Land Use Policy 67 (2017) 608–624

Contents lists available at ScienceDirect

Land Use Policy

journal homepage: www.elsevier.com/locate/landusepol

Explaining rural land use change and reforestation: A causal-historical MARK

approach
Bradley B. Walters
Department of Geography & Environment, Mount Allison University, 144 Main Street, Sackville, N.B., E4L 1A7, Canada

A R T I C L E I N F O A B S T R A C T

Keywords: Research on human-environment interactions is bedeviled by two key analytical challenges: integrating natural
Forest transition and social science information and demonstrating causal connections between proximate and distant influences.
Research methodology These challenges can be met by adopting an event-focused, causal-historical approach to research methodology,
Abductive causal eventism (ACE) referred to here as Abductive Causal Eventism (ACE). With ACE, researchers construct causal histories of in-
Land change science
terrelated social and/or biophysical events backward in time and outward or inward in space through a process
Political ecology
Socio-ecological systems
of eliminative inference and reasoning from effects to causes, called abduction. ACE is contrasted with three
leading approaches to human-environment research: Land Change Science (LCS), Socio-ecological Systems
(SES), and Political Ecology (PE). For illustration, ACE is applied to a study of post-War environmental change in
two rural watersheds in Saint Lucia, West Indies. Findings reveal that the most consequential change has been
the widespread reforestation of lands abandoned from farming. This change occurred irrespective of the type of
land tenure, but was especially commonplace on lands with steeper slopes and further from roads. Reforestation
during the 1960s and 1970s was caused by a combination of commodity market challenges, abandonment of
subsistence cultivation in response to smaller family sizes, and sizable out-migrations of younger adults overseas.
The expansion of banana cultivation in the 1960s and then again in the 1980s slowed and in places reversed this
trend. But an especially large wave of farmland abandonment swept the island from the mid-1990s to early-
2000s because the banana export market collapsed as a result of preferential market access being eroded by a
series of WTO trade rulings. These effects have been reinforced by a surge in investment from return migrants
and the tourism industry which has drawn labour out of farming while also creating economic incentive and
political support for protecting more forests on both private estates and public lands. Yet, the post-War trend in
reforestation may have ended as agriculture displays signs of rebounding and residential and tourism devel-
opment expands unabated into the countryside. This study demonstrates the advantages of using ACE where
explanations entail diverse types of causes operating across space and over time.

1. Introduction
Calls for interdisciplinary human-environment research have mul-
tiplied as human impacts on the biosphere have grown (Lambin et al.,
2001; Ludeke et al., 2004; Newell et al., 2005; IGBP, 2006; Young et al.,
2006; Rudel, 2008; Holm et al., 2013; Ogden et al., 2013; Palsson et al.,
2013; Agnoletti and Rotherham, 2015). Despite many advances,
human-environment researchers continue to wrestle with two central
analytical challenges: first, how to integrate natural and social science
information coherently and rigorously and, second, how to confirm
causal connections between environmental changes and causal influ-
ences both proximate and distant in space and time (GLP, 2005:3;
Turner et al., 2007:20667; Pooley et al., 2013). The analytical metho-
dology, Abductive Causal Eventism (an expanded version of what was
formerly called Event Ecology), has been proposed as a way to address
these twin challenges (Walters and Vayda, 2009, 2011; Walters,
2012a).
Abductive Causal Eventism—hereafter ‘ACE’—is a form of causal-
historical analysis that entails constructing causal histories of inter-
related social and/or biophysical events backward in time and outward
or inward in space through a process of eliminative inference and
reasoning from effects to causes, called abduction. In this paper, ACE
will be described and its application illustrated with an empirical case
study of research that seeks to explain recent, widespread reforestation
in rural Saint Lucia, West Indies. Small island states like Saint Lucia
provide an opportune, living-laboratory for doing human-environment
research given the decisive historical relationships between local socio-
economic and environmental changes and wider international influ-
ences wrought by colonialism, migration and the agricultural com-
modities trade.

E-mail address: bwalters@mta.ca.

http://dx.doi.org/10.1016/j.landusepol.2017.07.008
Received 7 March 2017; Received in revised form 28 June 2017; Accepted 4 July 2017
0264-8377/ © 2017 Elsevier Ltd. All rights reserved.
B.B. Walters
The paper is organized as follows. First, I will review three metho-
dological approaches commonly applied in leading subfields of human-
environment research. Abductive Causal Eventism (ACE) will then be
described and contrasted with these. This is followed by an empirical
case study of the application of ACE to explain reforestation in Saint
Lucia. The paper concludes with a discussion of research findings and
reflections on how the case study relates to different research meth-
odologies.
2. Analytical methodologies in human-environment research
Research on the human environment is diverse and influenced by
many academic fields. In this section, I review three leading sub-fields
of human-environment research: Land Change Science (LCS), Socio-
Ecological Systems (SES) and Political Ecology (PE), highlighting their
respective theoretical orientations and analytical approaches, in each
case considering how these inform research methodology. This is
challenging given the range of research being done within each sub-
field and because there is some crossing-over between them (e.g.,
Brannstrom and Vadjunec, 2013a). Nonetheless, there are important,
broad distinctions that reflect both explicit theoretical and methodo-
logical commitments as well as underlying assumptions (Turner and
Robbins, 2008; Cote and Nightingale, 2011; Watts, 2011).
Since the focus here is methodology, it is useful first to distinguish
this from methods, theories and hypotheses. Methods are the tools and
techniques of research. Methodology is the logic and justifications
guiding deployment of methods and interpretation of research results
(Vayda and Walters, 2011:2). Theory has diverse meanings (Abend,
2008). For the purposes here, theories are explanations of empirical
phenomena that have attained some degree of generalizability by virtue
of their prior confirmation elsewhere (theories that attain extremely
high levels of confirmation may become established ‘facts’ or ‘laws’).
Hypotheses are conjectured explanations for the case at hand, so may or
may not be recognized theories.
2.1. Land change science (LCS)
Land change science (LCS) emerged as a coherent program of
human-environment research during the 1990s, propelled by a series of
international research collaborations—the Land-Use and Land-Cover
Change Project (LUCC) and then Global Land Project (GLP)—conducted
under the auspices of the International Geosphere-Biosphere Program
(Lambin et al., 1999; GLP, 2005). In short, LCS entails “the linking of
natural, social and geographic information sciences to study land sur-
face changes and their consequences” (IGBP, 2006:29).
A stated priority of LCS is advancing understanding of the causes of
land change (Geist et al., 2006:42). To do this, researchers employ a
range of methods and analytical tools, but typically approach in-
vestigations inductively by searching for general patterns within
quantitative data sets using inferential statistics and modeling (e.g.,
Chowdhury and Turner, 2005; Mena et al., 2006; Mendoza et al., 2011;
Redo et al., 2012; Silva et al., 2016; see Turner and Robbins, 2008:6.7-
6.8). As empirical findings have multiplied, however, general theories
and models have emerged and introduced a more explicitly deductive1
bent to some LCS work (Veburg et al., 2006:120). Forest Transition
Theory (FTT) is perhaps the best known example of this (Mather, 1992;
Rudel et al., 2005; Turner et al., 2007). Most LCS researchers now
presume elements of FTT are credible and some design studies, not to
explain changes in forest cover per se, but rather to test the validity of
FTT as it is currently understood (Perz, 2007; Barbier et al., 2010;
1 4
Inductive reasoning is the drawing of generalizations or probabilities from empirical
observations (including experiments); e.g., every farmer surveyed in this village had planted
tree crops; therefore, all farmers plant tree crops in this village. Deductive reasoning uses ex-
isting generalizations or theories to account for empirical observations; e.g., all farmers
plant tree crops; this is a farmer; therefore, she plants tree crops.
609
Land Use Policy 67 (2017) 608–624
Lambin and Meyfroidt, 2010; Costa et al., 2017).
LCS researchers recognize that causes of land use change are often
complex and site-specific, with both proximate and underlying factors
(‘drivers’) needing to be considered (e.g., Lambin et al., 2001; Geist
et al., 2006; Turner et al., 2007; Dalla-Nora et al., 2014; Shaver et al.,
2015; Plieninger et al., 2016). Qualitative methods are not categorically
ruled-out, but the guiding research strategy for LCS is explicitly
founded on the application of quantitative methods and data collection,
preferably standardized across case studies (Geist et al., 2006; Lambin
et al., 2006). This approach has encouraged use of social science
methods common to quantitative sociology and economics, notably
regression modeling of discrete variables (i.e., factors) using data de-
rived from national statistics or large sample, pre-structured household
surveys. Using these methods, LCS studies have shown statistical as-
sociations between land use/land-cover changes and various socio-
economic and demographic factors.
However, such statistical approaches can be severely limiting where
explanations of change entail multiple, conjunctural causes, as they
often do (Ragin, 1987; Gaddis, 2002; Goba, 2008; Ylikoski and
Kuorikoski, 2010; Efroymson et al., 2016). In the absence of in-depth,
contextual knowledge of the actual cases under study, such analyses are
best viewed as exploratory as they tend to generate findings which beg
more questions than they answer about why identified factors actually
correlate or not with land use decisions, practices or land-cover changes
in specific places (Walters and Vayda, 2009; see also Redo et al.,
2012:799-802).2 Thus, Geist et al. (2006:45) acknowledge the obstacles
created for LCS where factors “crucially important in explaining change
in one place may be irrelevant in other nearby places” and where “a
given factor may be implicated in opposite land-cover outcomes” (see
also Meyfroidt, 2015).3 Modeling is also challenged by the fact that
causal dynamics of land change in a given place change over time
(Aspinall, 2004). As argued below, LCS researchers facing confounding
or ambiguous results like this might benefit from adopting a more ex-
plicitly, causal-historical approach to their analysis (cf. Freedman,
1991; Goba, 2008; Ebach et al., 2016).
2.2. Socio-ecological systems (SES)
Socio-ecological Systems (SES) thinking emerged in the late 1990s,
the result of collaboration between ecologists and social scientists
seeking to bring a more holistic, ecosystem-inspired perspective to re-
search on the human environment (Berkes and Folke, 1998; Adger,
2000; Gunderson and Holling, 2002; Folke, 2006). The SES approach is
founded on appeals to a loose cluster of conceptual heuristics and
theoretical propositions, adapted mostly from systems thinking in the
natural sciences (Holling, 1996; Turner et al., 2003a; Walker et al.,
2006; Liu et al., 2007; Leslie and McCabe, 2013).4 In this respect, SES
differs from LCS in being more deductively-oriented in its analytical
approach, although among researchers there is little consensus re-
garding the ontological status of theory in SES. For some, SES concepts
and ideas comprise the building blocks of a solid theoretical foundation
(actual or in-the-making), but others appear to view them as just ana-
lytical heuristics (Folke, 2006; Walker et al., 2006; Ostrom, 2009).
Consider, for example, the SES concept of resilience which has
2
Quantitative-statistical methods suffer from other limitations beyond the scope of this
paper to discuss (see Ragin, 1987; Freedman, 1991, 1997; McKim, 1997; Gaddis, 2002:
chap. 4; Goba, 2008; Barnes, 2013; Nuzzo, 2014).
3
For example, the influence of tenure on land use and tree cover is surprisingly am-
biguous: tenure insecurity may encourage or discourage tree planting and other land
investments depending on wider circumstances or historical contingencies (Walters et al.,
1999; Walters, 2012b).
SES researchers advocate a diverse array of systems concepts to aid in analysis,
among these resilience, adaptive capacity, adaptive cycle, response diversity, biocomplexity,
reciprocal effects, feedbacks, non-linearity, indirect effects, emergent properties, path de-
pendency, vulnerability, thresholds, cross-scale interactions, heterogeneity, time lags and legacy
effects.
B.B. Walters
attained prominence in many academic and policy circles (Downes
et al., 2013; Neocleous, 2013). Like the concept of sustainable devel-
opment, resilience serves as a ‘boundary object’, drawing to it a variety
of scholars seeking to reconcile the social and ecological within an in-
tegrated analytical framework (Brand and Jax, 2007).5 Resilience has
much normative appeal, but there is little scientific or policy consensus
on its actual meaning (Grimm and Wissel, 1997; Adger, 2000; Brand
and Jax, 2007; Downes et al., 2013; Baggio et al., 2015). The more
widespread its use, the less consistent are the meanings for which it is
used (Neocleous, 2013). In fact, the conceptual vagueness of resilience,
while analytically dubious, is arguably a source for much of its appeal
(Brand and Jax, 2007; Strunz, 2012; Basken, 2013).
Another prominent conceptual proposition within SES is the
‘adaptive cycle’, a model of organic succession whereby ecosystems and
social systems (or socio-ecological systems) are thought to pass in se-
quence through four distinct phases: (i) growth/expansion; (ii) con-
servation; (iii) disturbance/collapse; and (iv) reorganization, which is
then followed by a return to growth, etc. (Gunderson and Holling,
2002; Folke, 2006). As with resilience, there are disagreements as to
whether the adaptive cycle constitutes a heuristic schema or something
more solidly theoretical and explanatory (Walker et al., 2006; Burns
and Rudel, 2015). Either way, the adaptive cycle has a teleological
quality and is also conceptually vague in that it lacks specific criteria by
which to clearly define and demarcate different phases of the cycle
(Bunce et al., 2009:223). What constitutes a system ‘collapse’, ‘re-
organization’, etc., thus seems too much in the eyes of the beholder
(e.g., Winkel et al., 2016).
SES approaches have also been criticized for their inability to
meaningfully incorporate history, politics and other qualitative factors
whose influences are difficult to measure discretely and usually origi-
nate beyond the temporal or spatial boundaries of the designated
‘system’ (Armitage and Johnson, 2006; Nadasdy, 2007; Bunce et al.,
2009; Fraser and Stringer, 2009; Hornborg, 2009:251-55; Davidson,
2010; Kirchhoff et al., 2010; Jennings, 2011; Watts, 2011; Clement,
2012; Cote and Nightingale, 2012; Widgren, 2012a, 2012b; Galaty,
2013; Homewood, 2013; Leslie and McCabe, 2013; Moritz, 2017). SES
also offers little concrete guidance beyond its broad conceptual heur-
istics and general frameworks for practicing researchers. An extensive
review of resilience studies found no consistency in research methods or
methodology (Downes et al., 2013; see also Binder et al., 2013).
Arguably, the most explicit articulation of methodology in SES en-
tails the application of a ‘coupled human and natural systems’ approach
(CHANS), wherein researchers are encouraged to compile empirical
data to build models that incorporate natural and social factors so that
reciprocal interactions and feedbacks between variables can be mapped
and assessed (Turner et al., 2003a; Liu et al., 2007; McConnell et al.,
2011; Carter et al., 2014). CHANS is to some extent an attempt to
synthesize SES principles and frameworks with the formal modeling
approaches common to LCS. CHANS is explicit in recognizing that im-
portant interactions often occur between the local and non-local
(‘across scales’). Yet, CHANS remains analytically unfocused and is
notable for its lack of clear criteria with which to assess the relative
boundedness, frequency and recurrence of specified human-environ-
ment relationships. It thus remains unduly arbitrary how systems are
bounded and components selected for inclusion or not within them.
Also, distinctions are not readily made between genuinely systemic
interactions (‘couplings’) and more contingent interactions that may
constitute only singular or irregular (but nonetheless important) causal
events (Vayda and Walters, 2011:17).
SES and CHANS have heuristic value by virtue of their fostering
5
Resilience is arguably also a ‘buzzword’, a term whose power “… derives from a
combination of ambiguous meaning and strong normative resonance” (Cairns and
Kryzwoszynska, 2016). The editors of the journal, Nature, caution that buzzwords may
“… obfuscate even as they pretend to enlighten” (Editors, 2016:140).
610
Land Use Policy 67 (2017) 608–624
integrative, interdisciplinary thinking and analysis. They may also be
appropriate for research where detailed, descriptive understanding is
sought and where resources permit development of complex data sets
and models. But such initiatives are costly and complex to implement,
and so impractical for most researchers.6 And as the case study pre-
sented below illustrates, they may simply be ill-suited for answering
explanatory why questions where relevant causal information is not
meaningfully captured or quantifiable within a systems framework of
analysis (Walters, 2012a).
2.3. Political ecology (PE)
Political ecology (PE) includes a range of theoretical perspectives
and analytical approaches that are variously informed by Marxist po-
litical economy, dependency theory and poststructuralism (Bryant and
Bailey, 1997; Watts, 2003; Watts and Peet, 2004; Zimmerer and Bassett,
2003). As such, it views peoples’ relationships with the environment as
decisively influenced by capitalism and mediated by political structures
and socially-constructed ideas that re-enforce unequal access and con-
trol over environmental resources (Robbins, 2004; Forsyth, 2008;
Mann, 2009). PE is skeptical of the post-positivist/reductionist view of
explanation7 that underlies LCS and SES, preferring structuralist and
constructivist explanatory approaches instead (Turner and Robbins,
2008; Aldrich et al., 2012).8 Quantitative methods and models so
central to LCS and SES are also typically eschewed by PE in favor of
qualitative analysis of case studies (Turner and Robbins, 2008). Many
PE studies are empirically detailed, but their analytical orientation is
typically more deductive or theory-driven than either LCS or SES.
PE has championed the idea that local environmental changes often
need to be understood as the outcome of non-local, political processes,
structures and events. Influential early advocates for PE (Blaikie and
Brookfield, 1987) argued explicitly for linking the local with the non-
local using nested ‘chains of explanation’, albeit this approach has been
difficult to operationalize in practice and has been criticized for being
overly hierarchical in its explanatory structure (Batterbury and
Bebbington, 1999; Blaikie, 1999:140; Robbins and Bishop, 2008; see
Walters and Vayda, 2009:546). Political ecologists have since for the
most part steered clear of this sort of methodological thinking (but see
Shaver et al., 2015). Instead, PEs can be roughly divided into those who
embrace a fairly general theory of the political economy of the en-
vironment and those guided by a more eclectic set of analytic-theore-
tical perspectives.
Either way, the heavily theoretical character of much PE scholarship
can be seen as both a strength and weakness. PE views LCS and SES
approaches as inadequate for dealing with society and politics, yet re-
search within PE has been justly criticized for its inadequate attention
to the empirical complexities of environmental change (Vayda and
Walters, 1999; Walker, 2005; Nygren and Rikoon, 2008; Pollini, 2010;
Moritz, 2017). More generally, PE’s emphasis on politics above all else
is a recipe for confirmation bias in causal explanations; i.e., the human
tendency to seek or interpret evidence “in ways that are partial to ex-
isting beliefs, expectations or a hypothesis at hand” (Nickerson,
6
For example, a team of leading CHANS researchers admitted that, “…a full vulner-
ability assessment following the framework developed by the Sustainable Systems
Program may lie well beyond the capacities of most research efforts” (Turner et al.,
2003b:8085). See also Turner et al. (2016) for a description and cost-breakdown of an
LCS project.
7
Postpositivism views science as the search for general causal relationships, accom-
plished by way of isolating and measuring variables (usually quantitatively) and then
evaluating their relation to one-another using experimental, modeling or statistical
treatments (Turner and Robbins, 2008).
8
Structuralism situates understanding of human-environment interactions in the
context of relations of power, social differentiation and inequality. Constructivist ex-
planations focus on human-environment interactions as these are understood in thought,
communicated in language, and established as shared theories or narratives (Turner and
Robbins, 2008; Mann, 2009).
B.B. Walters
1998:175; Vayda, 2009:29-43). One consequence of these shortcomings
is that political ecologists have been quick to embrace ‘revised’ narra-
tives about human-environment relationships that are enlightening and
yet arguably as over-simplified as the original narratives they replaced
(Grossman, 1998; Nyerges, 2008; Pollini, 2010; Behnke and Mortimore,
2016).
2.4. Abductive Causal Eventism (ACE)
Scholars have long acknowledged an epistemological kinship be-
tween history and the evolutionary sciences (biology, ecology, geology,
paleontology, etc.) (Gould, 1989; Kelly, 2016). These fields have largely
advanced by resisting law-like generalities and teleological thinking
and, instead, by embracing diversity, contingency and change as fun-
damental features of the human and natural world (McCullagh, 1998;
Gaddis, 2002; Sagoff, 2016). While historians are rarely explicit about
their methodologies, insights derived from the causal analysis of his-
torical events are viewed increasingly by philosophers as central more
generally to explanations in both the social and natural sciences (Dray,
1957, 1964; Scriven, 1966, 2008; McCullagh, 1984, 1998; Lewis, 1986;
Hawthorn, 1991; Roberts, 1996; Vayda and Walters, 2011). The im-
plication is that it is simply unwarranted making firm distinctions be-
tween historical and supposedly ‘scientific’ explanations: all explana-
tions are ultimately causal-historical.
It is therefore not just a matter of scientists adopting an historical
perspective or incorporating historical methods and information in
their research. Historians approach research in different ways and not
all entail causal explanation as a goal (McCullagh, 1998; Gaddis, 2002).
For example, much historical research is primarily descriptive or in-
terpretive. Even where explanatory, it does not always follow the
causal-analytic approach described here. The name, ‘abductive causal
eventism’, was thus adopted to emphasize critical elements of historical
analysis that are often, but not always given priority by historians,
themselves. But these are elements well-suited for the analysis of en-
vironmental change.
Specifically, ACE is an explanation-oriented methodology, based on
a pragmatic view of research methods and explanation that places at
the center of research inquiry the answering of ‘why’ questions about
events, including human actions and environmental changes, rather
than evaluating causal theories, models or factors that are thought in
advance to influence such changes. ACE entails constructing causal
histories of inter-related social and/or biophysical events backward in
time and outward or inward in space through a process of eliminative
inference and reasoning from effects to causes. Avoiding rigid a priori
assumptions about which events (or kinds of events) will do the ex-
plaining, the researcher may seek whatever socio-economic and bio-
physical information is expected to be relevant to answering specific
questions of interest. Diverse types of evidence are then effectively in-
tegrated by virtue of focus, not on what is prescribed by some general
theory or model, but rather on clear, concrete events as possible, si-
tuation-specific causes. The approach and underlying rationale for ACE
is detailed elsewhere (Vayda, 2009, 2013; Walters and Vayda, 2009;
Vayda and Walters, 2011; Walters, 2012a). This section highlights key
distinctions between ACE and the aforementioned research approaches.
First, ACE makes events the primary, but not exclusive, focus of
analysis. Put simply, an event is something that happens somewhere
during a particular interval of time (Lombard, 1991; Priest, 2017).9 The
causal analysis of events may entail consideration of particular events
or, at a more general level, ‘events of a given kind’ (Lewis, 1986; Vayda
and Walters, 2011:1-6). More durable structures or conditions may also
form parts of explanations, as the below case study will show, but ACE
9
For more detailed discussion of events and the justifications for using these, see
Walters and Vayda (2009:540-41), Vayda and Walters (2011:2-6), Priest (2017), and
references therein.
611
Land Use Policy 67 (2017) 608–624
researchers should where possible identify effects and causes in terms of
events. This is because, for explanation, events are generally regarded
as more precise analytically and less problematic ontologically than
systems, structures or processes.10 By focusing on events, ACE can
readily incorporate the influence of causes that are singular or in-
frequent (contingent) in their effects, as well as causes that are re-
current and thus possibly indicative of the presence of ‘systems’,
‘structures’, ‘networks’, ‘processes’ or ‘event sequences’ (cf. Perramond,
2007:502).11
A second distinction is the central place of abductive reasoning in
ACE. Many philosophers have long-recognized abduction as a form of
analytic reasoning distinct from induction and deduction (see especially
Peirce, 1932), but there has been a revival of interest in it (Hintikka,
1998; Thomas, 2010; Campos, 2011). Abductive reasoning is inference
to an explanatory hypothesis: faced with novel and perhaps puzzling
observations, a researcher conceives of plausible conjectures to explain
these. Abductions may be ‘habitual’ (drawing from conjectures already
known to the inquirer) or ‘creative’ (conjectures previously unknown to
the inquirer but conceived of in response to puzzling observations).
Seemingly mundane, abduction is an act of intellectual insight and
arguably the basis for scientific creativity and discovery (Campos,
2011:425-431). Abductive reasoning has only recently garnered the
attention it deserves by philosophers, but for most research scholars it
remains unknown as a concept and under-appreciated in its sig-
nificance.
Critical for ACE, abduction entails reasoning from effects to causes
(rather than from causes to effects) and thus provides the analytical
basis for following chains of event causation backward in time and
either outward or inward in space. This step-wise process entails an
eliminative strategy whereby alternative plausible hypothesis are first
derived abductively to explain an event of interest. Next, these hy-
potheses are evaluated inductively or deductively using counterfactual
reasoning and whatever empirical methods and evidence can be
brought to bear on the matter. Less plausible hypotheses are then
progressively eliminated in favor of those more plausible (cf.
Chamberlin, 1965 [orig. 1890]; Scriven, 2008:21-23). This eliminative
strategy may reveal singular or multiple, converging causes, and it can
then be repeated to follow specific chains of event causation further
steps backward.
Abduction is indispensable to doing science, so all researchers en-
gage in at least some effects-to-causes reasoning, but for most it is
secondary to causes-to-effects reasoning (examples of each are pre-
sented in Table 1). ACE suggests reversing this emphasis, and offers the
following advantages for doing so:
2.4.1. ACE is iterative and adaptive
Causal inquiries evolve as evidence is revealed and interrogated;
novel or unexpected findings can be readily incorporated and in-
vestigations modified accordingly.
2.4.2. ACE is unbounded and scale-independent
Causal chains can be traced backwards to proximate and/or distant
origins as interest and evidence warrants. That is, researchers follow
causal chains where they lead, not to pre-determined boundaries, lo-
cations, levels or scales of analysis (Walters and Vayda, 2009:543-
10
Many philosophers, even if by no means all, regard events as the relata of causal
explanation (for a clear exposition of this view, see Lewis, 1986) and see so-called systems
or structures as having been constituted by recurrently connected events. Some ‘critical
realists’ have argued for also attributing causal powers to social structures (e.g., Elder-
Vass, 2010), but this view is much contested (Wahlberg, 2013; see also Dore, 1961).
11
For example, remittance income transfers from Salvadorians living abroad con-
tributed to reduced cultivation and increased reforestation back home (Hecht, 2010).
Remittance transfers are singular events, but ‘income remittance’ could also be described
in this case as a ‘process’ because transfers tend to occur repeatedly over time between the
same individuals and it is the expectation of continued receipt of remittance income that
in large part alters on-farm decisions about cultivation.
B.B. Walters Land Use Policy 67 (2017) 608–624

Table 1
Examples of questions for human-environment researchers using causes-to-effects reasoning vs. effects-to-causes reasoning.

Causes-to-effects reasoning Effects-to-causes reasoning

How are globalization and population pressure affecting land use decisions and practices? (GLP,
2005:10)
What effects did the collapse of the coffee market have on peasant farmers and the forests in
which it is harvested? (Robbins, 2004:5)
What are the implications of response diversity for resilience and adaptive capacity of East
African pastoralists? (Leslie and McCabe, 2013)
What effects do land tenure, population density and other factors have on reforestation in Nepal?
(Nagendra, 2007)
What are the effects of bioenergy policy on land use change? (Efroymson et al., 2016)
Why did the North Atlantic Cod population collapse in the early 1990s?
(McGuire, 1997)
Why did fires in Indonesia’s wet-tropical forests start, but only some of them
spread? (Vayda, 2006)
Why are the lands north of Binangun, Java severely degraded (deforested and
eroding)? (Lukas 2014)
Why are people in the Philippines planting mangroves in some coastal
villages, but not others? (Walters, 2004)

545).12 As such, ACE avoids reification of ‘systems’ and ‘scale’ and es-
capes the hierarchical, causal determinism that is found, for example, in
Blaikie and Brookfield’s (1987) ‘chains of explanation’ approach (see
Robbins, 2004:210; Rangan and Kull, 2009). There is also no a priori
need to distinguish between ‘proximate’ and ‘underlying’ causes/dri-
vers.

2.4.3. ACE is interdisciplinary

Events and their causes may be socio-cultural, political, economic or
bio-physical in nature, but regardless, focusing on events encourages
analytical clarity and precision, which fosters cross-disciplinary com-
munication. Likewise, both qualitative and quantitative empirical evi-
dence can be drawn-upon to evaluate the plausibility of causal con-
nections between events. This encourages an eclectic use of methods,
models and theoretical ideas, and fosters integrative, interdisciplinary
analysis without being committed to either a reductionist-analytical
paradigm or to systems as ontological entities or to specific holistic
analytical frameworks.

3. Case study: land use and forest change in Saint Lucia

3.1. Research context and questions

Saint Lucia is a small, English- and French Creole-speaking nation in

the Windward Islands of the Eastern Caribbean (Fig. 1). It has a wet
tropical climate and mostly rugged, mountainous topography with
fertile soils of relatively recent volcanic origin. For nearly 250 years,
Saint Lucia’s economy has been founded on export-oriented agriculture
Fig. 1. Map of study watersheds, Saint Lucia.
(sugar, cotton, cocoa, coffee, coconuts, citrus and bananas), much of it
concentrated along the flat bottom lands and lower slopes of two dozen
island was subsequently sub-divided into large estates where more acces-
major river valleys. Since the 1960s, other economic sectors have
sible lands were cleared for plantation agriculture, with smallholder colo-
grown in significance, notably tourism, but also small-scale manu-
nists and slaves farming smaller, often subsistence-focused plots along ad-
facturing, construction, real estate, and administrative services
jacent hillsides. Emancipation in 1838 triggered expansion of smallholder
(Walters, 2016a, 2016b).
farming by former slaves onto unclaimed lands located primarily on slopes,
Landscapes of Saint Lucia have been modified by settlement and agri-
ridge tops and interior forested sites. Some estates that later fell into decline
culture for centuries. Pre-contact indigenous people cultivated some of these
were also sub-divided into smaller holdings (Moberg, 2008; Harmsen et al.,
lands and introduced many plants that are still found today (Kimber 1988;
2014).
Harmsen et al., 2014). Early European settlement of Saint Lucia was slow
The agricultural sector in Saint Lucia struggled between the late
and incremental (Breen, 1844). Not until the mid-18th century did French
1800s and mid-1900s, the result of increasingly competitive export
colonists establish significant commercial agriculture there.13 Much of the
markets and inadequate capital investment in its primary industry,
sugar.14 Many agricultural estates fell into chronic decline or were
12
It makes sense to distinguish proximate from distant or local from non-local causes, abandoned outright. Some of the unused lands eventually returned to
but these terms are not commensurate with ‘scaling’ (see GLP, 2005:3; Neumann, 2009), forest; others were settled and farmed by former estate workers; and
which entails the re-description of data at different levels of aggregation. Scaling might
others still were acquired by the government and sub-divided and sold
aid in causal analysis, for example, by enabling researchers to interrogate empirical
patterns not initially observable at higher or lower levels of aggregation. But it is pro- as smallholdings. The emergence of a significant banana export in-
blematic to attribute causal powers to ‘scale’ as an ontological entity, albeit socially- dustry following WWII reinvigorated the island’s moribund farming
constructed ideas about scale may have them (Brown and Purcell, 2005).
13
In contrast to neighboring islands, Barbados and Martinique, Saint Lucia’s terrain
14
was rugged and its’ interior relatively impenetrable and inhabited by Amerindians hostile Small-island Caribbean producers faced growing competition on international sugar
to colonists. It was less appealing for agriculture and was, instead, viewed by early co- markets from European sugar beets and larger-scale, lower-cost sugarcane producers in
lonial governments as valuable primarily for military-strategic reasons because of its Trinidad, Cuba, Guyana, Brazil, etc. According to analysts at the time, this problem was
sheltered bays and one especially deep, sheltered harbor (the current port of Castries) compounded by a general reluctance of West Indian estate owners to invest capital to
(Breen, 1844). modernize sugar production (Cox, 1897; WIRC, 1897; Lowenthal, 1972)

612
B.B. Walters
sector. Several of the remaining estates transitioned from sugar to ba-
nanas. But from the late 1950s onward, the smallholder sector came to
dominate banana production (Romalis, 1975). From the late 1950s
until the late 1990s, bananas were the dominant agriculture crop in
Saint Lucia and its effects on the economy, landscape and environment
were pervasive (Moberg, 2008).
This study was motivated by observations of Saint Lucia’s landscape
that spanned a 15-year period. Specifically, I lived and worked with an
environmental non-government organization in Saint Lucia in 1990/91,
a time when agriculture, especially bananas, still dominated the rural
landscape. Upon my return in 2005, however, things had changed
dramatically. Banana production had seemingly collapsed and ex-
tensive rural lands formerly under cultivation appeared now to be ei-
ther abandoned, returning to forest or converted to residential housing
or other forms of ‘hard’ development. This presented a superb research
opportunity to field-test ACE because dramatic changes had clearly
occurred, yet the specific character and causes of these changes were by
no means clear. My research thereby began with two over-riding
questions: (i) what specific environmental changes had actually un-
folded in recent years and (ii) why had these changes happened?
3.2. Study areas
To capture key elements of environmental change at a landscape
level, I chose the watershed as general geographic unit of study. Saint
Lucia is essentially bisected—north-to-south—by a steep range of
mountains, so its geography and human settlement are largely struc-
tured around watersheds draining from the center eastward and west-
ward, respectively. I thus included one study watershed on each of the
two sides of the island: Soufriere (1570 ha) on the west and Mamiku
(790 ha) on the east (Fig. 1).
Soils within Soufriere are geologically younger and so especially
fertile, and local climatic conditions are particularly amenable to di-
verse agricultural production. Lying in the western shadow of the tallest
mountains in Saint Lucia, Soufriere receives the highest rainfall on the
island and is significantly protected from easterly-prevailing winds that
often damage crops during storms. These factors, combined with the
locating of the first center of French colonial administration in
Soufriere, established this area as an early epicenter of plantation
agriculture on the island (Breen, 1844; Harmsen et al., 2014). The
rugged but otherwise favorable environmental conditions also en-
couraged widespread planting of tree crops like cocoa, coffee, coconut
and lime.
But Soufriere’s rugged terrain and relative geographic isolation from
the rest of the island have posed challenges. Through most of its history,
Soufriere was only accessible by sea or inland foot/mule paths. This
posed little problem during the days of sailing ships, but lack of good
road access disadvantaged Soufriere with the advent of steam ships and
the locating of centralized shipping terminals in the north and south of
the island. These challenges were especially pronounced with bananas
because they cannot tolerate extended travel over rugged terrain
without bruising. Farmers in Soufriere were thus often reluctant to
convert existing tree crops wholesale to banana production. Banana
cultivation became important in Soufriere, but never dominated to the
extent it did elsewhere on the island.
In contrast, watersheds on Saint Lucia’s east coast, including
Mamiku, are less rugged, but drier and more directly exposed to
Atlantic trade winds. Farming has thrived there during periods of robust
market demand for less water-demanding crops like cotton, sugar, co-
conuts and livestock. These same conditions also made Mamiku fa-
vorable for intensive banana cultivation, which by the 1970s dominated
farming in the watershed.
3.3. Methods
Detailed descriptions of methods can be found in Walters (2012b,
613
Land Use Policy 67 (2017) 608–624
2016a) and Walters and Hansen (2013:213-214). Briefly, vegetation of
the two watersheds was surveyed in 2006 using a quadrat-census plot
method. To enhance aerial coverage and reduce site selection bias, plots
(100 m2) were located in a stratified grid across each watershed
(1000 m by 500 m apart), with pre-programmed GPS coordinates used
to locate sampling points. A total of 34 plots were sampled in Soufriere
and 22 in Mamiku for a total of 56 plots. Altitude, slope and aspect were
measured at each plot location. Other plot characteristics were noted,
including evidence of human or natural disturbance. Information about
land tenure, land holding size, and land use history of each site was
sought through interviews with local farmers and by consulting official
land registry maps. The distance from the site to the nearest drivable
road was estimated using air photos and maps. Within each of the 56
sample plots, all trees > 2.9 cm diameter at breast height (dbh) were
measured, mapped and identified by species (n = 1699 trees) based on
Howard (1974), Graveson (2005) and Saint Lucia Herbarium collec-
tions. Trees were classified as being planted or natural, where planted
referred to trees which were known to be traditionally cultivated by
Saint Lucians and natural referred to trees which were not. Measures of
canopy height and structure as well as ground cover were also taken.
Near-comprehensive, aerial-photo coverage for both watersheds
(1966, 1977, 1992 and 2004) plus recent LANDSAT (GoogleEarth) sa-
tellite imagery (2009/13) were combined with field observations from
2006 and 2015 to assess trends in five general land status categories
over time: ‘developed’ (human habitations, etc.), ‘annual agriculture’
(short-lived crops like bananas or vegetables), ‘agro-forest’ (tree crops),
‘secondary forest’ (young, post-agricultural forest), and ‘mature forest’
(long-established forest). To do this, air photos were geo-referenced
using ARC-View and the 56 plot-sites were used as discrete, spatial
reference data points with which to compare by back-casting differ-
ences between successive air-photo and satellite surveys. The 56 plot-
sites were re-visited (on foot) in 2015 to visibly assess their most recent
condition. The sampling protocol for the vegetation survey was both
systematic and fairly comprehensive in its coverage of the two water-
sheds. As such, cumulative changes across the 56 points are assumed to
be reasonably representative of broader trends in watershed landscape
change.
Archival and ethnographic investigations followed the vegetation
survey work so that these could be informed by its findings. Most in-
terviews and primary archival work were conducted in 2007 with
follow-up in 2015. Thirty-seven farmers (31 smallholder; 6 estate) and
25 key informants (government officials, NGO reps, etc.) were inter-
viewed using a semi-structured format to document their respective
knowledge about agriculture, land use, etc., and to interrogate their
interpretations of changes in these. Sampling included multiple small-
holder farmers from each of the six rural settlements within the two
watersheds and either the owner or manager (or both) of each of the six
active agricultural estates.
3.4. Research findings
3.4.1. Rural reforestation
Evidence from the watershed surveys, air photos and interviews
confirmed the significance of reforestation as the most consequential
change in recent decades to the island’s rural environment (Walters and
Hansen, 2013). The vegetation surveys revealed landscapes of both
watersheds to be dominated by forest and tree cover, a mix of mature
and young natural forests, agro-forests and otherwise scattered in-
dividual trees dispersed across cultivated fields. Eighty-two percent (46
of 56) of vegetation plots fell in forest habitats (natural or agro-forest),
of which nearly half were post-agricultural forests under varying stages
of natural succession. The survey identified a cumulative 125 tree
species (99 natural and 26 planted) in the 56 plots (Table 2).
Air photos (1977–2004) revealed a striking increase in forests and
decline in agriculture (both agro-forests and annual crops), but these
trends moderated after 2004 (Fig. 2). In 1966, 33 of 56 plot sites were
B.B. Walters Land Use Policy 67 (2017) 608–624

Table 2 Table 3
Cumulative species richness of natural and planted trees in vegetation plots by primary Reasons cited by farmers in 2007 for the presence of forest or bush on their lands.
habitat type (2006 field surveys).
Estate farmer Smallholder farmer Total
Habitat No. plots Tree species richness (cumulative) (n = 6) (n = 30) (n = 36)

Planted trees Natural trees All trees Land abandoned from 3 21 24

farming
Annual Crops 6 3 1 4 Land too steep to farm 3 5 8
Agro-forest/agro-secondarya 19 25 19 44 Environmental 5 1 6
Secondary forest 13 3 69 72 protection
Mature forest 14 0 57 57 Tourism opportunities 3 – 3
Other 4 0 0 0 Swidden fallow – 2 2
All habitats 56 26 99 125 Land owner retired – 2 2
Land too inaccessible – 1 1
a
Includes both managed agro-forests and agro-forests abandoned and infilling with
successional forest.
Table 4
under cultivation while 21 were in forest. In 2015, these numbers were Recent changes to crop cultivation as cited by farmers in 2007.
nearly reversed, with 35 under forest and 15 being actively farmed.
Estates Smallholders Total
Interviews from 2007 confirmed these trends. All but one farmer (36 of (n = 6) (n = 28) (n = 34)
37) cited having at least some forest or bush on their land. Of these,
80.6% (29 of 36) said the amount of their lands under forest or bush Reductions in cultivation:
had increased in recent years, with no difference in responses between Bananas 3 15 18
Coconuts 5 1 6
watersheds or between estate and smallholder farmers. Farmland
provisions (root crops) 1 4 5
abandonment was by far the most commonly cited reason for the pre- Cocoa 1 2 3
sence of forest or bush on a farmer’s land (Table 3). fruit crops 0 2 2
‘Developed’ land in the form of residential housing, tourism facil- Increases in cultivation:
ities and public infrastructure became increasingly significant during Vegetables 3 7 10
this period (Fig. 2). Of four plots identified as developed in 2015, three provisions (root crops) 0 8 8
entailed conversion of agro-forest to residential housing and public fruit crops 1 4 5
ornamentals (cut 4 0 4
infrastructure, while the fourth involved the conversion of natural flowers)
forest to residential housing. Cocoa 2 0 2

3.4.2. Land use change

Reduced cultivation of bananas was the land use change most fre-
quently cited by farmers (Table 4), an observation consistent with na-
tional statistics showing a dramatic decline in banana exports since the
mid-1990s (Fig. 3). Estate farmers, in particular, also cited reducing
their cultivation of coconuts (used primarily for copra). This change in
part accounts for the especially sharp decline in agro-forests (and
concomitant increase in forests) from 1977 to 1992 as one particularly
large estate owner in Mamiku abandoned an expanse of coconut land in
the early 1980s equal in area to about 20% of the entire watershed.
Expansions in cultivation also occurred, notably of mixed vegetables
and cut flowers, but these are intensive, relatively small-scale opera-
tions that have nowhere near offset the area abandoned from bananas
and other crops. As such, 74.3% (26 of 35) of farmers cited a net overall
reduction in land under cultivation, whereas only 5.7% (2 of 35) cited
an increase. Since these interviews in 2007, a turn-around in the for-
tunes of the cocoa industry has prompted some farmers to restore ex-
isting or establish new cocoa groves. This is one factor contributing to
the stabilization of farming since 2007 (Fig. 2).
Farmers cited a variety of reasons for their reduced cultivation,
including declining farm-gate prices for crops, especially bananas
(46%); growing cost/scarcity of farm labour (42%); ageing (23%); and
increased cost of farm inputs like fertilizers and pesticides (19%). Key
informants echoed the same causes, but included the loss of farmland to
residential and tourism development, improved enforcement against
illegal farming in forest reserves, and uncertainties over land tenure
(Table 5). More will be said about these below.
3.4.3. Local topography
The terrain of both watersheds is relatively flat near the coast, but
grades into hills and then steep, rugged slopes inland. Forests are
usually found at higher altitudes, on steeper slopes and more distant
from roads than are cultivated areas (Walters and Hansen, 2013:5-7).
These patterns reflect access constraints and land-use decisions by
farmers and the national government. Specifically, it is commonplace
for individual farm holdings to include a mix of flatter and steeper lands
at varied distances from roads. Some farmers maintain portions of their

Fig. 2. Back-casting land use on census plots over time using field
census and air photos. Data from Mamiku and Soufriere watersheds
combined.

614
B.B. Walters Land Use Policy 67 (2017) 608–624

Fig. 3. Saint Lucia banana exports (tonnes), 1950-2015.

Sources: McFarlane, 1964; World Bank, 1979; GOSL, 1991, 2002,
2012, 2016.

Table 5 3.4.4. Land tenure

Primary reasons cited in 2007 for decline in farming in Saint Lucia, according to farmers Land tenure has been a central concern of scholarship on West
and key informants. Indian land use because of the sharp contrasts between agricultural
Farmers Key Informants
estates and the smallholder farm sector and the diversity of tenure ar-
(n = 26) (n = 25) rangements involving the two (Rubenstein, 1975; Rojas, 1984; Besson
and Momsen, 1987; Rojas and Meganck, 1987; Barrow, 1992; Brierley,
Declining farm-gate prices 12 14 1992; Dujon, 1997). This study found a range of interactions between
Growing shortage of farm labour 11 16
Ageing of farmer population 6 5
land tenure, land use and the arboreal landscape, but there was little
Increased cost of farm inputs 5 6 evidence that land tenure per se was an important causal influence on
Loss of farmland to residential & tourism – 9 the recent surge in farmland abandonment and reforestation.15 Post-
development agricultural, secondary forests were common on private-estate, private-
Improved enforcement against illegal – 7
smallholder and communal ‘family’ lands (Walters, 2012b).16 The re-
farming in Forest Reserves
Uncertain tenure over family lands – 6 lative proportion of vegetation plots on actively farmed vs. abandoned
Absentee ownership of farmlands – 1 land was also similar across tenure types (Table 7). Interviews con-
firmed these patterns, with equal numbers of private (8 of 10) and fa-
mily land (14 of 18) farmers having recently reduced their land under
Table 6 cultivation. In fact, security of land tenure was never cited by farmers as
Summary of watershed vegetation survey comparing currently farmed vs. recently a factor influencing their decision to increase or decrease cultivation.
abandoned plots by altitude, slope and distance from nearest road.
These various findings thus rule out the influence of land tenure as a
Farmed plots Abandoned plots t-value significant cause of recent farmland abandonment and reforestation.
(n = 20) (n = 20) (d.f. = 38) That said, land tenure does indirectly influence land use in ways
that conserve forests. Specifically, strong growth in tourism and re-
Altitude (m) 182.4 (115.4) 300.9 (187.7) 2.40*
Slope (degrees) 12.7 (11.2) 26.4 (12.8) −3.60**
sidential home construction since the late 1980s has led to development
Distance to road 114.4 (117.4) 286.0 (299.4) −2.08* of large housing estates and numerous tourism facilities, especially in
(m) Soufriere (Walters, 2016b). But, legal complexities associated with re-
solving diverse claims to family land, in particular, have discouraged
* p < 0.05. their sale and development. This has kept some sizable tracts of prime
** p < 0.005.
land (located near the coasts in each watershed) under forest where
they would otherwise not be (Walters, 2012b).17
land under more-or-less permanent forest cover, especially on steep
slopes that are difficult to farm and at high risk of erosion and land-
slides. But economic considerations may influence whether such mar- 3.4.5. Banana booms and busts
ginal lands are brought under cultivation. Given recent market turmoil Evidence unequivocally points to troubles in the banana export
in bananas and other broader challenges facing the agricultural sector market as a major cause of reduced cultivation and reforestation in
(see below), farmers have tended to downsize by abandoning cultiva- Saint Lucia. Bananas have not been the only sector of agriculture ex-
tion on more marginal sites first. For example, 38.5% (10 of 26) of periencing difficulty, but they attained such dominance in the second
farmers who had reduced cultivation specifically cited having aban- half of the 20th century that any large changes to bananas were likely
doned steeper and/or more remote lands on their respective holdings. to have major consequences for land use on the island. Since its
Confirming this, vegetation plots on abandoned vs. cultivated land were emergence as a significant industry in the 1940s, the banana sector has
found to be at significantly higher altitude, on steeper-slopes, and more
distant from roads (Table 6). 15
Among notable effects of tenure were the greater abundance of planted trees on
The government has re-enforced this trend of reforesting marginal
private smallholder lands and the higher frequency of mature forest on public lands.
lands by strategically targeting selected areas deemed critical for wa- 16
The effects of land tenure are potentially confounded by the effects of topography
tershed protection, in some cases acquiring lands outright from private given that family and public lands are more likely than private lands to be on steeper
owners through purchase or land-swaps, but more generally by en- slopes and at higher elevation (Walters, 2012b). However, because most individual farms
couraging tree planting on steep-sloped, private lands and by stepping- include a variety of topography, the relative influence of one vs. the other could be
generally discerned through a combination of observation and interviews. In short, evi-
up enforcement against illegal farming of Forest Reserve land. While
dence pointed to topographic considerations having a pervasive influence on decisions to
not trivial, the cumulative effect on the landscape of these various abandon cultivation both within and between farms, but tenure had no consistent effect.
government initiatives is relatively small compared to that of voluntary 17
By law, all co-owners of family land (or someone legally designated to act on their
farmland abandonment. behalf) must agree to its sale. This creates onerous, often insurmountable challenges
because there may be dozens of co-owners, including many who reside out of the country.

615
B.B. Walters Land Use Policy 67 (2017) 608–624

Table 7 Table 8
Summary of vegetation plots (n) comparing land tenure by land use (farmed vs. aban- Selected population and agricultural statistics, National Agricultural Census Reports,
doned). 1961–2007. Soufriere District on the west coast corresponds closely to the boundaries of
the Soufriere watershed, whereas Micoud District on the east coast encompasses Mamiku
Land Tenure plus several adjacent watersheds.

Private Private Family Government Census Year

Estate Smallholder Smallholder
1961 1973/4 1986 1996 2007
Farmed 5 5 10 0
Abandoned 4 4 12 0 Population living in – 52,283 58,000 51,553 32,919
farm households (48.6%) (47%) (37%) (21%)
(% of National)a
experienced two distinct boom-bust periods (Fig. 3).18 Explaining these - Soufriere District – 4522 – 2684 1460
only
changes requires one to follow causal chains backward and far-out-
- Micoud District only – 3670 – 9919 6742
ward. Total no. farmers 10,772 11,501 13,052 9800
Nearly all banana exports from Saint Lucia have gone to the UK, Full-time farmingb 6960 7455 7625 5723
aided decisively by a series of preferential trade policies, based on so- Part-time farmingb 3812 4046 5427 4077
Median age of farmers 45.0 43.8 46.0 50.3
called ‘Imperial Preference’, intended to shore-up the UK’s weakened
Total no. farm 13,008 10,938 11,551 13,366 9972
currency and support post-War economic development of existing and holdingsc
former British colonies (Clegg, 2002).19 Banana cultivation is especially - Soufriere District 999 855 792 455
well-suited to small farmers because it is labour-intensive and produces only
year-round (O’Neil, 1964; Rojas, 1984). As well, in response to ongoing - Micoud District only 1086 1693 2462 2008
Cumulative area of 87,375 72,001 58,017 51,323 30,204
labour and related political unrest that had emerged in the 1930s, the
farm holdings
British Colonial administration implemented programs of land reform (acres)c
and rural development that targeted the small-farm sector, including - Soufriere District 6953 5,988 3,784 1490
agricultural extension and marketing services and support for banana only
- Micoud District only 11,301 12,416 10,811 7047
grower associations (Romalis, 1975; Grossman, 1994; Clegg, 2002).
These diverse government supports combined with secure export mar- a
Numbers compiled from GOSL (1996: Table 10) and GOSL (2007: Chart 17).
kets encouraged widespread adoption and expansion of banana culti- b
‘Full-time’ are farmers whose primary income source is farming; otherwise, they are
vation during the late 1950s/early 1960s onto increasingly marginal, considered ‘part-time’.
c
often forested lands (Welch, 1994, 1996). Numbers complied from GOSL (1996: Tables 1, 2 & 3) and GOSL (2007: Table
A13 & Charts 1 & 3).
Problems emerged in the late 1960s, however, when a sharp deva-
luation of British Sterling drove-up the cost of imported fertilizers and
pesticides, resulting in a marked reduction in their use by farmers and construction of the east coast highway that linked the new international
consequent declines in banana production and quality (Clegg, 2002). airport in the south with the capital city and major tourism destinations
These changes precipitated the first major collapse (Fig. 3) which was in the north (Welch, 1996). Thus, the farm household population within
compounded through the 1970s by financial problems facing the Saint the East-coast District of Micoud (which includes Mamiku watershed)
Lucia Banana Growers Association (SLBGA) and back-to-back hurri- almost tripled in size from 1974 to 1996, even as it declined in Soufriere
canes in 1979 (David) and 1980 (Allen). In response to these difficul- over the same period (Table 8).
ties, the British Government provided financing to re-structure and Unlike the 1960s decline which was due primarily to currency
modernize the banana industry through adoption of field-packing and changes and structural challenges within the Saint Lucian industry, the
other quality control measures (Romalis, 1975). The export market also bust that followed in the late 1990s was largely caused by changes in
markedly improved during the 1980s as UK-consumer demand for ba- export markets. Specifically, the formal entry of the UK into the
nanas grew and the British Sterling appreciated.20 European Union and Common Market in 1992 precipitated a series of
This confluence of highly favorable events and conditions resulted challenges against trade provisions of the Lome Convention, an inter-
in a stunning, nearly three-fold increase in banana production and ex- national agreement that since 1975 had enshrined the Imperial
port from 1982 to 1986 (Fig. 3). With parallels to the late 1950s, ex- Preference policies of the UK (as well as France) for agricultural pro-
isting farmers expanded areas under cultivation and many new farmers ducts from their former colonies. The first challenges came from fellow
entered the sector, often as part-timers. Also like before, this expansion EU members (especially Germany) whose import markets for bananas
was aided by the SLBGA which expanded support for small farmers by were more openly competitive. Negotiations within the EU led in 1993
subsidizing agro-chemicals, providing crop insurance, and establishing to agreement of a special ‘Banana Protocol’ that maintained most of the
a network of inland buying depots (GOSL, 1985, 1987). This second original Lome protections for West Indian producers, including Saint
boom in cultivation was especially marked on the east side of the island Lucia. But while members of the EU were placated, others outside it
where extensive road up-grades inland were made in tandem with were not.
The major multi-national banana companies, US-based Chiquita and
Dole, sourced most of their product from Latin America (so-called
18
Bananas were grown for domestic consumption in Saint Lucia long before an export ‘dollar bananas’) and viewed the formation of the EU in 1992 as a
trade developed. Early attempts were made between the Wars to establish an export trade unique opportunity to expand market share there. The failure of the
in bananas to North America and the UK, but these exports remained modest in size
new EU-Banana Protocol to enhance access led Chiquita to lobby for
(typically less than 1000 tonnes per year) and were ended altogether by wartime cur-
tailments of shipping (McFarlane, 1964; Moberg, 2008; Wiley, 2008).
intervention by the US and several Latin American countries by in-
19
Heavy debt loads associated with the War and post-War reconstruction weakened itiating formal trade challenges under the GATT and newly-formed
UK Sterling which led policy-makers to favor trade with partners in the Commonwealth WTO.21 The WTO subsequently ruled against provisions of the Banana
that used UK currency. In the case of bananas, preferential trade policy was based on both
tariff and quota controls that restricted importation of bananas from non-Commonwealth
countries (Grossman, 1994; Clegg, 2002; Myers, 2004).
20 21
The contract price for banana exports to the UK was negotiated in British Sterling, This fascinating, complicated story has been documented in detail elsewhere so only
but beginning in 1976 the Commonwealth States of the Eastern Caribbean had tied their a summary of events is provided here (see Read, 1994; Clegg, 2002; Myers, 2004; Wiley,
currency to $US. 2008; Fridell, 2011).

616
B.B. Walters
Protocol that had maintained colonial preference. Ensuing efforts to
settle the matter diplomatically failed and were followed by a succes-
sion of WTO challenges, rulings and re-negotiations. This produced a
step-by-step erosion of market protections for West Indian bananas;
declining and more volatile farm-gate prices for banana growers; and
widespread loss of confidence about the industry’s future (Clegg, 2002;
Myers, 2004; Moberg, 2008). Making matters worse, a once-again, in-
debted SLBGA was forced to reduce marketing supports and subsidies
for farmers, driving-up their production costs (GOSL, 1998).22 Com-
bined, these changes led to an exodus of farmers from the industry and
downsizing amongst those who remained.
Attempts to stem this decline have continued, most notably through
the adoption of Fair Trade (FT) certification of bananas, first started in
2000.23 By 2007, over 90% of remaining commercial banana farmers in
Saint Lucia had achieved FT certification and 100% of exports to the UK
were FT certified, with the remainder sold onto local and regional
markets (Fingal, 2008; GOSL, 2008). While initially promising, recent
downward price pressures from lower-cost FT-producers elsewhere
have combined with damaging weather events and outbreaks of black
leaf spot disease to further hobble the struggling industry. In 2011,
more Saint Lucian bananas were sold within the region to neighboring
islands than to the traditional UK market (GOSL, 2012).
3.4.6. Out-migration, return-migration and farm labour scarcity
The weakened state of agricultural markets, especially for bananas,
encouraged many to depart farming and discouraged others (esp. the
young) from entering it. But evidence also points to changes in the
wider economy and labour markets in Saint Lucia as contributing to the
movement of people out of farming and the growing scarcity of farm
labour. The departure of people from farming is thereby both an effect
and a contributing cause of agricultural decline. But this part of the
story has played out somewhat differently in the two watersheds.
Recent demographic and land use trends in Mamiku have closely
tracked the fate of the banana industry. Interviewees cited growing
scarcity of farm labour as among the most important causes of agri-
cultural decline (Table 5). Banana farming is especially labour-intensive
with labour costs estimated to account for two-thirds of total costs of
production (GOSL, 1990). Full-time banana farmers typically rely on
hired labour during critical stages, including harvest, so labour scarcity
has clearly constrained overall production levels among many who
continued to farm. Farmers are also ageing and increasingly not being
replaced by younger generations (Table 8). For example, only 6% of the
children of interviewed farmers had themselves entered farming
whereas 46% were living outside the country. Most had out-migrated
either to urban areas in Saint Lucia or to overseas destinations (other
Caribbean islands, UK, Canada, US).24
Trends in Soufriere—where bananas were never dominant—are
more reflective of socio-demographic and economic changes unfolding
outside the farm industry sector. For one, declining fertility rates have
led to smaller family sizes and thus reduced pressure to grow sub-
sistence crops. This change is evident in a notable reduction in
22
The SLBGA’s debt was absorbed by the Saint Lucian government in 1998 and the
organization restructured as a private company, the Saint Lucia Banana Corporation
(SLBC), with shares allocated based on production volumes. Subsequent in-fighting led to
splintering-off and formation of four other rival companies (GOSL, 1999; Moberg, 2008;
Wiley, 2008).
23
Fair Trade (FT) involves selling certified bananas at a premium price directly to
major UK supermarket retailers. In exchange for the FT price premium, farmers are re-
quired to meet a series of social and environmental standards in their practices and a
portion of profits from FT sales are allocated directly to community projects (Raynolds,
2000; Lamb, 2008).
24
Elderly farmers lamented the absence of younger people to assist on the farm, yet
most admitted they had encouraged their children to pursue other occupations. In fact,
out-migration has been so commonplace throughout West Indian history that it is seen as
an unremarkable, almost expected thing for a person to do at some point in their life
(Lowenthal and Comitas, 1962; Carnegie, 1982; Momsen, 1986; Conway, 1999; Conway
and Potter, 2006; Walters, 2016a).
617
Land Use Policy 67 (2017) 608–624
smallholder farming (especially in remote and rugged sites) comparing
the earliest air phots (1966) with those taken in 1977 and 1992. Old-
timers speak of walking long distances along rugged paths, often deep
into the mountains, to tend subsistence gardens. Many of these sites
have long been abandoned and returned to forest in Soufriere, whereas
in Mamiku these sites were often reclaimed and converted to banana
cultivation during the 1980s boom and only returned to forest once this
sector collapsed.
Substantial employment growth in domestic construction and
tourism since the late 1980s has also drawn many out of farming and
contributed to farm labour scarcity. This effect has been especially
pronounced in Soufriere, one of the growth centers for island tourism
and for residential housing development, the latter being driven pri-
marily by remittance income and a huge wave of (mostly retired) re-
turn-migrants who first departed the island in the 1960s/70s (Walters,
2016a).25 At the same time, hotels, cruise ship arrivals and related
tourism services and infrastructure have expanded dramatically, gen-
erating much employment in construction, landscaping, security, taxi
transportation, and hospitality services (Fig. 4). Some farmers have
responded to these changes by shifting cultivation to non-traditional
crops in demand by tourism markets. But most simply quit farming (or
have chosen not to enter into it) to pursue these alternatives, most of
which require no advanced formal education and, in contrast to farm
work, entail less physical labour and generate predictable, stable in-
comes (Walters, 2016a).
3.4.7. Tourism development and forest conservation policy
In addition to its indirect influence via the labour market, tourism
development has also encouraged reforestation in some cases by
creating incentives for the government and some of the larger estates to
provide for specific eco-tourism opportunities and to enhance protec-
tion of environmental amenities critical for tourism (Walters, 2016b).
Thus, the greater priority now placed by the Government on upland
forest protection and forest lands acquisition is motivated to a large
degree by the desire to protect fresh water supplies critical to the
continued development of the tourism sector. As well, the tourism in-
dustry in Saint Lucia has with the support of Government programs
increasingly positioned itself as a heritage and eco-tourism destination.
This has manifest in the creation of private botanical gardens on estates
in each watershed, plus nature trails that cross recently-reforested es-
tate lands. In fact, all nine estates within the two watersheds have re-
positioned themselves to focus all or in-part on tourism. These estates
continue to cultivate crops, but all but one have significantly reduced
their land area under cultivation in the past 20 years. Commodity
market challenges and labour scarcity have contributed to this trend,
but the shift to a tourism-focused business model has enabled or en-
couraged at least some of these estates to downsize cultivation more
than they otherwise would have (Walters, 2016b).
25
About 10% of Saint Lucia’s population—mostly young adults—migrated to the UK
between 1955 and 1962, encouraged by worker shortages in the UK and by immigration
policies highly favorable to Commonwealth citizens. As UK immigration became more
restrictive following passage of the Commonwealth Immigration Act of 1962, migrants
increasingly ventured to the US and Canada, albeit in fewer numbers on a per capita basis.
While away, out-migrants typically maintained ties to family members and family lands.
Upon retirement, many have chosen to return home, often making substantial invest-
ments in new houses (Peach, 1967, 1968; Momsen, 1986; Abenaty, 2000; Byron, 2000;
Byron and Condon, 1996, 2008). The resulting growth in residential housing can be in-
ferred from national statistics showing a 24% increase in the number of private house-
holds from 2001 to 2010 (GOSL, 2011a). These return migrants often maintain gardens
and orchards, but rarely farm for sale. However, the houses and yards of return migrants
are relatively large and their development typically encroaches onto former farmland and
occasionally into forests (Walters, 2016a, 2016b). In this regard, residential housing
developments within the study areas do not (yet) contribute significantly to deforestation,
but they preclude sizable areas from reforesting.
B.B. Walters
4. Discussion
4.1. Research findings
To sum-up the findings: the most consequential change in recent
decades to Saint Lucia’s rural environment has been widespread re-
forestation of lands abandoned from farming. This change has occurred
for the most part irrespective of the type of land tenure, but is especially
commonplace on lands with steeper slopes and further from roads.
Marginal lands like these have been under abandonment since the
1960s (in some cases to be later reclaimed for farming), but an espe-
cially large wave of abandonment swept the island from the mid-1990s
to early-2000s because of price uncertainties for exportable bananas
related to the erosion of preferential market access to the UK due to a
series of WTO trade challenges. The effects on land abandonment of this
and other commodity market-related challenges have been re-enforced
by wider demographic and economic changes. Specifically, many sub-
sistence-focused farms in the countryside were abandoned in the
1960s/70s (and probably earlier) in response to the gradual shift to
smaller family sizes plus large out-migrations of younger adults to the
UK, US and Canada. More recently, the movement of labour from
farming to fast-growing construction, tourism and services sectors has
sapped the agricultural sector of cheap labour, accelerating land
abandonment. One major source of alternative employment for farmers
has been an island-wide residential construction boom caused by re-
mittances from abroad and a protracted wave of return migrants, many
themselves former farmers. The fast-growth of tourism since the early
1990s has also drawn much labour out of farming and created economic
incentive and political support for protecting more forests on both
private estates and public lands to sustain eco-tourism and freshwater
supplies.
These findings are broadly consistent with the growing literature on
‘forest transitions’ in the global south (Rudel et al., 2005). Specifically,
they support propositions about post-agrarian transitions being linked
to societal modernization, notably the significance of rural-to-urban
migration and globalization of agricultural commodity markets as
causes of marginal farmland abandonment and reforestation (Mather
and Needle, 1998; Rudel, 1998; Aide et al., 2013). These findings are
not surprising given that West Indian economies have been inextricably
linked to the international agricultural trade for several centuries and
its people are among the world’s most migratory (Richardson, 1992). In
fact, the Caribbean region has the highest rate of net reforestation
among tropical regions and evidence from other studies suggests Saint
Lucia’s experience may be mirrored on other Caribbean islands (Grau
et al., 2003; Helmer et al., 2008; Alvarez-Berrios et al., 2013; Timms
et al., 2013; Atkinson and Marin-Spiotta, 2015; Keenan et al., 2015; van
Andel et al., 2016). Particular details of this case study are nonetheless
618
Land Use Policy 67 (2017) 608–624
Fig. 4. Tourism visitor arrivals, 1978–2015 (GOSL, 1983, 1991, 2002,
2011b, 2016).
unique and revealing in their own right, including the convoluted story
of the so-called ‘banana wars’ that ultimately led to the erosion of
preferential trade arrangements between West Indian producers and the
UK consumer market, and the subsequent adoption (with only limited
success) of Fair Trade certification as a way to respond to this (Clegg,
2002; Myers, 2004; Moberg, 2008).
In fact, there is more to the story of bananas than was presented
here. For example, the investigation could have gone even further
backwards into details of the political and economic origins of the ba-
nana trade dispute, but this was felt unnecessary because the topic has
already received so much in-depth attention from other scholars (e.g.,
Read, 1994; Clegg, 2002; Striffler and Moberg, 2003; Myers, 2004;
Lamb, 2008; Moberg, 2008; Wiley, 2008; Fridell, 2011). As well, re-
search could have probed more deeply into the causal influence of price
changes to agricultural inputs, like pesticides and fertilizers, and the
related role of governing bodies like the Ministry of Agriculture and
banana growers associations. These topics have been examined in Saint
Lucia and neighboring Dominica and St. Vincent by researchers who
showed how the expansion of banana farming was significantly enabled
by subsidies and supports provided by these institutions (Romalis,
1975; Grossman, 1994, 2003; Moberg, 2008; Welch, 1993, 1996). Ac-
tions taken by these bodies also influenced the rate of subsequent de-
cline, but it seems doubtful they could have prevented its eventual
collapse given the dominance of causes out of their control.
The relationship between farmland quality—notably steepness and
distance from roads—and abandonment is significant, but not sur-
prising. It is widely accepted in the literature that marginal lands are
more likely to be abandoned, but few studies have actually empirically
confirmed this, let alone drawn from both biophysical and socio-eco-
nomic evidence to do so (Walters and Hansen, 2013; Aide et al., 2013).
As such, patterns of land use and abandonment are influenced by bio-
physical and socio-economic considerations. For example, the fairly
common practice of growing ganja confounds the aforementioned re-
lationship because farmers intentionally seek out marginal locations to
cultivate this crop in light of its illegal status (pers. observ.). Un-
authorized cultivation of food crops on remote lands in the National
Forest Reserve was widespread in the 1960s and 1970s, but ganja is
today the only crop sometimes grown there.
Notable too are findings that suggest Saint Lucia’s supposed forest
transition probably began over 50 years ago but was interrupted—not
once, but twice—as bananas took-hold and spread widely under the
unique circumstances of a highly protected export market, an estate
farm sector in decline, and a smallholder sector in ascendency. In fact,
agriculture in Saint Lucia has been characterized over three centuries
by recurrent booms and busts and associated expansions and contrac-
tions in the area of land under cultivation. Such fluctuations were often
linked to changes in export markets (cotton, sugar, bananas), but
B.B. Walters
sometimes to colonial-territorial conflicts, domestic political turmoil,
and extreme weather events (WIRC, 1897; Harmsen et al., 2014).
Viewed from this vantage, one cannot but wonder whether the recent
wave of land abandonment and reforestation has already run its course
and be showing signs of yet another turnaround (Fig. 2).
The link made between earlier out-migrants and later return mi-
grants is also significant. In both cases, migrants contributed to less
farming: the first directly as young farmers departing the country and
the second indirectly as elderly return migrants drawing farmers out of
the countryside to build and service their relatively lavish, new homes
(Walters, 2016a). Out-migration and ensuing income remittances have
been found elsewhere to reduce farming pressures (Hecht, 2010), but
few studies have clarified the specific causal dynamics and made the
appropriate distinctions between out- vs. return-migration (Hecht et al.,
2015; Rigg et al., 2016; Taylor et al., 2016). In Saint Lucia, the net
effect of all this migration has been to increase reforestation of the
rugged interior while spurring development on flatter lowlands and
lands near the coast (Walters, 2016a, 2016b; see also Aide and Grau,
2004).
The finding that land tenure was relatively unimportant as a cause
of land abandonment and reforestation was unexpected given that lit-
erature on West Indian land use has often emphasized its’ significance
(Rubenstein, 1975; Rojas, 1984; Besson and Momsen, 1987; Rojas and
Meganck, 1987; Barrow, 1992; Brierley, 1992; Dujon, 1997; see also
Robinson et al., 2014). Yet, this wider literature is also sharply divided
as to which forms of tenure—estate, smallholder private, communal
family land, etc.—are more or less conducive to successful agricultural
development (Walters, 2012b). Such ambiguity suggests either that
tenure’s significance for land use in the West Indies has been generally
over-estimated or its influence is highly context-specific. The influence
of tenure in this study was not trivial, but nonetheless was largely
swamped by other factors shaping agriculture and land use. By contrast,
the impact on land use of earlier agricultural declines has indeed varied
depending on land ownership. For example, the collapse of sugar in the
late 1800s and again in the 1930s/40 s hit the estate farm sector
especially hard and led to calls for land redistribution and more support
of smallholder farming as a viable alternative (WIRC, 1897; Harmsen
et al., 2014).
4.2. Reflections on research methodology
Scholars from a range of academic disciplines have adopted historical
approaches to study land use and forest change. There is now an extensive
literature on these topics within history (e.g., Cronon, 1983; Thirgood,
1989; Kirby and Watkins, 1998; Agnoletti and Anderson, 2000; Williams,
2003), geography (e.g., Blaikie and Brookfield, 1987; Kummer, 1992;
Peluso, 1992; Leach and Mearns, 1996; Reenberg, 2001; Klepeis, 2003;
Lukas, 2014), anthropology (e.g., Moran, 1993; McGuire, 1997; Balee,
1998; Acheson and McCloskey, 2008), sociology (e.g., Rudel, 1993) and
ecology (e.g., Foster, 1992). This paper complements but also adds to this
wider literature by describing and illustrating an analytical methodology
(ACE) that is well-suited for doing interdisciplinary, causal-historical re-
search on such things as land use and forest change.
Political ecology (PE) and land change science (LCS) typically differ
in their approaches to studying environmental change (Turner and
Robbins, 2008; Brannstrom and Vadjunec, 2013b), yet both have taken
steps to incorporate more history in their analysis. For example, land
change scientists have proposed ‘integrated land history’ as way to in-
corporate longer time frames and deeper engagement with historical
events and processes (Klepeis and Turner, 2001). This is a positive
development, but its advocates continue to treat historical and scientific
research components as qualitatively distinct endeavors, the former
focused on creating qualitative narratives and the latter focused on
compiling and modeling quantitative data. History thus provides con-
textual information, but ultimately science delivers the hard explana-
tions. In contrast, a causal-historical perspective views such distinctions
619
Land Use Policy 67 (2017) 608–624
as inaccurate and ultimately counter-productive because qualitative
evidence may be equally or even more compelling than quantitative
evidence for assessing causal claims.
Political ecology has sought to incorporate history and historical
analysis in a variety of ways, some heavily theoretical (Watts, 2003;
Watts and Peet, 2004) and others more empirical, based on the appli-
cation of historical methods and case-by-case construction of historical
narratives (Peluso, 1992; Leach and Mearns, 1996; Batterbury and
Bebbington, 1999; Shaver et al., 2015). As with historical research more
generally, so-called ‘historical political ecology’ is often not explicit
about analytical methodology and, where it is, may approach research
and analysis differently than ACE. A notable example of this is Blaikie
and Brookfield’s (1987) ‘chains of explanation’ approach which has
commonalities with ACE, but differs in placing emphasis on social
structures (rather than events) and in insisting that lower-level phe-
nomena (e.g., land users) are causally nested and thereby determined
by higher-level phenomena (e.g., government policies). But this rigid
prescription is simply not consistent with the indeterminate nature of
historical causation.
In contrast, Lukas (2014, 2015) provides a compelling illustration of
PE research on land use change that is consistent with ACE. Briefly,
Lukas sought to explain the causes of downstream sedimentation and
upstream land degradation in a watershed in Java. Using a variety of
methods, including historical cartography and interviews with land
users and policy makers, Lukas began his research by clearly doc-
umenting historical changes to lagoon sedimentation and from there he
followed various lines of causal inquiry backward in time and outward
in space. By doing this, he learned that a wide range of causal factors
had at different times contributed to upland degradation and soil ero-
sion, including high-impact cultivation practices, land tenure in-
security, in-migration of people, resource conflicts, government policies
and a volcanic eruption (Lukas, 2015).
What insights can be drawn from the Saint Lucia case study re-
garding the different approaches to human-environment research?
Firstly, LCS and ACE are alike (but different from SES and PE) in their
insistence on being clear in the beginning about the ‘explanandum’
event—i.e., the environmental change(s) for which an explanation is
being sought (Meyfroidt, 2015; Efroymson et al., 2016). But the two
approaches are likely to diverge on the next steps. In LCS, researchers
typically proceed (using causes-to-effects reasoning) to compile large
data sets and build models to test for statistical associations between
variables which may suggest possible causation.26 Explanations that
may be derived from this work still depend on qualitative reasoning
and are, ultimately, only as credible as established theory and em-
pirical knowledge of the cases at hand permit. For this reason, LCS
tends to be at its best when quantitative-statistical work is well-in-
tegrated with in-depth, qualitative research of the cases under study
(e.g., Shaver et al., 2015; Turner et al., 2016; see also Meyfroidt,
2015).
With ACE, having identified the explanandum event, research then
proceeds to build causal explanations using abductive (effects-to-
causes) reasoning as a guide to inquiry, drawing on diverse sources of
evidence and selecting methods appropriately suited for evaluating
alternative causal hypothesis as these emerge through the course of
subsequent investigations (Walters and Vayda, 2009). In the research
presented here, quantitative data were also compiled and subject to
statistical analysis, but this was done in a more targeted manner to
evaluate specific causal hypotheses that had been conjectured as
plausible on the basis of general knowledge of the situation and qua-
litative observations and investigations already made on-the-ground.
This approach enabled a precise, efficient use of statistical tools that
26
Efroymson et al. (2016) argue for a more pluralistic approach to causal analysis in
LCS studies. Nonetheless, their overall analytical approach remains based on causes-to-
effects, rather than an effects-to-causes reasoning (see Table 1).
B.B. Walters
generated reliable results.27 For example, statistical tests confirmed an
already suspected causal relationship between topographical conditions
(slope, elevation, road access) and reforestation, but refuted—rather
unexpectedly—land tenure as a decisive cause. At the same time, bio-
ecological and socio-economic information were gathered and analyzed
in tandem, not treated as separate study components with attempts at
integration made after-the-fact. Rather, with an eye kept on the guiding
research questions, different sources of evidence were integrated and
causal hypotheses formulated and evaluated in an iterative, step-wise
manner through the course of the research.28
Second, it is not easy relating the case study findings writ large to
SES concepts of resilience and the adaptive cycle because, as discussed
earlier, these ideas are so conceptually and ontologically ill-defined
(Walters, 2012a). This poses a serious problem for explanation-oriented
research because clarity and precision are core criteria for evaluating
the success of explanations (McCullagh, 1998; Gaddis, 2002; Hedstrom
and Ylikoski, 2010; Ylikoski and Kuorikoski, 2010; Priest, 2017; see
also Walters and Vayda, 2009; Vayda and Walters, 2011; Meyfroidt,
2015:12-13; Efroymson et al., 2016).
For many of its advocates, SES concepts and frameworks serve
primarily as analytical heuristics (Folke, 2006; Walker et al., 2006;
Carter et al., 2014; Burns and Rudel, 2015). Heuristics have value as
guides to intellectual inquiry, but they do not explain phenomena, and
so are of little concrete interest here. But it is also not apparent what
insights are gained in this case by granting theoretical-explanatory
weight to SES concepts and schema because these simply do not readily
incorporate the decisive causal influence of such disparate and his-
torically contingent phenomena as the rise of tourism, out- and return-
migration, and fluctuating international commodity markets (see also
Bunce et al., 2009). As such, it is hard to conceive how concepts like
‘resilience’, ‘stable states’ and ‘adaptive cycles’ could in any but the
most facile theoretical way impart real understanding of a society and
environment that has for so long been in such dramatic, open-ended
flux (Harmsen et al., 2014; Walters, 2016a).
That said, there are sub-elements of the larger case study that
showed evidence of tight-couplings and genuinely system-like behavior.
Specifically, banana farming and its gradual development in Saint Lucia
reveal a clear trend in increased complexity (and arguably, ‘rigidity’) as
farmers were required over time to adhere to ever-more elaborate on-
farm management and off-farm marketing protocols intended to boost
productivity (fertilizers, pest management) and deliver optimum pro-
duct quality (e.g., field-packing, harvest scheduling) (see Romalis,
1975; Fingal, 2008; Moberg, 2008). In fact, Grossman’s (1994, 1998,
2003) research on the banana industry in neighboring St. Vincent,
which he characterizes as a ‘contract-farming system’, elucidates im-
portant insights of a systemic nature about its evolution and response to
internal pressures and externally-imposed shocks.
In Saint Lucia, an argument could be made that the first banana bust
in the late 1960s was reflective of systemic ‘collapse’ brought on by a
combination of external shocks and internal management breakdown,
which was then followed by a ‘re-organization’ phase and return to
‘growth’ of the system (i.e., an adaptive cycle). But a similar argument
27
In a pithy critique of contemporary science’s obsession with statistical P values,
Nuzzo (2014) reminds us that the validity of a statistical test to a null hypothesis depends,
not just on the result of the test, itself, but most importantly, on the plausibility of the
original hypothesis being tested. Many others have similarly argued that qualitative
reasoning and evidence play a central and indispensable role in causal analysis (e.g.,
Ragin, 1987; Freedman, 1991, 1997; Goba, 2008; Vayda, 2009; Ylikoski and Kuorikoski,
2010; Walters and Vayda, 2011; Clarke et al., 2014; Meyfroidt, 2015).
28
This research project was also completed with a modest budget and without the
efforts of a large, multi-disciplinary team. The entire project budget was about $US
100,000 and this included field support for 3 graduate students to pursue their own
parallel projects. The author has been broadly trained in both the social and ecological
sciences and so was able to perform most of the research and analysis, himself. Ecological
investigations involved the assistance of several students and local field guides and, in
particular, botanical identification was performed by others with that specific expertise.
620
Land Use Policy 67 (2017) 608–624
cannot be made about the recent banana collapse, which was caused
largely by events and trends that lay well-beyond and often had nothing
to do with the farming system per se. In short, this recent decline re-
flected not a ‘collapse’ of the banana farming system, but rather a
gradual, voluntary abandonment of farming from less productive sites
in response to wider socio-economic changes. In fact, banana farming
persists today in Saint Lucia, but in much reduced form and on the most
productive sites only (personal observation). And what production has
been lost from Saint Lucia has only been transferred to other, likely
even more management-intensive farming ‘systems’ elsewhere.
Concepts of resilience and adaptive cycle are just not helpful for un-
derstanding, yet alone explaining, these recent changes.
Finally, this study should put to rest concerns raised by some poli-
tical ecologists that the methodological approach used with ACE fails to
account for politics and structural factors in explanations (e.g., Robbins,
2004:208; Watts and Peet, 2004:18; Perramond, 2007:503; Penna-
Firme, 2012:200). As illustrated here, the influence of both structural
factors (e.g., land tenure) and political events (e.g., changes in trade
policy and international banana markets) were rigorously evaluated as
possible causes of rural reforestation, in each case with attention to the
actual evidence of causation and being as clear and precise as possible
about the causal chains involved. By doing this abductively (reasoning
from effects to causes rather than from causes to effects), existing the-
ories and hypotheses about the political economy of environmental
degradation were considered, but not prioritized. In fact, two major
study findings—that reforestation was caused in part by increased trade
liberalization and was equally widespread on large estates and on small
farms—contradict widely-held assumptions in PE about inequality and
trade liberalization being dominant causes of environmental degrada-
tion in the global south.29 Much social and environmental damage
across the Caribbean can be causally linked to the region’s long history
of colonial and capitalist exploitation, but the specific socio-economic
and environmental changes captured in this study do not conform to
this general narrative.
Acknowledgements
This research was supported by grants from the Social Sciences and
Humanities Research Council of Canada and Marjorie Young Bell
Faculty Fund of Mount Allison University. I am grateful to the support
of the Government of Saint Lucia and, in particular, members of
Forestry Division, including Michael Bobb, Adams Toussaint, Michael
Andrew and Rebecca Rock. Special thanks also to Gregor Williams, Yves
Renard, Lisa Hansen, Melvin Smith, Roger Graveson, Kai Wulf, Jennifer
Sargent, Nigel Selig, Marshall Symons, Frances Ross, Shannon White,
Peter Jackson, Jolien Harmsen, Christina Tardif and Mariana Baptista. I
much appreciate comments on earlier drafts of this paper from Tom
Rudel, Pete Vayda and anonymous reviewers.
References
Abenaty, F.K., 2000. Saint Lucians and Migration: Migrant Returnees, Their Families and
Saint Lucian Society. PhD. South Bank University, London (313 pp).
Abend, G., 2008. The meaning of ‘theory’. Sociol. Theory 26, 173–199.
Acheson, J.M., McCloskey, J., 2008. Causes of deforestation: the Maine case. Hum. Ecol.
36, 909–922.
Adger, W.N., 2000. Social and ecological resilience: are they related? Prog. Hum. Geogr.
24, 347–364.
Forest History: International Studies on Socio-economic and Forest Ecosystem Change.
29
One could argue that the shift in banana production away from Saint Lucia to other
countries as a result of the WTO rulings simply transfers environmental degradation from
one location to another. However, the major loss of banana production in Saint Lucia
occurred from marginal, hillside lands that typically achieved far lower production levels
than are attained by competitors who typically grow their bananas on flatter lands
(Welch, 1996; Moberg, 2008). As such, the shift almost certainly resulted in net land-
sparing and a reduction in the sorts of impacts (soil erosion, leaching of agro-chemicals)
typically associated with steep-hillside farming of bananas.
B.B. Walters
Report No. 2 of the IUFRO Task Force on Environmental Change. In: Agnoletti, M.,
Anderson, S. (Eds.), CAB International, New York.
Agnoletti, M., Rotherham, I.D., 2015. Landscape and biocultural diversity. Biodivers.
Conserv. 24, 3155–3165.
Aide, T.M., Grau, H.R., 2004. Globalization, migration, and Latin American ecosystems.
Science 305, 1915–1916.
Aide, T.M., Clark, M.L., Grau, H.R., Lopez-Carr, D., Levy, M.A., Redo, D., Bonilla-Moheno,
M., Riner, G., Andrade-Nunez, M.J., Muniz, M., 2013. Deforestation and reforestation
of Latin America and the Caribbean (2001–2010). Biotropica 45, 262–271.
Aldrich, S., Walker, R., Simmons, C., Caldas, M., Perz, S., 2012. Contentious land change
in the Amazon’s arc of deforestation. Ann. Assoc. Am. Geogr. 102, 103–128.
Alvarez-Berrios, N.L., Redo, D.J., Aide, M.T., Clark, M.L., Grau, R., 2013. Land change in
the Greater Antilles between 2001 and 2010. Land 2, 81–107.
Armitage, D., Johnson, D., 2006. Can resilience be reconciled with globalization and the
increasingly complex conditions of resource degradation in Asian coastal regions?
Ecol. Soc. 11 (1), 19pp.
Aspinall, R., 2004. Modelling land use change with generalized linear models—a multi-
model analysis of change between 1860 and 2000 in Gallatin Valley, Montana. J.
Environ. Manag. 72, 91–103.
Atkinson, E.E., Marin-Spiotta, E., 2015. Land use legacy effects on the structure and
composition of subtropical dry forests in St. Croix, U.S. Virgin Islands. For. Ecol.
Manag. 335, 270–280.
Baggio, J.A., Brown, K., Hellebrandt, D., 2015. Boundary object or bridging concept? A
citation network analysis of resilience. Ecol. Soc. 20 (2), 2 10/5751/ES-07484-
200202.
Advances in Historical Ecology. In: Balee, W. (Ed.), Columbia University Press, New York.
Barbier, E.B., Burgess, J.C., Grainger, A., 2010. The forest transition: towards a more
comprehensive theoretical framework. Land Use Policy 27, 98–107.
Barnes, T.J., 2013. Big data, little history. Dialog. Hum. Geogr. 3, 297–302.
Barrow, C., 1992. Family Land and Development in Saint Lucia. Monograph Series No. 1.
Institute of Social & Economic Research, University of the West Indies, Cave Hill,
Barbados (83 pp).
Basken, P., 2013. That elastic term. Chron. High. Ed. Chron. Rev. 2013 (May).
Batterbury, S., Bebbington, A., 1999. Environmental histories, access to resources and
landscape change: an introduction. Land Degrad. Dev. 10, 279–290.
Behnke, R., Mortimore, M., 2016. Introduction: the end of desertification? In: Behnke,
R.H., Moritmore, M. (Eds.), The End of Desertification? Disputing Environmental
Change in the Drylands. Springer Earth Systems Science, pp. 1–34. http://dx.doi.org/
10.1007/978-3-642-16014-1_1.
Linking Social and Ecological Systems: Management Practices and Social Mechanisms for
Building Resilience. In: Berkes, F., Folke, C. (Eds.), Cambridge University Press,
Cambridge.
Land and Development in the Caribbean. In: Besson, J., Momsen, J. (Eds.), Macmillan,
London.
Binder, C.R., Hinkel, J., Bots, P., Pahl-Wost, l., 2013. Comparison of frameworks for
analyzing social-ecological systems. Ecol. Soc. 18 (4), 26. http://dx.doi.org/10.5751/
es-05551-180426.
Blaikie, P., Brookfield, H., 1987. Land Degradation and Society. Methuen, New York.
Blaikie, P., 1999. A review of political ecology: issues, epistemology and analytical nar-
ratives. Zeitschrift fur Wirtschaftsgeogr. 43, 131–147.
Brand, F.S., Jax, K., 2007. Focusing the meaning(s) of resilience: resilience as a de-
scriptive concept and a boundary object. Ecol. Soc. 12 (1), 23 pp.
Land Change Science, Political Ecology and Sustainability. In: Brannstrom, C., Vadjunec,
J.M. (Eds.), Routledge, New York.
Brannstrom, C., Vadjunec, J.M., 2013b. Notes for avoiding a missed opportunity in sus-
tainability science: integrating land change science and political ecology. In:
Brannstrom, C., Vadjunec, J.M. (Eds.), Land Change Science, Political Ecology and
Sustainability. Routledge, New York, pp. 1–23.
Breen, H.H., 1844. St. Lucia: Historical, Statistical and Descriptive. Longman, Brown,
Green and Longman’s, London 423 pp.
Brierley, J.S., 1992. A study of land redistribution and the demise of Grenada’s estate
farming system 1940–1988. J. Rural Stud. 8 (1), 67–84.
Brown, J.C., Purcell, M., 2005. There’s nothing inherent about scale: political ecology, the
local trap, and the politics of development in the Brazilian Amazon. Geoforum 36,
617–624.
Bryant, R.L., Bailey, S., 1997. Third World Political Ecology. Routledge, London.
Bunce, M., Mee, L., Rodwell, L.D., Gibb, R., 2009. Collapse and recovery in a remote small
island: a tale of adaptive cycles or downward spirals? Glob. Environ. Change 19,
213–226.
Burns, T.J., Rudel, T.K., 2015. Metatheorizing structural human ecology at the down of
the third millennium. Hum. Ecol. Rev. 22, 13–33.
Byron, M., Condon, S., 1996. A comparative study of Caribbean return migration from
Britain and France: towards a context-dependent explanation. Trans. Inst. Brit. Geogr.
N.S. 21, 91–104.
Byron, M., Condon, S., 2008. Migration in Comparative Perspective: Caribbean
Communities in Britain and France. Routledge, London 288 pp.
Byron, M., 2000. Return migration to the Eastern Caribbean: comparative experiences
and policy implications. Soc. Econ. Stud. 49 (4), 155–188.
Cairns, R., Kryzwoszynska, A., 2016. Anatomy of a buzzword: the emergence of the
‘water-energy-food nexus’ in UK natural resources debates. Environ. Sci. Policy 64,
164–170.
Campos, D.G., 2011. On the distinction between Peirce’s abduction and Lipton’s inference
to the best explanation. Synthese 180, 419–442.
Carnegie, C.V., 1982. Strategic flexibility in the West Indies: a social psychology of
Caribbean migration. Caribb. Rev. 11 (1), 10–13.
Carter, N.H., Vina, A., Hull, V., McConnell, W.J., Axinn, W., Ghimire, D., Liu, J., 2014.
621
Land Use Policy 67 (2017) 608–624
Coupled human and natural systems approach to wildlife research and conservation.
Ecol. Soc. 19 (3), 43. http://dx.doi.org/10.5751/ES-06881-190343.
Chamberlin, T.C., 1965. The method of multiple working hypotheses. Science 148, 754-
59. Republished (2011) In: Vayda, A.P., Walters, B.B. (Eds.), Causal Explanation for
Social Scientists: A Reader. AltaMira Press, Lantham, MD, pp. 168–178.
Chowdhury, R.R., Turner, B.L.I.I., 2005. Reconciling agency and structure in empirical
analysis Smallholder land use in the Southern Yucatan, Mexico. Ann. Assoc. Am.
Geogr. 96, 302–322.
Clarke, B., Gillies, D., Illari, P., Russo, F., Williamson, J., 2014. Mechanisms and the
evidence hierarchy. Topoi 33, 339–360.
Clegg, P., 2002. The Caribbean Banana Trade: From Colonization to Globalization.
Palgrave MacMillan, New York 213 pp.
Clement, F., 2012. For critical socio-ecological systems studies: integrating power and
discourses to move beyond the right institutional fit. Environ. Conserv. 40, 1–4.
Conway, D., Potter, R., 2006. Caribbean transnational return migrants as agents of
change. Geogr. Compass 1, 1–21.
Conway, D., 1999/2000. The importance of migration for Caribbean development, Glob.
Develop. Stud, 2(1–2), 73–105.
Costa, R.L., Prevedello, J.A., de Souza, B., Cabral, D.C., 2017. Forest transitions in tropical
landscapes: a test in the Atlantic forest biodiversity hotspot. Appl. Geogr. 82, 93–100.
Cote, M., Nightingale, A.J., 2012. Resilience thinking meets social theory: situating social
change in socio-ecological systems (SES) research. Progr. Hum. Geogr. 36, 1–15.
Cox, H., 1897. West Indian sugar. Econ. J. 7 (28), 599–605.
Cronon, W., 1983. Changes in the Land: Indians, Colonists and the Ecology of New
England. Hill & Wang, New York.
Dalla-Nora, E.L., Aguiar, A.P.D., Lapola, D.M., Woltjer, G., 2014. Why have land use
change models for the Amazon failed to capture the amount of deforestation over the
last decade? Land Use Policy 39, 403–411.
Davidson, D.J., 2010. The applicability of the concept of resilience to social systems: some
sources of optimism and nagging doubts. Soc. Nat. Resour. 32, 1135–1149.
Dore, R.F., 1961. Function and cause. Am. Soc. Rev. 26 (6), 843–853.
Downes, B.J., Miller, F., Barnett, J., Glaister, A., Ellemor, H., 2013. How do we know
about resilience? An analysis of empirical research on resilience, and implications for
interdisciplinary praxis. Environ. Res. Lett. 8. http://dx.doi.org/10.1088/1748-
9326/8/1/014041.
Dray, W.H., 1957. Laws and Explanation in History. Oxford University Press, Oxford.
Dray, W.H., 1964. Philosophy of History. Prentice-Hall, Englewood Cliffs, New Jersey.
Dujon, V., 1997. Communal property and land markets: agricultural development policy
in St Lucia. World Develop. 25, 1529–1540.
Ebach, M.C., Michael, M.S., Shaw, W.S., Goff, J., Murphy, D.J., Matthews, S., 2016. Big
data and the historical sciences: a critique. Geoforum 71, 1–4.
Editors, 2016. Buzzword off. Nature 538, 140.
Efroymson, R.A., Kline, K., Angelsen, A., Verburg, P.H., Dale, V., Langeveld, J., McBride,
A., 2016. A causal analysis framework for land-use change and the potential role of
bioenergy policy. Land Use Policy 59, 516–527.
Elder-Vass, D., 2010. The Causal Power of Social Structures: Emergence, Structure and
Agency. Cambridge University Press, Cambridge.
Fingal, C., 2008. Fair ‘trade not aid’: globalization, fair trade and the Saint Lucia banana
industry. Caribb. Geogr. 15, 118–129.
Folke, C., 2006. Resilience: the emergence of a perspective for socio-ecological systems
analysis. Glob. Environ. Change 16, 253–267.
Forsyth, T., 2008. Political ecology and the epistemology of social justice. Geoforum 39,
756–764.
Foster, D.R., 1992. Land-use history (1730–1990) and vegetation dynamics in Central
New England, USA. J. Ecol. 80, 753–772.
Fraser, E.D.G., Stringer, L.C., 2009. Explaining agricultural collapse: macro-forces, micro-
crises and the emergence of land use vulnerability in southern Romania. Glob.
Environ. Change 19, 45–53.
Freedman, D.A., 1991. Statistical models and shoe leather. In: Marsden, P.V. (Ed.),
Sociological methodology, vol. 21. Basil Blackwell for the American Sociological
Association, Oxford, pp. 291–313. Republished (2011) In: Vayda, A.P., Walters, B.B.
(Eds.), Causal Explanation for Social Scientists: A Reader. AltaMira Press, Lantham,
MD, pp. 151–167.
Freedman, D.A., 1997. From association to causation via regression. In: McKim, V.R.,
Turner, S.P. (Eds.), Causality in Crisis? Statistical Methods and the Search for Causal
Knowledge in the Social Sciences. University of Notre Dame Press, Notre Dame, pp.
113–161.
Fridell, G., 2011. The case against cheap bananas: lessons from the EU-Caribbean Banana
Agreement. Crit. Sociol. 37 (3), 285–307.
GLP, 2005. Science Plan and Implementation Strategy. IGBP Report No. 53/IHDP Report
No. 19. IGBP Secretariat, Stockholm 64 pp.
GOSL, 1983. Saint Lucia Economic Review: 1982. Ministry of Finance, Planning and
Statistics. Government of Saint Lucia, Castries.
GOSL, 1985. Saint Lucia Economic Review: 1984. Ministry of Finance, Statistics and
Negotiating. Government of Saint Lucia, Castries.
GOSL, 1987. Saint Lucia Economic and Social Review: 1986. Ministry of Finance, Statistics
and Negotiating. Government of Saint Lucia, Castries.
GOSL, 1990. Saint Lucia Economic and Social Review: 1989. Ministry of Finance, Statistics
and Negotiating. Government of Saint Lucia, Castries.
GOSL, 1991. Saint Lucia Economic and Social Review: 1990. Ministry of Finance, Statistics
and Negotiating. Government of Saint Lucia, Castries.
GOSL, 1996. Saint Lucia Census of Agriculture, Final Report. Ministry of Agriculture, Lands,
Fisheries and Forestry, Government of Saint Lucia, Castries.
GOSL, 1998. Economic and Social Review, 1997. Ministry of Finance and Planning,
Government of Saint Lucia, Castries.
GOSL, 1999. Economic and Social Review, 1998. Ministry of Finance and Planning,
B.B. Walters
Government of Saint Lucia, Castries.
GOSL, 2002. Economic and Social Review, 2001. Ministry of Finance, International
Financial Services and Economic Affairs, Government of Saint Lucia, Castries.
GOSL, 2007. 2007 Saint Lucia Census of Agriculture, Final Report. Ministry of Agriculture,
Forestry & Fisheries, Government of Saint Lucia, Castries.
GOSL, 2008. Economic and Social Review, 2007. Ministry of Finance, Economic Affairs
and National Development, Government of Saint Lucia, Castries.
GOSL, 2011a. 2010 Population and Housing Census, Preliminary Report. Castries.
GOSL, 2011b. Economic and Social Review, 2010. Ministry of Finance, Economic Affairs
and National Development, Government of Saint Lucia, Castries.
GOSL, 2012. Economic and Social Review, 2011. Ministry of Finance, Economic Affairs
and National Development, Government of Saint Lucia, Castries.
GOSL, 2016. Review of the Economy, 2015. Ministry of Finance, Economic Affairs,
Planning and Social Security, Government of Saint Lucia, Castries 62pp. + app.
Gaddis, J.L., 2002. The Landscape of History: How Historians Map the Past. Oxford
University Press, Oxford.
Galaty, J.G., 2013. Comment on Response diversity and resilience socio-ecological sys-
tems. Curr. Anthr. 54 (2), 130–131.
Geist, H., McConnell, W., Lambin, E.F., Moran, E., Alves, D., Rudel, T., 2006. Causes and
trajectories of land-use/cover change. In: Lambin, E.F., Geist, H. (Eds.), Land-Use and
Land-Cover Change: Local Processes and Global Impacts. Springer-Verlag, Berlin, pp.
41–70.
Goba, G., 2008. Re-conceptualizing generalization: old issues in a new frame. The Sage
Handbook of Social Research Methods. Sage, New York, pp. 193–213.
Gould, S.J., 1989. Wonderful Life: The Burgess Shale and the Nature of History. W.W
Norton & Co, New York.
Grau, H., Aide, T., Zimmerman, J., Thomlinson, J., Helmer, E., Zou, X., 2003. The eco-
logical consequences of socio-economic land-use changes in postagriculture Puerto
Rico. BioSci 53 (12), 1159–1168.
Graveson, R., 2005. Plants of Saint Lucia. Unpublished electronic data base: saintlucia-
plants com/.
Grimm, V., Wissel, C., 1997. Babel, or the ecological stability discussions: an inventory
and analysis of terminology and a guide for avoiding confusion. Oecologia 109,
323–334.
Grossman, L.S., 1994. British aid and Windwards bananas: the case of St. Vincent and the
Grenadines. Soc. Econ. Stud. 43, 151–179.
Grossman, L.S., 1998. The Political Ecology of Bananas. University of North Carolina
Press, Chapel Hill.
Grossman, L.S., 2003. The St. Vincent Banana Growers’ Association, contract farming, and
the peasantry. In: Striffler, S., Moberg, M. (Eds.), Banana Wars: Power, Production
and History in the Americas. Duke University Press, Durham, pp. 286–315.
Panarchy: Understanding Transformations in Human and Natural Systems. In: Gunderson,
L., Holling, C.S. (Eds.), Island Press, Washington, D.C.
Harmsen, J., Ellis, G., Devaux, R., 2014. A History of Saint Lucia. Lighthouse Road
Publications, Vieux Fort, Saint Lucia 413 pp.
Hawthorn, G., 1991. Plausible Worlds: Possibility and Understanding in History and the
Social Sciences. Cambridge University Press, Cambridge, Mass.
Hecht, S., Yang, A.L., Basnett, B.S., Padoch, C., Peluso, N.L., 2015. People in Motion,
Forests in Transition: Trends in Migration, Urbanization, and Remittances and Their
Effects on Tropical Forests. Occasional Paper 142. Center for International Forestry
Research (CIFOR), Bogor, Indonesia (37 pp).
Hecht, S., 2010. The new rurality: globalization, peasants and the paradoxes of land-
scapes. Land Use Policy 27, 161–169.
Hedstrom, P., Ylikoski, P., 2010. Causal mechanisms in the social sciences. Ann. Rev.
Sociol. 36, 49–67.
Helmer, E., Kennaway, T., Pedreros, D., Clark, M., Marcano, H., Tieszen, L., Ruzycki, T.,
Schill, S., Carrington, S., 2008. Land cover and forest formation distributions for St.
Kitts, Nevis, St. Eustatius, Grenada and Barbados from decision tree classification of
cloud-cleared satellite imagery. Caribb. J. Sci. 44 (2), 175–198.
Hintikka, J., 1998. What is abduction? The fundamental problem of contemporary epis-
temology. Trans. Charles S Peirce Soc. 34 (3), 503–533.
Holling, C.S., 1996. Engineering resilience versus ecological resilience. Engineering
Within Ecological Constraints. National Academy of Sciences, Washington, D.C, pp.
31–43.
Holm, P., Goodsite, M.E., Cloetingh, S., Agnoletti, M., Moldan, B., Lang, R.D., Leemans,
R., Moeller, J.O., Buendia, M.P., Pohl, W., Scholz, R.W., Sors, A., Vanheusden, B.,
Yusoff, K., Zondervan, R., 2013. Collaboration between the natural, social and human
sciences in global change research. Environ. Sci. Policy 28, 25–35.
Homewood, K., 2013. Comment on Response diversity and resilience socio-ecological
systems. Curr. Anthr. 54, 131–132.
Hornborg, A., 2009. Zero-sum world: challenges in conceptualizing environmental load
displacement and ecologically unequal exchange in the world-system. Intern. J.
Comp. Sociol. 50 (3–4), 237–262.
Howard, R., 1974. Flora of the Lesser Antilles: Leeward and Windward Islands. Arnold
Arboretum, Jamaica Plains, Mass 6 Volumes.
IGBP, 2006. Science Plan and Implementation Strategy IGBP Report No. 55. IGBP
Secretariat, Stockholm 76 pp.
Jennings, T.L., 2011. Transcending the adaptation/mitigation climate change science
policy debate: unmasking assumptions about adaptation and resilience. Weather
Clim. Soc. 3, 238–248.
Keenan, R.J., Reams, G., Achard, F., de Freitas, J.V., Grainger, A., Lindquist, E., 2015.
Dynamics of global forest area: results from the FAO global forest resources assess-
ment 2015. For. Ecol. Manag. 352, 9–20.
Kelly, A.M., 2016. The Discovery of Chance: The Life and Thought of Alexander Herzen.
Harvard University Press, Cambridge, Mass.
Kimber, C.T., 1988. Martinique Revisited: The Changing Plant Geographies of a West
622
Land Use Policy 67 (2017) 608–624
Indian Island. Texas A & M University Press, College Station 480 pp.
The Ecological History of European Forests. In: Kirby, K.J., Watkins, C. (Eds.), CAB
International, New York.
Kirchhoff, T., Brand, F.S., Hoheisel, D., Grimm, V., 2010. The one-sidedness and cultural
bias of the resilience approach. GAIA: Ecol. Perspect. Sci. Soc. 19 (1), 25–32.
Klepeis, P., Turner II, B.L., 2001. Integrated land history and global change science: the
example of the Southern Yucatan Peninsular Region Project. Land Use Policy 18,
27–39.
Klepeis, P., 2003. Development policies and tropical deforestation in the Southern
Yucatan Peninsula: centralized and decentralized approaches. Land Degrad. Dev. 14,
541–561.
Kummer, D.M., 1992. Deforestation in the Post-War Philippines. University of Chicago
Press, Chicago.
Lamb, H., 2008. Fighting the Banana Wars and Other Fair Trade Battles. The Fairtrade
Foundation, Rider London 245 pp.
Lambin, E.F., Meyfroidt, P., 2010. Land use transitions: socio-ecological feedback versus
socio-economic change. Land Use Policy 27, 108–118.
Lambin, E.F., Baulies, X., Bockstael, N., Fischer, G., Krug, T., Leemans, R., Moran, E.F.,
Rindfuss, R.R., Sato, Y., Skole, D., Turner, B.L., II, Vogel, C., 1999. Land-use and land-
cover change (LUCC) implementation strategy. International Geosphere-Biosphere
Programe and International Human Dimensions Programme on Global
Environmental Change. IGBP Report 48/IHDP Report 10 (http://www.geo.ucl.ac.be/
LUCC/implstrategy/, downloaded 4/17/07).
Lambin, E.F., Turner, B.L., Geist, H., Agbola, S., Angelsen, A., Bruce, J., Coomes, O.,
Dirzo, R., Fischer, G., Folke, C., George, P., Homewood, L., Imbernon, J., Leemans, R.,
Li, X., Moran, E., Mortimore, M., Ramakrishnan, P., Richards, J., Skanes, H., Steffan,
W., Stone, G., Svedin, U., Veldkamp, T., Vogel, C., Xu, J., 2001. The causes of land-
use and land-cover change: moving beyond myths. Glob. Environ. Change 11,
261–269.
Lambin, E.F., Geist, H., Rindfuss, R.R., 2006. Introduction: local processes with global
impacts. In: Lambin, E.F., Geist, H. (Eds.), Land-Use and Land-Cover Change: Local
Processes and Global Impacts. Springer-Verlag, Berlin, pp. 1–8.
The Lie of the Land: Challenging Received Wisdom on the African Environment. In:
Leach, M., Mearns, R. (Eds.), International African Institute in association with James
Curry, Oxford.
Leslie, P., McCabe, J.T., 2013. Response diversity and resilience socio-ecological systems.
Curr. Anthro. 54, 114–143.
Lewis, D., 1986. Philosophical Papers, Vol. 2, Oxford University Press, Oxford.
Republished (2011) In: Vayda, A.P, Walters, B.B. (Eds.), Causal Explanation for Social
Scientists, Altamira Press, Lantham, MD, pp. 25–39.
Liu, J., Dietz, T., Carpenter, S., Folke, C., Alberti, M., Redman, C., Schneider, S., Ostrom,
E., Pell, A., Lubchenco, J., Taylor, W., Ouyang, Z., Deadman, P., Kratz, T.,
Provencher, W., 2007. Coupled human and natural systems. Ambio 36 (8), 639–649.
Lombard, L.B., 1991. Events. In: In: Burkhardt, H., Smith, B. (Eds.), Handbook of
Metaphysics and Ontology, vol. 1. Philosophia Verlag, Munich, pp. 256–259.
Lowenthal, D., Comitas, L., 1962. Emigration and depopulation: some neglected aspects
of population geography. Geogr. Rev. 52, 195–210.
Lowenthal, D., 1972. West Indian Societies. Oxford University Press, London.
Ludeke, M.K.B., Petschel-Held, G., Schellnhuber, H.-J., 2004. Syndromes of global
change: the first panoramic view. GAIA 13 (1), 42–49.
Lukas, M.C., 2014. Eroding battlefields: land degradation in Java reconsidered. Geoforum
56, 87–100.
Lukas, M.C., 2015. Reconstructing Contested Landscapes: Dynamics, Drivers and Political
Framings of Land Use and Land Cover Change, Watershed Transformations and
Coastal Sedimentation in Java, Indonesia. PhD Dissertation. University of Bremen.
Mann, G., 2009. Should political ecology be Marxist? A case for Gramsci’s historical
materialism. Geoforum 40, 335–344.
Mather, A.S., Needle, C.L., 1998. The forest transition: a theoretical basis. Area 30,
117–124.
Mather, A.S., 1992. The forest transition. Area 24, 367–379.
McConnell, W.J., Millington, J., Reo, N., Alberti, M., Asbjornsen, H., Baker, L., Brozovic,
N., Drinkwater, L., Drzyzga, S., Fragaso, J., Holland, D., Jantz, C., Kohler, T.,
Maschner, H., Monticino, M., Podesta, G., Pontius Jr., R., Redman, C., Sailor, D.,
Urquhart, G., Liu, J., 2011. Research on coupled human and natural systems
(CHANS): approach, challenges and strategies. Bull. Ecol. Soc. Am. 92, 218–228.
McCullagh, C.B., 1984. Justifying Historical Descriptions. Cambridge University Press,
Cambridge.
McCullagh, C.B., 1998. The Truth of History. Routledge, London.
McFarlane, D., 1964. The future of the banana industry in the West Indies. Soc. Econ.
Stud. 13, 38–93.
McGuire, T., 1997. The last northern cod. J. Pol. Ecol. 4, 41–54.
McKim, V.R., 1997. Introduction. In: McKim, V.R., Turner, S.P. (Eds.), Causality in Crisis?
Statistical Methods and the Search for Causal Knowledge in the Social Sciences. Notre
Dame, University of Notre Dame Press, pp. 1–19.
Mena, C.F., Barbieri, A.F., Walsh, S.J., Erlien, C.M., Holt, F.L., Bilsborrow, R., 2006.
Pressure on the Cuyabeno Wildlife Reserve: development and land use/cover change
in the northern Ecuadorian Amazon. World Develop. 34, 1831–1849.
Mendoza, M.E., Grandos, E.L., Geneletti, D., Perez-Salicrup, D.R., Salinas, V., 2011.
Analyzing land cover and land use change processes at watershed level: a multi-
temporal study in the Lake Cuitzeo Watershed, Mexico (1975–2003). Appl. Geogr.
31, 237–250.
Meyfroidt, P., 2015. Approaches and terminology for causal analysis in land systems
science. J. Land Use Sci. 11, 501–522. http://dx.doi.org/10.1080/1747423X.2015.
1117530.
Moberg, M., 2008. Slipping Away: Banana Politics and Fair Trade in the Eastern
Caribbean. Berghan Books, New York 250pp.
B.B. Walters
Momsen, J.D., 1986. Migration and rural development in the Caribbean. Tijdschrift voor
Econ. En Soc. Geogr. 77, 50–58.
Moran, E.F., 1993. Deforestation and land use in the Brazilian Amazon. Hum. Ecol. 21,
1–21.
Moritz, M., 2017. Misreading a pastoral property regime in the Lagone floodplain,
Cameroon. Ecol. Soc. 22 (1), 13.
Myers, G., 2004. Banana Wars: The Price of Free Trade. Zed Books, London.
Nadasdy, P., 2007. Adaptive co-management and the gospel of resilience. In: Armitage,
D., Berkes, F., Doubleday, N. (Eds.), Adaptive Co-Management: Collaboration,
Learning and Multi-Level Governance. University of British Columbia Press,
Vancouver, pp. 208–226.
Nagendra, H., 2007. Drivers of reforestation in human-dominated forests. Proc. Natl.
Acad. Sci. 104 (39), 15218–15223.
Neocleous, M., 2013. Resisting resilience. Radical Phil. 178, 2–7.
Neumann, R.P., 2009. Political ecology: theorizing scale. Prog. Hum. Geogr. 33, 398–406.
Newell, B., Crumley, C., Hassan, N., Lambin, E.F., Pahl-Wostle, C., Underdal, A., Wasson,
R., 2005. A conceptual template for integrative human-environment research. Glob.
Environ. Change 15, 299–307.
Nickerson, R.S., 1998. Confirmation bias: a ubiquitous phenomenon in many guises. Rev.
Gen. Psych. 2, 175–220.
Nuzzo, R., 2014. Statistical errors. Nature 506, 150–152.
Nyerges, A.E., 2008. Orthodoxy and revision in West African Guinea savanna ecology. In:
Walters, B.B., McCay, B.J., West, P., Lees, S. (Eds.), Against the Grain: The Vayda
Tradition in Human Ecology and Ecological Anthropology. AltaMira Press, Lanham,
MD, pp. 81–98.
Nygren, A., Rikoon, S., 2008. Political ecology revisited: integration of politics and
ecology does matter. Soc. Nat. Resour. 21, 767–782.
O’Neil, H.W., 1964. The economy of Saint Lucia. Soc. Econ. Stud. 13, 440–470.
Ogden, L., Heynen, N., Oslender, U., West, P., Kassam, K.-A., Robbins, P., 2013. Global
assemblages, resilience, and earth stewardship in the Anthropocene. Front. Ecol.
Environ. 11 (7), 341–347.
Ostrom, E., 2009. A general framework for analyzing sustainability of social-ecological
systems. Science 325, 419–422.
Palsson, G., Szerszynski, B., Sorlin, S., Marks, J., Avrisl, B., Crumley, C., Hackmann, H.,
Holm, P., Ingram, J., Kriman, A., Buendia, M.P., Weehuizen, R., 2013.
Reconceptualizing the ‘Anthropos’ in the Anthropocene: integrating the social sci-
ences and humanities in global environmental change research. Environ. Sci. Policy
28, 3–13.
Peach, G.C.K., 1967. West indian migration to britain. Int. Migr. Rev. 1, 34–45.
Peach, C., 1968. West Indian Migration to Britain: A Social Geography. Institute of Race
Relations, Oxford University Press, London.
Peirce, C.S., 1932. In: In: Hartshorne, C., Weiss, P. (Eds.), Collected Papers of Charles
Saunders Peirce, vol. 2 Harvard University Press, Cambridge, Mass.
Peluso, N., 1992. Rich Forests, Poor People: Resource Control and Resistance in Java.
University of California Press, Berkeley.
Penna-Firme, R., 2012. Political and event ecology: critiques and opportunities for col-
laboration. J. Polit. Ecol. 20, 199–216.
Perramond, E.P., 2007. Tactics and strategies in political ecology research. Area 39,
499–507.
Perz, S.G., 2007. Grand theory and context-specificity in the study of forest dynamics:
forest transition theory and other directions. Prof. Geogr. 59, 105–114.
Plieninger, T., Draux, H., Fagerholm, N., Bieling, C., Burgi, M., Kizos, T., Kuemmerle, T.,
Primdahl, J., Verburg, P.H., 2016. The driving forces of landscape change in Europe:
a systematic review of the evidence. Land Use Policy 57, 2014–2214.
Pollini, J., 2010. Environmental degradation narratives in Madagascar: from colonial
hegemonies to humanist revisionism. Geoforum 41, 711–722.
Pooley, S.P., Mendelsohn, J.A., Milner-Gulland, E.J., 2013. Hunting down the chimera of
multiple disciplinarity in conservation science. Conserv. Biol. 28, 22–32.
Priest, G., 2017. Framing causal questions about the past: the Cambrian explosion as case
study. Stud. Hist. Phil. Biol. Biomed. Sci. 63, 55–63.
Ragin, C.S., 1987. The Comparative Method. University of California Press, Berkeley.
Rangan, H., Kull, C.A., 2009. What makes ecology ‘political’?: rethinking ‘scale’ in poli-
tical ecology. Prog. Hum. Geogr. 33, 28–45.
Raynolds, L.T., 2000. Re-embedding global agriculture: the international organic and fair
trade movements. Agric. Hum. Values 17, 297–309.
Read, R., 1994. The EC internal banana market: the issues and the dilemma. World Econ.
17, 219–235.
Redo, D.J., Aide, T.M., Clark, M.L., 2012. The relative importance of socioeconomic and
environmental variables in explaining land change in Bolivia, 2001–2010. Ann.
Assoc. Am. Geogr. 102, 778–807.
Reenberg, A., 2001. Agricultural land use pattern dynamics in the Sudan-Sahel—towards
an event-driven framework. Land Use Policy 18, 309–319.
Richardson, B.C., 1992. The Caribbean in the Wider World, 1492–1992: A Regional
Geography. Cambridge University Press, Cambridge.
Rigg, J., Salamanca, A., Thompson, E.C., 2016. The puzzle of East and Southeast Asia’s
persistent smallholder. J. Rural Stud. 43, 118–133.
Robbins, P., Bishop, K.M., 2008. There and back again: epiphany, disillusionment, and
rediscovery in political ecology. Geoforum 39, 747–755.
Robbins, P., 2004. Political Ecology: A Critical Introduction. Blackwell, London.
Roberts, C., 1996. The Logic of Historical Explanation. The Pennsylvania State University
Press, University Park, Pennsylvania.
Robinson, B.E., Holland, M.B., Naughton-Treves, L., 2014. Does secure land tenure save
forests? A meta-analysis of the relationship between land tenure and tropical defor-
estation. Glob. Environ. Change 29, 281–293.
Rojas, E., Meganck, R.A., 1987. Land distribution and land development in the Eastern
Caribbean. Land Use Policy 157–167.
623
Land Use Policy 67 (2017) 608–624
Rojas, E., 1984. Agricultural land in the Eastern Caribbean: from resources for survival to
resources for development. Land Use Policy 1, 39–54.
Romalis, R., 1975. Economic change and peasant political consciousness in the
Commonwealth Caribbean. J. Commonw. Comp. Politics 13, 219–241.
Rubenstein, H., 1975. The utilization of arable land in an eastern Caribbean valley. Can.
J. Sociol. 1, 157–167.
Rudel, T.K., Coomes, O.T., Moran, E., Achard, F., Angelsen, A., Xu, J.C., Lambin, E., 2005.
Forest transitions: towards a global understanding of land use change. Glob. Environ.
Change 15, 23–31.
Rudel, T.K., 1993. Tropical Deforestation: Small Farmers and Land Clearing in the
Ecuadorian Amazon. Columbia University Press, New York.
Rudel, T.K., 1998. Is there a forest transition? Deforestation, reforestation, and devel-
opment. Rural Sociol. 63, 533–552.
Rudel, T.K., 2008. Meta-analysis of case studies: a method for studying regional and
global environmental change. Glob. Environ. Change 18, 18–25.
Sagoff, M., 2016. Are there general causal forces in ecology? Synthese 193, 3003–3024.
Scriven, M., 1966. Causes, connections and conditions in history. In: Dray, W.H. (Ed.),
Philosophical Analysis in History. Harper and Row, New York, pp. 238–264.
Scriven, M., 2008. A summative evaluation of RCT methodology and alternative approach
to causal research. J. Multidisc. Eval. 5 (9), 11–24.
Shaver, I., et al., 2015. Coupled social and ecological outcomes of agricultural in-
tensification in Costa Rica and the future of biodiversity conservation in tropical
agricultural regions. Global Environ. Change 32, 74–86 12 co-authors.
Silva, R.F.B., Batistella, M., Moran, E.F., 2016. Drivers of land change: human-environ-
ment interactions and the Atlantic Forest transition in the Paraiba Valley, Brazil. Land
Use Policy 58, 133–144.
Banana Wars: Power, Production and History in the Americas. In: Striffler, S., Moberg, M.
(Eds.), Duke University Press, Durham.
Strunz, S., 2012. Is conceptual vagueness an asset? Arguments from philosophy of science
to the concept of resilience. Ecol. Econ. 76, 112–118.
Taylor, M.J., Aguilar-Stoen, M., Castellanos, E., Moran-Taylor, M.J., Gerkin, K., 2016.
International migration, land use change and the environment in Ixcan, Guatemala.
Land Use Policy 54, 290–300.
Thirgood, J.V., 1989. Man’s impact on the forests of Europe. J. World For. Resour. Manag.
4, 127–167.
Thomas, G., 2010. Doing case study: abduction not induction, phronesis not theory. Qual.
Inq. 16, 575–582.
Timms, B., Hayes, J., McCracken, M., 2013. From deforestation to reforestation: applying
the forest transition to the Cockpit Country of Jamaica. Area 45, 77–87.
Turner, B.L., Robbins II, P., 2008. Land-change science and political ecology: similarities,
differences, and implications for sustainability science. Ann. Rev. Environ. Resour. 33
(6), 1–6 22.
Turner, B.L.I.I., Kasperson, R., Matson, P., McCarthy, J., Corell, R., Chistensen, L., Eckley,
N., Kasperson, J., Luers, A., Martello, M., Polsky, C., Pulsipher, A., Schiller, A., 2003a.
A framework for vulnerability analysis in sustainability science. Proc. Natl. Acad. Sci.
100 (14), 8074–8079.
Turner, B.L.I.I., Matson, P., McCarthy, J., Corell, R., Chistensen, L., Eckley, N., Hovelsrud-
Broda, G., Kasperson, J., Kasperson, R., Luers, A., Martello, M., Mathiesen, S., Naylor,
R., Polsky, C., Pulsipher, A., Schiller, A., Selin, H., Tyler, N., 2003b. Illustrating the
coupled human-environment system for vulnerability analysis: three case studies.
Proc. Natl. Acad. Sci. 100 (14), 8080–8085.
Turner, B.L., Lambin II, E.F., Reenberg, A., 2007. The emergence of land change science
for global environmental change and sustainability. Proc. Natl. Acad. Sci. 104,
20666–20671.
Turner, B.L.I.I., Geoghegan, J., Lawrence, D., Radel, C., Schmook, B., Vance, C., Manson,
S., Keys, E., Foster, D., Klepeis, P., Vester, H., Rogan, J., Chowdhury, R.R., Schneider,
L., Dickson, R., Ogenva-Himmelberger, Y., 2016. Land system science and the social-
environmental system: the case of Southern Yucatan Peninsular Region (SYPR) pro-
ject. Curr. Opin. Environ. Sustain. 19, 18–29.
Vayda, A.P., Walters, B.B., 1999. Against political ecology. Hum. Ecol. 27, 167–179.
Vayda, A.P., Walters, B.B., 2011. Introduction: pragmatic methods and causal-history
explanations. In: Vayda, A.P., Walters, B.B. (Eds.), Causal Explanation for Social
Scientists: A Reader. AltaMira Press, Lanham, MD, pp. 1–21.
Vayda, A.P., 2006. Causal explanation of Indonesian forest fires: concepts, applications,
and research priorities. Hum. Ecol. 34, 615–635.
Vayda, A.P., 2009. Explaining Human Actions and Environmental Changes. AltaMira
Press, Lantham, MD.
Vayda, A.P., 2013. Causal explanation for environmental anthropologists. In: Kopnina,
H., Ouimet, E. (Eds.), Environmental Anthropology: Future Trends. Routledge,
London.
Veburg, P.H., Kok, K., Pontius, R.G., Veldkamp, A., 2006. Modeling land-use and land-
cover change. In: Lambin, E.F., Geist, H. (Eds.), Land-Use and Land-Cover Change:
Local Processes and Global Impacts. Springer-Verlag, Berlin, pp. 117–135.
WIRC, 1897. Norman Commission on the Sugar Industry of Saint Lucia Report of the West
Indian Royal Commission. HMSO, London.
Wahlberg, T.H., 2013. Elder-Vass on the causal power of social structures. Phil. Soc. Sci.
20 (10), 1–18.
Walker, B., Gunderson, L., Kinzig, A., Folke, C., Carpenter, S., Schultz, L., 2006. A handful
of heuristics and some propositions for understanding resilience in social-ecological
systems. Ecol. Soc. 11 (1), 13p.
Walker, P.A., 2005. Political ecology: where is the ecology? Prog. Hum. Geogr. 29, 73–82.
Walters, B.B., Hansen, L., 2013. Farmed landscapes, trees and forest conservation in Saint
Lucia. West Indies Environ. Conserv. 40, 211–221.
Walters, B.B., Vayda, A.P., 2009. Event ecology, causal historical analysis and human-
environment research. Ann. Assoc. Am. Geogr. 99, 534–553.
Walters, B.B., Cadelina, A., Cardano, A., Visitacion, E., 1999. Community history and
B.B. Walters
rural development: why some farmers participate more readily than others. Agric.
Syst. 59, 93–214.
Walters, B.B., 2004. Local management of mangrove forests in the Philippines: successful
conservation or efficient resource exploitation? Hum. Ecol. 32, 177–195.
Walters, B.B., 2012a. An event-based methodology for climate change and human-en-
vironment research. Danish J. Geogr.-Geografisk Tidsskrift 112 (2), 135–143.
Walters, B.B., 2012b. Do property rights matter for conservation? Family land, forests and
trees in Saint Lucia. West Indies Hum. Ecol. 40, 863–878.
Walters, B.B., 2016a. Migration, land use and forest change in St. Lucia. West Indies Land
Use Policy 51, 290–300.
Walters, B.B., 2016b. Saint Lucia’s tourism landscapes: economic development and en-
vironmental change in the West Indies. Caribb. Geogr. 21, 5–23.
Watts, M., Peet, R., 2004. Liberating political ecology. In: Peet, R., Watts, M. (Eds.),
Liberation Ecologies, 2nd ed. Routledge, London, pp. 3–41.
Watts, M., 2003. Political ecology. In: Sheppard, E., Barnes, T.J. (Eds.), A Companion to
Economic Geography. Blackwell, London, pp. 257–274.
Watts, M., 2011. On confluences and divergences. Dialog Hum. Geogr. 1, 84–89.
Welch, B., 1993. Challenging economic irrelevance: the role of banana growers’ asso-
ciations in Saint Lucia and Martinique. Caribb. Geogr. 4, 102–115.
Welch, B., 1994. Banana dependency: albatross or liferaft for the Windwards. Soc. Econ.
Stud. 43, 123–149.
Welch, B., 1996. The evolution of export banana cultivation in Dominica and Saint Lucia:
implications for sustainability. Caribb. Geogr. 7, 62–74.
Widgren, M., 2012a. Landscape research in a world of domesticated landscapes: the role
of values, theory and concepts. Quat. Intern. 251, 117–124.
Widgren, M., 2012b. Resilience thinking vs. political ecology: understanding the
624
Land Use Policy 67 (2017) 608–624
dynamics of small-scale, labour-intensive farming landscapes. Ch. 5. In: Plieninger,
T., Bieling, C. (Eds.), Resilience and the Cultural Landscape: Understanding and
Managing Change in Human-Shaped Environments. Cambridge University Press,
Cambridge.
Wiley, J., 2008. The Banana: Empires, Trade Wars & Globalization. University of
Nebraska Press, London 267 pp.
Williams, M., 2003. Deforesting the Earth: From Prehistory to Global Crisis. University of
Chicago Press, Chicago.
Winkel, T., Bommel, P., Chevarria-Lazo, M., Cortes, G., Del Castillo, C., Gasselin, P.,
Leger, F., Nona-Laura, J., Rambal, S., Tichit, M., Tourrand, J., Vacher, J., Vassas-
Toral, A., Viera-Pak, M., Joffre, R., 2016. Panarchy of an indigenous agroecosystem in
the globalized market: the quinoa production in the Bolivian Altiplano. Glob.
Environ. Change 39, 195–204.
World Bank, 1979. Current Economic Position and Prospects for Saint Lucia. Report No.
2440-CRB. Latin America and the Caribbean Regional Office.
Ylikoski, P., Kuorikoski, J., 2010. Dissecting explanatory power. Philos. Stud. 148,
201–219.
Young, O.R., Berkhout, F., Gallopin, G.C., Janssen, M.A., Ostrom, E., van der Leeuw, S.,
2006. The globalization of socio-ecological systems: an agenda for scientific research.
Glob. Environ. Change 16, 304–316.
Political Ecology: An Integrated Approach to Geography and Environment-Development
Studies. In: Zimmerer, K., Bassett, T. (Eds.), Guildford Press, New York.
van Andel, T., van der Hoorn, B., Stech, M., Arostegui, S.B., Miller, J., 2016. A quanti-
tative assessment of the vegetation types on the island of St Eustatius, Dutch
Caribbean. Glob. Ecol. Conserv. 7, 59–69.