S P E C I A L F E A T U R E:

SPECIAL FEATURE: PERSPECTIVE

PERSPECTIVE

Archaeology, environmental justice, and climate

change on islands of the Caribbean and
southwestern Indian Ocean
Kristina Douglassa,b,1 and Jago Cooperc

Edited by Daniel H. Sandweiss, University of Maine, Orono, ME, and accepted by Editorial Board Member Dolores R. Piperno
December 11, 2019 (received for review August 15, 2019)

Climate change impacts island communities all over the world. Sea-level rise, an increase in the frequency
and intensity of severe weather events, and changes in distribution and health of marine organisms are
among the most significant processes affecting island communities worldwide. On islands of the Caribbean
and southwestern Indian Ocean (SWIO), however, today’s climate change impacts are magnified by historical
environmental injustice and colonial legacies, which have heightened the vulnerability of human and other
biotic communities. For some islands, archaeological and paleoecological research offers an important re-
cord of precolonial climate change and its interplay with human lives and landscapes. The archaeological
record suggests strategies and mechanisms that can inform discussions of resilience in the face of climate
change. We detail climate-related challenges facing island Caribbean and SWIO communities using archae-
ological and paleoecological evidence for past climate change and human response and argue that these
cannot be successfully addressed without an understanding of the processes that have, over time, disrupted
livelihoods, reshaped land- and seascapes, threatened intergenerational ecological knowledge transfer, and
led to increased inequality and climate vulnerability.
| |
archaeology paleoclimate islands

Island communities are on the front line of global most vulnerable macroregion. Macroregional synthe-
change. Positioned at the interface of continental ses, however, tend to incorporate island groups into
geopolitical exchange, islanders have often been the analyses of nearby continental landmasses, making
first to experience the rapid socioenvironmental trans- it difficult to assess and compare the vulnerability of
formations of the planet. This is increasingly apparent island regions with one another.
when voices such as the Association of Small Island Here, we expand the IPCC’s systems-based frame-
States articulate the existential threats to island commu- work for assessing vulnerability and exposure to climate-
nities posed by sea-level rise and extreme weather driven threats by integrating deeper time perspectives
events (1). Although all island communities in the world on changing human–environment–climate dynamics.
face hazards linked to climate change, their vulnerability We have selected two island theaters as case studies,
and exposure to these hazards vary. The Intergovern- the Caribbean and southwestern Indian Ocean (SWIO),
mental Panel on Climate Change (IPCC) estimates vul- each with distinct cultural, historical, and environmental
nerability and exposure by assessing a community’s narratives but which are united, being home to some of
stress across physical (e.g., floods, drought, erosion/ the world’s most vulnerable and exposed island commu-
sea level, etc.), biological (e.g., ecosystems, biological nities and to the majority of the island African diaspora.
productivity, biodiversity, etc.), and human (food pro- Currently, the impacts of climate change are accel-
duction, health, livelihoods, economy, etc.) systems erating. Long-term historical models and short-term
(2). Using this framework to assess vulnerability to climate predictive models show that the Caribbean and SWIO
change, the IPCC has recognized Africa as the world’s will experience increased intensity of cyclonic activity,

a
Department of Anthropology, The Pennsylvania State University, University Park, PA 16802; bInstitutes of Energy and the Environment, The
Pennsylvania State University, University Park, PA 16802; and cAmericas Section, Department of Africa, Oceania and the Americas, British Museum,
WC1B 3DG London, United Kingdom
Author contributions: K.D. produced the figures and K.D. and J.C. wrote the paper.
The authors declare no competing interest.
This article is a PNAS Direct Submission. D.H.S. is a guest editor invited by the Editorial Board.
Published under the PNAS license.
Downloaded by guest on October 3, 2020

1
To whom correspondence may be addressed. Email: [email protected].
First published April 13, 2020.

8254–8262 | PNAS | April 14, 2020 | vol. 117 | no. 15 www.pnas.org/cgi/doi/10.1073/pnas.1914211117

 relative sea-level rise (RSLR), and unpredictability of precipitation inequality; disenfranchisement from autonomous land, sea, and
patterns (3, 4). These climate dynamics are closely intertwined with resource use; and loss of indigenous and local knowledge are
socioecological hazards, including biodiversity loss, pollution, de- social, economic, and political threats that exacerbate climate
forestation, coastal erosion and sedimentation, overfishing, and impacts. These cannot be ignored in favor of efforts to create
freshwater salinization (5) (Fig. 1 B–E and G–J). As island commu- top-down technological solutions for improved sustainability.
nities face these mounting challenges, understanding past human Moreover, the predominant top-down model of climate mitiga-
experience of specific climate hazards is important for building tion policy and concurrent losses and devaluing of indigenous
more resilient communities, and archaeologists have contributed and local knowledge severely constrain Caribbean and SWIO
many insights in this regard (6). communities’ ability to innovate sustainable solutions to climate
It is not enough, however, to simply look at past human change.
societies for inspiration in developing technological solutions to In a landmark special report, the IPCC has recently recognized
today’s challenges. This has been the focus of many projects to (8) that indigenous land management plays a vital role in miti-
apply past knowledge to present problems with ancient water gating climate change impacts and protecting biodiversity (9).
management systems recreated or resilient indigenous architec- This paper underscores the critical importance of assessing the
tural designs rebuilt. We must also recognize more widely the physical, biological, social, and historical dimensions of climate
historical disruptions to livelihoods, landscapes, ecologies, and vulnerability and prioritizing mitigation planning in close collab-
intergenerational knowledge transfer that have made contempo- oration with local, indigenous, and descendant (LID) communi-
rary communities more vulnerable to climate change impacts and ties. Through our case studies, we highlight the global importance
less flexible in their response (7). of a panoptical understanding of the past and the preservation
Here, we trace these disruptions, linked colonial legacies, and and revitalization of indigenous and local knowledge for effec-
persistent inequalities to highlight how they exacerbate climate tive and just action to mitigate future impacts of climate
impacts in the Caribbean and SWIO today. We emphasize that change.

Fig. 1. Maps of the case study areas highlighting some climate-related challenges in these regions: (A) map of Caribbean islands; (B) sargassum
pollution (Caribbean); (C) storm surge flooding in Havana, Cuba; (D) damage from Hurricane Maria on Dominica; (E) sunset at a Caribbean beach
resort; (F) map of SWIO islands; (G) green sea turtle (Chelonia mydas) ingesting plastic waste; (H) severe erosion and sedimentation of the
Downloaded by guest on October 3, 2020

Betsiboka River, Madagascar; (I) dense urban development on Unguja Island, Zanzibar; (J) traditional Vezo fishers preparing nets, southwest
Madagascar; and (K) global position of each case study region.

Douglass and Cooper PNAS | April 14, 2020 | vol. 117 | no. 15 | 8255
 Case Study 1—Caribbean Islands with cyclonic storm surges (13) (Fig. 1C). Past human communities
Humans have lived on the islands of the Caribbean since ∼6 ka have sometimes abandoned these vulnerable coastal areas before
(10) (Fig. 2), providing an opportunity to study the interaction the flooding events. This is because coastal freshwater aquifers in
between cultural lifeways and island ecologies contextualized low-lying coastal settlements are often salinized by slowly rising
within Holocene climate variability and contemporary Global relative sea levels, which force people to abandon the marginal
Change (11). The landscapes and material records of Caribbean littoral zone years before a hurricane storm surge then permanently
islands reveal the 16th century genocide of millions of indigenous floods the area (14). Today, sea level continues to rise and to exert
peoples with a corresponding loss of thousands of years of tradi- similar pressures on coastal communities. However, instead of
tional ecological knowledge; the 17th and 18th century trans- seeing salinization of coastal freshwater sources as an early warning
location of European and African lifeways (including plants and of vulnerability that requires settlement relocation, contemporary
animals) transforming island ecologies; and the 19th century in- strategies are focused on geoengineering (15): construction of
dustrialization of landscape modification and biodiversity loss desalination plants, piped water systems, and flood defense bar-
before a 20th century revolution in political and economic models riers (Fig. 3 A–C). The scale of past flooding events with up to 8-m
for cultural development and sovereign determination (12) storm surges underlines the extreme vulnerability of communities
(Fig. 2). and the unreliability of these short-term engineered solutions,
The Caribbean Community Climate Change Center (CCCCC) which will fail to cope with the forecasted changes in the lifetime of
has long identified the greatest hazards of global climate change the current generation (16).
in the Caribbean as 1) sea-level rise, 2) hurricane events, and 3)
precipitation variability (3). Human communities have experienced Hurricane Events. Recent hurricanes have consistently destroyed
all of these impacts over time, necessitating a review of divergent livelihoods, ended major infrastructure investment projects, and
human experiences based on the interplay between ways of life, crippled island economies (Fig. 1D). Paleotempestological re-
ecological context, and environmental hazard. Archaeological cords demonstrate that human communities have always lived
and historical investigations demonstrate how human communi- with the threat of hurricanes in the Caribbean (17). Variability in
ties have lived through the impacts of climatic hazards over the frequency and intensity is linked to movements of the Intertropical
past 6 ka. Convergence Zone (ITCZ) and sea surface temperatures (SSTs) in
the North Atlantic, but hurricanes have always posed a significant
Sea-Level Rise. The melting of polar ice and related isostatic hazard to Caribbean settlements (Fig. 3D). Contemporary climate
forcing have driven extensive RSLRs in the Caribbean, which vary change is increasing the intensity of cyclonic events in the Caribbean
based on local geomorphology and tectonics. Some island pop- (18), and this is of great concern to Caribbean nations. Comparative
ulations have experienced >6 m of RSLR during the last 6 ka, with studies between 5 ka of pre-Columbian household architecture,
one-sixth of Cuba’s land area submerged during the human oc- settlement locations, and food systems and the last 500 y of co-
cupation of the island (10). Archaeological and paleoenvironmental lonially introduced lifeways highlight distinctly different vulnera-
case studies suggest that the human experience of RSLR is not in bilities and disaster management strategies (14). The adoption of
increments of millimeters per year but in sudden events with per- European settlement patterns for new towns in river valleys and
manent salt water flooding of low-lying coastal plains associated estuaries from the 16th century onward as well as European

Fig. 2. Timeline of significant periods of change in interlinked human–environment dynamics: (A) island colonization, (B) megafauna extinction,
(C) animal translocations, (D) plant translocations, (E) population numbers in the millions, (F) genocide, (G) transoceanic slave trade, (H) boom in
Downloaded by guest on October 3, 2020

transoceanic trade (Columbian Exchange in the Caribbean and Swahili Trade in the SWIO), (I) urbanization, (J) industrial-scale resource extraction,
and (K) international tourism.

8256 | www.pnas.org/cgi/doi/10.1073/pnas.1914211117 Douglass and Cooper

 the islands of the Caribbean. This is often linked to the history of
an island’s geopolitical context, ecological vulnerability, and
economic capacity to plan for and recover from climate change
impacts.

Case Study 2—SWIO Islands

The SWIO includes continental islands off the coast of East Africa,
such as the islands of the Zanzibar Archipelago, and oceanic is-
lands, such as Madagascar and the Mascarene islands (Mauritius,
Réunion, and Rodrigues) (Fig. 1). The region’s oceanic islands
formed when Gondwana split apart around 90 Ma ago. The
SWIO’s continental islands were part of mainland Africa during
Pleistocene glacial periods. During interglacials, most recently
after the Last Glacial Maximum, warming trends drove RSLR and
reshaped the biogeography of the region. With the exception of
Fig. 3. Sea-level rise, hurricane events, and precipitation variability
case study illustrations: (A) water desalination plant, St. Thomas, US Madagascar and Zanzibar, which have both yielded limited Early
Virgin Islands; (B) waves crashing against sea wall, Havana, Cuba; (C) to Mid-Holocene cultural deposits (23, 24), the islands of the
vintage postage stamp from the Dutch Antilles featuring a water SWIO do not have records of human settlement and activity
treatment facility; (D) hurricane damage to a concrete house, earlier than the Late Holocene (Fig. 2). This short human history is
Baracoa, Cuba; and (E) traditional wood house made primarily of
particularly surprising given the region’s physical proximity to
locally available materials, Las Terrenas, Samana, Dominican
Republic. eastern and southern Africa’s paleoanthropological record and
recurrent physical connection to the mainland. In the case of some
SWIO islands, like Mauritius, the human record is remarkably re-
household architecture and building materials have increased the cent, with initial human settlement ∼300 y ago (25). In addition to
relative vulnerability of post-Columbian communities to hurricane- the surprising timing of SWIO island colonization events, the
related flooding and wind shear. Current disaster management region experienced migration from around the Indian Ocean
strategies are predominantly based on robustness and resistance to rim, including the westernmost expansion of Austronesian
hurricane wind shear, storm surges, and flooding events. This speakers (26).
approach gives primacy to preventing damage through a strong Islands of the SWIO sit at the crossroads of one of the world’s
built environment in total contrast with preexisting indigenous most dynamic maritime trading spheres, connecting at various
frameworks for resilience that focused on speed of recovery. points in time the ports of China and Indonesia to those of East
Today, reinforced concrete is normally considered hurricane- Africa and the Persian Gulf. Long distance Indian Ocean trade
resilient architecture in the Caribbean (Fig. 3D). A thousand years networks emerged at least as early as 2 ka (27), but evidence for
ago, resilient architecture facilitated swift posthurricane recon- population growth and the appearance of large villages and
struction using locally available materials, privileging the ability to towns on SWIO islands seems connected to the emergence of
swiftly rebuild and recover over attempts to be robust and with- Swahili trade beginning ∼1 ka (Fig. 2H). The rise of Swahili trade
stand a hurricane’s destructive impact (19) (Fig. 3E). thus marks a turning point in the SWIO human–environment story.
Along with urbanism, this trade resulted in the translocation of
Precipitation Variability. Precipitation variation has significant plants, animals, and people on a scale not yet seen in the region
implications for human occupation in the Caribbean, particularly (28, 29). By ∼500 y B.P., European disruption of trade along the
as many communities live with limited freshwater supplies (15). East African coast and Madagascar marks another turning point,
Water is often a key tipping point in forcing island abandonment involving forced translocation of people across the Indian and
as archaeological examples demonstrate (20). The Caribbean has Atlantic Oceans. Subsequent European colonialism and the
seen considerable long-term variation in precipitation rates, which postcolonial 20th and 21st centuries have seen the advent of
is particularly relevant for islands with small water catchment massive extractive industries (fisheries, logging, and mining) and
areas. The 16th century translocation of monoculture agriculture efforts to “modernize” indigenous communities, all of which
later exacerbated by island-dependent cash crop economies has have exacted a heavy toll on local livelihoods and environments
reduced water retention, increased soil erosion, and exposed (Fig. 2J).
people to greater vulnerability to droughts (Fig. 2D). At the same Today, the island communities of the SWIO are experiencing
time, increasingly urbanized communities with high water con- significant and interlinked threats to environments and liveli-
sumption have depleted freshwater supplies and require engineered hoods. According to the Indian Ocean Observing System (IndOOS),
water management systems (Fig. 3A). Long-term indigenous strat- key threats across the region are 1) food insecurity due to deg-
egies for effective water management included rainwater capture, radation of coastal and marine environments and declining ag-
managed exploitation of coastal freshwater aquifers, and use of ricultural productivity, 2) water scarcity due to freshwater
lower water-dependent food crops. salinization and pollution, and 3) rising energy demands due to
Environmental vulnerability is also linked to environmental population growth and urbanization (5). Increasingly rapid cli-
justice for these historically contingent and economically diver- mate change exacerbates all of these threats. The Indian Ocean
gent communities. The impacts of climate change exacerbate is warming more rapidly than any other ocean in the world, with
existing risks, including earthquakes, tsunamis, landslides, and significant consequences for regional and global climate re-
volcanic activity, contextualized within island-specific socio- gimes, ecosystems, and people (4). Observations from the SWIO
Downloaded by guest on October 3, 2020

environmental vulnerabilities (21, 22). The opportunities for post- indicate a century-long SST warming trend that is poised to alter
disaster migration and costly reconstruction vary greatly between biological productivity dramatically in the region (30). Although

Douglass and Cooper PNAS | April 14, 2020 | vol. 117 | no. 15 | 8257
 there is growing evidence of significant climate change in the
region, two overarching issues constrain our ability to miti-
gate the impacts of change. First, despite recent efforts to im-
prove Indian Ocean observation systems (e.g., IndOOS), a lack of
basin-wide observational data limits our ability to assess current
climate variability and predict future change across this ocean (4).
Second, although SWIO island communities have experienced,
influenced, and responded to many extreme shifts in climate and
environment over the course of the Holocene, current economic
inequality and political marginalization, often rooted in the region’s
relatively recent colonial history, compound the impacts of climate
change. In this context, archaeological and historical insights into
past lived experiences in the region can inform current climate and
environmental policy debates.
Fig. 4. Food insecurity, water scarcity, and rising energy demands
Food Insecurity. With the exception of the southern tip of case study illustrations: (A) a southern Malagasy herder provisions his
Madagascar, the islands of the SWIO fall within the tropics, and cattle with prickly pear (Opuntia spp.) fodder; (B) southwest
several important forcing mechanisms shape today’s SWIO cli- Madagascar’s Vezo seminomadic fishers sail traditional laka
mate. The Southwest Indian Ocean Monsoon system brings heavy (outrigger canoes) at dusk; (C) freshwater infinity pool at a luxury
tourist resort on the Zanzibar Archipelago; (D) vintage postage
rains from December to March, with lighter rainfall from October stamp from Réunion featuring a plantation worker harvesting sugar
to December (31). Influenced by the north–south shifting of the cane; and (E) an aerial view of sugar cane plantations on Mauritius
ITCZ, rainfall patterns are highly variable (32). The SWIO is par- (Saccharum officinarum).
ticularly susceptible to variability in SSTs, which also influence
regional rainfall (33–37). SSTs, in turn, are sensitive to and influ-
decimation of cattle herds, and the economic crippling of south-
ence the El Ni~ no–Southern Oscillation (ENSO) and the Pacific
ern communities, the effects of which continue today to make
Decadal Oscillation (38). The ENSO’s influence likely extends to
southern communities less resilient to drought.
the frequency and intensity of extreme weather events, such as
cyclones, in the region (37, 39).
Water Scarcity. Linked to the threat of food insecurity is water
As the Indian Ocean continues to warm, variability in pre-
scarcity, a challenge that SWIO communities have faced for mil-
cipitation is expected to increase, suggesting that severe
lennia. The SWIO Holocene paleoclimate record has been
droughts and floods may become more frequent (40). This poses a
reconstructed from cave stalagmites, lake sediment and coral
great threat to food security as the majority of SWIO communities
are reliant on rain-fed agriculture and pastoralism (41). The cores, and archaeological evidence (48). These records show that
interlinked threats of precipitation variability, colonial legacies, the Pleistocene–Holocene transition was characterized by signif-
and food insecurity are clear in the case of southern Madagascar, icant hydrological shifts toward more mesic conditions through-
where successive years of crop failures due to insufficient rainfall out the SWIO. Wetter conditions during the Early Holocene likely
precipitated a catastrophic famine in 2016 that directly affected drove social and ecological changes in the broader Indian Ocean
nearly half a million people (42). Although periodic famine is an region (49, 50). Despite an overall shift toward wetter conditions,
accepted risk in the lives of southern Malagasy communities, successive dry and wet phases characterize the SWIO Holocene
centuries of coevolution with the landscapes of the south have led record (51–55).
to important agropastoral adaptations that mitigate disastrous At the time of the major 4-ka desiccation event, only Mada-
incidences of climate-driven famine. The cultivation and wide- gascar and the continental islands of East Africa were inhabited by
spread dispersal of raketa, the nonnative prickly pear cactus small-scale foraging and fishing communities. Archaeological
(Opuntia spp.), as fodder for cattle; as a source of water for cattle, evidence from coastal southwest Madagascar spanning the last
people, and other plants; and as a defensive barrier against in- 2 ka suggests that early communities were highly mobile, leaving
truders, were two of the most important adaptations of southern relatively ephemeral traces of their presence at individual sites
Malagasy pastoralists to climate stochasticity, turning the dry (56). Likewise, oral histories of the southwest indicate that the
south into a subterranean anthropogenic wetland (43) prior to razana (ancestors) responded to the adaptive challenge of limited
prickly pear eradication (see below). freshwater through a high degree of settlement mobility. Today,
Since the prickly pear’s introduction to Madagascar from the the combined effect of multiple factors exacerbates the climate-
Americas by French colonialists in 1769, southern “cactus” pas- driven threat of SWIO water scarcity. Postcolonial resource man-
toralists have tightly integrated it into their ecology (44) (Fig. 4A). agement approaches, an emphasis on international tourism for
In the 1930s, however, the French colonial administration waged economic development, and the vulnerability of many islands’
biological warfare on southern Malagasy by dispersing parasitic shallow freshwater aquifers pose an underappreciated threat to
cochineal larvae (Dactylopius tomentosus) to eradicate the prickly indigenous livelihoods. This is particularly problematic in the dry
pear (45). This attack was framed by the French as a measure to coastal zones of southwest Madagascar and on the SWIO’s
civilize the south, disrupt a way of life deemed unproductive and smaller islands. In coastal southwest Madagascar today, settle-
prone to failure, and force southerners to adopt “modern” agro- ment mobility is constrained by conservation and development
pastoral practices through the cultivation of cash crops, irrigation, initiatives that seek to regulate resource use within recently
and improvement of grasslands (46, 47). As a result of the eradi- established zones managed by spatially fixed associations of local
Downloaded by guest on October 3, 2020

cation of raketa gasy (Opuntia monocantha), the preferred prickly communities (57). Although positive outcomes have emerged
pear, imminent drought was followed by widespread famine, from these efforts, such as increased catch of commercially valuable

8258 | www.pnas.org/cgi/doi/10.1073/pnas.1914211117 Douglass and Cooper

 taxa (58), tensions have also arisen as zone-based management inequality; energy insecurity; and more rapid deforestation,
approaches conflict with established migratory practices (59) (Fig. erosion, and sedimentation of waterways and fragile coastal
4B). The social and ecological consequences of rapid sedentari- habitats (5, 68).
zation of long nomadic or seminomadic communities have not
been evaluated carefully. Discussion
As conservationists in the SWIO emphasize the need to shift The archaeological and paleoclimatic records of the Caribbean
communities toward alternative livelihoods less reliant on mobility and SWIO offer glimpses of past human responses to climatic and
and the use of wild aquatic and terrestrial resources, SWIO gov- environmental change and the human actions that continue to
ernments and development agencies have looked to international interact with climate and environment today. Looking at the past
tourism as a sustainable, alternative source of income. On Zanzi- highlights how periods of rapid environmental change were often
bar, however, tourism has brought with it unsustainable use of the precipitated by social, political, and economic transformations,
archipelago’s thin lenses of freshwater as a growing number of such as the emergence of transoceanic trading networks, the
foreign travelers consume water with little awareness or concern translocation of nonnative plants and animals, and urbanization.
for its scarcity or the extreme vulnerability of the water table to Our two case studies also reveal commonalities in how patterns of
rainfall variability (60) (Fig. 4C). Similar pressure from international cultural and ecological transformation are entangled with issues of
tourism is likely impacting other SWIO island communities. On environmental justice that create contingent vulnerabilities to
Zanzibar, the shallow nature of freshwater aquifers is compound- climate change impacts today. Even simple ambitions to create
ing the ecological and public health hazards of human waste and goals for sustainability require an appreciation of the great
garbage, which are mounting as the tourism industry expands and transformations that all of the islands discussed above have ex-
disproportionately impact impoverished communities (5). perienced over time. Our case studies underscore 1) the variety of
different environmental baselines that could be used to chart
Rising Energy Demands. Biological productivity in the SWIO is ecological restoration and sustainable development targets; 2)
shaped on land by the availability of rainfall and in the ocean by the dialectical relationship between local and global desires and
SSTs that influence the degree of upwelling of cool, nutrient- socioecological entanglements; 3) the irreversible thresholds that
dense waters. Recent observations suggest that rapid warming of many island ecologies have already undergone; and 4) the value
the SWIO will have negative impacts on biological productivity (4). and vulnerability of intergenerational ecological knowledge
As population growth and globalization cause energy demands to transfer.
rise, pressure on resources such as fuel will intensify.
Meeting rising energy demands in the context of climate- Baselines and Restoration. The introduction of cattle to Mada-
driven decreases in biological productivity is a multifaceted gascar and the subsequent threats to pastoral livelihoods de-
challenge. Wood charcoal, for example, remains an essential fuel scribed above provide an important example of the challenge of
for cooking in the SWIO, with Kenya, Tanzania, and Madagascar identifying baselines for ecological restoration and future sus-
ranking among the heaviest users of charcoal in the world (61). tainability planning in the face of climate change. Zebu cattle (Bos
Efforts to transition communities in the region to stoves powered indicus) were introduced to the island ∼1 ka (69) and likely con-
by cleaner and more sustainable sources are motivated by con- tributed to declines in endemic megafauna populations through
cerns about the negative health consequences of cooking with ecological competition. Despite these consequences, the ar-
charcoal, especially indoors; extensive deforestation; and carbon chaeological and historical records suggest that ancient Malagasy
dioxide emissions. Energy policies aimed at encouraging the use communities maintained large herds of zebu that may have
of more sustainable fuel sources in the SWIO, however, have often buffered against crop failure, drought, and political instability (70).
been implemented abruptly without considering legacies of co- Colonial efforts to disrupt pastoral lifeways were accompanied by
lonial era land grabs and resource disenfranchisement (5) and the the introduction of cash crops, like corn (Zea mays), that have high
cultural importance of timber products as sources of fuel. As a water requirements and are maladapted to local climate (66).
result, rigid policies banning or restricting the use of charcoal Despite these coercive measures and the violent illicit trade in
have often resulted in increased deforestation by driving up fuel zebu today, cattle remain a central pillar of southern Malagasy
prices (61). cultures (71).
Past expansion of transoceanic trade and colonial era land
policies similarly drove deforestation, exacerbating the energy Local vs. Global Practices. As transoceanic trade networks con-
crisis today. SWIO islands were sparsely populated by forager– nected the islands of the Caribbean and SWIO to broader regions,
fisher communities until a boom in trade ∼1 ka sparked rapid nonlocal technologies, practices, and aesthetics were imposed on
urbanization and population growth along the Swahili corridor indigenous communities. The same dialectic between local and
(62) (Fig. 2H). The emergence of Swahili port cities like Mahilaka global exists today. For example, archaeological evidence of a
(63) coincided with indications of deforestation (64), suggesting pre-Columbian mode of domestic architecture across the Carib-
that urban expansion required substantial timber for fuel and bean island region suggests that semipermanent houses were
construction. A spike in deforestation is next seen beginning in well adapted to climate hazards (72). The labor-intensive com-
the 17th century, when European powers sought to maximize the ponents of houses were designed to withstand hurricane winds
profitability of SWIO colonies by extracting resources and farming and other hazards, while other components were easy to rebuild.
cash crops on an industrial scale (65, 66). Colonial administrations This approach to domestic architecture contrasts with modern
dispossessed indigenous communities of the most productive houses, which are built from rigid materials, are more dangerous
land and supplanted indigenous management with cash cropping during natural disasters, and are costlier to rebuild. Perceptions of
and commercial logging (65, 67) (Fig. 4 D and E). Indigenous modernity and the simultaneous devaluing of indigenous forms
Downloaded by guest on October 3, 2020

farmers and herders were pushed onto marginal lands where soils and practices are trends throughout the Caribbean and SWIO as
were quickly depleted, a process that led to increasing economic well other island communities around the world. The Caribbean

Douglass and Cooper PNAS | April 14, 2020 | vol. 117 | no. 15 | 8259
 and SWIO, however, have been less well represented in historical concretely identified and integrated into assessments of climate
ecological studies than other island groups (5, 73, 74) and require vulnerability alongside the physical, biological, and human systems
further research on the adaptive role of indigenous forms and stresses currently outlined by the IPCC framework. Integrating
practices alongside efforts to disseminate research findings to a deeper time data on human–environment–climate dynamics and
broad audience, including policy makers. When indigenous forms past injustices and disruptions into climate vulnerability assess-
and practices are celebrated in these regions today, it is typically ments will more accurately reveal regions with the greatest
in a fetichized context of ecotourism, where foreign visitors are vulnerability. Meanwhile the revitalization and preservation of
offered opportunities to experience “traditional” lifeways, rather
intergenerational LID knowledge and practices can highlight
than in a context of developing and revitalizing successful solu-
potential ways to build resilience. The Caribbean and SWIO are
tions to climate hazards (72).
well placed to demonstrate the role of the past in creating ef-
Tipping Points. The human colonization of islands of the Carib- fective plans for cultural adaptation to the impacts of climate
bean and SWIO significantly impacted endemic island fauna and change as well as revealing a global context of environmental
flora. Subsequent human activity and the translocation of people, injustice for those who bear the greatest burden of global
animals, and plants throughout these regions precipitated further change (84, 85).
irreversible ecological shifts, such as extinction. Several faunal and The impacts of climate change in the Caribbean and SWIO are
floral extinction events are documented in both regions at dif- profound, tangible, and intensifying. The human experience of
ferent times. The Mid-Holocene extinction of megafauna on Ca- more extreme climate events combined with the increased vul-
ribbean islands (75) and Late Holocene animal and plant nerability of communities is the source of great public concern. In
extinctions in the Caribbean and SWIO have irrevocably changed recent years, entire islands in the Caribbean have been depopu-
local ecologies (76–79). Despite ongoing debate as to the precise lated (86), many people have died, and island economies have
cause and timing of extinction on some islands (24), in cases like been devastated (87). In the SWIO, communities face a future of
Mauritius, direct anthropogenic pressures led to swift population more frequent droughts, declines in biological productivity, and
crashes of plant and animal taxa such that the contemporary growing economic inequality (4). These threats are rooted in
and ancient landscapes bear little resemblance to one another complex, oftentimes long-term human–environment interactions
(80–82).
that must be carefully parsed if we are to develop effective solu-
tions for sustainable, just, and resilient futures.
Intergenerational LID Knowledge. Loss of LID knowledge cor-
relates closely with ecological change (83). The histories of human– Methods
environment dynamics in the Caribbean and SWIO are punctu- We aim to demonstrate the importance of deeper time perspectives in assessing
ated by periods of significant ecological change that also resulted vulnerability and innovating just and effective solutions for mitigating the
impacts of climate change. Using the United Nations Development Program’s
in loss of LID ecological knowledge. The most extreme case of
Human Development Index (HDI) and Multidimensional Poverty Index (MPI) as
knowledge and culture loss in these regions was caused by the proxies for climate vulnerability across physical, biological, and human systems
genocide of indigenous Caribbean peoples by European colo- (88), we selected the Caribbean and SWIO island groups as our case studies
nists beginning in AD 1492 (Fig. 2F). Although the consequences (Fig. 1) as these regions are home to several island nations with the lowest HDI
of this genocide extend far beyond knowledge loss, knowledge scores and highest MPI scores. Moreover, our work enhances recent studies of
long-term human–environmental interactions in other groups, including the
accumulated and transmitted over millennia by communities that
Pacific (77, 78).
had long coevolutionary relationships with island environments In each case study, we focus on three climate-driven threats identified by the
was destroyed, severely constraining communities’ capacity to CCCCC and the IndOOS, the leading climate observation and action organi-
draw on generations of past experience to innovate future adap- zations in each region. Each case study describes diachronic human–environment
interactions in the region and historical and archaeological evidence of past and
tive strategies. The revitalization and preservation of indigenous
present human responses to the three highlighted climate-driven threats. This
languages, knowledge, and oral histories represent a critical di- approach emphasizes the historical contingency of today’s most urgent climate-
mension of planning for a just and sustainable future. driven threats and the role of environmental injustice in influencing island com-
munities’ vulnerability and resilience to climate change.
Conclusion
Data Availability. All data synthesized in this paper are published and acces-
We have highlighted how the integration of historical, archaeo-
sible through consultation of cited works.
logical, paleoenvironmental, and paleoclimatic research adds a
critical missing dimension to the current IPCC framework for
Acknowledgments
assessing communities’ vulnerability to climate impacts. Put sim- We thank Drs. Daniel H. Sandweiss and Torben Rick for the invitation to
ply, past environmental injustices have created underlying vul- contribute a paper to the Special Feature on Archaeology, Climate, and Global
nerabilities to present and future climate change. These must be Change and for their constructive comments on the manuscript.

1 AOSIS, Rising Tides, Rising Capacity: Supporting a Sustainable Future for Small Island Developing States (Association of Small Island States, United Nations
Development Programme, New York, NY, 2017).
2 IPCC, Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change, R. K. Pachauri, L. Meyer, Eds. (Intergovernmental Panel on Climate Change, Geneva, Switzerland, 2014), p. 151.
3 Caribbean Community Climate Change Centre, Climate Change and the Caribbean: A Regional Framework for Achieving Development Resilient to Climate
Change (2009-2015) (Caribbean Community Climate Change Centre, Belmopan, Belize, 2009).
4 J. C. Hermes et al., A sustained ocean observing system in the Indian Ocean for climate related scientific knowledge and societal needs. Front. Mar. Sci. 6, 1–21
(2019).
5 K. Douglass et al., Historical perspectives on contemporary human-environment dynamics in southeast Africa. Conserv. Biol. 33, 260–274 (2019).
Downloaded by guest on October 3, 2020

6 J. H. Altschul et al., Opinion: Fostering synthesis in archaeology to advance science and benefit society. Proc. Natl. Acad. Sci. U.S.A. 114, 10999–11002 (2017).

8260 | www.pnas.org/cgi/doi/10.1073/pnas.1914211117 Douglass and Cooper

 7 P. Nunn, R. Kumar, Understanding climate-human interactions in Small Island Developing States (SIDS). Int. J. Clim. Chang. Strateg. Manag. 10, 245–271 (2018).
8 IPCC, Climate change and land: An IPCC Special Report on climate change, desertification, land degradation, sustainable land management, food security, and
greenhouse gas fluxes in terrestrial ecosystems. https://www.ipcc.ch/srccl/. Accessed 23 September 2019.
9 R. Schuster, R. R. Germain, J. R. Bennett, N. J. Reo, P. Arcese, Vertebrate biodiversity on indigenous-managed lands in Australia, Brazil, and Canada equals that in
protected areas. Environ. Sci. Policy 101, 1–6 (2019).
10 R. Valcárcel Rojas, J. Ulloa Hung, O. Feria Garcia, “Levisa 1. Studying the earliest indigenous peoples of Cuba in multicomponent archaeological sites” in Insular
Caribbean: Dearchaising the Archaic, C. L. Hofman, A. T. Antczak, Eds. (Sidestone Press, Leiden, the Netherlands, 2019), pp. 177–190.
11 J. Cooper, M. C. Peros, The archaeology of climate change in the caribbean. J. Archaeol. Sci. 37, 1226–1232 (2010).
12 B. W. Higman, A Concise History of the Caribbean (Cambridge University Press, Cambridge, UK, 2011).
13 G. A. Milne, M. C. Peros, Data-model comparison of Holocene sea-level change in the circum-Caribbean region. Global Planet. Change 107, 119–131 (2013).
14 J. Cooper, “Fail to prepare then prepare to fail: Re-thinking threat, vulnerability and mitigation in the pre-Columbian Caribbean” in Surviving Sudden
Environmental Change: Answers from Archaeology, J. Cooper, P. Sheets, Eds. (University Press of Colorado, Boulder, CO, 2012), pp. 91–114.
15 A. C. Cashman, Water policy development and governance in the Caribbean: An overview of regional progress. Water Policy 14, 14–30 (2011).
16 M. A. Mycoo, Beyond 1.5 °C: Vulnerabilities and adaptation strategies for Caribbean Small Island Developing States. Reg. Environ. Change 18, 2341–2353 (2018).
17 B. Malaize et al., Hurricanes and climate in the caribbean during the past 3700 years BP. Holocene 21, 911–924 (2011).
18 K. T. Bhatia et al., Recent increases in tropical cyclone intensification rates. Nat. Commun. 10, 635 (2019).
19 J. Cooper, “Building resilience in Island communities: A paleotempestological perspective” in AGU Chapman Conference on Climates, Past Landscapes, and
Civilizations, L. Giosan, P. Clift, D. Fuller, R. Flad, S. VanLaningham, J. Aimers, Eds. (Geophysical Monograph Series, Geopress, Washington, DC, 2012), vol. 198,
pp. 43–49.
20 H. Dawson, Mediterranean Voyages. The Archaeology of Island Colonisation and Abandonment (Left Coast Press Inc., Walnut Creek, CA, 2014).
21 S. Fitzpatrick, On the shoals of giants: Natural catastrophes and the overall destruction of the Caribbean’s archaeological record. J. Coast. Conserv. 16, 173–186
(2012).
22 S. R. Scheffers, J. Haviser, T. Browne, A. Scheffers, Tsunamis, hurricanes, the demise of coral reefs and shifts in prehistoric human populations in the Caribbean.
Quat. Int. 194, 69–87 (2009).
23 M. E. Prendergast et al., Continental island formation and the archaeology of defaunation on zanzibar, Eastern Africa. PLoS One 11, e0149565 (2016).
24 K. Douglass et al., A critical review of radiocarbon dates clarifies the human settlement of Madagascar. Quat. Sci. Rev. 221, 1–11 (2019).
25 K. Seetah, The archaeology of Mauritius. Antiquity 89, 922–939 (2015).
26 A. Crowther et al., Ancient crops provide first archaeological signature of the westward Austronesian expansion. Proc. Natl. Acad. Sci. U.S.A. 113, 6635–6640
(2016).
27 L. Casson, The Periplus Maris Erythraei: Text with Introduction, Translation, and Commentary (Princeton University Press, Princeton, NJ, 1989).
28 D. Q. Fuller, N. Boivin, T. Hoogervorst, R. Allaby, Across the Indian ocean: The prehistoric movement of plants and animals. Antiquity 85, 544–558 (2011).
29 P. Beaujard, Histoire et Voyages des Plantes Cultivées à Madagascar (Karthala, Paris, France, 2017).
30 M. K. Roxy, K. Ritika, P. Terray, S. Masson, The curious case of Indian ocean warming. J. Clim. 27, 8501–8509 (2014).
31 M. R. Jury, The Climate of Madagascar, S. M. Goodman, J. P. Benstead, Eds. (The University of Chicago Press, Chicago, IL, 2003), pp. 75–88.
32 R. E. Dewar, A. F. Richard, Evolution in the hypervariable environment of Madagascar. Proc. Natl. Acad. Sci. U.S.A. 104, 13723–13727 (2007).
33 C. A. Grove et al., Madagascar corals reveal pacific multidecadal modulation of rainfall since 1708. Clim. Past 9, 641–656 (2013).
34 C. A. Grove et al., Confounding effects of coral growth and high SST variability on skeletal Sr/Ca: Implications for coral paleothermometry. Geochem. Geophys.
Geosyst. 14, 1277–1293 (2013).
35 J. E. Tierney et al., Northern hemisphere controls on tropical southeast African climate during the past 60,000 years. Science 322, 252–255 (2008).
36 J. E. Tierney, S. C. Lewis, B. I. Cook, A. N. LeGrande, G. A. Schmidt, Model, proxy and isotopic perspectives on the East African humid period. Earth Planet. Sci.
Lett. 307, 103–112 (2011).
37 J. Zinke, M. Pfeiffer, O. Timm, W. C. Dullo, G. J. A. Brummer, Western Indian ocean marine and terrestrial records of climate variability: A review and new concepts
on land–ocean interactions since AD 1660. Int. J. Earth Sci. 98, 115–133 (2009).
38 T. Crueger, J. Zinke, M. Pfeiffer, Patterns of Pacific decadal variability recorded by Indian Ocean corals. Int. J. Earth Sci. 98, 41–52 (2008).
39 A. F. Mavume, L. Rydberg, M. Rouault, J. R. E. Lutjeharms, Climatology and landfall of tropical cyclones in the South-West Indian Ocean. Western Indian Ocean J.
Marine Sci. 8, 15–35 (2010).
40 R. d’Arrigo, R. Wilson, El Ni~no and Indian Ocean influences on Indonesian drought: 1342 implications for forecasting rainfall and crop productivity. Int. J. Climatol.
28, 611–618 (2008).
41 C. Funk et al., Warming of the Indian Ocean threatens eastern and southern African food security but could be mitigated by agricultural development. Proc. Natl.
Acad. Sci. U.S.A. 105, 11081–11086 (2008).
42 T. Healy, The Deep South: Constraints and Opportunities for the Population of Southern Madagascar towards a Sustainable Policy of Effective Responses to
Recurring Droughts/Emergencies, C. Gevers, M. Lundell, D. Ringold, Eds. (The World Bank, Washington, DC, 2017).
43 J. C. Kaufmann, Prickly pear cactus and pastoralism in Southwest Madagascar. Ethnology 43, 345–361 (2004).
44 J. C. Kaufmann, The non-modern constitution of famines in Madagascar’s spiny forests: “Water-food” plants, cattle and Mahafale landscape praxis. Environ. Sci. 5,
73–89 (2008).
45 P. Binggeli, “Cactaceae, Opuntia spp., prickly pear, raiketa, rakaita, raketa” in The Natural History of Madagascar, S. M. Goodman, J. P. Benstead, Eds. (University
of Chicago Press, Chicago, IL, 2003), pp. 335–339.
46 H. P. de la Bâthie, Les pestes végétales à Madagascar. Rev. Bot. Appl. Agric. Colon.8, 36–43 (1928).
47 H. P. de la Bâthie, Les famines du sud-ouest de Madagascar: Causes et remèdes. Rev. Bot. Appl. Agric. Trop. 14, 173–186 (1934).
48 K. Douglass, J. Zinke, Forging ahead by land and by sea: Archaeology and paleoclimate reconstruction in Madagascar. Afr. Archaeol. Rev. 32, 267–299 (2015).
49 V. Prasad et al., Mid–late Holocene monsoonal variations from mainland Gujarat, India: A multi-proxy study for evaluating climate culture relationship.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 397, 38–51 (2014).
50 G. W. Preston, A. G. Parker, Understanding the evolution of the Holocene pluvial phase and its impact on neolithic populations in South-East Arabia. Arab.
Archaeol. Epigr. 24, 87–94 (2013).
51 B. E. Crowley, K. E. Samonds, Stable carbon isotope values confirm a recent increase in grasslands in Northwestern Madagascar. Holocene 23, 1066–1073 (2013).
52 E. J. de Boer et al., Climate variability in the SW Indian Ocean from an 8000-yr long multi-proxy record in the Mauritian lowlands shows a middle to late Holocene
shift from negative IOD-state to ENSO-state. Quat. Sci. Rev. 86, 175–189 (2014).
53 D. Fleitmann et al., Holocene ITCZ and Indian Monsoon dynamics recorded in stalagmites from Oman and Yemen (Socotra). Quat. Sci. Rev. 26, 170–188 (2007).
54 F. Gasse, E. Van Campo, A 40,000-yr pollen and diatom record from lake tritrivakely, Madagascar, in the Southern Tropics. Quat. Res. 49, 299–311 (1998).
55 L. Wang et al., The African Humid Period, rapid climate change events, the timing of human colonization, and megafaunal extinctions in Madagascar during the
Holocene: Evidence from a 2m Anjohibe Cave stalagmite. Quat. Sci. Rev. 210, 136–153 (2019).
56 K. Douglass, “An archaeological investigation of settlement and resource exploitation patterns in the velondriake marine protected area, Southwest Madagascar,
ca. 900 BC to AD 1900,” PhD dissertation, Yale University, New Haven, CT (2016).
Downloaded by guest on October 3, 2020

57 A. Harris, “To live with the sea” development of the velondriake community-managed protected area network, Southwest Madagascar. Madag. Conserv. Dev. 2,
43–49 (2007).

Douglass and Cooper PNAS | April 14, 2020 | vol. 117 | no. 15 | 8261
 58 S. Benbow et al., Lessons learnt from experimental temporary octopus fishing closures in southwest Madagascar: Benefits of concurrent closures. Afr. J. Mar. Sci.
36, 31–37 (2014).
59 G. Cripps, C. J. Gardner, Human migration and marine protected areas: Insights from Vezo fishers in Madagascar. Geoforum 74, 49–62 (2016).
60 S. Gössling, The consequences of tourism for sustainable water use on a tropical island: Zanzibar, Tanzania. J. Environ. Manage. 61, 179–191 (2001).
61 S. Batchelor et al., Solar electric cooking in Africa: Where will the transition happen first? Energy Res. Soc. Sci. 40, 257–272 (2018).
62 C. M. Radimilahy, Z. Crossland, Situating Madagascar: Indian ocean dynamics and archaeological histories. Azania 50, 495–518 (2015).
63 C. Radimilahy, “Mahilaka: An archaeological investigation of an early town in Northwestern Madagascar,” PhD thesis, University of Uppsala, Uppsala, Sweden
(1998).
64 S. J. Burns et al., Rapid human-induced landscape transformation in Madagascar at the end of the first millennium of the Common Era. Quat. Sci. Rev. 134, 92–99
(2016).
65 L. Jarosz, Defining and explaining tropical deforestation: Shifting cultivation and population growth in colonial Madagascar (1896-1940). Econ. Geogr. 69, 366–
379 (1993).
66 I. R. Scales, Farming at the forest frontier: Land use and landscape change in Western Madagascar, 1896-2005. Environ. Hist. 17, 499–524 (2011).
67 G. Sodikoff, Forest and Labor in Madagascar: From Colonial Concession to Global Biosphere (Indiana University Press, Bloomington, IN, 2012).
68 J. Maina et al., Linking coral river runoff proxies with climate variability, hydrology and land-use in Madagascar catchments. Mar. Pollut. Bull. 64, 2047–2059 (2012).
69 K. Douglass et al., Multi-analytical approach to zooarchaeological assemblages elucidates Late Holocene coastal lifeways in southwest Madagascar. Quat. Int.
471, 111–131 (2018).
70 M. Parker Pearson, Pastoralists, Warriors and Colonists: The Archaeology of Southern Madagascar (British Archaeological Reports, Oxford, UK, 2010).
71 G. Heurtebize, Les anciennes cultures de l’Androy central. Taloha 10, 171–182 (1986).
72 A. V. Samson, C. A. Crawford, M. L. Hoogland, C. L. Hofman, Resilience in pre-columbian caribbean house-building: Dialogue between archaeology and
humanitarian shelter. Hum. Ecol. Interdiscip. J. 43, 323–337 (2015).
73 S. F. Fitzpatrick, W. F. Keegan, Human impacts and adaptations in the Caribbean islands: An historical ecology approach. Trans. R. Soc. Edinb. 98, 29–45 (2007).
74 P. E. Siegel, Ed., Island Historical Ecology: Socionatural Landscapes of the Eastern and Southern Caribbean (Berghahn Books, New York, NY, 2018).
75 S. B. Cooke, L. M. Dávalos, A. M. Mychajliw, S. T. Turvey, N. S. Upham, Anthropogenic extinction dominates Holocene declines of West Indian mammals. Annu.
Rev. Ecol. Evol. Syst. 48, 301–327 (2017).
76 S. Federman et al., Implications of lemuriform extinctions for the Malagasy flora. Proc. Natl. Acad. Sci. U.S.A. 113, 5041–5046 (2016).
77 T. Braje, T. P. Leppard, S. M. Fitzpatrick, J. M. Erlandson, Archaeology, historical ecology and anthropogenic island ecosystems. Environ. Conserv. 44, 286–297
(2017).
78 T. Rick, P. V. Kirch, J. M. Erlandson, S. M. Fitzpatrick, Archeology, deep history, and the human transformation of island ecosystems. Anthropocene 4, 33–45
(2013).
79 L. A. Newsom, E. S. Wing, On Land and Sea: Native American Uses of Biological Resources in the West Indies (The University of Alabama Press, Tuscaloosa, AL,
2004).
80 D. Angst, A. Chinsamy, L. Steel, J. P. Hume, Bone histology sheds new light on the ecology of the dodo (Raphus cucullatus, Aves, Columbiformes). Sci. Rep. 7,
7993 (2017).
81 K. F. Rijsdijk et al., Mid-Holocene (4200 kyr BP) mass mortalities in Mauritius (Mascarenes): Insular vertebrates resilient to climatic extremes but vulnerable to
human impact. Holocene 21, 1179–1194 (2011).
82 V. F. B. Florens, “Conservation” in Mauritius and Rodrigues: Challenges and Achievements from Two Ecologically Devastated Oceanic Islands. Conservation
Biology: Lessons from the Tropics, P. H. Raven, N. S. Sodhi, L. Gibson, Eds. (John Wiley & Sons, Ltd, Oxford, UK, 2013), pp. 40–50.
83 L. Maffi, “On the interdependence of biological and cultural diversity” in On Biocultural Diversity: Linking Language, Knowledge, and the Environment, L. Maffi,
Ed. (Smithsonian Institution Press, Washington, DC, 2001), pp. 1–50.
84 A. M. Bauer, E. C. Ellis, The anthropocene divide: Obscuring understanding of social-environmental change. Curr. Anthropol. 59, 209–227 (2018).
85 E. Ronneberg, “Small islands and the big issue: Climate change and the role of the alliance of small Island States” in The Oxford Handbook of International
Climate Change Law, K. R. Gray, R. Tarasofsky, C. Carlarne, Eds. (Oxford University Press, Oxford, UK, 2016), pp. 761–778.
86 A. K. Baptiste, H. Devonish, The manifestation of climate injustices: The post-Hurricane Irma conflicts surrounding Barbuda’s communal land tenure. J. Extr. Even.
06, 1940002 (2019).
87 C. E. Willison, P. M. Singer, M. S. Creary, S. L. Greer, Quantifying inequities in US federal response to hurricane disaster in Texas and Florida compared with Puerto
Rico. BMJ Glob. Health 4, e001191 (2019).
88 UNDP, Human development indices and indicators: 2018 statistical update. http://hdr.undp.org/sites/default/files/2018_human_development_statistical_update.
pdf. Accessed 2 October 2019.
Downloaded by guest on October 3, 2020

8262 | www.pnas.org/cgi/doi/10.1073/pnas.1914211117 Douglass and Cooper