Content uploaded by Simon James Pittman
Author content
All content in this area was uploaded by Simon James Pittman on Dec 17, 2021
Content may be subject to copyright.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 391
k
k k
k
391
13
Human Ecology at Sea: Modelling and Mapping
Human-Seascape Interactions
Steven Saul and Simon J. Pittman
13.1 Introduction
All life forms, including humans, modify the environment in which they live in order to
accommodate their needs and desires. A recent archaeological study suggests that there
are no longer any places on earth unaffected by human activity and that this has been the
case for thousands of years (Boivin et al. 2016). Yet, the true size and scope of our eco-
logical footprint remains largely unknown (White 1967). Studies that have mapped the
global footprint of humans on the ocean suggest that most of the global ocean space is
now affected by cumulative impacts from human activities generated from both land
and sea-based activities (Figure 13.1; Halpern et al. 2008). All too often, insufficient
emphasis and funding is placed on understanding the human dimensions of marine
ecosystems (Branch et al. 2006; Fulton et al. 2011). is is an ongoing challenge in
marine spatial planning where ‘the social landscape of the marine environment is undoc-
umented and remains a missing layer in decision-making’ (Martin & Hall-Arber 2008).
Insufficient attention to the socioeconomic drivers of systems structure and function
has sometimes led to either the failure of, or unintended consequences of, management
measures meant to balance human benefit with resource sustainability (Hanna 2001;
Acheson 2006; Clark 2006; Milner-Gulland 2012).
Where human-seascape information is available it is typically accompanied by high
uncertainty due to data-collection methods and patchy coverage in time and space.
When managing natural resources, the costs of uncertainty are derived from how
well managers can predict human response to regulations under consideration. e
challenges of working with uncertainty are exemplified by the debate over the amount of
warming degrees the earth may experience in response to anthropogenic atmospheric
carbon emissions. For instance, the change in warming degrees is estimated by the
Intergovernmental Panel on Climate Change to be between 1.5 and 4.5 ∘Coverthe
next century (IPCC 2014). As a result of uncertainties, there is great deliberation over
the economic and social effects such warming may have on the costs and benefits
humans receive from nature (Wagner & Weitzman 2015). Costs can manifest through
overpreparing for potential environmental changes that never occur (e.g. reinforcing
buildings to withstand stronger storms, which never occur in an area), or by under-
preparing and suffering damages that result from changing climatic conditions (e.g.
flooding and coastal erosion). Predictability is complicated by adaptive system feedback
Seascape Ecology, First Edition. Edited by Simon J. Pittman.
© 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 392
k
k k
k
392 Seascape Ecology
Very low impact (<1.4) Medium impact (4.95–8.47) High impact (12–15.52)
Very high impact (>15.52)
Medium high impact (8.47–12)Low impact (1.4–4.95)
Figure 13.1 Map from Halpern et al. (2008) illustrating cumulative human impact across 20 ocean
ecosystem types. Source: Reproduced with permission from the American Association for
Advancement of Science.
interactions including human behavioural responses and how behaviour affects the
environment (Bath 1998; Wagner & Weitzman 2015). As a result, management of
seascapes is largely concerned with evaluating risks to ecosystem structure and func-
tion with a focus on directly managing human activities rather than the resources. In
fisheries-focused EBM, human behaviour is often considered a major driver of change
and a source of uncertainty facing fisheries managers (McCay 1978; Fulton et al. 2011;
Berkes 2012).
Understanding what people do to impact seascapes and how they respond to
seascape structure, function and change is integral to a holistic approach to seascape
ecology (see Chapter 12 in this book). e well-known microbiologist, Robert L.
Starkey, argued that whatever people do, whether considered to be good or bad for our
environment, happens in the course of natural events (Starkey 1976). Starkey argued
that humans evolved as part of nature such that all human activity can be considered
natural, even if we sometimes don’t like or agree with the actions or their ecosystem
effects. is inclusive perspective, which will later be defined as a key principle in a
holistic systems-based approach, recognizes humans as animal members of the system,
interacting with the biotic and abiotic features of their ecosystem (see Chapter 12 in
this book). In marine ecosystem-based management (EBM), people are considered a
part of nature, with human activities recognized as important and influential ecological
processes within a complex adaptive and coupled social-ecological system (Berkes &
Folke 1998; Liu et al. 2007; see also Chapter 12 in this book).
In this chapter, we discuss tools and concepts from spatial ecology that can be used
to investigate geographic patterns of human activity and influence across the seascape.
e premise of this discussion will focus around the fact that humans are a part of the
ecosystem rather than an external force operating on it. is chapter focuses on provid-
ing the reader with quantitative spatial techniques, including some familiar to landscape
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 393
k
k k
k
Human Ecology at Sea 393
ecologists, which can be used to measure human use across the seascape to advance a
holistic understanding of marine systems and to support effective marine stewardship.
In doing so, the dynamic coupling that exists between people and the surrounding envi-
ronment is highlighted and integrated approaches that involve interdisciplinary teams
(ecology, economics, sociology, anthropology) are advocated. Emphasis is placed on
techniques that help researchers and managers understand the spatial patterning of
seascapes and the distributions of human activities at sea. As such, many of the tech-
niques discussed in this chapter will be applicable to quantifying both human activity, as
well as other ecosystem processes. is chapter will first discuss the importance of eval-
uating spatial patterns and considering spatial scale when conducting natural resource
evaluations. Second, the chapter introduces several quantitative approaches that can
be applied across the seascape to evaluate the spatial and temporal patterns of human
use. Each approach is accompanied by examples to demonstrate how the technique was
applied in the context of marine spatial management.
13.2 Seascape Ecology, Spatial Patterns and Scale
e concepts and methods in this chapter are discussed within the context of seascape
ecology, which refers to the application of the theoretical and analytical frameworks
of landscape ecology to the marine environment (Pittman et al. 2011). Landscape
ecology, now a mature discipline, has been defined in different ways throughout its
evolution (Wu 2013; Turner 2015). A common thread, as the discipline evolved, has
been the integration and quantification of human use patterns. Naveh & Lieberman
(1994) defined landscape ecology as a ‘branch of modern ecology that deals with the
interrelationship between man and his open and built-up landscapes’ (Naveh & Lieber-
man 1994). In search of principles to integrate human cultural aspects into landscape
ecology, Nassauer (1995) proposed that ‘human landscape perception, cognition and
values directly affect the landscape and are affected by the landscape’. On land, much
work has focused on integrating landscape ecology into urban planning because of its
emphasis on the interrelationship between landscape patterns and socio-ecological
processes at different scales, while encouraging place-based research that integrates
ecology with planning, design and other social sciences (Leitão & Ahern 2002; Wu
2008; Pickett et al. 2016). ese holistic socio-ecological perspective in landscape
ecology, with a focus on understanding how landscape patterns influence wellbeing
and sustainability, also hold promise for application to the marine environment.
Applying the principles of landscape ecology to human use of the seascape involves
studying the spatial patterns, interactions and bidirectional feedbacks that occur
between people and the marine ecosystems they use and affect. is extension of
landscape ecology also acknowledges that human use patterns are influenced by
people’s knowledge of the spatial and temporal distribution of marine resources and,
conversely, in many locations, human use patterns directly affect the spatial and tem-
poral distributions of marine resources. Given the complexities of these processes and
their interactions, the techniques outlined in this chapter will revolve around applying a
systems-based approach to seascape ecology. Application of a systems-based approach
has long been advocated for and applied in the terrestrial ecology realm (Naveh 2000;
Oreszczyn 2000; Atkins et al. 2011), but has more recently found favour with the
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 394
k
k k
k
394 Seascape Ecology
marine community under the banner of the ‘ecosystem approach’ to assessment and
management. In theory, the ecosystem approach is meant to ‘address the various
natural and anthropogenic pressures faced by the key components of marine systems
simultaneously’ (Link & Browman 2014).
Over the years, the majority of ecosystem-based research has disproportionally
focused on studying and modelling the nonhuman components of the ecosystem,
with less emphasis placed on understanding and modelling human behaviour. Many
of the modelling efforts that have been developed as part of the ecosystem modelling
paradigm have not quantitatively represented human behaviour and its nuances in
a spatially explicit and ecologically meaningful way. For example, ecosystem-based
fisheries models have gone to great lengths to develop biological realism, such as
the interaction of predator-prey relationships, together with planktonic and detrital
dynamics and although complex spatial dynamics in fisher behaviour has long been
recognized (Schaefer 1957; McCay 1978), conventional fisheries models have typically
represented fishing effects very simply, often just by applying a uniform mortality
value across spatially and temporally heterogeneous populations. is is sometimes
due to high uncertainty in the reliability of the fisheries data and an unfamiliarity with
the importance of spatial and temporal heterogeneity in fishing effort. ough the
emphasis towards replicating relevant complex biological dynamics in the ocean is
critically important, such models are increasingly being sought after to provide resource
management advice. As a result, equal attention toward representing the complexities
of human behaviour, as that given to the other biotic and abiotic components is becom-
ing increasingly important. A number of innovative approaches have recently been
explored to improve our understanding and representation of the human component
within ecological models such as in developing more realistic bioeconomic fisheries
fleet models and ‘whole of system’ models integrating social-ecological systems that
incorporate aspects of human behaviour and implications for human wellbeing (e.g.,
Fulton 2010; Griffith et al. 2012; Fulton et al. 2015; Kroetz & Sanchiro 2015; see also
Chapter 12 in this book). In spatially and temporally heterogeneous systems, which
experience fishing, the spatial dispersal of the harvesting sector is just as important to
model as other dynamic ecosystem processes such as biological dispersal processes.
A movement to integrate social and ecological systems is underway in ecology, but
rarely do these models examine the consequences of spatially explicit seascape patterns
for human behaviour at sea. An unusual early exception is that of Sanchirico & Wilen
(1999), whereby a spatially explicit bioeconomic model was developed that incorporated
complex spatial patch dynamics. To address questions such as: How does the spatial
pattern of effort depend upon the type of biological system the industry is exploiting? the
authors characterized a resource population structure using concepts from metapopu-
lation biology with a focus on patchiness, heterogeneity and interconnections among
and between patches. In this three-patch model system, a patch was defined as a loca-
tion in space that contains, or has the potential to contain, an aggregation of biomass.
Patches are connected by dispersal processes and are affected by the spatial distribu-
tion of harvesting effort. Dispersal or diffusion mechanisms linking space and time are
a critical component of spatial-dynamic systems (Smith et al. 2009). is modelling
approach allowed the model to examine edge effects where patches existed in a lin-
ear arrangement and the flow of biomass between patches at the edges and interior
could be examined, as well as implications for economic parameters including the cost
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 395
k
k k
k
Human Ecology at Sea 395
of exploitation. By quantifying patchiness and connectivity, the model revealed that
the spatial pattern of effort for a mobile and economically responsive fishing fleet is
driven fundamentally by patch-specific cost-price ratios. Of additional interest is that
the model recognizes gradient structures existing between patches in both economic
(i.e., movements of fishing effort) and biological variables (i.e., movements of marine
biomass) created by density. Sanchirico & Wilen (2005) compared spatial (patchy) and
nonspatial (spatially uniform) solutions finding significant differences suggesting that a
spatially differentiated approach allowing interaction between spatial gradients in eco-
nomic and biological variables will reduce errors in estimates of fishing effort, har vesting
and biomass levels across space. In addition to consideration of spatial heterogene-
ity in human behaviour, studies are now also beginning to consider heterogeneity in a
wide variety of linked variables such as income, prices and ecological impacts. Work-
ing on the red sea urchin fishery in California, Wilen et al. (2002) demonstrated an
economics-based model of fisher behaviour (participation and spatial choice) linked to a
biological model of metapopulation dynamics to show that modelling spatial behaviour
strongly affects the predicted outcomes of management policies. To this end, more work
is needed to improve the representation and quantification of human dynamics and the
linkages to other structural patterning of the seascape in more realistic spatially complex
environment, both as part of already built ecosystem models and as standalone analyses
in support of management across the seascape.
13.2.1 Scale and Scaling
Scale refers to the spatial, temporal, quantitative, or analytical dimensions used to
measure and study any phenomenon (Gibson et al. 2000). Scale effects and scaling is
an equally important consideration in the study of human-focused spatial patterns as it
is in any other topic in seascape ecology (see Chapter 4 of this book). Scale mismatches
between dynamic ecological systems and social scales informing human behaviour and
decision making in resource management are widespread and can result in suboptimal
performance in management objectives (Gibson et al. 2000; Cumming et al. 2006;
Wilson 2006). e spatial distribution of a natural resource typically determines how
humans are distributed geographically (Luck 2006; Boivin et al. 2016), although the
correlation between the spatial distribution of human uses and the spatial distribution
of the targeted resource is contingent on how much information resource users have
about the spatial distribution of the resource (Alessa et al. 2008). Some landscape
ecologists focusing on human dimensions have recommended that environmental
patterning be scaled by human processes (Nassauer 1995). is can be achieved by
understanding the space-use patterns over time (daily home range, monthly, seasonal,
annual etc.) and can be measured and analysed with telemetry and home range esti-
mators much the same as we would do for any other animal in movement ecology (see
also Chapter 12 in this book). In this way we can map the boundaries of an ecological
neighbourhood for a person or group of people. Metcalfe et al. (2016) collaborated with
fishers in the Republic of Congo to map and characterize human space-use behaviour
(distance offshore and water depth) using Global Positioning System tracking devices
equipped with motion detectors.
It is important to quantify and report the scale(s) selected (geographical / temporal
extent and spatial / temporal resolution) to anchor the study in a specific space and
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 396
k
k k
k
396 Seascape Ecology
time domain, although this is rarely done. is will avoid ambiguity and facilitate
comparative analyses and to help investigate scale dependency in ecological relation-
ships (Schneider 2001; see also Chapter 4 in this book). Ultimately, scale of a study
should not be selected arbitrarily but instead must be guided by a process of interest,
the research question(s) and information need. Additional important fundamental
considerations related to scale and scaling are: (i) how does scale, extent and resolution
affect the identification of patterns? (Wu 2004); (ii) how does scale affect the explanation
of social phenomena? (iii) can theories and interpretations of system function derived
from observations at one scale be generalized to another scale?
Matching the spatial scales at which sampling, assessment and management of a
resource occurs is critical for producing appropriate regulatory actions that best fit
both the managed resource and the users of that resource (Sale 1998). is is especially
relevant, for example, when establishing spatial regulations such as marine protected
areas and marine spatial planning, yet rarely considering in a meaningful, quantitative
way (Claudet et al. 2010). us, there is a great need to carefully identify and define
the appropriate spatial and temporal scales relevant to a specific process or research
question as an initial step in any seascape ecology study.
13.3 Human Use Data Types and Geographical Information
Systems
ere are generally two types of data used when considering human behaviour and
decision making: stated preference and revealed preference. Stated preference data
refers to information that was collected from individuals under hypothetical scenarios.
is could be data collected by implementing a questionnaire, or a choice experiment
where respondents must choose what they prefer from a series of hypothetical options.
Revealed preference data is information collected by observing people’s actual deci-
sions in the real world and the conditions under which they make these decisions.
Experiments can be conducted to collect this kind of information, but must occur
in real life settings (i.e., real time and space). Here we discuss some techniques for
gathering information on human use patterns and the benefits and pitfalls of some of
these methodologies. Tools designed to collect information for other purposes can,
at times, also be used to quantify human use patterns across the seascape. Finally, in
many cases, existing data can be leveraged in creative ways to understand human use.
e benefits and challenges to working with existing data (often referred to as ‘data of
opportunity’) are discussed.
e first question one needs to consider is how does one know that they have the ‘right’
data? is is largely dependent on the research and management questions under con-
sideration and the time and spatial scales that they encompass. e data used to answer a
research question or develop management strategies should reflect the important com-
ponents of the system (both human and natural) that are affected and addressed by the
research question or management action under consideration. Attempting to represent
and account for every single process within a system doesn’t make sense and is typically
limited by the available data.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 397
k
k k
k
Human Ecology at Sea 397
13.3.1 Mapping Human Behaviour across the Seascape
Often, the best predictors of future human behaviour can be understood by studying the
past decisions made under a similar set of circumstances (Aarts et al. 2006; Ouellette &
Wood 1998). us, forecasting human behaviour, such as how people may respond to a
proposed regulation, can be understood and modelled from past observations of human
decision making made under a variety of conditions. Later sections of this chapter will
discuss how decision-making behaviour can be modelled. Panel datasets, those with
repeated observations of people’s decisions made under different environmental condi-
tions, are often used to develop models of human behaviour. Often, panel datasets (often
referred to as revealed preference data) may not be directly available, but can be devel-
oped by combining and rearranging several different existing datasets. For example, as
mentioned above, the fishing industry is legally required to submit their logbook data
to the government, which contains daily information on when and where they fished,
what and how much they caught, as well as the gear they used, how much and how long
they spent fishing with that gear (revealed preference data). Although the purpose for
collecting this information as legally mandated is to calculate trends in catch and popu-
lation abundance, the data can also be used to address three important socioeconomic
questions: when people fish, where they fish and when they return to port. In order to
forecast decision making under different sets of conditions, one must compute the prob-
ability that such decisions were made in the past, across a range of different conditions.
In order to accomplish this, a panel dataset must be constructed by combining the log-
book data with information on weather conditions, economic conditions (such as fuel
and fish prices), regulatory measures and their spatial knowledge of relative resource dis-
tribution. e information from these different data sources should be merged together
based on common spatial and temporal criteria (Saul & Die 2016). Revealed preference
data (that representing observed human decision making) is often not available, or does
not exist. In this case, stated preference data can be collected to help quantify human
behaviour. Although conducting fieldwork in coastal communities to gather this kind
of information is typically informative and fun, like any other field campaign, it poses a
number of challenges. First, such work needs to be funded, which can be costly, espe-
cially if your study site is in a remote location, or far from your home. Second, any
work with human subjects, even if it is just asking them to complete a survey, must
be approved by an institutional review board to be sure it complies with ethical stan-
dards. ird, it can be difficult to recruit participants to your study, as they may be
reluctant to share their personal information, even if they are ensured anonymity. is is
especially true if mistrust exists between resource users and management agencies and
the information requested in your study is perceived by resource users as being used
against them in some way. It can also be difficult to find mutually available times and
connect with resource users, especially if their work requires them to be at sea for long
periods. Fourth, data provided through questionnaires or interviews carry inherent lim-
itations such as recall bias, when respondents, who are asked to reflect back on their past
behaviour, tend to provide a more optimistic view of their past behaviours and decisions
(Trumbo et al. 2011; Loomis 2014; see Podsakoff et al. 2003 for review of questionnaire
limitations). Respondents also tend to recall events in the distant past, as if they had
occurred more recently (e.g., telescope bias, Dex 1995).
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 398
k
k k
k
398 Seascape Ecology
Despite these limitations, in many cases, gathering field data is sometimes the only
way to understand human use patterns across the seascape. Two instruments that can be
used to collect quantitative data on human behaviour include questionnaires and stated
preference experiments. Questionnaires are advantageous in that they are practical and
cost effective to implement, allow the researcher to collect a large volume of quantitative
informationinarelativelyshortperiodoftimeandcanbeusedasanexploratorytoolto
create new theories or develop hypotheses. Stated preference experiments involve using
game theory to present an individual with choices in an experimental setting (typically
using a computer simulation), under different sets of simulated environmental and eco-
nomic conditions and recording the decisions they make under those conditions. is
allows the researcher to understand how a respondent’s choice may change under dif-
ferent state conditions, or different attribute characteristics and is a way of developing a
de facto panel dataset when one does not exist (Train 2009). Discrete choice experiments
tend to be more expensive (compensation to incentivize participation and pay someone
to develop the software) and are more time consumptive of participants (usually taking
hours to complete), in comparison to questionnaires. However, the data collected dur-
ing discrete choice experiments are typically considered more robust then that collected
during a questionnaire because the responses are thought to be less biased by social
desirability (Fifer et al. 2014).
13.3.1.1 Remote Sensing
In recent years, remote sensing from satellites and aircraft have made it possible to
acquire large amounts of spatially explicit data in a reasonably fast and cost-effective
process. is is a particularly useful tool for remote locations (Douvere 2008). Satellite
data has been used creatively to map human use patterns and interactions with the envi-
ronment (Geoghegan et al. 1998). e most common use of the technology with respect
to mapping human resource use has been to observe spatial change detection over time
(for review articles see Singh 1989; Coppin et al. 2004; Tewkesbury et al. 2015; Willis
2015). Spatial change detection refers to the comparison of two, or more, georectified
images of the same location to determine how the distribution and use of the habitat
in that location has changed overtime. While chance detection studies and techniques
are useful towards understanding how human activity has affected the landscape or
seascape over time, they typically do not measure or observe human activity directly and
cannot unravel the mechanisms that drive human behaviour (Fox et al. 2003). Instead,
remote sensing data can measure the spatial context of social phenomena, by cata-
loguing the magnitude and direction of changes caused by a human activity (Patino &
Duque 2013). In order to understand the mechanisms behind these changes, some stud-
ies have linked remote sensing data with other social science research techniques such
as surveys and interviews in order to elucidate the anthropogenic mechanisms driving
observed changes (Dennis et al. 2005; Taubenböck et al. 2009; Herrmann et al. 2014).
e coupling of remotely sensed information with socioeconomic data has also been
used to track the spread of diseases such as malaria (Ohemeng & Mukherjee 2015)
and determine the risks of natural disasters on property values, such as from a flood
(Taubenböck et al. 2011), fire (Dennis et al. 2005), or erosion (Leh et al. 2013).
Remotely sensing human behaviour patterns from orbit in real time (not just how
human use patterns have shaped the landscape) is more challenging. is is because
the spatial resolution of most spaceborne sensors available for public use, the most
sensitive of which can be resolved down to about a meter, too coarse to track individual
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 399
k
k k
k
Human Ecology at Sea 399
people. Despite this, several marine-focused studies have used remote sensing tech-
nology creatively to map human use patterns. For example, Davies et al. (2016) used
satellite imagery to map night-time artificial lights from coastal development, offshore
infrastructure, shipping and fishing boats to determine exposure of marine protected
areas to light pollution. Rowlands et al. (2012) used night time light patterns detected
by spaceborne satellite to spatially characterize fishing activity. Other spectrally based
remote sensing technologies such as hyperspectral imaging from aircraft, which collects
at a finer resolution and can be further enhanced using spectral unmixing techniques,
could be very effective at measuring human activity however it has not convention-
ally been used for this purpose. One reason for this may be because deploying sensors
from aircraft are costly and gathering repeated observations over time can become very
expensive. In addition, aerial photography can also be a useful source of information,
especially if a high-resolution camera is used. With the increasing advancement of drone
technology and its availability to civilians at competitive prices, such airborne sensors
may become much easier and more cost effective to deploy. It is also noteworthy that
the military regularly uses drone technology to observe human behaviour for national
security purposes. ermal infrared sensors may also be a useful tool for sensing human
use of the seascape. While difficult to use over land since emissivity varies as land type
changes, over the oceans, emissivity is known and nearly constant. As a result, any addi-
tional heat signatures from people or ships can be detected.
A number of other remote sensors observe and quantify human use in different ways
such as micro electromechanical sensors, image sensors, radio frequencies identifiers
and pressure sensors (Amato et al. 2013). For example, some fisheries are required to
use video cameras on commercial fishing vessels in order to measure the catch and
monitor for illegal fishing activity. Mandatory satellite vessel monitoring systems (VMS)
are becoming more commonplace on ships and have improved the ability to moni-
tor marine protected area compliance and determine human use patterns across the
seascape (Pedersen et al. 2009). VMS units report a vessel’s location back to a satel-
lite at regular time intervals and these data are now becoming available to scientists
and resource managers for use mapping fishing patterns across the seascape (Witt &
Godley 2007). Another similar initiative is a project called Global Fishing Watch oper-
ated by SkyTruth, a not-for-profit organization that uses data transmitted to a satellite
from a ship’s Automatic Identification System (AIS), an electronic unit used by ships
to avoid collisions at sea during reduced visibility. Similar to VMS data, AIS has been
used to enforce spatial closures and for marine spatial planning in Kiribati (Figure 13.2;
McCauley et al. 2016). Here, six months of post-closure monitoring revealed only one
case of fishing activity in the Phoenix Island Protected Area (PIPA), and this vessel was
fined by regulatory authorities in Kiribati. Vessel characteristic information is typically
transmitted together with location information for both AIS and VMS. Organizations
(such as SkyTruth and NOAA Fisheries) are using this information, together with their
location, to develop algorithms that determine when a vessel is engaged in fishing activ-
ity with its gear deployed. Such spatially explicit information can be used to map marine
corridors for vessel activity, identify and measure edge effects, such as where fishing
intensity is concentrated along the boundary of no-take marine reserves (a spatial phe-
nomenon known as fishing the line) and evaluate the displacement of fishing effort
(Stelzenmüller et al. 2008).
Aerial surveys have been used to help quantify the magnitude and spatial distribution
of recreational human use across the seascape. In south Florida, aerial surveys are
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 400
k
k k
k
400 Seascape Ecology
PIPA
EEZ
Fishing days
1. 0
(a)
0°
5°S
175°W170°W
175°W170°W
(b)
0°
5°S
8.0 ≥15.0
Figure 13.2 Automatic
Identification System (AIS) data
showing fishing intensity (purse
seine and long line) patterns before
(a) and after (b) establishing a
marine reserve Phoenix Island
Protected Area (PIPA) in Kiribati.
Source: McCauley et al. 2016.
regularly used to collect data on seascape use by private and commercial boats.
Use categories included fishing, diving and snorkeling, sightseeing, picnicking and
sunbathing, among others (Gorzelany 2005a, 2005b, 2009; Behringer & Swett 2011).
Information collected by these surveys in Florida helps to guide the development of
management measures and focus their emphasis on the areas that receive the most
use (for example, installation of mooring buoys in areas with most use). In addition,
when overlaid with habitat maps in a geographic information system (to be discussed
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 401
k
k k
k
Human Ecology at Sea 401
below), such information on human use patterns can provide valuable insight into the
types of habitats that are visited and their relative use levels. For example, aerial surveys
determined that vessels were often found to cluster around benthic features that con-
tained reef hard bottom (Behringer & Swett 2011). Similarly, another study that mapped
the spatial location of fishing boats correlated their spatial occurrence with a higher
incidence of marine debris (Bauer et al. 2008). Debris on the seafloor was associated
with the presence of ledges because the more complex seafloor relief attracted recre-
ational fishers and scuba divers, therefore linking benthic seascape structure to human
space-use patterns, behaviour at sea and consequently to measurable and undesirable
impacts.
13.3.1.2 Participatory Mapping and Spatial Analysis
Spatially explicit data is increasingly required to support spatial decision making in
marine planning (Purkis & Klemas 2011). Geographical Information Systems (GIS) are
useful ways to compile and visualize the data available, understand its spatial extent and
where it overlaps with other datasets and determine the appropriate analytical approach
for addressing research. GIS have statistical tools built in to conduct modelling and
analysis and characterize spatial uncertainty. Some spatial decision support tools have
been developed exclusively for marine spatial planning. For example, SeaSketch, an
interactive web-based application, allows users to participate in the ocean planning
process (McClintock 2013; McClintock et al. 2016). Both spatial and nonspatial data
are uploaded to the software from any participating user for inclusion in the project.
Using simple sketching tools, stakeholders can develop spatial plans and share their
plans with the group. e strength of this tool is that it allows a diverse group of stake-
holders to contribute their local knowledge and management advice for consideration.
Contingent on the kinds of data included in the project, reports can be easily generated
for each proposed management scenario that provide information on the types of
habitats protected, social and economic costs and benefits and other metrics. One of
itsmorenotableapplicationshasbeentheSafePassageProject,insouthernCalifornia,
which has worked to develop marine spatial planning solutions to balance commercial
shipping needs with conservation priorities, such as endangered whales vulnerable to
ship strikes (McClintock et al. 2016).
e SeaSketch framework is an example of participatory GIS (pGIS) which seeks
to involve stakeholders, community members, managers and scientists alike in the
production and use of geographical information (Dunn 2007; Radil & Junfeng 2016).
Examples of successful implementation of pGIS across the seascape include: deter-
mining the placement of wind farms (Mekonnen & Gorsevski 2015); resolving gear
conflicts between small-scale and large-scale fishing boats in Vietnam (inh et al.
2016); designing a marine protected area (Aswani & Lauer 2006); and resolving
conflicts among different uses of a coral reef in the coastal zone (Levine & Feinholz
2015). In the US Virgin Islands (eastern Caribbean), Loerzel et al. (2017) developed
a web-based mapping tool to harness local expert knowledge from the occupational
SCUBA diving community on coral reef characteristics, health, human uses, threat
and perceived resilience to change. e pGIS process enabled the project to compre-
hensively identify coral reefs used for commercial, recreational and scientific purposes
providing valuable information, which contributed to a spatial prioritization for
conservation actions.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 402
k
k k
k
402 Seascape Ecology
13.3.1.3 Social Sensing
A variant on pGIS, called social sensing, acknowledges that the spectral reflectance
data provided by most remote sensing platforms cannot be used to infer socioeconomic
information about the features that they sense. For example, although spectral sensors
can detect landscape and infrastructure features, such as roads, they lack the capacity
to extract the socioeconomic attributes and human dynamics about these features,
such as the movement patterns and daily activities of people when they use these
roads (Liu et al. 2015). As a result, the approach advocates combining typical remote
sensing information (such as satellite imagery) with unconventionally considered,
remotely sensed data that contains social information. Such data sources have included
taxi trajectories, mobile phone records, social media or social networking data, or
the activity on ‘smart cards’ scanned to board public transportation (Liu et al. 2015).
Studies of human mobility using cell phone locations suggested that, in general, our
movement patterns are relatively simple and predictable with a high degree of temporal
and spatial regularity characterized by return visits to a few highly frequented locations
(González et al. 2008). In the nearshore marine environment, cell phone activity could
be used to map human use across the seascape. In Estonia, a study was conducted
using cell phone data to assess the coastal locations favoured by tourists (Ahas et al.
2007), while cellphone-based reporting has been considered to gather self-report
data on recreational fishing activity to improve fisheries management (Baker &
Oeschger 2009).
13.3.1.4 Mapping Ecosystem Services
Economists and policy makers have recently embraced a construct of natural capital
accounting called ecosystem services, a mechanism to define and quantify the direct
and indirect benefits that wildlife, biodiversity, and ecosystems provide to people
for human wellbeing (Millennium Ecosystem Assessment 2005; Boyd & Banzhaf
2007). ese ecosystem services are often spatially mapped and categorized using an
accounting system as part of a valuation exercise in order to assist decision makers
in evaluating the tradeoffs between alterative management measures (Millennium
Ecosystem Assessment 2005). Quantifying and mapping ecosystem services helps
managers identify the ecosystem components and their locations that provide sub-
stantial services and merit either protection or sustainable management (Posner et al.
2016). In addition, conducting these evaluations at regular time intervals can enable
managers and business interests to identify and prepare for changes to ecosystem
service provisions (Petter et al. 2012). Spatially mapping ecosystem services contributes
to a systems-based approach toward evaluating management strategies because it
demonstrates to stakeholders and managers alike the full suite of ways that components
of the ecosystem can contribute to human wellbeing. When considering the manage-
ment question to be addressed, ecosystem services should be evaluated and mapped
at the appropriate spatial and temporal scales and can be analysed at multiple scales
(Gr̂
e
.t-Regamey et al. 2014).
e literature is replete with references to ecosystem services and varying approaches
on how to characterize and map such services (i.e., Ott & Staub 2009; Carpenter et al.
2009; Haines-Young & Potschin 2011; Staub et al. 2011; Martínez-Harms & Balvanera
2012). A review of various different decision-support tools that exist for quantifying
and valuing ecosystem services is provided by Bagstad et al. (2013). Due to the number
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 403
k
k k
k
Human Ecology at Sea 403
of disparate tools that exist, there is a need for the ecosystem services community to
develop and embrace a more standardized process to map and model ecosystem services
(Boyd & Banzhaf 2007; Crossman et al. 2013). Rarely are insights from landscape ecol-
ogy considered such as the influence of spatial configuration of the seascape on the
type, quality and value of services. Barbier (in Chapter 15 in this book) discusses the
importance of connectivity between patches of mangroves, seagrasses and coral reefs
on ecosystem service values.
One popular tool for mapping and categorizing ecosystem services is the Integrated
Valuation of Ecosystem Services and Tradeoffs (InVEST) modelling toolbox. InVEST
is a suite of models and algorithms used to map, model and value multiple ecosys-
tem services, as well as project the status of these ecosystem services under different
future scenarios. e tool can be used across various environments (terrestrial, fresh-
water and marine) and contains a number of marine specific models for understanding
how ecosystem benefits can be realized under different human use scenarios across the
seascape (Guerry et al. 2012). For example, the tool can be used to examine the effect
of altering land use on the spatial patterns of terrestrial runoff to marine waters (see
Chapter 11 in this book). In Colombia, South America, InVEST was applied to evalu-
ate the risk to coastal property due to inundation from storm surge if mangrove areas
were converted to development, and valuing the service that the mangroves provided to
local subsistence fishers. Spatial data on land elevation, bathymetry and coastal habitat
types, together with historical storm information on wind speeds, directions and wave
heights were used to model inundation scenarios. Model results showed that areas cur-
rently occupied by mangrove forest would experience flooding during a typical storm,
however water would not infiltrate beyond the mangrove forest. A 20% reduction in
mangrove cover predicted minimal inundation on lands adjacent to the mangrove forest,
with land loss valued at 2.4 million US dollars (Figure 13.3). e 40% and 60% reduction
in mangrove coverage scenarios, however, forecast a substantial portion of land in the
northeast part of the property to be at least partially flooded.
13.4 Modelling Human-Seascape Interactions with a
Systems Approach
e use of quantitative and spatially explicit modelling tools to understand and quantify
the dynamics within a system and how components interact with one another consti-
tutes a systems-based approach to ecology. e systems-based philosophy emphasizes
how ecosystem function can be influenced by human intervention and use and strongly
encourages interdisciplinary collaboration between biologists, ecologists, economists,
anthropologists and all relevant stakeholder groups (Berkes & Folke 1998; Glaeser et al.
2009; see also Chapter 12 in this book). is section will introduce the reader to different
types of modelling frameworks, connect the data sources discussed in previous sections
to their use and application in these modelling frameworks and provide some examples.
For purposes of discussing seascape ecology and human behaviour, the term ‘model’
(and by extension modelling) will be used to refer to a computer-based program that
imitates or represents the operations of real-world processes within a system. What is
meant by the term ‘system’ is largely dependent on the objectives and location of a study.
In the context of seascape ecology, a system usually refers to the collection of entities and
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 404
k
k k
k
404 Seascape Ecology
10°18ʹN
10°18ʹN
10°18ʹN
10°18ʹN
40% Mangrove reduction
Status quo 20% Mangrove reduction
60% Mangrove reduction
75°36ʹW
012
km
N
75°36ʹW
75°36ʹW75°36ʹW
Legend
Land
Water
Mangroves
Inundation due to mangrove removal
Flood potential – No mangrove removal
Figure 13.3 InVEST model results mapping the predicted effect of three different mangrove removal
scenarios on coastal inundation during storm events. Mangroves are located in all flood potential
areas, coloured in dark blue, under the status quo scenario of no mangrove removal.
processes (i.e., fish, currents, seagrass, fishing vessels, oil rigs, etc.)thatinteractwithone
another (Law 2007). e state of a system is defined as a set of variables that describe the
characteristics of a system at a given point in time (Law 2007). For example, on a given
day these can represent the weather conditions such as wind speed, oceanic conditions
(e.g., wave height, sea temperature), economic conditions (e.g., unemployment rate or
annual income), or regulatory conditions.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 405
k
k k
k
Human Ecology at Sea 405
Systems modelling is an important analytical approach for scientists and resource
managers to understand and use, as it is one of the most effective ways to communi-
cate system complexity and interdependencies to a broad audience. Scenario modelling
is useful for addressing different management hypotheses, quantifying the costs and
benefits of each management alternative and understanding how they may affect the
management of ecosystem services. For example, systems modelling can inform and
quantify the degree that ecosystem restoration may re-establish the functions of direct
value to humans and the ability of such ecological systems to cope with future distur-
bance (Moberg & Rönnbäck 2003; Weijerman et al. 2015).
13.4.1 Custom-built Statistical Models
Ecological resource assessments typically use custom-built statistical modelling
approaches to represent the ecological and human components of a system. e strat-
egy behind employing this class of model for resource assessment and management
is to estimate the parameters that describe the natural and human components of the
system and their interactions, by finding a numerical solution to the suite of equations
governing these processes. e collections of equations, which represent the ecological
and human processes (i.e., animal growth, animal reproduction, resource harvest), are
often linked together using a likelihood function. e parameters in the model are
numerically adjusted by the computer in an iterative process (maximum likelihood),
until the likelihood value cannot be further maximized, at which point an estimated
solution is reached. Data providing observations of the natural system are incorporated
into the model and the computer algorithm adjusts the parameter values in order to
most closely replicate (i.e., fit) the patterns seen in the data. Once a solution is reached,
the resulting estimated parameters are assumed to represent the current state of that
system. Using these parameter estimates, the possible effect of various management
scenarios can be forecast into the future under different assumed state conditions
(Fournier et al. 2011).
Many fisheries stock assessment models are custom-built statistical models. An
example is the tool, Stock Synthesis, most frequently used by NOAA Fisheries to
evaluate the status of fish populations. is tool includes a population simulation
model to represent abundance and mortality, an observation model that relates the
simulated population to the observable data (landings, fish size and abundance trends)
by simulating the fishing process and fitting the simulated fishery observations to
the real data and a statistical model that adjusts the parameters to achieve the best
fit (Methot & Wetzel 2012). e final set of estimated parameters is then used to
forecast the fish harvest levels and population trends one could expect under different
proposed management measures. e book published by Quinn & Deriso (1999) is a
comprehensive reference on this subject.
ough the utility of sophisticated stock assessment tools such as Stock Synthesis have
proven useful in successfully managing fisheries resources around the world, they often
poorly represent the seascape ecology discipline in several important ways. Firstly, most
stock assessment models only represent the dynamics of a single species without consid-
ering its predator-prey interactions. Secondly, most stock assessments do not consider
the spatial dimensions of either the fish or the fishers, though Stock Synthesis can handle
large-scale spatial partitions. Predator-prey dynamics and spatial considerations in fish-
eries are typically reserved for ecosystem-based models (such as Atlantis, Ecopath with
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 406
k
k k
k
406 Seascape Ecology
Ecosim) and not built into regular single species assessments. is is largely due to
the lack of available data to parameterize such processes and the time that it takes to
put together such elaborate ecosystem models. Furthermore, ecosystem-based fisheries
models are still in their infancy. eir growth and development has largely been made
possible in recent years through the availability of inexpensive computing power. As
such, they often take years to parameterize and can be difficult and unstable during
optimization due to the sheer magnitude of parameters to estimate and the processes
they represent.
irdly and most importantly, mainstream fisheries modelling platforms have over-
simplified their representation of human behaviour. Research on the behaviour and
decision-making processes of fishers has been given considerably less attention in
comparison to advances in our understanding of marine ecology (Branch et al. 2006).
In most cases, the intricacies of the fishing process are reduced down to a function
representing the size of fish a particular fishing gear can capture (selectivity), the
probability of capture by the gear (catchability) and a fishing mortality parameter. In
reality however, we know that fishers are heterogeneous and don’t necessarily operate
as rational profit maximizers (Eggert & Kahui 2013; Nguyen & Leung 2013). Gaps in
understanding of fisher behaviour have sometimes led to unintended and surprising
responses to management interventions and have been identified as a key impediment
to progress in ecosystem-based fishery management (Fulton et al. 2011).
13.4.2 Predefined Statistical Routines
e second category of models to discuss relative to human use of the seascape is pre-
defined statistical models. is category is very broad and encompasses models that
may or may not be spatial, may fit to time series or spatial data, or both and can range
in complexity from large likelihood maximization exercises, to simple t-tests, ANOVA,
or linear regression. Predefined statistical models assume that the observed data you
are analysing (i.e., the size of 100 snapper collected from the Atlantic Ocean) was col-
lected from and is representative of a greater population (i.e., the size distribution of
all the snapper fish in the Atlantic Ocean). is greater population is meant to repre-
sent all of the items in the universe from which you have a small sample (McCullagh
2002). e assumptions behind predefined statistical models are usually quite strict and
assume that your data (and the variability contained in your data) sufficiently approxi-
mates a particular probability distribution, which is the same distribution as that found
in the greater population. Statistical inference is defined as the process of fitting scien-
tific data to a variety of statistical models and, comparing the fit of these models to select
the one that best approximates the process you are trying to represent (Burnham and
Anderson 1998). is can be used to compare different parameterizations of both pre-
defined statistical routine models, as well as custom-built models as described above.
Akaike’s Information Criterion (AIC) has emerged as the current standard for compar-
ing competing statistical models. e text by Burnham and Anderson (1998) provides
an excellent guide for making inferences from scientific data using statistical modelling.
ere are numerous software tools available for statistical analysis and each have their
strengths and weaknesses. Further, many of these tools have spatial components that
can be used to evaluate spatially explicit data collected across the seascape (i.e., ESRI’s
ArcMap,eR-Project,MATLAB,SPSS,Stata,etc.). A seascape specific, statistically
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 407
k
k k
k
Human Ecology at Sea 407
driven decision support tool for marine spatial planning is MARXAN, which stands
for Spatially Explicit Annealing. MARXAN is standalone software, which uses statisti-
cal optimization routines to generate different possible systems of marine reserves. Its
strength lies in its ability to achieve user specified conser vation and socioeconomic goals
by objectively determining the size, shape and geographic placement of protected areas.
e statistical algorithm accomplishes this by minimizing the costs and maximizing the
benefits across many potential planning options (Esfandeh et al. 2015). MARXAN and
it’s built in statistical optimizer, has been used in a variety of settings to determine the
effective size and placement of marine protected areas, given the ecosystem services
provided though human use of the seascape (Fraschetti et al. 2009).
e field of spatial statistics is an area where seascape ecology has greatly benefited
from its land-based counterpart. Much of the methodologies that make up the current
suite of spatial statistics arose out of necessity to answer natural resource related ques-
tions across different terrestrial disciplines. ese included mining engineering and soil
sciences, which facilitated the development of geostatistics and spatial prediction to
identify the location of mineral resources and agriculture and forestry, which facilitated
the development of randomization and block designs (Gelfand et al. 2010). An increased
interest in spatial and space-time problems, likely ushered in through the availability
of inexpensive computers, has led to the growth and application of spatial statistics
to marine spatial planning. Where point data are patchy and a continuous surface is
more desirable, spatial prediction techniques can rapidly provide spatial information
on patterns of human use across the seascape using techniques such as inverse distance
weighting, variogram modelling and kriging , generalized linear modelling-based regres-
sion and regression kriging (Cressie 1993; Dale & Fortin 2014).
13.4.3 Discrete Choice Models
A special group of predefined statistical models used to quantify and predict human
behaviour are discrete choice models (also referred to as random utility models). ese
econometric models use binomial and multinomial logit and probit functions to rep-
resent and describe the decision-making process of people. Such models rely on either
revealed or stated preference data in the form of a panel dataset for estimation as dis-
cussed earlier in this chapter. Like all statistical models, the factors included in discrete
choice models to describe the state variables that affect an individual’s decision mak-
ing typically don’t include all of the factors that go into the decision-making process.
Humans are complex creatures and the factors that motivate our behaviours and deci-
sions, are not always apparent or observable. As a result, the unexplained factors that
go into an individual’s decision making are grouped into an error term, which is not
observed. us, the probability that an individual making a decision will select a par-
ticular outcome from the set of all possible outcomes is calculated as a probability.
Information detailing how to structure and estimate these types of models is best pro-
vided by Train (2009).
Specific to our human relationship with the seascape, these models have been used
in a variety of ways including to assess the nonmarket benefits of seagrass restoration
(Börger & Piwowarczyk 2016), value recreational beach use (Lew & Larson 2008),
manage tourism use of marine resources (Mejía & Brandt 2015), understand spatial
fishing patterns (Abbott & Wilen 2011; Alvarez et al. 2014; Davies et al. 2014; Saul &
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 408
k
k k
k
408 Seascape Ecology
Die 2016) and study how multiple seascape uses can best be managed (Tidd et al. 2015).
Recently, Saul & Die (2016) used discrete choice models to quantify the drivers behind
the decision-making behaviours of commercial fishers who participate in the reef fish
fishery on the West Florida Shelf. ree main decisions were modelled: participation,
site choice and trip termination.
A panel dataset was constructed by merging daily logbook observations with other
datasets containing state information for a given day or time period. Vessel logbook data
contained information on when a vessel started a fishing trip, the location they fished,
what they caught and when they returned to port. Data on vessel characteristics were
incorporated from NOAA’s vessel operating unit data. Landing sites and port locations
for each trip were obtained from corresponding NOAA dealer records. Spatially explicit
daily wind speed came from the NOAA National Data Buoy Center. Real weekly diesel
fuel price came from the Energy Information Administration and was corrected by the
Consumer Price Index (CPI). e daily price of fish and expected revenue were calcu-
lated from landings data for the five reef fish species that affect markets and corrected
by the CPI. Expected revenue at each potential fishing location choice was calculated for
these five fish species as the product CPI adjusted fish price, monthly variation in abun-
dance and each vessel’s relative fishing power. Regulatory history was obtained from
the most recent NOAA Fisheries Southeast Data Assessment and Review process stock
assessments for commercially important reef fish species and was incorporated into the
decision models (Saul & Die 2016).
Results show that some factors identified as statistically significant influence the
spatial and temporal distribution of fisher effort (i.e., weather such as wind speed, price
changes, fuel costs, expected revenue). For example, the top two panels in Figure 13.4
depict the partial probabilities that different wind speeds contribute to the participation
decision and the partial probability that filling up the fish hold contributes to a decision
to return to port. e bottom two panels in Figure 13.4 show the effect of fuel price
and wind speed on fishing location choice, with red circles depicting the locations
that were statistically significant for each effect. Estimated model results were used to
parameterize human behaviour in an agent-based simulation model (further discussed
below). Behavioural studies such as this are important to help inform scientists and
managersaboutthehumanusepatternsthatoccuracrosstheseascapeinresponseto
new regulations or changing environmental conditions. Awareness of the magnitude
and direction of such human behaviours can help reduce management uncertainty and
avoid surprising and unexpected responses to newly implemented fishing regulations
(Branch et al. 2006; Fulton et al. 2011; Saul & Die 2016).
13.4.4 Simulation Modelling
In many cases, the system being studied may be highly complex and, as such, analytical
or numerical solutions may not be attainable. When this is the case, the system must
then be studied using simulation modelling. In simulation modelling, the model is iter-
atively run with different sets of inputs to study how these different input combinations
affect a priori specified output performance measures. Like statistical models, simula-
tion models can be broken down into numerical (mathematic or statistical) simulation
models and agent-based simulation models. Simulation modelling is an effective tool in
operations research as it allows the user to recreate a problem and explore alternative
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 409
k
k k
k
Human Ecology at Sea 409
Wind speed
Wind speed (knots)
Fuel price
Longitude Longitude
Wind speed
Multiple lines represent weekdays and weekends
Ratio
10
–88 –86 –84 –82 –88 –86 –84 –82
20
24 26 28 30
24 26 28 30
30 40 50
ProbabilityLatitude
Latitude Probability
0.5 1.0 1.5
Regulated fisheries open
Regulated fisheries closed
Regulated fisheries open
Regulated fisheries closed
Trip choice HL: Panhandle
Site choice longline Site choice HL: Florida West Coast
Return to port HL: West Coast
Ratio catch to fish hold
0.00
0.0 0.5 1.0 1.5
0.04 0.08
Figure 13.4 Discrete choice model fit results showing some of the factors that influenced various
modelled fisher decisions off the west coast of Florida. The top two panels show the partial
probabilities that different wind speeds and catches relative to fish hold capacity affect the decisions
to take a fishing trip and return to port respectively. Red points represent the observed values, while
black lines are model fits. Bottom panels spatially depict the fishing locations for which parameters for
fuel price and wind speed were significant using red circles. Source: Saul & Die (2016).
ways that the system may respond to different possible interventions. In many cases, it
is not physically possible, moral, legal, or cost effective to conduct such experiments in
the real world (i.e., test what would happen if all the sharks were removed from a sys-
tem, test whether adding an extension to a factory would increase productivity). As a
result, simulation modelling is increasingly recognized as an important decision sup-
port tool. Due to their complexity, simulation modelling lends itself well to representing
the dynamics of coupled human-natural systems across the seascape.
e book by Law (2007) is one of the most comprehensive guides devoted to all
aspects of simulation modelling. Law (2007) recognizes that developing a simulation
model is not a linear process and in many cases, the model development process itself is
iterative, where the developer may need to go back a previous step before progressing.
In this spirit, Law (2007) outlines ten steps practitioners follow when developing a
simulation model (Figure 13.5). is workflow, when used to develop models to test
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 410
k
k k
k
410 Seascape Ecology
Collect data and
define a model
Conceptual
model
valid?
No
Ye s
Construct a
computer program
and verify
Programmed
model
valid?
Make pilot runs
Design
experiments
Make production
runs
Analyse output
data
Document, present
and use results
No
Formulate
problem and plan
study
Figure 13.5 Steps to conduct a sound simulation study. Source:
Adapted from Law (2007).
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 411
k
k k
k
Human Ecology at Sea 411
potential marine management measures, is sometimes referred to as management
strategy evaluation (Butterworth & Punt 1999; Sainsbury et al. 2000). Note that the two
main points of revision occur in validating the model assumptions and validating the
model itself. A clear list of model assumptions should be produced when the available
data reflecting the processes being modelled has been reviewed and when the simu-
lation model structure has been defined. is should take place before programming
begins to avoid the need to reprogram later. Validation of the model itself is another
critical step and one that is often inquired about in great detail by those the results
of the model may affect (i.e., policy makers and other stakeholders). Model validation
simply means comparing the patterns that the model produces and forecasts, with
actual human patterns across the seascape.
ere are various ways to validate a simulation model. If data are available from the
real world that represents aspects of the system you are trying to model, then you could
compare performance measures and patterns produced by the simulation with those of
the actual system. When validating a model in this way, the data from the system being
compared to the simulation model should not have been used in any way to parameter-
ize the simulation model. A model’s robustness can also be evaluated by conducting
sensitivity analyses, which explores how a range of input values for each parameter
affects the model’s performance. ose factors found during sensitivity analysis to sig-
nificantly impact performance measures, need to then be considered very carefully and
where possible, verified against data for accuracy. Finally, the analyst together with stake-
holders and subject matter experts should review the model carefully for correctness.
Involvement of stakeholders early and often throughout model development, validation
and implementation is critically important to achieving the goals behind developing the
model in the first place.
13.4.5 Agent-based Models
Agent-based modelling (also sometimes referred to as individual-based modelling or
IBMs) is an approach to simulating systems (see also Chapter 8 in this book). ese
models comprise autonomous, interacting individuals, with characteristics and
behaviours, parameterized at the individual level. Applications of agent-based mod-
elling (ABM) have been used in a variety of disciplines (from economics and social
science to biology and ecology) and commercial applications (military, transportation,
etc.) to model complex systems and study the emergent properties that result from
individuals’ interactions and behaviours (Bankes 2002; Bonabeau 2002). With ABMs,
an individual’s behaviours and interrelationships are defined by rule-based expressions.
Agents that are similar may be grouped into the same class, with which they share a
collection of behaviours. e state of a particular agent is defined by discrete values
contained within attribute variables. ough an agent may belong to a particular class,
each agent possesses a unique set of attribute values. Patterns emerge from system
dynamics as agents interact with one another and their surrounding environment at
discrete time intervals (Grimm et al. 2005).
is modelling approach is well suited to studying complex ecological systems both
in the terrestrial and aquatic realms, where overall system dynamics are highly complex,
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 412
k
k k
k
412 Seascape Ecology
nonlinear and contain many parameters to estimate (Uchmanski & Grimm 1996; Grimm
1999). Such systems can be very difficult to represent using equation-based, numeri-
cal / statistical simulation models due to their nonlinearity. is may cause the optimizer
to find a local maxima or minima and think a solution has been reached, when in real-
ity, a better solution actually exists but was not found by the algorithm. e agent-based
modelling approach is not meant to replace equation-based modelling, but is comple-
mentary and can incorporate fitted parameters from equation-based models to repre-
sent individual agent behaviours and ecosystem function (Fahse et al. 1998; Parunak
et al. 1998; Wilson 1998). Key challenges to address relate to the type of information
used to parametrize decision making, which can be a set of rule-based decisions based
on theory or on observations of real decision making. Whether the parametrization
is entirely statistical versus using adaptive behavioural rules or some hybrid will affect
model outcomes (Filatova et al. 2013).
13.4.6 Pattern-oriented Modelling
Validation of agent-based models is often disproportionately criticized compared
to their numerical counterparts, because agent-based models do not fit to data
or optimize a likelihood function. erefore, parameter fitting and model rigor can-
not be evaluated using typical fitting metrics such as likelihood ratio tests or AIC.
Despite this, however, Grimm et al. (2005) developed a framework for validating
such models called pattern-oriented modelling. Although developed specifically for
agent-based models, this technique can and should be used to validate all models. In
addition, the pattern-oriented modelling approach is well suited to the discussion in
this chapter, which has considered human use spatial patterns across the seascape.
e pattern-oriented approach involves comparing alternative models of the same
system and process using what Grimm (2005) refers to as ‘inference’. e process
involves implementing alternative theories of an agent’s decisions using various
characteristic patterns at both the individual and higher levels and then testing how
well these models reproduce the patterns seen in the real world. Models that fail to
reproduce the characteristic patterns observed in nature are rejected and improved
upon, while models that replicate the human and natural patterns observed in the
system are considered to be robust to the rules and assumptions used (Grimm et al.
2005, 2006).
Comparison of the patterns emergent from the simulation model, with those observed
in the system for which there is data, can be done in various different ways depending
on the data that is available. Traditional practitioners who are new to agent-based
modelling may feel more comfortable with a statistical comparison. For example, an
agent-based model developed to represent the migration of reef fish in the Gulf of
Mexico was validated using a tagging study conducted in the same location. Firstly,
a potential model of fish location was developed and, in the simulation, a simulated
tagging study was conducted analogous to that conducted in the real world for which
data was available. Secondly, results from the simulation were statistically evaluated
against the tagging data by using Wilcoxon rank sum or Student’s ttests to compare
the following emergent patterns represented as probability distributions: linear-
movement speed between the tagging and recapture locations, the distance moved
between the tagging and recapture locations, the distribution of fish lengths between
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 413
k
k k
k
Human Ecology at Sea 413
Empirical linear speed Simulated linear speed
Empirical tagging distance Simulated tagging distance
1.0 2.0 3.00.01. 0
Grid cells/day
Grid cells Grid cells
Grid cells/day
2.0
α = 0.7
β = 0.2
α = 0.8
β = 0.1
α = 0.7
β = 13.6
α = 0.8
β = 13.9
3.00.0
0 50 100 150 200 0 50 100 150 200
00 100 200 300 400
15005000 0 500
Frequency
FrequencyFrequency
Frequency
1500
100 300
Figure 13.6 Statistical comparison of population level patterns emerging from an agent-based model,
with population level patterns from field collected empirical data. Simulated linear fish speeds
(Wilcoxon rank-sum test: p=0.2933) and tagging distance (Wilcoxon rank-sum test: p=0.780)
emergent from the simulation model for red grouper are compared to the same metrics calculated
from a field tagging experiment. Alpha and beta are the parameters of the fitted gamma distributions
used to approximate the simulated and empirical probability density functions depicted in the graphs.
Source: Saul et al. (2012).
tagging and recapture and the spatial distribution of recaptured fish (Figure 13.6;
Saul et al. 2012).
If data comparing simulated output with real world observations is not available or
accessible to the investigator, a more visual approach may be used. For example, the
agent-based simulation model of reef fish in the Gulf of Mexico that was described
above, also modelled the two primary fishing fleets that capture reef fish. As alluded
to earlier in this chapter, individual vessel behaviour in the simulation model was
parameterized using fitted discrete choice models. As the model was developed and
refined, spatial results mapping fisher use of the seascape in the simulation were visually
compared with plots of vessel monitoring system (VMS) data using pattern-oriented
modelling (Figure 13.7). Statistical comparison was not able to be conducted because
we were not permitted access to the VMS data itself due to confidentiality of the
data. Consequently, note that as a result, VMS data was not used to parameterize
the simulation in any way. Visual comparison of these two maps and their spatial
patterns show that the underlying processes programmed and parameterized that
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 414
k
k k
k
414 Seascape Ecology
Simulated red grouper catch
30° N
87° W 84° W
0
81° W
27° N
24° N
87° W 84° W 81° W
Red grouper catch (Ibs.)
320 plus
24° N
27° N
30° N
0 50 100 200
km
N
88°W
GM_A32_Gag_VMS9 mxd; NMFS_SERO
Red grouper VMS tracks (2008–2009)
Time-area closures landings (ww)
25°N
26°N
27°N
28°N
29°N
30°N
87°W 86°W
< 10 TP
FL
< 25 TP
< 50 TP
< 100 TP
< 150 TP
0 12.525 100 Miles7550
N
EW
S
85°W 84°W 83°W 82°W 81°W
25°N
26°N
27°N
28°N
29°N
30°N
88°W 87°W 86°W 85°W 84°W 83°W 82°W 81°W
U.S. EEZ
Figure 13.7 Visual comparison of simulated emergent spatial catch patterns of red grouper predicted
by an agent-based model, with actual spatial catch observations of red grouper provided by vessel
monitoring system (VMS) data. Source: Saul (2012).
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 415
k
k k
k
Human Ecology at Sea 415
govern individual fish and vessel behaviours and decisions, when they interact, recreate
the spatial patterns observed in the real system. us, one can conclude that the
individual level behaviours are reflective of what may occur on an individual level in
the real system (Saul 2012).
ere are many other examples from fishery science in which agent-based modelling
was the approach of choice. One such model is a spatially explicit ABM for modelling
vessel movements and alternative effort allocation patterns of Danish fishing vessel
activities (Bastardie et al. 2010, 2014). e effects of effort reallocation on total fuel
consumption and energy efficiency in relation to the quantity and value of landings
were investigated for alternative scenarios of fuel-saving behaviour. Another ABM was
used to model the effects of an individual transferable quota system in a multispecies,
multisector fishery and the model was applied to the Coral Reef Fin Fish Fishery
in Queensland, Australia (Little et al. 2009). Other examples include (i) a model
of a fishery targeting different species in different areas, developed to analyse the
implications of taking into account the response of fishing fleets to regulatory controls
(Soulie & ebaud, 2006); (ii) an interactive model designed to better understand the
interactions between regional and local drivers strongly influencing the health of coral
reefs in the Yucatan peninsula, Mexico (Perez et al. 2009); and (iii) an agent-based
fishery management model of Hawaii’s longline fishery (Yu et al. 2009).
13.5 Conclusions and Future Research Priorities
e knowledge gap on human use patterns and decision-making behaviour across the
seascapes and the ecological and economic consequences of these human-seascape
interactions presents both a problem hindering effective spatial management and an
exciting frontier for new interdisciplinary research. Clearly, more work is needed to
better operationalize the human dimension in seascape ecology (Samhouri et al. 2013;
Link & Browman 2014; Dolan et al. 2016). Unprecedented opportunity now exists for
the study of human-seascape relationship through theoretical advances in modelling
socio-ecological systems together with advances in geospatial technologies for spatial
data acquisition and sophisticated spatial modelling algorithms. ese improvements
are critical steps toward advancing the field of seascape ecology and progressing
ecosystem-based management, which is dependent on the inclusion of human use
pattern information.
Special attention must be focused on understanding and communicating uncertainty.
As discussed in the introduction to the chapter, uncertainty is expensive, particularly
when it leads to management failures. e cost of uncertainty can manifest itself either
in foregone use of ecosystem services provided by the seascape, which could have been
more aggressively exploited, or conversely in the unsustainable overuse of resources
and their depletion. Cressie (1993) begins his book on spatial statistics by referring to
statistics as the ‘science of uncertainty’, which ‘attempts to model order in disorder’
(Cressie 1993). We can extend this mantra to include the suite of quantitative tools
discussed in this chapter. It is important to be aware that increasing the complexity
of a model may not necessarily reduce uncertainty because uncertainty in one com-
ponent can propagate and multiply as it cascades through the whole system. As pointed
out by Oreskes (2003) ‘a complex model may be more realistic, yet more uncertain’.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 416
k
k k
k
416 Seascape Ecology
In some cases, a modular approach may be appropriate for building systems models
where greater attention is given to the development of specialist submodels, which are
then integrated to create the whole system. Some suggestions for how this can be done
and associated discussion of advantages and disadvantages are provided by Voinov &
Shugart (2013).
Managers and scientists often have a disproportionate understanding of the human
use patterns that occur within marine systems, in comparison to our understanding
of the biological population dynamics and oceanographic dynamics. As a result, the
complexity of human use across the seascape is often overly simplified down to one or
several simple parameters. Implementing management action without a comprehen-
sive understanding of human use patterns and how they interact with natural system
dynamics may occur for several reasons: in response to a governmental mandate or
law, pressure from lobby groups or lawsuits, or because local scientists and managers
lack the technical expertise or funding to conduct such analyses (Turner et al. 2016).
is sometimes results in the implementation of management measures that fall short
of achieving their desired conservation or management objectives. is can result in
surprising, unconsidered outcomes as resource users adapt and respond to the newly
implemented regulation in unanticipated ways (Hilborn 1985; Lane 1988; Hilborn &
Walters 1992; Branch et al. 2006; Fulton et al. 2011). us, an important recommen-
dation is to develop capacity-building initiatives for resource scientists and managers,
to train them on the use of quantitative techniques for evaluating human use patterns
across the seascape and using the results to develop management.
e quantitative tool one selects for resource and policy evaluation must match the
data available and the spatial and temporal scale of the study. In data limited locations,
analysts may only be able to view several layers of coarsely defined spatial data using
GIS software, with spatial planning based simply on where the spatial layers overlap
and therefore where competing interests intersect. In locations where sufficient infor-
mation exists, an estimation or simulation model is advised, even if it is simple, to help
scientists and managers quantify uncertainty and test the implementation of different
hypothesized management measures. Further, even a highly stylized simulation model,
meaning one that is mostly theoretical due to the absence of sufficient data on the sys-
tem, can go a long way towards helping managers and resource users better comprehend
the complexities of a system and the direction that their actions may perturb system
dynamics (Carpenter & Gunderson 2001).
Data collection campaigns and data mining initiatives should be structured to address
management questions, with model development and parameterization as the end
goal if possible. is will help identify gaps where additional information may need
to be gathered, to which data collection efforts can be refocused. Future research
priorities should focus on developing improved data collection programs that capture
information on the behaviours and decision making of human seascape use patterns
across different sectors (i.e., fisheries, mining, shipping, tourism and recreation). is
can include socioeconomic surveys, as well as theoretical and applied experiments (i.e.,
game theory). is is important as models and statistical analyses and the predictions
they provide are only as good as the data used to parameterize or fit them (i.e., garbage
in, garbage out mantra).
Collecting field data, whether biological or socioeconomic, is difficult and expensive.
All too often, data from field work ends up not being analysed, or fully exploited.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 417
k
k k
k
Human Ecology at Sea 417
is is typically due to a lack of time, personnel, computational resources, desire
and / or capacity (i.e., statistical knowledge). In a sense, the next generation of seascape
ecologists needs to be prepared to be a sort of ‘analytical jack of all trades and master
of some’ to be successful. e ability to evaluate existing data, identify gaps, collect
data to fill missing information and synthesize diverse information using statistical
analyses, GIS systems, or spatially explicit models. Furthermore, seascape ecologists
need to have a basic interdisciplinary knowledge of (and respect for) the terminology
and methodology used by other experts of a range of complimentary disciplines such
as ecologists, economists, sociologists, geographers, psychologists and others. e best
way to achieve this synergy is by establishing interdisciplinary teams of individuals
from appropriate disciplines.
Finally, where seascape ecology is being applied to management (e.g., marine spatial
planning) a transdisciplinary approach will be required whereby managers and policy
makers and possibly a range of stakeholders are included from the very beginning, dur-
ing problem conceptualization all the way through to the implementation of regulations
(Caldow et al. 2015). Many of the tools that this chapter discussed are spatial, from GIS
datasets to spatially explicit agent-based models. Spatial tools and the maps they pro-
duce are excellent ways to engage stakeholders in the mapping and modelling process.
For example, individuals watching a spatial simulation that shows fishing vessels leav-
ing a port and operating in different locations can relate to what they are seeing on the
computer screen and provide valuable feedback on the parameterization of the pro-
cesses in the model and the realism it may or may not represent. Finally, stakeholders
who have participated in the process often become important allies and help to educate
their resource using colleagues on the importance of the conservation objectives under
consideration.
Here we provide several priority areas for future research that represent new areas
of research with potential to advance the integration of human spatial dynamics into
seascape ecology and bridge the information gap from seascape ecology to marine stew-
ardship strategies such as marine spatial planning:
•As highlighted in our chapter, continued effort is required to improve the detailed
representation of spatial heterogeneity in socio-economic systems models to boost
predictive performance of human use patterns and the relationships between human
decision making and biophysical conditions.
•As model complexity increases so models become increasingly hard to analyse, under-
stand and interpret (Voinov & Shughart 2013). Consequently, research is required to
discover the optimal spatial complexity and spatial scales necessary to develop reliable
model systems.
•An important and largely unexplored focus of investigation in human-seascape rela-
tionships is the application of spatial pattern metrics to quantify human use patterns
across the seascape. Landscape ecology offers a plethora of algorithms that can be
applied to maps regardless of what they represent including socioeconomic data (i.e.,
corridors and connecting, edge effects, patchiness, patch geometry). Human distribu-
tions and activities have quantifiable patterning that can be linked to patterning in the
surrounding seascape. Taking a landscape ecology approach to investigating human
ecology will provide fresh insights into the human dimensions of seascape ecology,
its structure, function and change.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 418
k
k k
k
418 Seascape Ecology
References
Aarts H, Verplanken B, Knippenberg A (2006) Predicting behaviour from actions in the
past: repeated decision making or a matter of habit? Journal of Applied Social
Psychology 28: 1355–1374.
Abbott JK, Wilen JE (2011) Dissecting the tragedy: a spatial model of behaviour in the
commons. Journal of Environmental Economics and Management 62:
386–401.
Acheson JM (2006) Institutional failure in resource management. Annual Review of
Anthropology 35: 117–134.
Ahas R, Aasa A, Mark U, Pae T, Kull A (2007) Seasonal tourism spaces in Estonia: case
study with mobile positioning data. Tourism Management 28: 898–910.
Alessa L, Kliskey A, Brown G (2008) Social-ecological hotspots mapping: a spatial approach
for identifying coupled social-ecological space. Landscape and Urban Planning 85:
27–39.
Alvarez S, Larkin S, Whitehead J, Haab T (2014) A revealed preference approach to valuing
non-market recreational fishing losses from the deepwater horizon oil spill. Journal of
Environmental Management 145: 199–209.
Amato A, Di Lecce V, Piuri V (2013) Sensors for human behaviour analysis. In Amato A,
Di Lecce V, Piuri V (eds) Semantic Analysis and Understanding of Human Behaviour in
Video Streaming. Springer Science and Business Media, New York, NY.
Aswani S, Lauer M (2006) Incorporating fishermen’s local Knowledge and behaviour into
geographical information systems (GIS) for designing marine protected areas in
Oceania. Human Organization 65: 81–102.
Atkins JP, Burdon D, Elliott M, Gregory AJ (2011) Management of the marine
environment: integrating ecosystem services and societal benefits with the DPSIR
framework in a systems approach. Marine Pollution Bulletin 62: 215–226.
Bagstad KJ, Semmens DJ, Waage S, Winthrop R (2013) A comparative assessment of
decision-support tools for ecosystem services quantification and valuation. Ecosystem
Services 5: e27–e39.
Baker MS, Oeschger I (2009) Description and initial evaluation of a text message based
reporting method for marine recreational anglers. Marine and Coastal Fisheries:
Dynamics, Management and Ecosystem Science 1: 143–154.
Bankes SC (2002) Agent-based modelling: a revolution? Proceedings of the National
Academies of Science USA 99: 7199–7200.
Bastardie F, Nielsen JR, Andersen BS, Eigaard OR (2010) Effects of fishing effort allocation
scenarios on energy efficiency and profitability: an individual-based model applied to
Danish fisheries. Fisheries Research 106: 501–516.
Bastardie F, Nielsen JR, Miethe T. (2014) DISPLACE: a dynamic, individual-based model
for spatial fishing planning and effort displacement-integrating underlying fish
population models. Canadian Journal of Fisheries and Aquatic Sciences 71:
366–386.
Bath AJ (1998) e role of human dimensions in wildlife resource research in wildlife
management. Ursus 10: 349–355.
Bauer LJ, Kendall MS, Jeffrey CFG (2008) Incidence of marine debris and its relationships
with benthic features in Gray’s Reef National Marine Sanctuary, southeast USA. Marine
Pollution Bulletin 56: 402–413.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 419
k
k k
k
Human Ecology at Sea 419
Behringer DC, Swett RA (2011) Determining Vessel Use Patterns in the Southeast Florida
Region. Southeast Florida Coral Reef Initiative, Florida Department of Environmental
Protection, Miami Beach, FL, pp 1–87.
Berkes F (2012) Implementing ecosystem-based management: evolution or revolution?
Fish and Fisheries 13: 465–476.
Berkes F, Folke C (1998) Linking Social and Ecological Systems: Management Practices and
Social Mechanisms for Building Resilience. Cambridge University Press, Cambridge.
Boivin NL, Zeder MA, Fuller DQ, Crowther A, Larson G, Erlandson JM, Denham T,
Petraglia MD (2016) Ecological consequences of human niche construction: examining
long-term anthropogenic shaping of global species distributions. Proceedings of the
National Academies of Science USA 113: 6388–6396.
Bonabeau, E (2002) Agent-based modelling: methods and techniques for simulating human
systems. Proceedings of the National Academies of Science USA 99: 7280–7287.
Börger T, Piwowarczyk J (2016) Assessing non-market benefits of seagrass restoration in
the Gulf of Gda ´
nsk. Journal of Ocean and Coastal Economics 3(1): 1–20.
Boyd J, Banzhaf S (2007) What are ecosystem services? e need for standardized
environmental accounting units. Ecological Economics 63: 616–626.
Branch TA, Hilborn R, Haynie AC, Fay G, Flynn L, Griffiths J, Marshall KN, Randall JK,
Scheuerell JM, Ward EJ, Young M (2006) Fleet dynamics and fishermen behaviour:
lessons for fisheries managers. Canadian Journal of Fisheries and Aquatic Sciences 63:
1647–1668.
Burnham KP, Anderson DR (1998) Model Selection and Multimodel Inference, A Practical
Information-eoretic Approach (2nd edition). Springer Science and Business Media,
New York, NY.
Butterworth DS, Punt AE (1999) Experiences in the evaluation and implementation of
management procedures. ICES Journal of Marine Science 56: 985–998.
Caldow C, Monaco ME, Pittman SJ (2015) Biogeographic assessments: a framework for
information synthesis in marine spatial planning. Marine Policy 51: 423–432.
Carpenter SR, Gunderson LH (2001) Coping with collapse: ecological and social dynamics
in ecosystem management. BioScience 51: 451–457.
Carpenter SR, Mooney HA, Agard J, Caplstrano D, DeFries RS, Diaz S, Dietz T, Duralappah
AK, Oteng-Yeboah A, Pereira HM, Perrings C, Reid WV Sarukhan J, Scholes RJ, Whyte
A (2009) Science for managing ecosystem services: beyond the Millennium Ecosystem
Assessment. Proceedings of the National Academies of Sciences 106: 1305–1312.
Clark CW (2006) e Worldwide Crisis in Fisheries: Economic Models and Human
Behaviour. Cambridge University Press, New York, NY.
Claudet J, García-Charton JA, Lenfant P (2010) Combined effects of levels of protection
and environmental variables at different spatial resolutions on fish assemblages in a
marine protected area. Conservation Biology 25: 105–114.
Coppin PR, Jonckheere I, Nackaerts K, Muys B, Lambin E (2004) Review article digital
change detection methods in ecosystem monitoring: a review. International Journal of
Remote Sensing 25: 1565–1596.
Cressie NAC (1993) Statistics for Spatial Data. John Wiley & Sons, New York, NY.
Crossman ND, Burkhard B, Nedkov S, Willemen L, Petz K, Palomo I, Drakou EG,
Martín-Lopez B, McPhearson T, Boyanova K, Alkemade T, Egoh B, Dunbar MB, Maes J
(2013) A blueprint for mapping and modelling ecosystem services. Ecosystem Services 4:
4–14.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 420
k
k k
k
420 Seascape Ecology
Cumming GS, Cumming DH, Redman CL (2006) Scale mismatches in social-ecological
systems: causes, consequences, and solutions. Ecology and Society 11(1): 14.
Dale MRT, Fortin MJ (2014) Spatial Analysis: A Guide for Ecologists. Cambridge University
Press, Cambridge.
Davies TW, Duffy JP, Bennie J, Gaston KJ (2016) Stemming the tide of light pollution
encroaching into marine protected areas. Conservation Letters 9: 164–171.
Davies TK, Mees CC, Milner-Gulland EJ (2014) Modelling the spatial behaviour of a
tropical tuna purse seine fleet. PLOS One 9: e114037.
Dennis RA, Mayer J, Applegate G, Chokkalingam U, Colfer CJP, Kurniawan I, Lachowski
H, Maus P, Permana RP, Ruchiat Y, Stolle F, Suyanto, Tomich TP (2005) Fire, people and
pixels: linking social science and remote sensing to understand underlying causes and
impacts of fires in Indonesia. Human Ecology 33: 465–504.
Dex S (1995) e reliability of recall data: a literature review. Bulletin of Sociological
Methodology 49: 58–89.
Dolan TE, Patrick WS Link JS (2016) Delineating the continuum of marine
ecosystem-based management: a US fisheries reference point perspective. ICES Journal
of Marine Science 73: 1042–1050.
Douvere F (2008) e importance of marine spatial planning in advancing
ecosystem-based sea use management. Marine Policy 32: 762–771.
Dunn CE (2007) Participatory GIS – a people’s GIS. Progress in Human Geography 31:
616–637.
Eggert H, Kahui V (2013) Reference-dependent behaviour of paua (abalone) divers in New
Zealand. Applied Economics 45: 1571–1582.
Esfandeh S, Kaboli M, Eslami-Andargoli L (2015) A Chronological review on application of
MARXAN tool for systematic conservation planning in landscape. International Journal
of Engineering and Applied Sciences 2: 6–17.
Fahse L, Wissel C, Grimm V (1998) Reconciling classical and individual-based approaches
in theoretical population ecology: a protocol for extracting population parameters from
individual-based models. e American Naturalist 152(6): 838–852.
Fifer S, Rose J, Greaves S (2014) Hypothetical bias in stated choice experiments: is it a
problem? And if so, how do we deal with it? Transportation Research Part A 61:
164–177.
Filatova T, Verburg PH, Parker DC, Stannard CA (2013) Spatial agent-based models for
socio-ecological systems: challenges and prospects. Environmental Modelling and
Software 45: 1–7.
FournierDA,SkaugHJ,AnchetaJ,IanelliJ,MagnussonA,MaunderMN,NielsenA,Sibert
J (2011) AD Model Builder: using automatic differentiation for statistical inference of
highly parameterized complex nonlinear models. Optimization Methods and Software
27: 233–249.
Fox J, Mishra V, Rindfuss R, Walsh S (eds) (2003) People and the Environment: Approaches
for Linking Household and Community Surveys to Remote Sensing and GIS. Kluwer
Academic Publishers, Boston, MA.
Fraschetti S, d’Ambrosio P, Micheli F, Bussotti S, Pizzolante F, Terlizzi A (2009) Planning
marine protected areas in a human-dominated seascape. Marine Ecology Progress Series
375: 13–24.
Fulton EA (2010) Approaches to end-to-end ecosystem models. Journal of Marine Systems
81(1): 171–183.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 421
k
k k
k
Human Ecology at Sea 421
Fulton EA, Boschetti F, Sporcic M, Jones T, Little LR, Dambacher JM, Gray R, Scott R,
Gorton R (2015) A multi-model approach to engaging stakeholder and modellers in
complex environmental problems. Environmental Science and Policy 48: 44–56.
Fulton EA, Smith ADM, Smith DC, van Putten IE (2011) Human behaviour: the key source
of uncertainty in fisheries management. Fish and Fisheries 12: 2–17.
Gelfand AE, Diggle PJ, Fuentes M, Guttorp P (eds) (2010) Handbook of Spatial Statistics.
CRC Press, Boca Raton, FL.
Geoghegan J, Prichard L, Ogneva-Himmelberger Y, Chowdhury RR, Sanderson S, Turner
BL (1998) ‘Socializing the pixel’ and ‘pixelizing the social’ in land-use and land-cover
change. In Liverman DM, Moran E, Rindfuss R, Stern P (eds) People and Pixels: Linking
Remote Sensing and Social Science. Committee on the Human Dimensions of Global
Environmental Change. National Research Council, National Academies Press,
Washington, DC.
Gibson CC, Ostrom E, Ahn T (2000) e concept of scale and the human dimensions of
global change: a survey. Ecological Economics 32 (2): 217–239.
Glaeser B, Bruckmeier K, Glaser M, Krause G (2009) Social-ecological Systems analysis in
coastal and marine areas: A path toward integration of interdisciplinary knowledge.
Current Trends in Human Ecology 183(203): 183–203.
González MC, Hidalgo CA, Barabasi AL (2008) Understanding individual human mobility
patterns. Nature 453(7196): 779–782.
Gorzelany, JF (2005a) Recreational Boat Traffic Surveys of Broward County, Florida. Mote
Marine Lab, Mote Technical Report No. 1017. Sarasota, FL.
Gorzelany, JF (2005b) Assessment of Boat Traffic Patterns and Vessel Wake Effects at
Blowing Rocks Marina, Martin County, Florida, Final report. Mote Marine Lab, Mote
Technical Report No. 1058, Sarasota, FL.
Gorzelany, JF (2009) Recreational Boating Activity in Miami-Dade County. Mote Marine
Lab, Mote Technical Report No. 1357, Sarasota, FL.
Gr̂
e
.t-Regamey A, Weibel B, Bagstad KJ, Ferrari M, Geneletti D, Klug H, Schirpke U,
Tappeiner U (2014) On the effects of scale for ecosystem services mapping. PloS One 9:
e112601.
Griffith GP, Fulton EA, Gorton R, Richardson AJ (2012) Predicting interactions among
fishing, ocean warming, and ocean acidification in a marine system with
whole-ecosystem models. Conservation Biology 26(6): 1145–1152.
Grimm V (1999) Ten years of individual-based modelling in ecology: what have we learned
and what could we learn in the future? Ecological Modelling 115: 129–148.
Grimm V, Berger U, Bastiansen F, Eliassen S, Ginot V, Giske J, Goss-Custard J, Grand T,
Heinz SF, Huse G, Huth A, Jepsen JU, Jørgensen C, Mooij WM, Muller B, Pe’er G, Piou
C, Railsback SF, Robbins AM, Robbins MM, Rossmanith E, Ruger N, Strand E, Souissi S,
Stillman RA, Vabø R, Visser U, DeAngelis DL (2006) A standard protocol for describing
individual-based and agent-based models. Ecological Modelling 198: 115–126.
Grimm V, Revilla E, Berger U, Jeltsch F, Mooij WM, Railsback SF, ulke H, Weiner J,
Wiegand T, DL DeAngelis (2005) Pattern-oriented modelling of agent-based complex
systems: lessons from ecology. Science 310: 987–991.
Guerry AD, Ruckelshaus MH, Arkema KK, Bernhardt JR, Guannel G, Kim C, Marsik M,
PapenfusM,ToftJE,VerutesG,WoodSA,BeckM,ChanF,ChanKMA,GelfenbaumG,
Gold BD, Halpern BS, Labiosa WB, Lester SE, Levin PS, McField M, Pinsky ML,
Plummer M, Polasky S, Ruggiero P, Sutherland DA, Tallis H, Day A, Spencer J (2012)
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 422
k
k k
k
422 Seascape Ecology
Modelling benefits from nature: using ecosystem services to inform coastal and marine
spatial planning. International Journal of Biodiversity Science, Ecosystem Services, and
Management 8: 107–121.
Haines-Young R, Potschin M (2011) Common International Classification of Ecosystem
Services (CICES) 2011 Update. Paper prepared for discussion at the expert meeting on
ecosystem accounts organized by the UNSD, the EEA and the World Bank, London,
December 2011. EEA/BSS/07/007. European Environment Agency, Nottingham.
Halpern BS, Walbridge S, Selkoe KA, Kappel CV, Micheli F, D’Agrosa C, Bruno JF, Casey
KS, Ebert C, Fox HE, Fujita R, Heinemann D, Lenihan HS, Madin EMP, Perry MT, Selig
ER, Spalding M, Steneck R, Watson R (2008) A global map of human impact on marine
ecosystems. Science 319: 948–952.
Hanna S (2001) Managing the human-ecological interface: marine resources as example
and laboratory. Ecosystems 4: 736–741.
Herrmann SM, Sall I, Sy O (2014) People and pixels in the Sahel: A study linking
coarse-resolution remote sensing observations to land users’ perceptions of their
changing environment in Senegal. Ecology and Society 19: 29.
Hilborn R (1985) Fleet dynamics and individual variation: why some people catch more fish
than others. Canadian Journal of Fisheries and Aquatic Sciences 42: 2–13.
Hilborn R, Walters C (1992) Quantitative Fisheries Stock Assessment. Choice, Dynamics
and Uncertainty. Chapman & Hall, New York, NY.
IPCC (2014) Climate Change 2014 Synthesis Report. Contribution of Working Groups I, II
and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change (Core Writing Team, R.K. Pachauri and L.A. Meyer (eds)). IPCC, Geneva.
Kroetz K, Sanchirico J (2015) e bioeconomics of spatial-dynamical systems in natural
resource management. Annual Review of Resource Economics 7: 16.1–16.19.
Lane DE (1988) Investment decision-making by fishermen. Canadian Journal of Fisheries
and Aquatic Sciences 45: 782–796.
Law AM (2007) Simulation Modelling and Analysis. McGraw-Hill, New York, NY.
Leh M, Bajwa S, Chaubey I (2013) Impact of land use change on erosion risk: an integrated
remote sensing, geographic information system, and modelling methodology. Land
Degradation and Development 24: 409–421.
Leitão AB, Ahern J (2002) Applying landscape ecological concepts and metrics in
sustainable landscape planning. Landscape and Urban Planning 59(2): 65–93.
Levine AS, Feinholz CL (2015) Participatory GIS to inform coral reef ecosystem
management: mapping human coastal and ocean uses in Hawaii. Applied Geography 59:
60–69.
Lew DK, Larson DM (2008) Valuing a beach day with a repeated nested logit model of
participation, site choice and stochastic time value. Marine Resource Economics 23:
233–252.
Link JS, Browman HI (2014) Integrating what? Levels of marine ecosystem-based
assessment and management. ICES Journal of Marine Science 71: 1170–1173.
Little LR, Punt AE, Mapstone BD, Begg GA, Goldman B, Williams AJ (2009) An
agent-based model for simulating trading of multi-species fisheries quota. Ecological
Modelling 220: 3404–3412.
LiuJ,DietzT,CarpenterSR,AlbertiM,FolkeC,MoranE,PellAN,DeadmanP,KratzT,
Lubchenco J, Ostrom E (2007) Complexity of coupled human and natural systems.
Science 317(5844): 1513–1516.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 423
k
k k
k
Human Ecology at Sea 423
Liu Y, Liu X, Gao S, Gong L, Kang C, Zhi Y, Chi G, Shi L (2015) Social sensing: a new
approach to understanding our socioeconomic environments. Annals of the Association
of American Geographers 105: 512–530.
Loerzel J, Goedeke TL, Dillard MK, Brown G (2017) SCUBA divers above the waterline:
Using participatory mapping of coral reefs conditions to inform reef management.
Marine Policy 76: 79–89.
Loomis JB (2014) Strategies for overcoming hypothetical bias in stated preference surveys.
Journal of Agriculture and Resource Economics 39: 34–46.
Luck GW (2006) e relationships between net primary productivity, human population
density and species conservation. Journal of Biogeography 34: 201–212.
Martin KS, Hall-Arber, M (2008) e missing layer: geo-technologies, communities, and
implications for marine spatial planning. Marine Policy 32: 779–786.
Martínez-Harms M, Balvanera P (2012) Methods for mapping ecosystem service supply: a
review. International Journal of Biodiversity Science, Ecosystem Services, and
Management 8: 17–25.
McCay BJ (1978) Systems ecology, people ecology, and the anthropology of fishing
communities. Human Ecology 6(4): 397–422.
McCauley DJ, Woods P, Sullivan B, Bergman B, Jablonicky C, Roan A, Hirshfield M,
Boerder K, Worm B (2016) Ending hide and seek at sea: new technologies could
revolutionize ocean observation. Science 351: 1148–1150.
McClintock W (2013) GeoDesign: optimizing stakeholder-driven marine spatial planning.
Proceedings of the Marine Safety and Security Council, e Coast Guard Journal of
Safety and Security at Sea, Fall 2013: 63–67.
McClintock W, Paul E, Burt C, Bryan T (2016) McClintock Lab: SeaSketch. Designing for
Our Oceans, http://www.seasketch.org (accessed 15 June 2017).
McCullagh P (2002) What is a statistical model? e Annals of Statistics 30: 1225–1310.
Mejía CV, Brandt S (2015) Managing tourism in the Galapagos Islands through price
incentives: a choice experiment approach. Ecological Economics 117: 1–11.
Mekonnen AD, Gorsevski PV (2015) A web-based participatory GIS (PGIS) for offshore
wind farm suitability within Lake Erie, Ohio. Renewable and Sustainable Energy Reviews
41: 162–177.
Metcalfe K, Collins T, Abernethy KE, Boumba R, Dengui J-C, Miyalou R, Parnell RJ,
Plummer KE, Russell DJF, Safou GK, Tilley D, Turner RA, VanLeeuwe H, Witt MJ,
Godley BJ (2016) Addressing uncertainty in marine resource management; combining
community engagement and tracking technology to characterise human behaviour.
Conservation Letters 10.1111/conl.12293.
Methot RD, Wetzel CR (2012) Stock synthesis: a biological and statistical framework for
fish stock assessment and fishery management. Fisheries Research 142: 86–99.
Millennium Ecosystem Assessment. (2005) Ecosystems and Human Well-being: Synthesis.
Island Press, Washington, DC.
Milner-Gulland EJ (2012) Interactions between human behaviour and ecological systems.
Philosophical Transactions of the Royal Society B: Biological Sciences 367:
270–278.
Moberg F, Rönnbäck P (2003) Ecosystem services of the tropical seascape: interactions,
substitutions, and restoration. Ocean and Coastal Management 46: 27–46.
Nassauer JI (1995) Culture and changing landscape structure. Landscape Ecology 10:
229–237.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 424
k
k k
k
424 Seascape Ecology
Naveh Z (2000) What is holistic landscape ecology? A conceptual introduction. Landscape
and Urban Planning 50: 7–26.
Naveh Z, Lieberman AS (1994) Landscape ecology: theory and application, 2nd edition.
Springer, New York, NY.
Nguyen Q, Leung P (2013) Revenue targeting in fisheries. Environment and Development
Economics 18: 559–575.
Ohemeng FD, Mukherjee F (2015) Modelling the spatial distribution of the Anopheles
mosquito for malaria risk zoning using remote sensing and GIS: a case study in the
Zambezi Basin, Zimbabwe. International Journal of Applied Geospatial Research 6:
7–20.
Oreskes N (2003) e role of quantitative models in science. In Canham et al.(eds)
Models in ecosystem science. Princeton University Press, Princeton, NJ.
Oreszczyn S (2000) A systems approach to the research of people’s relationships with
English hedgerows. Landscape and Urban Planning 50: 107–117.
Ott W, Staub C (2009) Welfare-significant environmental indicators. A feasibility study on
providing a statistical basis for the resources policy. Summary. Environmental Studies
No. 0913. Federal Office for the Environment, Bern, Switzerland.
Ouellette JA, Wood W (1998) Habit and intention in everyday life: the multiple processes
by which past behaviour predicts future behaviour. Psychological Bulletin 124: 54–74.
Parunak HVD, Savit R, Riolo RL (1998) Agent-based modelling vs. equation-based
modelling: a case study and users guide. Proceedings of the First International Workshop
on Multi-agent Systems and Agent-based Simulation. Lecture Notes in Computer
Science 1534: 10–25.
Patino JE, Duque JC (2013) A review of regional science applications of satellite remote
sensing in urban settings. Computers, Environment and Urban Systems 37: 1–17.
Pedersen SA, Fock H, Krause J, Pusch C, Sell AL, Böttcher U, Rogers SI, Sköld M, Skov H,
Podolska M, Piet GJ (2009) Natura 2000 sites and fisheries in German offshore waters.
ICES Journal of Marine Science 66: 155–169.
Perez P, Dray A, Cleland D, Arias-González JE (2009) An agent-based model to address
coastal management issues in the Yucatan Peninsula, Mexico. Proceedings of the World
IMACS/MODSIM Congress 18: 72–79.
Petter M, Mooney S, Maynard SM, Davidson A, Cox M, Horosak I (2012) A methodology
to map ecosystem functions to support ecosystem services assessments. Ecology and
Society 18(1): 31.
Pickett STA, Cadenasso ML, Childers DL, McDonnell MJ, Zhou W (2016) Evolution and
futureofurbanecologicalscience:ecologyin,of,andforthecity.EcosystemHealthand
Sustainability 2(7): e01229.
Pittman SJ, Kneib R, Simenstad C, Nagelkerken I (2011) Seascape ecology: application of
landscape ecology to the marine environment. Marine Ecology Progress Series 427:
187–190.
Podsakoff PM, MacKenzie SB, Lee J, Podsakoff NP (2003) Common method biases in
behavioural research: A critical review of the literature and recommended remedies.
Journal of Applied Psychology 88: 879–903.
Posner SM, McKenzie E, Ricketts TH (2016) Policy impacts of ecosystem services
knowledge. Proceedings of the National Academies of Science of the USA 113:
1760–1765.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 425
k
k k
k
Human Ecology at Sea 425
Purkis S, Klemas V (2011) Remote Sensing and Global Environmental Change. John Wiley
& Sons, Ltd., Hoboken, NJ .
Quinn TJ, Deriso RB (1999) Quantitative Fish Dynamics. Oxford University Press,
New York, NY.
Radil SM, Junfeng, J (2016) Public participatory GIS and the geography of inclusion. e
Professional Geographer 68: 202–210.
Rowlands G, Purkis S, Riegl B, Metsamaa L, Bruckner A, Renaud P (2012) Satellite imaging
coral reef resilience at regional scale. A case-study from Saudi Arabia. Marine Pollution
Bulletin 64: 1222–1237.
Sainsbury KJ, Punt AE, Smith ADM (2000) Design of operational management strategies
for achieving fishery ecosystem objectives. ICES Journal of Marine Science 57:
731–741.
Sale PF (1998) Appropriate spatial scales for studies of reef-fish ecology. Australian Journal
of Ecology 23: 202–208.
Samhouri JF, Haupt AJ, Levin PS, Link JS, Shuford R (2013) Lessons learned from
developing integrated ecosystem assessments to inform marine ecosystem-based
management in the USA. ICES Journal of Marine Science 71(5): 1205–1215.
Sanchirico JN, Wilen JE (1999) Bioeconomics of spatial exploitation in a patchy
environment. Journal of Environmental Economics and Management 37(2): 129–150.
Sanchirico JN, Wilen JE (2005) Optimal spatial management of renewable resources:
matching policy scope to ecosystem scale. Journal of Environmental Economics and
Management 50(1): 23–46.
Saul S (2012) An Individual-based Model to Evaluate the Effect of Fisher Behaviour on Reef
Fish Catch per Unit Effort. PhD Dissertation. University of Miami, Miami, Florida, FL.
Saul S, Die D (2016) Modelling the decision making behaviour of fishers in the reef fish
fishery on the west coast of Florida. Human Dimensions of Wildlife 21: 567–586.
Saul S, Die D, Brooks EN, Burns K (2012) An individual-based model of ontogenetic
migration in reef fish using a biased random walk. Transactions of the American
Fisheries Society 141: 1439–1452.
Schaefer MB (1957) Some considerations of population dynamics and economics in
relation to the management of the commercial marine fisheries. Journal of the Fisheries
Research Board of Canada 14(5): 669–681.
Schneider DC (2001) e rise of the concept of scale in ecology e concept of scale is
evolving from verbal expression to quantitative expression. BioScience 51(7): 545–553.
Singh A (1989). Review article: digital change detection techniques using remotely-sensed
data. International Journal of Remote Sensing 10: 989–1003.
Smith MD, Sanchirico J, Wilen J (2009) e economics of spatial-dynamic processes.
Applications to renewable resources. Journal of Environmental Economics and
Management 57(1): 104–121.
Soulie J, ebaud O (2006) Modelling fleet response in regulated fisheries: an agent-based
approach. Mathematical and Computer Modelling 44: 553–564.
Starkey RL (1976) In the Course of Natural Events. Proceedings of the ird International
Biodegradation Symposium. Sharpley JM, Kaplan AM (eds), Applied Science Publishers
Ltd, London.
Staub C, Ott W, Heusi F, Klingler G, Jenny A, Hacki M, Hauser A (2011) Indicators for
Ecosystem Goods and Services: Framework, Methodology and Recommendations for a
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 426
k
k k
k
426 Seascape Ecology
Welfare-related Environmental Reporting. Environmental Studies No. 1102, Federal
Office for the Environment, Bern, Switzerland.
Stelzenmüller V, Maynou F, Bernard G, Cadiou G, Camilleri M, Crec’hriou R, Criquet G,
Dimech M, Esparza O, Higgins R, Lenfant P (2008) Spatial assessment of fishing effort
around European marine reserves: implications for successful fisheries management.
Marine Pollution Bulletin 56(12): 2018–2026.
Taubenböck H, Wurm M, Netzband M, Hendrik Z, Roth A, Rahman A, Dech S (2011)
Flood risks in urbanized areas – multi-sensoral approaches using remotely sensed data
for risk assessment. Natural Hazards and Earth System Sciences 11: 431–444.
TaubenböckH,WurmM,SetiadiN,GebertN,RothA,StrunzG,BirkmannJ,DechS
(2009) Integrating remote sensing and social science. Joint Urban Remote Sensing Event
2009: 1–7.
Tewkesbury AP, Comber AJ, Tate NJ, Lamb A, Fisher PF (2015) A critical synthesis of
remotely sensed optical image change detection techniques. Remote Sensing of
Environment 160: 1–14.
inh NA, Huan NC, anh NV, Tuyen LT, Ly TTP, Nam NM (2016) Spatial conflict and
priority for small-scale fisheries in near-shore seascapes of the central coast Vietnam.
Journal of Geography and Regional Planning 9: 28–35.
Tidd AN, Vermard Y, Marchal P, Pinnegar J, Blanchard JL, Milner-Gulland EJ (2015)
Fishing for space: fine-scale multi-sector maritime activities influence fisher location
choice. Public Library of Science One 10: e0116335.
Train KE (2009) Discrete Choice Methods with Simulation. Cambridge University Press,
New York, NY.
Trumbo C, Lueck M, Marlatt H, Peek L (2011) e effect of proximity to hurricanes Katrina
and Rita on subsequent hurricane outlook and optimistic bias. Risk Analysis 31:
1907–1918.
Turner BL, Esler KJ, Bridgewater P, Tewksbury J, Sitaws N, Abrahams B, Chapman FS,
Chowdhury RR, Christie P, Diaz S, Firth P, Knapp CN, Kramer J, Leemans R, Palmer M,
Pietri D, Pittman J, Sarukhán J, Shackleton R, Seidler R, van Wilgen B, Mooney H (2016)
Socio-environmental systems (SES) research: what have we learned and how can we use
this information in future programs. Current Opinion in Environmental Sustainability
19: 160–168.
Turner MG (2015) Twenty-five years of United States landscape ecology: looking back and
forging ahead. In Barrett GW, Barrett TL, Wu J (eds) History of Landscape Ecology in
the United States. Springer, New York, NY.
Uchmanski J, V Grimm (1996) Individual-based modelling in ecology: what makes the
difference? Trends in Ecology and Evolution 11: 437–441.
Voinov A, Shugart HH (2013) ‘Integronsters’, integral and integrated modelling. Journal of
Environmental Modelling and Software 39: 149–158.
Wagner G, Weitzman ML (2015) Climate Shock: e Economic Consequences of a Hotter
Planet. Princeton University Press, Princeton, NJ.
Weijerman M, Fulton EA, Janssen AB, Kuiper JJ, Leemans R, Robson BJ, van de Leemput
IA, Mooij WM (2015) How models can support ecosystem-based management of coral
reefs. Progress in Oceanography 138: 559–570.
White L (1967) e historical roots of our ecological crisis. Science 10: 1203–1207.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 427
k
k k
k
Human Ecology at Sea 427
Wilen JE, Smith MD, Lockwood D, Botsford LW (2002) Avoiding surprises: incorporating
fisherman behaviour in to management models. Bulletin of Marine Science 70(2):
553–575.
Willis KS (2015) Remote sensing change detection for ecological monitoring in United
States protected areas. Biological Conservation 182: 233–242.
Wilson WG (1998) Resolving discrepancies between deterministic population models and
individual-based simulations. American Naturalist 151: 116–134.
Wilson J (2006) Matching social and ecological systems in complex ocean fisheries. Ecology
and Society 11(1).
Witt MJ, Godley BJ (2007) A step towards seascape scale conservation: using vessel
monitoring systems (VMS) to map fishing activity. Public Library of Science One 2(10):
e1111.
Wu J (2004) Effects of changing scale on landscape pattern analysis: scaling relations.
Landscape Ecology 19(2): 125–138.
Wu J (2008) Making the case for landscape ecology an effective approach to urban
sustainability. Landscape Journal 27(1): 41–50.
Wu J (2013) Key concepts and research topics in landscape ecology revisited: 30 years after
the Allerton Park workshop. Landscape Ecology 28: 1–11.
Yu R, Pan M, Railsback SF, Leung P (2009) A prototype agent based fishery management
model of Hawaii’s longline fishery. 18th World IMACS/MODSIM Congress, Cairns,
Australia.
Trim Size: 170mm x 244mm Single Column Pittman c13.tex V2 - 08/28/2017 6:34pm Page 428
k
k k
k
