance between a given pair of patches is less than or equal to the distance the
organism can move, then the patches are considered to be connected or potentially
connectable.
Connectivity can be weighted using variables such as patch type, size, isolation,
and other measures of patch quality to create least-cost movement pathways (Bunn
et al. 2000, Urban 2005). The pair-wise connections are then scaled-up to measure
connectivity across the entire seascape or area of interest and graph theory provides
a set of metrics to summarize various attributes of the connections. For example,
patterns of connectivity can be evaluated and models can be constructed that exam-
ine the dysfunction that may result from changes to the spatial arrangement of the
seascape, i.e., the loss of a node or patch (Urban and Keitt 2001). Graph theory
has provided valuable insight into the spread of terrestrial invasive species (Urban
and Keitt 2001), and could prove similarly valuable for predicting the rate of spread
and spatial pathways of marine invasive species or disease. For species that undergo
distinct ontogenetic habitat shifts, graphs could identify or rank seascapes based on
their potential connectivity, and these models could then be evaluated using abun-
dance data or fish telemetry or tag–recapture/resight data. Furthermore, the tech-
nique allows one to calculate the area of connected habitat that falls within and
outside existing MPA boundaries and also provides connectivity surfaces to inform
the design of MPAs and MPA networks.
While relatively novel to marine systems, graph theory is well developed in ter-
restrial urban planning, computer science, and protected area design (Urban and
Keitt 2001, Rothley and Rae 2005). Treml et al. (2008) was the first to develop a
516 R. Grober-Dunsmore et al.
Fig. 14.7 A graph-theoretic illustration of marine connectivity. Coral reef habitat is represented by
nodes within the graph framework. When larvae from a source reef reach a downstream reef site,
a dispersal connection is made. This dispersal connection and direction is represented by an arrow,
or ‘edge’ within the graph. The thickness of the arrow reflects the strength of connection (from
Treml et al. 2008, with kind permission of Springer Science+Business Media)
marine application of graph theory. The authors applied a metapopulation concep-
tual framework and utilized an advection–diffusion biophysical model to develop
connectivity estimates between islands for dispersing coral larvae in the tropical
Pacific Ocean (Fig. 14.7). When combined with benthic habitat maps, the tech-
nique offers great promise for mapping connectivity and for the identification of
optimally-connected seascapes to support marine protected area designation efforts
and efficacy. Further studies are required to evaluate the utility of this technique for
marine systems.
14.2.3.3 Computer Simulation Models
Given constraints associated with broad-scale field manipulations and data col-
lection, simulation models are valuable tools for examining the potential influ-
ence of seascape structure on species distributions and individual movements.
When employed as an exploratory tool, model results can be used to construct
testable hypotheses to reveal ecological mechanisms underlying spatial patterns.
In terrestrial systems, a special suite of spatially-explicit landscape simulation
models, termed neutral models, have proven very effective in examining connec-
tivity (Gardner and O’Neill 1991, With 1997). Neutral models use a set of deci-
sion rules to create random structural patterns independent of ecological processes.
In these models, landscape structure is typically binary (suitable and unsuitable
patches), although more complex models can incorporate more of the natural vari-
ability in ecological systems such as hierarchical random landscapes and gradients
using fractal algorithms (techniques) to generate complex clustered spatial patterns
(With 1997). Patch-mosaic neutral models have also been developed. These models
simulate mosaic structure rather than the configuration of pixels in a raster grid and
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 517
focus on incorporating aspects of composition and spatial arrangement (Gaucherel
et al. 2006).
Ecological thresholds can be important phenomena in nature, and knowing when
and where a threshold will occur is extremely important information for resource
management. Neutral models have been used to identify thresholds in connectivity
particularly in relation to the loss of habitat that occurs along a fragmentation gra-
dient (With and Crist 1995, Pearson et al. 1996). Percolation theory proposes that
beyond a predictable threshold (approx. 60%) an abrupt change will occur in system
behavior (Plotnick et al. 1993, With and Crist 1995). In terrestrial landscape ecol-
ogy, studies of fragmentation effects have detected thresholds at approximately 30%
of remaining suitable habitat (70% loss) for a wide range of fauna (Andr´
en 1994),
although the exact effect will be both species-specific and scale-dependent. Such
rough guidelines can be useful for predicting the response of populations to degra-
dation or loss of habitat (Taylor et al. 2007), or determining how a reduction in
connectivity will influence the population dynamics of a species (With and Crist
1995).
14.2.3.4 Ecological Thresholds in Seascape Structure
In marine systems, less is known about critical ecological thresholds and how
they vary between species, though recent evidence suggests that they also occur
in shallow-water marine ecosystems (Table 14.2). In seagrass beds of Moreton Bay
(Australia) a gradual decline in resident fish abundance was detected, along spa-
tial gradients in seagrass cover, until approximately 15–20% seagrass cover, beyond
which many abundant species were absent (Pittman 2002, Pittman et al. 2004). In
the Caribbean (Virgin Islands, Florida Keys, and Turks and Caicos), an examina-
tion of fish communities on coral reefs along a spatial gradient in seagrass cover
revealed that fish diversity and abundance increased from 0 to 20–30% seagrass
coverage then plateaued out at 40% indicating a threshold-like response (Grober-
Dunsmore 2005). Thresholds for the amount of habitat available to support fish
appear to occur at lower percent cover values for seagrasses than those detected
for mammals in terrestrial systems (i.e., threshold at 15–30% for fish vs. 30% for
mammals) although much variability exists. Differences in thresholds for fish using
seagrasses may result from the highly dynamic patchiness of seagrass beds with
some patches relatively ephemeral due to die-offs and storms combined with the
highly mobile nature of many fish which enables them to traverse relatively large
distances to inhabit even small patches of seagrass, albeit at low abundance. Critical
thresholds imply that absence, loss, or degradation of habitat can have deleterious
effects on population dynamics, and are therefore crucially important to studies of
connectivity in tropical marine systems. Models for marine organisms can easily be
developed (Fig. 14.8) as data on the responses to varying levels of habitat availabil-
ity and movement becomes accessible.
Coupling neutral models with individual-based models of dispersal or models of
gene flow and population dynamics should also be developed for marine species
(Butler at al. 2005). Individual-based correlated random walk models (Schippers
518 R. Grober-Dunsmore et al.
R2 = 0.53
15
20
25
30
35
40
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Proportion seagrass habitat
Species richness
Fig. 14.8 Relationship of reef fish species richness (mobile invertebrate feeders) and seagrass
cover (proportion of habitat). Species richness declines considerably below 30% seagrass coverage.
Above 40% seagrass coverage, increases in seagrass do not result in additional species
et al. 1996) and other types of movement simulations reveal which spatial patterns
facilitate or impede movements across a landscape and the relative cost–benefits
associated with certain pathways (Tischendorf and Fahrig 2000). An individual-
based model was helpful to test how change in habitat area and spatial configu-
ration of seagrass beds influenced predator–prey interactions and cohort size for a
group of settling juvenile blue crabs (Callinectes sapidus) (Hovel and Regan 2008).
Prey cohort size was maximized in patchy seagrasses, which corresponded to results
from field experiments, whereas mobile prey able to detect and avoid predators had
higher survival in continuous seagrass beds (Hovel and Regan 2008). If individual-
based spatially explicit seascape models are to be useful in studying connectivity,
considerable effort will be required to identify and quantify behavioral responses
to seascape structure. Then, simulations can be parameterized with a meaningful
behavioral response or threshold effect.
14.3 Important Considerations
14.3.1 Data Needs
To investigate connectivity in terrestrial environments, landscape ecologists typi-
cally use a wide variety of spatial datasets often requiring integration and additional
digital processing within a GIS (e.g., vegetation maps, digital elevation models,
tracking data). In marine environments, studies of connectivity may also require
multiple spatial data sets including: (1) benthic habitat maps to capture informa-
tion on the distribution of patch types that can determine connectivity in a region,
(2) oceanographic characteristics (e.g., sea surface temperature, frontal boundaries,
upwelling zones, prevailing current patterns), (3) bathymetric surfaces (e.g., linear
features, canyons, continental shelf, banks, seamounts, promontories), and (4) eco-
logical factors (e.g., distribution of predators, prey, competitors) and human uses
(e.g., point or non-point source pollution, ship traffic, fishing areas) that may act as
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 519
facilitators, barriers, or modifiers to movement for marine organisms. For species
that are strongly linked to the benthos, a benthic habitat map alone may be sufficient
environmental data to begin studies of seascape connectivity.
Many of these environmental data are freely available via online data portals or
digital archives, however, for many regions of the earth additional data may need
to be collected or acquired. Several obstacles may inhibit the use of existing data
when applying landscape ecology approaches to connectivity including inadequate
spatial coverage, a mismatch of temporal coincidence, and most commonly, inap-
propriate or mismatched scales of data layers. An important first step in any broad
scale study of connectivity is to evaluate the availability and quality of data for the
study region. One of the most significant obstacles faced by marine spatial ecol-
ogists is the absence of appropriate spatial data (i.e., benthic habitat maps). Even
for data-rich areas, often the available data were acquired for an entirely differ-
ent purpose, and therefore may not necessarily represent the environmental reality
from the organism-based perspective. Given the paucity of suitable data, researchers
frequently proceed with environmental data with unknown accuracy and data that
often do not match the spatial and temporal resolution of the ecological processes
under investigation. This problem is made more challenging by that fact that very
little scientific information is available to guide scale selection such as identify-
ing the appropriate spatial grain and extent for a study. Also, very little is known
about the relative importance to marine species of different seascape features or
variables.
As a general rule, the environmental data must be available at a finer scale
(finer spatial grain) than the scale of the process under investigation. This then
provides an opportunity to coarsen the resolution of patterning so that the link-
age between organism and environment can be explored at multiple spatial
scales.
Even when the necessary data are readily available or easily collected, the value
of subsequent seascape analyses is compromised if researchers fail to assess the
accuracy of spatial data (Turner et al. 2001). Error associated with the remote sens-
ing and GIS data acquisition, processing, analysis, conversion, and final product pre-
sentation can have a significant impact on the confidence of decisions made using
the data (Lunetta et al. 1991). Potential sources of error include the age of data, com-
pleteness of aerial coverage, and map scale (Burrough 1986). Those that occur with
natural variation in original measurements include positional and content accuracy,
while other sources of error occur during processing (i.e., numerical computation,
classification) (Burrough 1986). Errors in spatial data can obscure or distort species–
habitat relationships and may even result in spurious correlations (Karl et al. 2000).
The saying ‘garbage in, garbage out’ applied to any analytical process, and typically
landscape ecology analyses are highly susceptible to data quality. Nevertheless, if a
strong ecological signal exists then even with relatively crude data (with some minor
errors) it may still be possible to detect the influence of pattern on process. Conse-
quently, a validation step will be essential when developing derived products such as
modeled outputs to enable appropriate statements of accuracy. In addition to testing
the accuracy of the original data, the use of multiple techniques to analyze relation-
ships or create models can verify the robustness of research results by providing
520 R. Grober-Dunsmore et al.
information on potential bias in the data, techniques, and interpretation of the final
model results.
14.3.2 Not all Habitat Patches are Created Equal
Many ecological processes operate to influence connectivity. Yet, relevant or
appropriately-scaled data often do not exist or are impractical to obtain at suf-
ficiently broad spatial scales. For instance, species interactions (e.g., predation,
competition) are known to significantly affect organism distribution and habitat uti-
lization patterns, but are difficult to incorporate, since spatial data that represent
their multidimensional complexity typically do not exist. The fact that predation
and competition can decouple or obscure species–environment data is rarely con-
sidered in landscape ecology analyses, where habitat patterns are the primary focus.
For example, within a life stage, the presence of a large number of predators or
prey may have a greater impact on determining the connectivity of a given species
to its environment, either by influencing the species attraction or avoidance of cer-
tain patches, or by introducing different mortality rates among patches. Similarly,
anthropogenic factors such as fishing pressure or pollution may influence distribu-
tion patterns and energy fluxes. Studies are now required that are capable of parti-
tioning the influence of species interactions from the influence of benthic seascape
structure.
14.4 Lessons Learned
14.4.1 Lessons from Terrestrial Landscape Ecology
Connectivity is a central theme in terrestrial landscape ecology (Turner et al. 2001,
Turner 2005, Crooks and Sanjayan 2006), providing specific concepts, analytical
tools, and unique insights in the linkage between ecological patterns and processes.
The general principles that have emerged from landscape ecology appear to be appli-
cable across a range of ecological subjects and natural resource management topics
(see Turner et al. 2001, Gutzwiller 2002, Taylor et al. 2006, Wiens 2006). The gener-
ality of these principles will now urgently require further testing in marine systems
and these efforts will set the stage for a deepening of the knowledge base of seascape
ecology. Here we list eight principles or ecological statements, some of which can
be formulated as testable hypotheses for future work in coral reef ecosystems and
the broader marine environment.
•Connectivity is a key feature of landscape structure. Actual connectivity of a
landscape is more complicated than simple corridors or proximity between two
patch types (Crooks and Sanjayan 2006). Understanding how the fabric of a land-
scape is woven together to facilitate or impede movement of organisms, materi-
als, or energy will be critical for conservation efforts.
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 521
•Landscape connectivity is species-specific. Different organisms will respond to
landscape structure in different ways (Taylor et al. 2006). The same landscape
will differ in connectivity for various processes, species, and life-history stages.
Species-specific responses must be recognized when managing entire ecosys-
tems across a range of spatial scales and taxa.
•Resource managers should manage the entire landscape mosaic. Managing the
landscape mosaic offers an effective means for preserving connectivity (Taylor
et al. 2006). Since single species-level approaches are difficult, and managing for
individual habitat patches creates challenges, managers must consider not only
the focal patch type but also the surrounding area (Turner et al. 2001).
•Real landscapes are not random because ecological patterns and processes are
not random (Forman 1995, Taylor et al. 2006). Landscapes contain barriers to
movement, detrimental habitat, areas of high predation risk, and areas that con-
tain patches with higher and lower quality habitat, which result from a variety
of causes including biotic and abiotic interactions, natural disturbances, and pat-
terns of human activities and stressors, and such heterogeneity will have pro-
found consequences on species distributions and ecological processes.
•Connectivity is a necessary, but not sufficient condition for species conservation
(Taylor et al. 2006). Landscape connectivity influences reproduction, mortality,
fitness, and access to resources; however, other landscape characteristics are also
critical, and must be considered for effective management.
•Connectivity is a dynamic concept. Landscapes are ever-changing and are being
modified by physical forces (e.g., hurricanes, climate change) and biological pro-
cesses (e.g., competition) over short and long time scales (Taylor et al. 2006).
Landscape connectivity will be related to behavioral characteristics, the degree
of natural and anthropogenic disturbance, and interactions among landscape ele-
ments. Variation in connectivity due to such effects must be recognized when
designing and interpreting results of landscape studies.
•Scale is crucial. There is no ‘right’ scale for studying landscapes (Wiens et al.
2002), but scale effects must be carefully considered when designing and evalu-
ating landscape connectivity.
•Consider potentially confounding effects. Because of the large spatial scale
over which landscape studies occur and the possible influence of non-measured
explanatory variables, the potential effects of fishing, predation, and other
non-measured attributes should be considered. Carefully designed experi-
ments that control or account for these potentially confounding factors are
recommended.
14.4.2 Insights for Seascape Ecology
Though few landscape ecology studies have been conducted in coral reef ecosys-
tems, a wealth of research provides insight into how various spatial elements of the
seascape may influence connectivity (Table 14.2). Such studies, though not always
522 R. Grober-Dunsmore et al.
designed with a landscape perspective, can provide a foundation for understanding
the consequences of connectivity (Table 14.2). The general principles derived from
major findings from examples of recent coral reef studies that address some aspect
of landscape structure (Table 14.2) are presented as a starting point for designing
seascape studies and interpreting results for resource managers.
•Patch size, distribution, and configuration are perhaps the most basic aspects of
seascape pattern, influencing the distribution of marine organisms. Patch size
may also affect a number of important ecological processes such as colonization,
reproduction, mortality, predator-prey interactions, and the transport of materi-
als, energy, and marine organisms across seascapes. The spatial configuration of
habitat patches (i.e., shape, clustering, edge:perimeter ratio, contiguity) may also
control the movement patterns of marine organisms.
•Habitat composition also influences the distribution of marine organisms. The
composition of habitat patches within an area and in surrounding areas can deter-
mine the presence, diversity, and abundance of marine organisms.
•Specific habitats can be crucial for certain marine organisms. While some species
are habitat generalists and have flexibility in their habitat requirements, other
species are habitat specialists that depend upon particular habitat types. These
dependencies may vary with life history stage.
•Movements interact with and are likely modified by seascape structure. Certain
features or configurations may function as facilitators or inhibitors of movement.
While some habitat types or features (e.g., extremely deep or shallow areas)
may facilitate movement, other features may impede it. If dispersal capacity of a
marine organism is low, connectivity may depend more heavily on the seascape
structure of features immediately surrounding them. If dispersal capacity is high,
seascape structure across larger spatial scales may have greater influence on con-
nectivity.
•Edge effects result from a combination of biotic and abiotic factors that alter
environmental conditions along patch edges compared to patch interiors, and the
presence and type of edges within a seascape may influence connectivity.
•Proximity to specific patch types or seascape features determines species rich-
ness, abundance, and density and therefore may influence colonization, survivor-
ship, or mobility across seascapes.
•Fragmentation (either through loss or degradation of particular habitats) can
influence important ecological processes, and may have possible consequences
on survivorship, or dispersal of marine organisms.
•Thresholds of habitat availability appear to occur in tropical seascapes, with habi-
tat becoming either connected or disconnected at some unknown threshold of
habitat abundance. As little is known about thresholds of habitat in marine sys-
tems, it should be considered as a potential factor in structuring marine commu-
nities.
•Interactions between seascape features may confound individual effects. Con-
nectivity may be determined by a combination of multiple seascape features, and
discerning the contribution of a single feature may prove challenging.
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 523
•Connectivity is increasingly being identified as a vital element of seascape struc-
ture, and has been shown to influence biomass, habitat use, site fidelity, and
movement of marine organisms. A few habitat types have been investigated, but
the importance of connectivity of other untested habitats and seascape features
should be considered.
With additional research, general principles useful for managing mobile marine
organisms in coral reef ecosystems may be further developed. Field researchers are
testing these concepts in coral reef ecosystems, and data obtained can be applied to
modeling efforts to solve complex resource management questions.
14.5 Implications for Resource Management
The ability to identify functionally well-connected seascapes and evaluate their rel-
ative importance to species and communities is of great value to resource man-
agers faced with the challenge of protecting an optimal subset of seascapes. The
knowledge of how spatial patterns in the environment influence connectivity will
also facilitate the design of habitat restoration or habitat creation plans that maxi-
mize organism survival, growth, productivity, and the species diversity of communi-
ties. This is a major knowledge-gap in applied marine ecology that requires urgent
attention since many millions of dollars are spent on selecting and implementing
MPAs and on restoration projects that do not have spatially explicit information on
connectivity.
By focusing specifically on ecological connectivity and carefully scaling inves-
tigations of seascape patterns to specific ecological processes, a landscape ecology
approach will allow us to determine the amount, type, configuration, and location
of patch types required to maintain ecological connectivity. These are the central
questions that must be addressed in tropical marine systems to successfully iden-
tify essential fish habitat, predict effects of habitat alteration, and prioritize among
management options. A landscape ecology approach also offers great potential for
the study of the spread of marine invasive species, with some seascapes being less
rapidly colonized than others seascapes. In terrestrial systems, landscape ecology
has made substantial contributions to our understanding of the direction and rate of
spread of invasive plants, wildfires, and climate-induced shifts in species distribu-
tions (With 2002).
Clearly, identifying optimal seascape composition and arrangement for marine
protected areas and networks of marine protected areas require the consideration of
interactions between structural and functional connectivity across multiple spatial
and temporal scales (Ward et al. 1999). Increasingly, resource managers will need to
manage mosaics of coral reef habitat within and among protected areas, rather than
focusing on individual habitat types or patches. Using landscape ecology principles,
concepts, and tools, connectivity for various species can be identified and evaluated
in relation to existing or planned jurisdictional boundaries to optimize conservation
efforts across broad spatial scales.
524 R. Grober-Dunsmore et al.
In terrestrial landscape ecology, several decision support tools have been suc-
cessfully applied to design reserves, create corridors of habitat, reduce the effects
of forest fragmentation, and optimize connectivity of important landscape features
for targeted organisms (Crooks and Sanjayan 2006). As ecological connectivity in
marine ecosystems is further investigated, similar applications can be developed
and incorporated into spatial tools to support resource management and conserva-
tion, especially marine spatial planning (Possingham et al. 2000, Mumby 2006).
The benefits of selecting one habitat type or patch over another or choosing among
alternate combinations of habitat patches can be evaluated using optimization algo-
rithms such as those used in software programs MARXAN (Possingham et al.
2000) and C-Plan (Pressey 1999, Margules and Pressey 2000). These approaches
are ideal for: (1) evaluating the costs and benefits of alternate protected area designs,
(2) predicting the impacts of degrading or excluding specific seascape features when
designating essential fish habitat, and (3) assessing the consequences of reducing or
increasing ecological connectivity for a wide spectrum of organisms. Such decision
support tools are now urgently required for enhanced management of the heavily
used and highly valued tropical marine seascapes worldwide. These approaches
can facilitate the selection and comparison of multiple candidate protected area
networks allowing resource managers to prioritize areas based on their ecologi-
cal connectivity (Mumby 2006). As landscape ecology approaches and tools are
increasingly applied in tropical marine ecosystems, the utility of such concepts
for improving our understanding of ecological connectivity and applying results
to make more informed decisions for conservation planning will be realized.
Acknowledgments The support of the National Marine Fisheries Service, Fisheries Ecology
Division at the Southwest Fisheries Science Center in Santa Cruz, CA, is greatly appreciated. In
particular, Churchill Grimes was instrumental in providing support for Dr. Rikki Dunsmore. In
addition, insights from Chris Jeffrey and Mark Monaco of NOAA’s Center for Coastal Monitoring
and Assessment were valuable. The Biogeography Branch coral reef monitoring and mapping
activities are supported by funding from the Coral Reef Conservation Program (CRCP).
References
Andreassen HP, Ims RA (2001) Dispersal in patchy vole populations: role of patch configuration,
density dependence, and demography. Ecology 82:2911–2926
Andrefouet S, Kramer P, Torres-Pulliza D et al (2003) Multi-site evaluation of IKONOS data for
classification of tropical coral reef environments. Remote Sens Environ 88:128–143
Andr´
en H (1994) Effect of habitat fragmentation on birds and mammals in landscapes with dif-
ferent proportions of suitable habitat: a review. Oikos 71:355–366
Appeldoorn RS, Friedlander A, Sladek Nowlis J et al (2003) Habitat connectivity in reef fish com-
munities and marine reserve design in Old Providence-Santa Catalina, Colombia. Gulf Caribb
Res 14:61–77
Ault TR, Johnson CR (1998) Spatially and temporally predictable fish communities on coral reefs.
Ecol Monogr 68:25–50
Beets J, Muehlstein L, Haught K et al (2003) Habitat connectivity in coastal environments: patterns
and movements of Caribbean coral reef fishes with emphasis on bluestriped grunt, Haemulon
sciurus. Gulf Caribb Res 14:29–42
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 525
Birkeland C (1985) Ecological interactions between mangroves, seagrass beds and coral reefs. In:
Birke-land C (ed) Ecological interactions between tropical coastal ecosystems. UNEP Regional
Seas Reports 73. Earth-print, Stevenage, UK
Bohnsack JA, Bannerot SP (1986) A stationary visual census technique for quantitatively assessing
community structure of coral reef fishes. NOAA Technical Report NMFS 41
Bohnsack JA, Harper DE, McClellan DB et al (1994) Effects of reef size on colonization and
assemblage structure of fishes at artificial reefs off southeastern Florida, USA. Bull Mar Sci
55:796–823
Brock VE (1954) A method of estimating reef fish populations. J Wildl Manage 18:297–308
Bunn AG, Urban DL, Keitt TH (2000) Landscape connectivity: a conservation application of graph
theory. J Environ Manage 59(SI 4):265–278
Burke N (1995) Nocturnal foraging habitats of French and bluestriped grunts, Haemulon flavolin-
eatum and H. sciurus at Tobacco Caye, Belize. Environ Biol Fish 42:365–374
Burrough PA (1986) Principles of Geographic Information Systems for land resources assessment.
Oxford University Press, Oxford, UK
Butler MJ, Dolan TW, Hunt JH et al (2005) Recruitment in degraded marine habitats: a spatially
explicit, individual-based model for spiny lobster. Ecol Appl 15:902–918
Calabrese JM, Fagan WF (2004) A comparison-shopper’s guide to connectivity metrics. Front Ecol
Environ 2:529–536
Carleton Ray G (1991) Coastal-zone biodiversity patterns. BioScience 41:490–498
Chapman MR, Kramer DL (2000) Movements of fishes within and among fringing coral reefs in
Barbados. Environ Biol Fish 57:11–24
Chateau O, Wantiez L (2007) Site fidelity and activity patterns of a humphead wrasse, Cheilinus
undulates (Labridae), as determined by acoustic telemetry. Environ Biol Fish 80:503–508
Chittaro PM, Fryer BJ, Sale R (2004) Discrimination of French grunts (Haemulon flavolineatum
Desmarest, 1823) from mangrove and coral reef habitats using otolith microchemistry. J Exp
Mar Biol Ecol 308:169–183
Christensen JD, Jeffrey CFG, Caldow C et al (2003) Cross-shelf habitat utilization patterns of reef
fishes in southwestern Puerto Rico. Gulf Caribb Res 14:9–27
Clark R, Monaco ME, Appeldoorn RS, Roque B (2005) Fish habitat utilization in a Puerto Rico
coral reef ecosystem. Proc Gulf Caribb Fish Inst 56:467–485
Cowen RK, Paris CB, Srinivasan A (2006) Scaling of connectivity in marine populations. Science
311:522–527
Crooks KR, Sanjayan M (2006) Connectivity conservation. Cambridge University Press, Cam-
bridge
Dahlgren CP, Eggleston DB (2000) Ecological processes underlying ontogenetic habitat shifts in a
coral reef fish. Ecology 81:2227–2240
Dorenbosch M, Grol MGG, Christianen MJA et al (2005) Indo-Pacific seagrass beds and man-
groves contribute to fish density and diversity on adjacent coral reefs. Mar Ecol Prog Ser
302:63–76
Dorenbosch M, Grol MGG, Nagelkerken I et al (2006a) Different surrounding landscapes may
result in different fish assemblages in East African seagrass beds. Hydrobiologia 563:45–60
Dorenbosch M, Grol MGG, Nagelkerken I et al (2006b) Seagrass beds and mangroves as potential
nurseries for the threatened Indo-Pacific humphead wrasse, Cheilinus undulatus and Caribbean
rainbow parrotfish, Scarus guacamaia. Biol Conserv 129:277–282
Dorenbosch M, van riel MC, Nagelkerken I et al (2004a) The relationship of reef fish densities to
the proximity of mangrove and seagrass nurseries. Estuar Coast Shelf Sci 60:37–48
Dorenbosch M, Verberk WCEP, Nagelkerken I et al (2007) Influence of habitat configuration on
connectivity between fish assemblages of Caribbean seagrass beds, mangroves and coral reefs.
Mar Ecol Prog Ser 334:103–116
Dorenbosch M, Verweij MC, Nagelkerken I et al (2004b) Homing and daytime tidal movements
of juvenile snappers (Lutjanidae) between shallow-water nursery habitats in Zanzibar, western
Indian Ocean. Environ Biol Fish 70:203–209
526 R. Grober-Dunsmore et al.
Dunning JB, Danielson BJ, Pulliam HR (1992) Ecological processes that affect populations in
complex landscapes. Oikos 65:169–175
Fagan WF, Calabreses JM (2006) Quantifying connectivity: balancing metric performance with
data requirements. In: Crooks KR, Sanjayan M (eds) Connectivity conservation. Conservation
biology 14, Cambridge University Press, Cambridge
Forman RTT (1995) Land mosaics: the ecology of landscapes and regions. Cambridge University
Press, Cambridge
Forman RTT, Godron M (1986) Landscape ecology. John Wiley and Sons, New York
Gardner RH, O’Neill RV (1991) Pattern, process and predictability: the use of neutral models
for landscape analysis. In: Turner MG, Gardner RH (eds) Quantitative methods in landscape
ecology: the analysis and interpretation of landscape heterogeneity. Springer-Verlag, New York
Garrabou J, Riera J, Zabala M (1998) Landscape pattern indices applied to Mediterranean subtidal
rocky benthic communities. Landsc Ecol 13:225–247
Gaucherel C, Fleury D, Auclair D (2006) Neutral models for patchy landscapes. Ecol Modell
197:159–170
Gillanders BM, Able KW, Brown JA et al (2003) Evidence of connectivity between juvenile and
adult habitats for mobile marine fauna: an important component of nurseries. Mar Ecol Prog
Ser 247:281–295
Gillis EA, Krebs CJ (2000) Survival of dispersing versus philopatric juvenile snowshoe hares: do
dispersers die? Oikos 90:343–346
Gladfelter WB, Ogden JC, Gladfelter EH (1980) Similarity and diversity among coral reef fish
communities: a comparison between tropical western Atlantic (Virgin Islands) and tropical
central pacific (Marshall Islands) patch reefs. Ecology 61:1156–1168
Grober-Dunsmore R (2005) The application of terrestrial landscape ecology principles to the
design and management of marine protected areas in coral reef ecosystems. Ph.D. disserta-
tion submitted to University of Florida, Department of Fisheries and Aquatic Sciences, Florida,
219 pp.
Grober-Dunsmore R, Beets J, Frazer T et al (2008) Influence of landscape structure on reef fish
assemblages. Landsc Ecol 23(SI):37–53
Grober-Dunsmore R, Bonito V (2009) Movement of reef fishes inside and outside of Votua MPA,
Fiji Islands. Report to NOAA Coral Reef International 2009 Coral Reef library, 24pp.
Grober-Dunsmore R, Frazer T, Beets J et al (2004) The significance of adjacent habitats on reef
fish assemblage structure: are relationships detectable and quantifiable at a landscape scale?
Proc Gulf Caribb Fish Inst 55:713–734
Grober-Dunsmore R, Frazer TK, Lindberg WJ et al (2007) Reef fish and habitat relationships in a
Caribbean seascape: the importance of reef context. Coral Reefs 26:201–216
Gustafson EJ, Parker GR (1992) Relationships between landcover proportion and indexes of land-
scape spatial pattern. Landsc Ecol 7:101–110
Gutzwiller KJ (2002) Applying landscape ecology in biological conservation. Springer-Verlag,
New York
Hanski I (1998) Metapopulation dynamics. Nature 396:41–49
Hargis CD, Bissonette JA, David JL (1998) The behavior of landscape metrics commonly used in
the study of habitat fragmentation. Landsc Ecol 13:167–186
Helfman GS, Meyer JL, McFarland WN (1982) The ontogeny of twilight migration patterns in
grunts (Pisces, Haemulidae). Anim Behav 30:317–326
Holland KN, Peterson JD, Lowe CG et al (1993) Movements, distribution and growth rates
of the white goatfish Mulloides flavolineatus in a fisheries conservation zone. Bull Mar Sci
52:982–992
Hovel KA, Lipcius RN (2002) Effects of seagrass habitat fragmentation on juvenile blue crab
survival and abundance. J Exp Mar Biol Ecol 271:75–98
Hovel KA, Regan HM (2008) Using an individual-based model to examine the roles of habitat
fragmentation and behavior on predator–prey relationships in seagrass landscapes. Landsc Ecol
23(S1):75–89
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 527
Irlandi EA, Ambrose WG, Orlando BA (1995) Landscape ecology and the marine environment-
how spatial configuration of seagrass habitat influences growth and survival of the Bay scallop.
Oikos 72:307–313
Irlandi EA, Crawford MK (1997) Habitat linkages: the effect of intertidal saltmarshes and adja-
cent sub-tidal habitats on abundance, movement, and growth of an estuarine fish. Oecologia
110:222–230
Jelbart JE, Ross PM, Connolly RM (2006) Edge effects and patch size in seagrass landscapes: an
experimental test using fish. Mar Ecol Prog Ser 319:93–102
Jelbart JE, Ross PM, Connolly RM (2007) Fish assemblages in seagrass beds are influenced by the
proximity of mangrove forests. Mar Biol 150:993–1002
Karl JW, Heglund PJ, Garton EO et al (2000) Sensitivity of species habitat-relationship model
performance to factors of scale. Ecol Appl 10:1690–1705
Kendall MS, Christensen JD, Hillis-Starr Z (2003) Multi-scale data used to analyze the spatial
distribution of French grunts, Haemulon flavolineatum, relative to hard and soft bottom in a
benthic landscape. Environ Biol Fish 66:19–26
Kendall MS, Kruer CR, Buja KR, Christensen JD, Finkbeiner M, Monaco ME (2002) Methods
used to map the benthic habitats of Puerto Rico and the U.S. Virgin Islands. NOAA/NOS Bio-
geography Program Technical Re-port. Silver Spring, MD, p 45
Kendall MS, Miller T (2008) The influence of thematic and spatial resolution on maps of a coral
reef ecosystem. Marine Geodesy 31:75–102
Kramer DL, Chapman MR (1999) Implications of fish home range size and relocation for marine
reserve function. Environ Biol Fish 55:65–79
Levin SA (1992) The problem of pattern and scale in ecology. Ecology 73:1943–1967
Li HB, Reynolds JF (1993) A new contagion index to quantify spatial patterns of landscapes.
Landsc Ecol 8:155–162
Li HB, Wu JG (2004) Use and misuse of landscape indices. Landsc Ecol 19:389–399
Lugendo BR, Nagelkerken I, Jiddawi N et al (2007b) Fish community composition of a tropical
non-estuarine embayment in Zanzibar (Tanzania). Fish Sci 73:1213–1223
Lugendo BR, Nagelkerken I, Kruitwagen G et al (2007a) Relative importance of mangroves as
feeding habitat for fish: a comparison between mangrove habitats with different settings. Bull
Mar Sci 80:497–512
Lugendo BR, Nagelkerken I, van der Velde G et al (2006) The importance of mangroves, mud and
sand flats, and seagrass beds as feeding areas for juvenile fishes in Chwaka Bay, Zanzibar: gut
content and stable isotope analyses. J Fish Biol 69:1639–1661
Lunetta RS, Congalton RG, Fenstermaker LK et al (1991) Photogramm Eng Remote Sens 57:
677–687
MacArthur RH, Wilson EO (1967) The theory of island biogeography. Princeton University Press,
Princeton, New Jersey
Malczewski J (1999) GIS and multicriteria decision analysis. John Whiley & Sons, New York
Margules CR, Pressey RL (2000) Systematic conservation planning. Nature 405:
243–253
McGarigal K, Cushman SA, Neel MC (2002) FRAGSTATS: spatial pattern analysis program
for categorical maps. University of Massachusetts Amherst, Massachusetts. http://www.umass.
edu/landeco/research/ fragstats/fragstats.html/
McGarigal K, McComb WC (1995) Relationships between landscape structure and breeding birds
in the Oregon Coast Range. Ecol Monogr 65:235–260
Meyer CG, Holland KN, Papastamatiou YP (2007) Seasonal and diel movements of giant trevally
Caranx ignobilis at remote Hawaiian atolls: implications for the design of marine protected
areas. Mar Ecol Prog Ser 333:13–25
Meyer CG, Holland KN, Wetherbee BM et al (2000) Movement patterns, habitat utilization, home
range size and site fidelity of whitesaddle goatfish, Parupeneus porphyreus, in a marine reserve.
Environ Biol Fish 59:235–242
Micheli F, Peterson CH (1999) Estuarine vegetated habitats as corridors for predator movements.
Conserv Biol 13:869–881
528 R. Grober-Dunsmore et al.
Mitchell MS, Lancia RA, Gerwin JA (2001) Using landscape-level data to predict the distribution
of birds on a managed forest: effects of scale. Ecol Appl 11:1692–1708
Moilanen A, Nieminen M (2002) Simple connectivity measures in spatial ecology. Ecology
83:1131–1145
Molles MC (1978) Fish species-diversity on model and natural reef patches-experimental insular
bio-geography. Ecol Monogr 48:289–305
Mumby PJ (2006) Connectivity of reef fish between mangroves and coral reefs: algorithms for the
design of marine reserves at seascape scales. Biol Conserv 128:215–222
Mumby PJ, Edwards AJ, Arias-Gonzalez JE et al (2004) Mangroves enhance the biomass of coral
reef fish communities in the Caribbean. Nature 427:533–536
Mumby PJ, Harborne AR (1999) Development of a systematic classification scheme of marine
habitats to facilitate regional management of Caribbean coral reefs. Biol Conserv 88:155–163
Nagelkerken I (2007) Are non-estuarine mangroves connected to coral reefs through fish migra-
tion? Bull Mar Sci 80:595–607
Nagelkerken I, Dorenbosch M, Verberk WCEP et al (2000a) Importance of shallow-water biotopes
of a Caribbean bay for juvenile coral reef fishes: patterns in biotope association, community
structure and spatial distribution. Mar Ecol Prog Ser 202:219–230
Nagelkerken I, Dorenbosch M, Verberk WCEP et al (2000b) Day-night shifts of fishes between
shallow-water biotopes of a Caribbean bay, with emphasis on the nocturnal feeding of Haemul-
idae and Lutjanidae. Mar Ecol Prog Ser 194:55–64
Nagelkerken I, Faunce CH (2007) Colonisation of artificial mangroves by reef fishes in a marine
seascape. Estuar Coast Shelf Sci 75:417–422
Nagelkerken I, Kleijnen S, Klop T et al (2001) Dependence of Caribbean reef fishes on mangroves
and seagrass beds as nursery habitats: a comparison of fish faunas between bays with and
without mangroves/seagrass beds. Mar Ecol Prog Ser 214:225–235
Nagelkerken I, Roberts CM, van der Velde G et al (2002) How important are mangroves and
seagrass beds for coral-reef fish? The nursery hypothesis tested on an island scale. Mar Ecol
Prog Ser 244:299–305
Nagelkerken I, van der Velde G (2002) Do non-estuarine mangroves harbour higher densities
of juvenile fish than adjacent shallow-water and coral reef habitats in Curacao (Netherlands
Antilles)? Mar Ecol Prog Ser 245:191–204
Ogden JC, Ehrlich PR (1977) Behavior of heterotypic resting schools of juvenile grunts
(Pomadasyidae). Mar Biol 42:273–280
Ogden JC, Gladfelter EH (1983) Coral reefs, seagrass beds and mangroves: their interaction in the
coastal zones of the Caribbean. UNESCO Rep Mar Sci 23:1–133
Parrish JD (1989) Fish communities of interacting shallow-water habitats in tropical oceanic
regions. Mar Ecol Prog Ser 58:143–160
Pearson SM, Turner MG, Gardner RH et al (1996) An organism-based perspective of habitat frag-
mentation. In: Szaro RC (ed) Biodiversity in managed landscapes: theory and practice. Oxford
University Press, California
Pither J, Taylor PD (1998) An experimental assessment of landscape connectivity. Oikos
83:166–174
Pittman SJ (2002) Linking fish and prawns to their environment in shallow-water marine land-
scapes. Ph thesis, Geographical Sciences Department and The Ecology Centre, University of
Queensland, Brisbane, Australia
Pittman SJ, Caldow C, Hile SD et al (2007b) Using seascape types to explain the spatial patterns
of fish in the mangroves of SW Puerto Rico. Mar Ecol Prog Ser 348:273–284
Pittman SJ, Christensen JD, Caldow C et al (2007a) Predictive mapping of fish species richness
across shallow-water seascapes in the Caribbean. Ecol Modell 204:9–21
Pittman SJ, McAlpine CA (2003) Movement of marine fish and decapod crustaceans: process,
theory and application. Adv Mar Biol 44:205–294
Pittman SJ, McAlpine CA, Pittman KM (2004) Linking fish and prawns to their environment: a
hierarchical landscape approach. Mar Ecol Prog Ser 283:233–254
14 A Landscape Ecology Approach for the Study of Ecological Connectivity 529
Plotnick RE, Gardner RH, O’Neill RV (1993) Lacunarity indexes as measures of landscape tex-
ture. Landsc Ecol 8:201–211
Possingham H, Ball I, Andelman S (2000) Mathematical models for identifying representative
reserve networks. In: Ferson S, Burgman M (eds) Quantitative methods for conservation biol-
ogy. Springer-Verlag, New York
Pressey RL (1999) Applications of irreplaceability analysis to planning and management problems.
Parks 9:42–51
Recksiek CW, Appeldoorn RS, Turningan RG (1991) Studies of fish traps as stock assessment
devices on a shallow reef in south-western Puerto Rico. Fish Res 10:177–197
Riitters KH, O’Neill RV, Hunsaker CT et al (1995) A factor analysis of landscape pattern and
structure metrics. Landsc Ecol 10:23–39
Robbins BD, Bell SS (1994) Seagrass landscapes: a terrestrial approach to the marine subtidal
environment. Trends Ecol Evol 9:301–304
Rothley KD, Rae C (2005) Working backwards to move forwards: graph-based connectivity met-
rics for reserve network selection. Environ Modell Assess 10:107–113
Rozas LP, Minello TJ (1998) Nekton use of salt marsh, seagrass, and nonvegetated habitats in a
south Texas (USA) estuary. Bull Mar Sci 63:481–501
Sale PF (2002) The science we need to develop for more effective management. In: Sale PF (ed)
Coral reef fishes: dynamics and diversity in a complex ecosystem. Academic Press, London
Saura S, Martinez-Millan J (2001) Sensitivity of landscape pattern metrics to map spatial extent.
Photogramm Eng Remote Sens 67:1027–1036
Schippers P, Verboom J, Knaapen JP et al (1996) Dispersal and habitat connectivity in complex
heterogeneous landscapes: an analysis with a GIS-based random walk model. Ecogeography
19:97–106
Schneider MF (2001) Habitat loss, fragmentation and predator impact: spatial implications for prey
conservation. J Appl Ecol 38:720–735
Schumaker NH (1996) Using landscape indices to predict habitat connectivity. Ecology 77:
1210–1225
Sheaves M (2005) Nature and consequences of biological connectivity in mangrove sytems. Mar
Ecol Prog Ser 302:293–305
Sisk TD, Haddad NM, Ehrlich PR (1997) Bird assemblages in patchy woodlands: modeling the
effects of edge and matrix habitats. Ecol Appl 7:1170–1180
Stamps JA, Buechner M, Krishnan VV (1987) The effects of edge permeability and habitat geom-
etry on emigration from patches of habitat. Am Nat 129:533–552
Starr RM, Sala E, Ballesteros E et al (2007) Spatial dynamics of the Nassau grouper Epi-nephelus
striatus in a Caribbean atoll. Mar Ecol Prog Ser 343:239–249
Taylor DS, Reyier EA, Davis WP et al (2007) Mangrove removal in the Belize cays: effects on
mangrove-associated fish assemblages in the intertidal and subtidal. Bull Mar Sci 80:879–890
Taylor PD, Fahrig L, With KA (2006) Landscape connectivity: a return to the basics. In: Crooks
KR, Sanjayan M (eds) Connectivity conservation. Cambridge University Press, Cambridge
Tewfik A, Bene C (2003) Effects of natural barriers on the spillover of a marine mollusc: implica-
tions for fisheries reserves. Aquat Conserv 13:473–488
Tischendorf L (2001) Can landscape indices predict ecological processes consistently? Landsc
Ecol 16:235–254
Tischendorf L, Fahrig L (2000) How should we measure landscape connectivity? Landsc Ecol
15:633–641
Treml E, Halpin P, Urban D et al (2008) Modeling population connectivity by ocean currents, a
graph-theoretic approach for marine conservation. Landsc Ecol 23(S1):19–36
Turner MG (1989) Landscape ecology the effect of pattern on process. Annu Rev Ecol Syst
20:171–197
Turner MG (2005) Landscape ecology: what is the state of the science? Annu Rev Ecol Evol Syst
36:319–344
Turner MG, Gardner RH, O’Neill RV (2001) Landscape ecology in theory and practice: pattern
and process. Springer-Verlag, New York
530 R. Grober-Dunsmore et al.
Turner SJ, Hewitt JE, Wilkinson MR et al (1999) Seagrass patches and landscapes: the influence
of wind-wave dynamics and hierarchical arrangements of spatial structure on macrofaunal sea-
grass communities. Estuaries 22:1016–1032
Underwood AJ, Chapman MG, Connell SD (2000) Observations in ecology: you can’t make
progress on processes without understanding the patterns. J Exp Mar Biol Ecol 250:97–115
Urban D, Keitt T (2001) Landscape connectivity: a graph theoretic perspective. Ecology
82:1205–1218
Urban DL (2005) Modeling ecological processes across scales. Ecology 86:1996–2006
Vanderklift MC, How J, Wernberg T et al (2007) Proximity to reef influences density of small
predatory fishes, while type of seagrass influences intensity of their predation on crabs. Mar
Ecol Prog Ser 340:235–243
Verweij MC, Nagelkerken I (2007) Short and long-term movement and site fidelity of juvenile
Haemulidae in back-reef habitats of a Caribbean embayment. Hydrobiologia 592:257–270
Vierweij MC, Nagelkerken I, Hol KEM et al (2007) Space use of Lutjanus apodus including move-
ment between a putative nursery and a coral reef. Bull Mar Sci 81:127–138
Ward TJ, Vanderklift MA, Nicholls AO et al (1999) Selecting marine reserves using habitats and
species assemblages as surrogates for biological diversity. Ecol Appl 9:691–698
Wiens J (1989) Spatial scaling in ecology. Funct Ecol 3:385–39
Wiens JA (2006) Connectivity research – what are the issues? In: Crooks KR, Sanjayan M (eds)
Connectivity conservation. Cambridge University Press, Cambridge
Wiens JA, Milne BT (1989) Scaling of landscapes in landscape ecology, or landscape ecology from
a beetle’s perspective. Landsc Ecol 3:87–96
Wiens JA, Stenseth NC, Vanhorne B et al (1993) Ecological mechanisms and landscape ecology.
Oikos 66:369–380
Wiens JA, Van Horne B, Noon BR (2002) Landscape structure and multi-scale management. In:
Liu J, Taylor WW (eds) Integrating landscape ecology into natural resource management. Cam-
bridge University Press, Cambridge
With KA (1997) The application of neutral landscape models in conservation biology. Conserv
Biol 11:1069–1080
With KA (2002) The landscape ecology of invasive spread. Conserv Biol 16:1192–1203
With KA, Crist TO (1995) Critical thresholds in species responses to landscape structure. Ecology
76:2446–2459
With KA, Gardner RH, Turner MG (1997) Landscape connectivity and population distributions in
heterogeneous environments. Oikos 78:151–169
Wu JG (2006) Landscape ecology, cross-disciplinarity, and sustainability science. Landsc Ecol
21:1–4
Zeller DC (1998) Spawning aggregations: patterns of movement of the coral trout Plectropomus
leopardus (Serranidae) as determined by ultrasonic telemetry. Mar Ecol Prog Ser 162:253–263
Zeller DC, Russ GR (1998) Marine reserves: patterns of adult movement of the coral trout (Plec-
tropomus leopardus (Serranidae)). Can J Fish Aquat Sci 55:917–924
