Luckeneder Et Al 2021

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/352118108
Surge in global metal mining threatens vulnerable ecosystems
Article in Global Environmental Change · July 2021

DOI: 10.1016/j.gloenvcha.2021.102303
CITATION READS
1 204
5 authors, including:
Sebastian Luckeneder Stefan Giljum

Wirtschaftsuniversität Wien Wirtschaftsuniversität Wien
5 PUBLICATIONS 25 CITATIONS 133 PUBLICATIONS 6,892 CITATIONS
SEE PROFILE SEE PROFILE
Anke Schaffartzik Victor Maus

Central European University Wirtschaftsuniversität Wien
80 PUBLICATIONS 1,597 CITATIONS 25 PUBLICATIONS 436 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Prosperous Pathways in the EU Bioeconomy (BIOWAY) View project
SCP-HAT (Sustainable Consumption and Production Hotspots Analysis Tool) View project
All content following this page was uploaded by Sebastian Luckeneder on 05 June 2021.
The user has requested enhancement of the downloaded file.

Global Environmental Change 69 (2021) 102303
Contents lists available at ScienceDirect
Global Environmental Change

journal homepage: www.elsevier.com/locate/gloenvcha
Surge in global metal mining threatens vulnerable ecosystems

Sebastian Luckeneder a, *, Stefan Giljum a, Anke Schaffartzik b, Victor Maus a, c, Michael Tost d
a
Institute for Ecological Economics, Vienna University of Economics and Business (WU), Austria
b
Institute of Social Ecology, University of Natural Resources and Life Sciences, Vienna, Austria
c
International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria
d
Chair of Mining Engineering and Mineral Economics, Montanuniversität Leoben, Austria
A R T I C L E I N F O A B S T R A C T
Keywords: Mining activities induce profound changes to societies and the environment they inhabit. With global extraction
Mining of metal ores doubling over the past two decades, pressures related to mining have dramatically increased. In this
Natural resources paper, we explore where growing global metal extraction has particularly taken effect. Using fine-grain data, we
Raw material extraction mapping
investigate the spatial and temporal distribution of mining of nine metal ores (bauxite, copper, gold, iron, lead,
Socio-environmental impact
manganese, nickel, silver and zinc) across approximately 3,000 sites of extraction worldwide between 2000 and
2019. To approach the related environmental implications, we intersect mining sites with terrestrial biomes,
protected areas, and watersheds categorised by water availability. We find that 79% of global metal ore
extraction in 2019 originated from five of the six most species-rich biomes, with mining volumes doubling since
2000 in tropical moist forest ecosystems. We also find that half of global metal ore extraction took place at 20 km
or less from protected territories. Further, 90% of all considered extraction sites correspond to below-average
relative water availability, with particularly copper and gold mining occurring in areas with significant water
scarcity. Our study has far-reaching implications for future global and local policy and resource management
responses to mitigate the negative effects of the expected expansion of metal mining.
1. Introduction the extractive sector are apparent, with countries and regions unequally
contributing to this global trajectory (Dorninger et al., 2021; Schaf
Mining plays an ambiguous role for society. It has become indis fartzik et al., 2016). Local mining expansion and intensified production,
pensable to the model of economic growth currently pursued in indus as well as related impacts in the immediate surroundings of the sites of
trialised, mineral-based societies. But it is also among the most extraction, are closely coupled to overarching global change, calling for
environmentally and socially hazardous human activities. Its harmful a “multilevel perspective” (Gibson et al., 2000) to understand the
consequences for the environment and its catalysing association with environmental and social implications of mining’s worldwide growth.
social conflicts are well-documented (Scheidel et al., 2020; Conde, 2017; By studying the local expressions of the global surge in mining in a
Bebbington et al., 2008; Bridge, 2004). Unprecedented and rapidly spatially explicit manner, we seek to advance the empirical under
growing extraction of metals and minerals during the past two decades standing and conceptual framing across levels of scale.
(Schandl et al., 2017; Schaffartzik et al., 2014) and the projected in This paper presents and contextualises a detailed assessment of how
crease in material demand (UN IRP, 2019; OECD, 2019) are alarming metal mining volumes are distributed across almost 3,000 mining pro
signs that associated impacts will intensify in the future. jects worldwide, covering nine metal ores (bauxite, copper, gold, iron,
The surge in global metal mining signifies an increased production of lead, manganese, nickel, silver and zinc) in the period 2000–2019. We
metal commodities through establishing new mining projects, physical explore whether the development of global metal mining has particu
expansion of existing sites and intensifying and optimising the extrac larly affected vulnerable ecosystems around the world and identify
tion process. It is part of an overarching trajectory of globally increasing hotspots of raw material extraction and ecosystem impact. We assume
resource use, referred to as the Great Acceleration (Steffen et al., 2015), extraction gains to be associated with additional pressures, because
which is pushing the global economy’s metabolism up against Planetary production volumes are likely related to the areal extent of mining sites
Boundaries (Rockström et al., 2009). Differences in growth dynamics of (Werner et al., 2020) and intensified use of heavy machinery. Based on
* Corresponding author at: Vienna University of Economics and Business (WU), Welthandelsplatz 1, 1020 Vienna, Austria.
E-mail address: [email protected] (S. Luckeneder).
https://doi.org/10.1016/j.gloenvcha.2021.102303
Received 11 December 2020; Received in revised form 8 April 2021; Accepted 13 May 2021
Available online 3 June 2021
0959-3780/© 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
S. Luckeneder et al. Global Environmental Change 69 (2021) 102303
our large-scale empirical assessment, we discuss the environmental Mining Database1 (SNL, 2020). Projects were considered for which SNL
implications of the expected increase in mining activities. We pave the reported any mining of bauxite, copper, gold, iron, lead, manganese,
way for in-depth local studies and future assessments of the cumulative nickel, silver or zinc within the time period 2000–2019, in total sum
magnitude and transmission of impacts, and summarise how mining ming up to 2935 individual mines. Focusing on the 21st century obvi
companies and policies can contribute to impact mitigation. ously presents a limitation, but it allows to cover the entire period of
In order to gauge the potential impact of mining on ecosystems, we what might be considered the “second great acceleration” (Görg et al.,
focus on ecosystem vulnerability. Vulnerability has been debated in 2020) of resource use at the global level.
sustainability science under definitions such as “the degree to which a Except for bauxite, iron and manganese, which are listed as ores
system, subsystem, or system component is likely to experience harm (gross weight), metal production is reported as metal content (net
due to exposure to a hazard, either a perturbation or a stress/stressor” weight). In order to construct a homogeneous measure of extracted
(Turner et al., 2003, p. 8074). In their extensive review, Weißhuhn et al. crude ore that is used as an input from the natural environment to the
(2018) suggest the term ecosystem vulnerability as being preferable to economic system, we applied country- and commodity-specific conver
ecological, environmental, or other notions of vulnerability and propose sion factors from UNEP’s Global Material Flow Database (UN IRP, 2017)
a framework with “exposure”, “sensitivity”, and “adaptive capacity” to all net weight commodities. This measure, referred to as “extraction”
defining the degree of vulnerability. They stress that such a perspective or “metal ore” in the following, corresponds to the actual amount of
on vulnerability is biocentric (Birkmann and Wisner, 2006) rather than extracted material exerting pressure on the environment instead of a
anthropocentric, because it understands environmental systems as being final product after several processing steps.2 Fig. 1 provides a summary
affected by natural and anthropogenic drivers, instead of being sources of the data and illustrates variations in extraction volumes and
of hazards that influence human systems. In employing this concept in increasing extraction rates for the nine metal ores. For more detail, see S
the context of intensified mining, we investigate whether the acknowl 1.1 and 1.2 in the supplementary material.
edgement of ecosystem vulnerability deters extraction. We considered this set of base and precious metals in our study
For the purposes of this paper, we assume that mining activities exert because of their extensive industrial use. Next to construction materials,
pressure on all ecosystems, but that certain areal characteristics can be coal and crude oil, iron, bauxite and copper ore feature the highest
identified which are considered to signal particular vulnerability. Our extracted mass among mineral resources (Murguía et al., 2016). Schaf
study uses three spatial layers as proxy indicators for ecosystem fartzik et al. (2016) name iron and aluminium (we consider bauxite)
vulnerability: terrestrial biome categorisations, protected areas, and being quantitatively most important. Together with other metals in
water scarcity. Subsequently, we connect these layers with spatio- smaller amounts including lead, manganese and copper they form the
temporal patterns of extraction at the mine level. In doing so, we “skeleton of industrial development” (ibid.: 103). Similar selections are
combine approaches from previous work dealing with single environ made, for example, by Durán et al. (2013) and Northey et al. (2017),
mental layers and their respective links to mining. Murguía et al. (2016) additionally including zinc and nickel. We furthermore consider the
and Sonter et al. (2018) demonstrate how mining relates to biodiversity precious metals gold and silver, because of the high prevalence of such
loss. Durán et al. (2013) find that metal mining activities undermine the mining facilities (Murguía et al., 2016) and persistent exploration
role of protected areas as a key policy tool for conservation. Regarding expenditure due to high market value (Ali et al., 2017). Other metals
an intersection of mining sites with water scarcity indicators, Northey with growing demand, such as cobalt or rare earth metals, were not
et al. (2017) find that the exposure of areas to water risk is especially covered in our analysis as they were poorly reported in the database at
high in the case of copper mining. Recently, studies have also investi hand.
gated the areal extent of mines, either based on estimations, such as Tost We are aware that assuming national averages as ore grades in
et al. (2020), or by the use of satellite imagery (Maus et al., 2020; troduces uncertainty in the data. As a validation of coverage and quality,
Werner et al., 2020). These studies focus on cross-sectional analyses of a we compared annual metal extraction according to our dataset (based on
specific selection of metals, such as four key metals in Durán et al. reports of production from individual mines) with official UNEP IRP
(2013) or of the locations of base metal resources in Northey et al. statistics (based on national accounts). In the supplementary material,
(2017). Our study builds upon these approaches to provide the first we demonstrate that our extraction data provides reliable coverage,
fine-grained assessment of metal mining regions on a worldwide scale, mimicking extraction trends reported by other sources (S 1.3.1). Largest
covering nine major metal commodities across a 20-year period. gaps between our modified SNL aggregates and UNEP IRP figures occur
We find that the rapidly expanding mining sector exerts increasing because of country-level irregularities, most notably regarding China.
pressures on ecosystems recognised as vulnerable. Our results show that For the case of iron, approximately 1,500 Megatonnes (Mt) of Chinese
there is a substantially skewed distribution in terms of extraction vol iron ore extraction in 2017 is missing in the SNL data compared to UNEP
umes, i.e. extraction per mine is not evenly distributed around a certain IRP. Moreover, we illustrate that assumptions about ore grades influence
value, but a minority of mines reports much higher figures than the extraction estimates for individual mines (S 1.3.2). For comparison, we
mass. Further, regional hotspots of mining growth have emerged during considered copper mines of four countries and mine-specific ore grades
the observed time period, in particular in Latin America, Central Africa, available from Mudd and Jowitt (2018). The exercise suggests that
India, and Western Australia. Due to the multilevel approach, our assumed national averages are conservative, as they rather lead to
findings have substantial implications for future global and local policy under-estimating extraction, and that estimates from both approaches
responses, supporting calls for a stricter set of rules for accessing primary correlate well. However, we must also note that we found substantial
resources. deviations for some extraction countries, such as Peru, where estimates
of the considered sample sum up to 350 Mt when utilising individual ore
2. Material and methods grades while they only amount to 90 Mt using the UNEP average grade.
2.1. Mining data and environmental spatial layers

1
This database, offered by Standard and Poor’s (S&P) Market Intelligence,
We utilised mine-specific production data from the SNL Metals and provides extensive operative and financial information on thousands of mining
projects based on company reports.
2
Waste rock is excluded because this information is not yet available for
global analyses, but it would be extremely valuable to researchers and policy
makers. We expect that an inclusion of over- and interburden might allow for a
better approximation of the potential disruption to the local system.
2
Fig. 1. Global extraction of bauxite, copper, gold, iron, lead, manganese, nickel, silver and zinc ores based on SNL (2020) and UN IRP (2017) conversion factors.
Please note the differences in scale on the y-axes. See Table S2 in the supplementary material for underlying data.
Difficulties to fully consider artisanal and informal mining on a in recognisably sensitive areas with regard to species richness, need of
global scale impose a limitation, again suggesting that the data we used protection, and water availability. In doing so, we also addressed dif
represent under- rather than over-estimations of material extraction. ferences in extraction patterns across the nine commodities. Third, to
The environmental pressures related to small-scale mining will be at detect the most critical developments and based on our findings from the
least as strong and possibly farther-reaching compared to larger mines layer analyses, we performed an assessment of hotspots among the three
considered in our study (Asner et al., 2013; Caballero Espejo et al., environmental layers. An illustration of our workflow is provided in
2018). Fig. 2.
Exposure to environmental complexity was assessed by spatially To facilitate visual interpretation, we aggregated annual metal
intersecting extraction data with terrestrial biome categorisations mining into 1 × 1 degree cells (corresponding to approximately 110 km
(accessed from Resolve, 2017 based on Dinerstein et al., 2017). Biomes at the equator) by summing up the total ore extraction across all nine
are defined as communities of plants and animals occurring together metals. This simplification serves as an initial, high-level picture of the
under certain climate conditions (Resolve, 2017). They systematically spatial distribution of global metal ore extraction. It cannot point out
subsume the richness of an area’s ecosystem in terms of variety in and potential environmental impacts that occur in direct and up to only a
number of species. As another proxy for anthropocentrically recognised few km proximity to mines, nor does it reflect the actual number and
rich, special, or endangered biodiversity, we considered a spatial layer specifics of mining projects such as type and scale within each grid cell.
on protected areas (UNEP-WCMC, 2020). We calculated the distance to However, recent studies show that assuming biodiversity (Sonter et al.,
the closest protected area for each mine. Third, we considered the 2020) and deforestation (Sonter et al., 2017) effects within at least a 50
Available Water Remaining (AWARE) index (WULCA, 2019) as an in km wide radius around mines is reasonable, and hence the map provides
dicator for water risk exposure (annual average at watershed level). It a first estimate of potentially negatively affected areas and indicates the
represents water availability after the demand of ecosystems and corresponding extraction volumes (for a more conservative grid of 0.5
humans is met. While a variety of spatial water indices exists with each and 0.1 degrees see supplementary material S 1.5). In order to intersect
providing a somewhat different perspective on the interactions between mining activities with regional environmental structure, we overlaid all
water resources and mining (Northey et al., 2017), we chose this index mine sites with the three different spatial layers mentioned above and
for two main reasons. First, it provides spatial coverage for the entire set aggregated annual extraction volumes into the respective layer
of considered mines except for Nalunaq mine in Greenland. Second, the categories.
AWARE index was designed to reflect potential water deprivation by Trends in the extraction volumes were estimated employing a
other users of water – a suitable indicator given that mines utilise water geographically weighted regression (GWR) model, modelling log-
in a number of processes. The index further features a convenient transformed extraction at the mine level as a function of time (see
interpretation: It is limited between 0.1 and 100, where 1 corresponds to supplementary material S 2 and Brunsdon et al. (1996) for more detail
the world average and 10, for instance, represents water availability that regarding GWR). GWR captures the spatial structure within the data and
is ten times less than the world average. yields spatially varying parameter estimates. In contrast to estimating a
trend for each single spatial observation, GWR incorporates the infor
mation of surrounding mines weighted by geographical distance and
2.2. Approaches to analysing global distribution and extraction trends
hence reflects potential compound effects of mining on the environment
in areas where multiple sites are close to each other. Extractive in
We implemented three analytical steps in order to evaluate the surge
dustries can have different relationships across regions such as networks
in global metal mining between the years 2000 and 2019. We first
of mines expanding in emerging mining regions, regional boosts in in
assessed annual spatial distribution and concentration of global metal
vestment and new technologies or multiple mines being closed in certain
ore extraction. Second, we performed spatial overlay analyses to
areas as a consequence of decreased (economic) feasibility of mining.
determine the extent to which increased production particularly occurs
3
Fig. 2. Workflow of this study, highlighting utilised spatial data, analytical steps 1–3, and respective operational level and scale of each stage. Dashed red lines
represent links across layers, and solid red lines indicate syntheses of information.
While the GWR approach makes it possible to depict sub-national with extraction occurring only in a few mines in the very south of the
densification clusters by reflecting heterogeneity within countries, it country. Likewise, in Brazil, 90% of all iron ore extracted in 2019 can be
comes at the cost that it smooths over individual outlier cases within the attributed to only ten mining sites within the states of Pará and Minas
clusters (using spatial weights assigned to proximate mining projects) Gerais.
that might also have critical impacts on the environment. The unequal distribution of extraction intensity across all mining
regions shows a similar pattern across commodities, but on different
3. Results scales regarding volumes, the number of mining regions and the degree
of concentration. For more detail, see supplementary material S 1.6,
The main findings of our study include global spatially explicit where we resampled extraction data to 0.25 degree resolution and or
extraction volumes, the distribution of mining sites, and their intersec dered the cells by their extraction volumes for all nine metals.
tion with terrestrial biomes, distance to protected areas and water risk In Figs. 4a-c, we map the three environmental layers and indicate
classifications over the period 2000–2019. We also identify particular how mining sites are distributed across them. With regard to terrestrial
hotspots of mining intensification and expansion. biome classifications, most mines are located in temperate broadleaf &
mixed forests (627 observations), tropical & subtropical moist broadleaf
forests (594) and deserts & xeric shrublands (448), while only 39 sites lie
3.1. Spatial distribution of metal mining in the tundra biome. The histogram in Fig. 4b illustrates that 92% of all
mines are located within a distance of 250 km to a protected area. More
Relative to the total earth surface, only a small area is used for than a third of the 2935 mines are in a range of 20 km, and 4.2% of all
mining. In a recent paper, Maus et al. (2020) estimated the global mining sites (i.e. 123 mines) are located within designated protected
mining area at around 57,300 km2 , approximately the size of Croatia or territories. Lastly, we detect that, on the one hand, only 280 mines, i.e.
Togo. In Fig. 3, we show all mining activities projected into 1 × 1 degree 9.5% of our full sample, are located in watersheds of above-average
cells.3 Among the mining cells, we observe strong heterogeneity water availability. On the other hand, 13% of mines lie within areas
regarding extraction volumes and a highly skewed distribution, i.e. a of an index score higher than 60 (corresponds to 4.1 on the figure’s
small fraction of mining areas reporting very large extraction volumes logarithmic scale), i.e. within high water risk regions such as various
and vice versa. For 2019, about 95% of all mining areas indicate less Central and Eastern Asian deserts, the Chihuahuan Desert, and South
than 25 Mt and half of the observations less than 1.3 Mt of extraction per east Australian temperate forests and savannas.
cell, while a small fraction of cells yields extraction of up to 279 Mt.
Mining activities are spatially concentrated in Western Australia,
Southern Africa, and along the Andes and into the Central and North 3.2. Surge in global metal mining in environmentally vulnerable regions
American ranges of the American Cordillera. Furthermore, we find sig
nificant activities in Brazil, West Africa, India, China and Southeast and Having explored the uneven distribution of mining activities across
Central Asia. For China, we observe a considerable number of mining the globe, we evaluated the intensification and expansion of mining
sites, but we also find that almost 80% of the observed cells show against the backdrop of perceivable regional vulnerability to the
extraction volumes not larger than 2 Mt. Maus et al. (2020) also high harmful consequences of extractive practices. Fig. 5 illustrates how
light that mining in China is characterised by many – on average smaller extraction volumes have developed in relation to the regional charac
– mining areas, while Australia reaches a comparable total areal extent teristics of the area in which they occur. It is important to note that
with fewer, but larger mining sites. Certain clusters appear in strong “regional” in this sense refers to a wider understanding than specific
concentrations within countries, such as in Brazil or along the Zambian localised features in the immediate surroundings of mines. Heteroge
and Congolese border. This fact is vital to consider for the discourse neity in space that occurs at more granular levels is not assessed here.
about mining and its impacts, as well as for global material flow ana Not only extraction volumes (see Section 3.1) but also biomes exhibit
lyses. Sizeable copper flows, for example, originate from the DR Congo, notable variation in size: the Mediterranean forests, woodlands & scrub
biome extends over only 1.5% of global land surface while boreal for
ests/taiga are the largest biome covering 12%. We hence evaluated both
3
2000, 2010, 2015 and 2019 maps separate by metal are available in the absolute numbers (left panels in Fig. 5; in Mt) and relative trends (right
supplementary material (see S 3). For an evaluation of how the counts of panels; using 2000 as the base year) for the three spatial layers.
mining cells have changed per commodity, see supplementary material S 1.4. Fig. 5a shows that, in absolute values, deserts and xeric shrublands
4
Fig. 3. Global 2019 metal ore extraction (in Mt) grouped by 8 quantiles on a 1 degree resolution (i.e. about 110 km × 110 km at the equator) using a Robinson map
projection. Based on SNL (2020) and UN IRP (2017) conversion factors.
(DesXS) were the most exploited terrestrial biome during the past mining in protected areas has risen from 225 Mt to 480 Mt, a 113%
twenty years. In 2019, 2,241 Mt of metal ores were mined there, fol increase. Further, the figure suggests that there was a significant surge in
lowed by tropical and subtropical moist broadleaf forests (TropSubMBF, mining in the 10–15 km range from protected areas.
911 Mt) and temperate broadleaf and mixed forests (TempBMF, 614 The large increase of mining activities in protected areas is partly
Mt). Upon ranking all biomes according to their richness in species4, we explained by new sites. Over the full period analysed in this study, 123
find that five of the six most complex terrestrial biomes jointly were the mines were detected within protected zones. The number of mining sites
origin of 79% of total metal ore extracted in 2019. Next to the three in such zones increased from 55 in 2000 to a peak of 96 in 2012 and
biomes mentioned above, these five include montane grasslands and since then oscillated between 80 and 90. One prominent example of a
shrublands (MontGS) and tropical and subtropical grasslands, savannas new project within a protected area is the expansion of mining in the
and shrublands (TropSubGSS). While this pattern has not altered over Carajás National Forest in the state of Pará, Brazil, located in the Xingu-
the past twenty years and the ranking across biomes remained mostly Tocantins-Araguaia moist forests ecosystem. In 2016, the Brazilian
unchanged since 2000, we see that extraction has changed in relative multinational corporation Vale opened the Serra Sul (also known as
terms within biomes. The right panel of Fig. 5a shows that mining has S11D) mine as the largest project in the company’s history (Vale, 2016).
intensified over time in all biomes except temperate conifer forests In 2019, 73 Mt of iron ore were mined at this site. Other mines within the
(TempCF) and MontGS. In TropSubMBF, the biome richest in species, area, which was declared national forest of the state Pará in 1998, have
metal ore extraction has increased by a factor of 2.1, due to, among already existed longer, yet increased their extraction significantly at the
others, the expansion of mining activities in the Central Range Papuan beginning of the 21st century. For example, the N5 iron ore mine
montane rain forests (New Guinea), the North Western Ghats moist increased its production from 11 Mt in 2001 to 54 Mt in 2013 and the
deciduous forests and the Malabar Coast moist forests (India), and the N4W mine from 15 Mt in 2001 to almost 40 Mt in 2012.
Borneo lowland rain forests (Indonesia). Some mines are not located within, but directly border on protected
Nature reserves are a political instrument for protecting specific areas. The Indonesian Grasberg copper and gold mine, one of the world’s
territories from anthropogenic environmental destruction. However, it largest mining projects, is such an example. Its concession area imme
is likely that the designation of conservation areas also accounts for diately neighbours Lorentz National Park, designated World Heritage
current and potential future minerals extraction in some countries. As a Site in 1999. The mine is related to the pollution of rivers and lakes in
second spatial layer, we examined the proximity of mining activities to that area due to riverine tailings disposal (Martinez-Alier, 2001). Lor
protected areas. Fig. 5b depicts global extraction up to 50 km from such entz National Park is the largest national park in South-East Asia. It has
areas. In 2019, 50% of all global metal ore extraction took place within a an outstanding biodiversity and comprises a number of fragile ecosys
20 km boundary around protected territories, and 480 Mt (8%) were tems, such as subalpine areas, tropical rainforest, and mangroves
mined within officially protected zones. Similar to mining within bi (UNESCO, 2020).
omes, we find that this is a pattern that has not changed fundamentally Supporting the choice of terrestrial biomes as one category indi
since 2000. But it has intensified. While the high concentration of metal cating ecologically complex and vulnerable regions, nine out of the ten
ore extraction within a buffer of 20 km around protected areas appears largest extraction projects that lie within protected areas (identified by
to be a stable finding for the period considered in our study, it is the accumulated mined volumes) are located in TropSubMBF or Trop
mining within protected areas that surged. Over the twenty-year period, SubGSS. Only the Grasberg mine in Indonesia belongs to the MontGS
biome. Eight of these nine mines in tropical regions are located in Brazil
(seven of which are iron ore mines and one is a bauxite mine).
The third layer intersection was conducted with regard to water
4
We based the ranking on the Millennium Ecosystem Assessment (2005), stress, using the normalised AWARE index with a score of 1 representing
where species richness (the number of species in a given area) is referred to as the available water remaining in a watershed as corresponding to world
the most common measure of biodiversity.
5
Fig. 4. Mining sites considered in this study (based on SNL, 2020) and their distribution across a) terrestrial biomes (based on Resolve, 2017), b) protected areas
(based on UNEP-WCMC, 2020), and c) AWARE water stress index classifications (log-transformed, based on WULCA, 2019) using Robinson map projections.
6
Fig. 5. Total (left) and relative (right) extraction volumes within a) biome classifications, b) 5 km proximity buffers to protected areas, and c) available water
remaining (AWARE) decile intervals; 2000–2019; based on SNL (2020) and UN IRP (2017) conversion factors.
average and scores above 1 corresponding to less water availability Escondida copper mine in the Chilean Atacama desert, iron ore pro
relative to the global average. Fig. 5c illustrates annual extraction in Mt duction in the Australian Pilbara shrublands, and copper mining in the
at the AWARE decile level. While in 2019, 1,080 Mt (18 %) of metal ore Kazakh semi-desert.
were extracted in the decile with high relative water availability, i.e.
watersheds with an AWARE score smaller or equal to 1, 90% of all
3.3. Differences across metal types
considered extraction sites are located within watersheds that have
below-average relative water availability. Considering that the
We next highlight differences across metals in Fig. 6. From top to
consumption-weighted average AWARE index is 43, meaning that the
bottom, this figure depicts the environmental layers and from left to
majority of economic activity and water consumption affects regions
right, it represents the nine metals considered in this study. We illustrate
where the index exceeds 1, we find that 1,034 Mt (17 %) were extracted
how the composition of extraction volumes per metal has developed
in watersheds with an AWARE score higher than this weighted average.
between 2000 and 2019. Regarding biomes (6a), the five categories with
In the most critical category, the 64 to 97 decile, 472 Mt were mined in
highest species richness – TropSubMBF, TropSubGSS, DesXS, tropical
2019, of which approximately one third was mined in the Chihuahuan
and subtropical dry broadleaf forests (TropSubDBF) and MontGS – are
desert in Mexico and the United States. Largest extraction volumes and
separately shown while all other biome categories are aggregated as
highest growth rates are measured for watersheds with AWARE scores
“other”. Charts in 6b show distance to protected areas in incremental
between 17 and 28. These include major mining hubs such as the
steps of 5 km each up to 20 km proximity and remaining sites grouped
7
Fig. 6. Extraction volumes by commodity and a) selected biomes, b) distance to protected area, and c) AWARE index; 2000–2019; based on SNL (2020) and UN IRP
(2017) conversion factors.
into a residual category. In 6c, AWARE index categorisations are mines in the USA. The fact that copper and gold ores are predominantly
depicted. We stress the five most critical water scarcity deciles and mined in desert ecosystems is reflected by water pressure indicators, as
subsume the other half of all observations under AWARE scores between becomes evident in Fig. 6c. Approximately 40% of all copper ore is
0 and 10.7. Note that the levels of extraction volumes differ widely, with mined inside the two highest AWARE deciles. While gold ore extraction
iron, copper and gold ore being mined in the largest amounts. shows a decreasing trend in the third highest decile of scores around 30,
Highest increases in global extraction are reported for iron and iron ore extraction substantially increased for regions with AWARE
manganese ores. In 2019, more than 75% of iron ore were mined either scores between 17 and 28. These include several Australian iron ore
in TropSubMBF, TropSubGSS or DesXS biomes. This share has grown mines located in the Pilbara shrublands, such as Hope Downs, the Sino-
since 2000, when 60% were mined in these three categories, mostly (but Iron project, and Roy Hill, which may have significant impact on
not only) because of extraction gains in DesXS. Manganese, which is groundwater and surface water in that area (WA Government, 2009).
predominantly used as an input for steel production, reaches an even Nickel mining tends to affect mostly areas with low AWARE scores.
higher share of production stemming from these critical biomes of about However, we also notice a rapid increase in the second decile since
90%. This proportion has not significantly changed since 2000, but 2016.
absolute volumes have risen from 16 Mt in 2000 to 46 Mt in 2019.
Furthermore, similar to iron ore and bauxite, manganese ore is, to a
3.4. Regional hotspots
notable extent, mined within protected areas. A prominent example are
manganese mining operations at Groote Eylandt: Australia’s fourth-
One of the innovations of our study is to show regional mining pat
largest island is entirely part of Anindilyakwa Indigenous Protected
terns over time based on spatially explicit accounts of metal ore
Area. Extraction volumes at Groote Eylandt have more than quadrupled
extraction. In order to estimate regional trends, we conducted a GWR
between 2000 and 2019, and in 2019, the mine extracted about three
analysis. As noted in Section 2.2, the framework of a GWR considers
quarters of Australian manganese ores.
both reported volumes for each mine and volumes for respective
Besides bauxite, iron, and manganese, nickel ores are most notably
neighbouring mining sites, with more weight given to closer operations.
mined in TropSubMBF. While bauxite mining decreased in TropSubMBF
We hence obtained trend estimates for each mining location not only
from 76 Mt in 2012 to 45 Mt in 2019, and shifted to TropSubGSS (49 Mt
accounting for the mine itself, but also for surrounding activities,
in 2012 and 72 Mt in 2019), the mining of nickel ores in TropSubMBF
enabling us to provide an overview of the trends across agglomerations
has been increasing during the past two decades. In 2019, about 55% of
of mining projects. In Fig. 7, we present all positive 2000–2019 trend
global nickel ore extraction took place in the species-richest tropical
estimates for total metal ore extraction. We limit results shown on the
biome, 66% of which can be attributed to only two mining sites in the
map to only positive coefficients in order to highlight hotspots of
Indonesian Sulawesi lowland rain forests, Sorowako and Pomalaa.
increased production, the focus of this study. A map including negative
Sulawesi island has been suffering from massive deforestation in recent
GWR trend coefficients and metal-specific maps are provided in the
decades, with nickel mining known as one among several significant
supplementary material (S 2). Table 1 lists selected hotspots with their
drivers (Supriatna et al., 2020). In addition, nickel mining is almost
respective GWR coefficients, biome, distance to the nearest protected
entirely conducted within a 20 km distance to protected areas, and
area, AWARE score, and average annual extraction.
about half of all ore is extracted at 5 km or less from protected areas.
On the one hand, GWR results reveal global hotspots of densification,
However, in contrast to bauxite and iron ore mining, nickel is hardly
i.e. positive trend coefficients for mining regions where extraction vol
mined within protected areas.
umes have on average increased. We find positive coefficients for 1272
The world’s largest copper and gold deposits are located in DesXS,
site observations. The regions with the highest extraction growth rates
and hence large mining projects are located in this biome, such as
(between 7% and 10% per year) are located in mining clusters in Peru,
Escondida and Chuquicamata in Chile, or Morenci and Bingham Canyon
the DR Congo, Zambia, India, China, and in Western Australia. Highest
8
Fig. 7. Mining sites with positive trends in metal ore extraction between 2000 and 2019. The linear coefficients were estimated for each mining location using GWR.
The map is a Robinson projection. Based on SNL (2020) and UN IRP (2017) conversion factors. Technical notes on GWR are provided in S 2 in the supplemen
tary material.
average annual growth is reported for iron ore and bauxite sites in extraction expansion) collapsed, causing 19 casualties and far-reaching
Odisha, Eastern India. On the other hand, 1659 site observations yield contamination of rivers, eventually spreading pollutants over more than
negative trend coefficients and can therefore be interpreted as regions 600 km (Fonseca do Carmo et al., 2017).
with decelerating extraction. In the following, we highlight some ex With average annual growth rates of approximately 7%, the stron
amples, connecting to other findings from the previous sections. gest intensification of mining activities in Africa is found in the border
Countries along the Andean Range tend to be strongly involved with region between Zambia and the DR Congo, known as the Central African
mining. For Peru, we can identify mining intensification in the south of Copperbelt. The highly mineralised region lies in the TropSubGSS
the country, whereas extraction volumes on average decreased in the biome, more precisely the Central Zambezian miombo woodlands. It is
north. The densification hotspots are represented in Table 1 by the known for its vast copper deposits, but is also the habitat for a great
example of the five mines with highest growth coefficients inside Peru. variety of wildlife such as large mammals. The adverse effects of mining
They are located in the Central Andean puna and Sechura desert ecor expansion in the Copperbelt, especially on forests and forest livelihoods,
egions, and lie within areas reaching a particularly high AWARE score of are evident. Moving millions of tonnes of earth, industrial copper mining
28.27. Significant mining intensification at these hotspots is cause for has directly and indirectly caused significant environmental change due
concern because Peruvian regions suffering from water scarcity are to extensive forest clearings, pollution of soil, air and water, and pop
shown to be particularly vulnerable to ecological distribution conflicts ulation pull effects of mining towns (Mwitwa et al., 2012; Peša, 2020).
related to mining and water competition (Salem et al., 2018; Bebbington Our results show that Zambian mining hotspots are located in particu
and Williams, 2008). larly close proximity to protected areas. Moreover, Lumwana mine with
For Brazil, we find substantial extraction growth in the state of Minas an average annual growth coefficient of 7.16% lies partly within Acres
Gerais (up to 5% p.a.), while mining regions outside Minas Gerais No. 105 National Forest.
stagnated on average. Four out of the five Brazilian mines with the The SNL (2020) data supports that metal mining has become a major
highest growth coefficients are located no farther than 5 km away from industry in India. Growth rates of metal ore extraction such as bauxite
natural reserves (e.g. Serra do Gandarela National Park). Furthermore, and iron ore are exceptionally high in the eastern territories and the
many projects of that hotspot are large-scale and situated in species-rich northwest. Yet, GWR results stress substantial extraction growth for
tropical biomes. One of the most serious threats to the environment and some of the country’s less prominent (and smaller in size) mining pro
people in that region are tailings dam failures of large mining projects. jects. While India’s largest projects, such as the Chhattisgarh Group,
Vale’s Mina Córrego do Feijão, for example, gained notoriety due to the Sesa Goa, and Noamundi iron mine, extract between 15 Mt and 25 Mt of
dam disaster near the municipality of Brumadinho in January 2019, metal ore per year (and results yield coefficients between 2% and 9%),
releasing around 12 million cubic meters of tailings and killing at least average annual ore extraction for the mines listed in Table 1 lies between
259 people (Freitas and Almeida, 2020). Brazil has hundreds of such 0.01 and 1.3 Mt. Extraction volumes of these iron ore and bauxite sites,
tailings dams, most of them in the state of Minas Gerais (Fonseca do however, grow at remarkably swift rates of about 10%, i.e. the highest
Carmo et al., 2017). Brucutu mine, extracting almost 18,000 Mt ore per rates observed globally. They are located in the tropical East Deccan
year on average and with an average extraction intensification of 5.2%, moist deciduous forests, which offer a spacious and rich habitat for a
is such an example, and it is located in proximity to the previous tailings great number of species, including endangered large vertebrates (Wik
dam failure. In 2019, as a response to the Brumadinho dam disaster, ramanayake et al., 2002). Extensive extraction gains in this area, as they
disposal of tailings at the Laranjeiras dam, part of Brucutu mine, was are observed, hence endanger conservation of a still-intact habitat.
temporarily suspended due to safety concerns. In November 2015, In the Global North, trend estimates for mines located in the Unites
Minas Gerais had already experienced a major environmental accident, States and eastern Canada indicate an average decline of mining activ
when two dams at Samarco minining complex (4.9% average annual ities, while slight intensification is found for the Canadian Rocky
9
Table 1
Mines selected by highest five coefficients (β) from GWR (average growth p.a.) per hotspot country. Dist. PA indicates distance to nearest protected area, μ indicates
average annual ore extraction based on SNL (2020) and UN IRP (2017) conversion factors. Empty cells in biome and AWARE column indicate constant characteristic
for all five mines.
Country Mine β [%] Biome Dist. PA [km] AWARE μ [kt]
Peru Ares 7.27 Montane Grasslands & Shrublands 19.80 28.27 1529.51
Santa Rosa 7.19 Deserts & Xeric Shrublands 20.18 95.54
Orcopampa 7.18 Deserts & Xeric Shrublands 21.52 2639.23
Arcata 7.11 Montane Grasslands & Shrublands 4.89 729.17
Shila-Paula 7.08 Montane Grasslands & Shrublands 53.93 295.81
Brazil Brucutu 5.22 Trop. & Subtrop. Moist Broadleaf Forests 14.69 3.14 17958.33
Sao Bento 5.12 Trop. & Subtrop. Moist Broadleaf Forests 4.08 367.65
Gongo Soco 5.05 Trop. & Subtrop. Moist Broadleaf Forests 1.49 5464.29
Samarco 4.87 Trop. & Subtrop. Grasslands, Savannas & Shrublands 1.92 17827.88
Alegria 4.84 Trop. & Subtrop. Grasslands, Savannas & Shrublands 3.05 12720.52
DR Congo Kamoto 7.21 Trop. & Subtrop. Grasslands, Savannas & Shrublands 18.15 1.67 2762.92
Metalkol RTR 7.21 22.47 1846.24
Mutoshi 7.20 33.32 248.28
Kolwezi 7.20 24.66 1359.43
Tilwizembe 7.16 48.99 48.87
Zambia Trident - Sentinel 7.42 Trop. & Subtrop. Grasslands, Savannas & Shrublands 5.22 7.43 16301.33
Lumwana 7.16 0.00 10431.14
Kansanshi 6.85 10.77 21100.96
Muliashi North 5.80 5.72 3224.72
Baluba 5.78 4.50 1621.13
India Malangtoli 10.83 Trop. & Subtrop. Moist Broadleaf Forests 59.15 16.90 573.33
Silijora-Kalimati 10.71 50.01 87.10
Unchabali 10.70 45.90 1284.75
Dubna 10.62 50.85 11.86
Khondbond 10.62 44.61 748.19
Australia Spinifex Ridge 9.14 Deserts & Xeric Shrublands 88.89 27.19 765.14
Corunna Downs 9.13 120.61 512.00
Bamboo Creek 9.11 79.34 5.12
Nullagine 9.06 147.96 1155.92
Mt Webber DSO 8.97 78.24 3783.89
Mountains. For Europe, we find almost entirely positive trend co 80% of global metal extraction in 2019 occurred in the world’s most
efficients, indicating growing metal extraction volumes within Europe, species-rich biomes, 90% of mining sites were in areas of relative water
although at generally low levels of absolute extraction. A geographical scarcity, and almost 50% of extraction occurred at less than 20 km
divide is apparent in Australia. Large iron ore sites in Western Australia distance or even within protected areas. By highlighting that these are
are the drivers for intensification rates up to almost 10% and make the especially vulnerable areas, it is not our intention to suggest that mining
region a global hotspot, while sites across the eastern half of the conti should be expanding “elsewhere”, in areas that are supposedly less
nent show an average decline by about 2.5% per year. Highest co vulnerable. Instead, we interpret the degree to which mining occurred in
efficients are reported for mines in Pilbara, which hosts some of the areas that - by indirect or direct stipulation - should be protected to
world’s largest iron ore mines. Among others, Mount Webber direct demonstrate the environmental unsustainability of the current mode of
shipping iron ore mine is located at that hotspot, with a growth coeffi expansion.
cient of almost 9%. The large mining complex, opened in 2014, The findings of this paper contribute to the literature by offering
continuously increased annual extraction volumes and mined more than annual estimates and trends for the nine metals at hand, but they also
7 Mt of ore in 2019. Australian mining activities also demonstrate that compare well with previous, often cross-sectional, studies. Northey et al.
the movement of vast amounts of earth not only causes pollution and (2017), for example, also find highest average AWARE scores for copper,
degradation of ecosystems and loss of biodiversity, but also destroys medium water risks for lead–zinc resources, and nickel to be mined
cultural and spiritual heritage. The mining expansion in Pilbara, for predominantly in areas that are exposed to less water risks. While the
example, disturbs significant Indigenous sites. In May 2020, Anglo- shares of copper, lead and zinc mining in critical categories are stable,
Australian multinational Rio Tinto was blamed for destroying the we do, however, detect a recent surge in nickel mining for higher
Aboriginal heritage of 46,000 year old Juukan Gorge rock shelters in AWARE scores. Durán et al. (2013) find that 7% of mines in their sample
order to expand Brockman 4 mine (EJatlas, 2020). overlap with protected areas and 27% lie within a 10 km boundary. Our
findings are slightly more conservative, but just as alarming. We
4. Discussion furthermore complement the findings of Durán et al. (2013) by the
observation that extraction volumes have considerably increased within
4.1. Mining’s socio-ecological impacts protected areas since 2000. A comparison with Murguía et al. (2016)
shows that results are sensitive to the choice how to proxy biodiversity.
In contrast to other frameworks that conceptualise the socio- Our findings are in line with their study regarding bauxite, while our
ecological crisis and possible responses at high levels of aggregation approach does not support their conclusions that silver mining showed a
and abstraction (e.g. Rockström et al., 2009), our results demonstrate high concentration in ”high diversity zones” (ibid: 416).
concretely where, from 2000 to 2019, the surge in global mining has The global approach applied in this study also entails some un
been implemented. What we demonstrate in particular is that increased certainties and limitations. Using crude ore extraction estimates instead
production occurs to a large extent in areas requiring protection. Almost of often reported net-metal contents more accurately quantifies
10
pressures that are exerted by the mining industry on the environment which is evidently affecting protected lands in many regions (see, e.g.,
and hence our estimates also serve as better indicators with regard to Asner and Tupayachi, 2016). However, our results suggest that public
potential impacts. A drawback to our crude ore approach is that as resistance against mining operations may rise due to increased pro
sumptions on average ore grades place substantial variances around duction within and close to protected areas and we hence highlight the
these estimates. However, we argue that such uncertainties affect the need for investigating and monitoring social dynamics around such
conclusions of this work less, because we can assume high correlation areas.
between our estimates and actual extraction volumes from conducted
robustness checks. Introducing mine-specific conversion factors can thus 4.2. Implications for mining companies and policies
be assumed to have level effects on our global-scale results rather than
causing substantial changes of our findings on the distributional struc Accelerated global extraction of metals and minerals particularly
ture and trends of global mining activities. threatens vulnerable ecosystems and selected regions emphasised in this
Further, by keeping a global scope while utilising broad categories of study. To reduce associated risk in the short and medium term, the
associated vulnerability, we provide only a crude assessment about impacts of mining itself need to decrease. There is no doubt that the
where environmental impacts may be more severe than elsewhere. This mining industry has already developed and implemented improvements
approach is beneficial as it serves as an early warning mechanism in environmental management processes and impact mitigation systems,
through a trend analysis of the global mining sector, helping to identify such as progressive rehabilitation throughout the life cycle of a mine.
potential high-impact areas. Nevertheless, we need to stress that we can While considerable advancements were also made in regulatory systems
only illustrate threats to the environment in terms of potential impacts or (e.g. cumulative environmental impact assessment policies and
a contextual risk. Estimating actual impacts would require more locally improved regional planning), environmental minimum standards and
specific data such as actual mine practices and more precise information better sustainability practises and performance must be realised at the
on (changes in) the mines’ immediate surroundings. Species-richness, mine sites (IRMA, 2020) and above and beyond corporate level com
for instance, serves as broad indicator, while biodiversity and hence mitments, as e.g. demanded by the International Council on Mining and
also impacts are heterogeneous across space within the same biome. Metals (ICMM, 2020).
Our study thus provides signposts to where more in-depth local Evaluating, challenging and improving this ongoing transition also
studies are needed that consider the interplay of regional circumstances includes critically reviewing and reforming national environmental
at the site- and case level. On the one hand, we support efforts to eval regulations. Our work contributes to such a discourse by pinpointing
uate the impacts of the global surge in mining based on investigations regions with expansive dynamics and alarming developments such as a
that are tied to specifically selected cases. On the other hand, there is surge in extraction within protected areas. These insights could be used
great need for more quantitative studies on the magnitude of impacts for more accurately targeting regional policies. As shown in Section 3.4,
induced by the entire mining industry. Any expansion of mining con areas with highest extraction growth rates are, with the exception of
stitutes a trade-off for other human and non–human uses or values Australia, located in low- and middle-income countries of the Global
attached to land and resources. More accurate assessments of mining South including Brazil, the DR Congo, India and China. These countries
projects’ propensities to exert additional pressures such as biodiversity score lower in the OECD Environmental Policy Stringency Index as
loss, deforestation, water and air pollution as well as social conflicts compared to industrialised countries (OECD, 2021). However, accord
could help constructing spatially varying impact measures. These in ing to this composite measure, nations such as China and India have put
dicators could then be used for global assessments of mining impacts and significant efforts in improving environmental standards, while the
industry monitoring. Importantly, more knowledge is needed about the index stagnates at very low levels for Brazil. The Central African Cop
nature and the extent of the spatial transmission and the temporal perbelt also marks a challenging region with substantial room for im
persistence of mining-induced environmental change. Such improve provements at the policy and company level. Even though forms of
ments will help not only to anticipate the consequences of increased protective laws and regulations were established in the 1990s in Zambia
metal production, but also to evaluate current progress towards more and in the 2000s in the DR Congo, this change in environmental man
sustainable technologies and better regulations, as well as to better agement practices remained rather a change on paper, while mining
consider competing stakeholder interests. companies intensified extractive operations based on a dominantly
Our research points mostly towards the potential environmental economic and technocratic rationale (Peša, 2020). Concrete opportu
impacts of mining expansion where it occurs. Simultaneously, desig nities for action include rethinking mining governance such that it
nating land (and resources in a much wider sense) to be used for mining avoids unnecessary large-scale infrastructure, averts opening up un
precludes many other human uses or the non-use of that land. Envi touched spaces to settlement, considers cumulative impacts across space
ronmental justice movements opposing mining have raised numerous (such as watershed regions) and over time and involves most affected
issues from the loss of land and water for subsistence uses to local air and populations in the decision making (Bebbington et al., 2020). National
water pollution and the destruction of cultural values (Martinez-Alier, and sub-national governments are integral parts of the International
2001; Temper et al., 2015). The categories we utilised may often proxy Resource Panel’s ”Sustainable Development Licence to Operate”
the risk of social disruption just as much as they indicate environmental framework (UN IRP, 2020), which makes a strong case for policy
vulnerability. Conflicting use and pollution of water is a frequent cause coherence along multiple levels that is grounded on robust laws and
for resistance of local communities against mining projects, even though regulations: National governments have the opportunity to define broad
mining operators insist that technologies would guarantee sustainable national development goals and to require mining activities being
use of water resources and concrete water monitoring plans would aligned to these. They can do so, for instance, through the use of
already be in place at major mining projects (Bebbington and Williams, auctioning, given that government policy objectives are clarified and
2008). Geological conditions determine that the deposits of some metals made publicly available well in advance of the auction. Sub-national and
are predominantly mined in areas where water is scarce. Our study il local governments, in turn, have the ability to actively collaborate in
lustrates the massive extraction of copper and iron ores in desert eco local development planning and to steer negotiations regarding
systems and areas of below-average water availability, which inevitably trade-offs between the environmental, economic and social dimensions
raises questions about potential conflicting uses of water and strategies of mining operations.
to prevent future social disruptions in affected regions. Similarly, pro The shared knowledge as to the distribution and the contextual risk
tected areas secure livelihoods and cultural values for many indigenous of increased mining, to which we have contributed with our research,
populations. We did not distinguish between indigenous lands and other turns the decision to expand mining anywhere into an informed decision
protected areas in this study and we did not consider informal mining, to inflict negative impacts on the environment and human communities.
11
Any expansion therefore arguably requires tight governance of which that indirect effects distant from mines may be extensive. We found that
metals can be extracted, not just when and where, but also for what the intensification of extraction shows distinct regional patterns, which
purpose and by what means. One possible solution how both of these increase the pressure on vulnerable ecosystems in several biomes across
aspects could be controlled better might be more vertical integration of the globe. In order to preserve both livelihoods and habitats to many,
mining in global supply chains, making these less complex and more often rare and endangered species, particularly tropical ecosystems of
transparent, which should be advantageous to consumer facing down vast biodiversity require stronger protection from interference through
stream companies, e.g. in the automotive or electronics sectors. mining. The increase of metal mining in vulnerable and protected areas
shown in our study points to the challenge of reversing current unsus
4.3. The global drivers of accelerated mining tainable trends in resource extraction.
Metals are point resources and do not occur ubiquitously. Rather
While our results have clear implications for mining companies and than implying that there is no choice but to mine them where they do
related regulation, the global framework in which mining actors operate occur, we have argued that this in fact supports an even stronger case for
needs to be equally considered, when discussing options to reduce the reducing resource consumption, first and foremost of the world’s
socio-ecological impacts of mining. Given that any expansion of mining wealthiest economies, in order to protect vulnerable ecosystems and
has detrimental environmental and social impacts, it seems straight their inhabitants. Further pursuing this agenda can be supported by the
forward to call for a halt to mining expansion, as has, in fact, been the type of information we have sought to develop for this article, including
request throughout various environmental justice movements (e.g. the spatially explicit mapping, the historical contextualisation and the
Temper et al., 2015). However, this claim is in stark contrast to expected assessment of the status quo of resource development and its trans
future trends. Global demand for metal ores is expected to significantly formative potential. Aggregate global conceptualisations and targets
increase in the coming decades (Elshkaki et al., 2018) mostly due to the must be integrated with in-depth knowledge of local-level consequences
build-up of global material stocks (Krausmann et al., 2017, 2020) and of mining expansion. For the examples of selected hotspots of mining
the expansion of low-carbon infrastructures, such as wind and solar expansion, we demonstrated that many of them are unambiguously
energy and battery storage capacities (Elshkaki and Shen, 2019; Watari related to local socio-environmental risk and disasters.
et al., 2020). It will be impossible to sustain increasing levels of con Further investigating and monitoring the spatial and temporal evo
sumption in those areas while simultaneously curbing the negative lution of metal mining can serve as an early warning mechanism and
environmental and social impacts of metal mining. will help to anticipate potentially hazardous developments and better-
The increasing metabolic inequalities of current growth trajectories inform mining management and policy making. Our results have im
also play an important role. Global supply and use chains currently plications for the way we organise the biophysical basis of our economic
direct the additional resources gained by metal mining to places of systems, because they underline that reoccurring local ecological dis
already high or rapidly increasing consumption and material wealth tribution conflicts all across the globe are not to be solved at the case
(Dorninger et al., 2021). High-income and some middle-income econo level. Instead, they are consequences of an expansion systematically
mies engage in net-appropriation of raw materials and without net- affecting species-rich, water-scarce, complex, fragile and hence vulner
imports would not be able to pursue their models of industrialised able ecosystems.
growth. While the mature industrialised economies have largely
exhausted their domestic resource base, countries in the earlier phases of CRediT authorship contribution statement
capitalist industrialisation continue to hinge their economic “develop
ment” on extractivist agendas (Gudynas, 2010), taking on roles of global Sebastian Luckeneder: Conceptualization, Methodology, Software,
suppliers of primary commodities while foregoing the higher value Formal analysis, Writing - original draft, Writing - review & editing.
added associated with refinement and manufacturing (UNCTAD, 2019). Stefan Giljum: Conceptualization, Writing - original draft, Writing -
As a consequence, final consumers are in many cases geographically review & editing. Anke Schaffartzik: Writing - original draft, Writing -
distant from resource extraction and the related impacts (Schaffartzik review & editing. Victor Maus: Methodology, Software, Writing - re
et al., 2016; Gudynas, 2010). Due to these “tele-connections” between view & editing. Michael Tost: Writing - original draft, Writing - review
production and consumption, the systemic character behind ecological & editing.
distribution conflicts needs to be addressed from a global perspective,
acknowledging pertinent patterns of ecologically unequal exchange
(Dorninger et al., 2021; Xu et al., 2020). Declaration of Competing Interest
5. Conclusion The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
To date, there is no reason to expect the expansion of metals and the work reported in this paper.
minerals extraction to halt in the near future. In contrast, the accelerated
build-up of global material stocks and the development of new and Acknowledgements
supposedly more sustainable technologies will create growing markets
for metal ores. The authors thank Nikolas Kuschnig for valuable comments and
In this study, we investigated and contextualised metal ore extrac discussions on a previous version of this manuscript, Jakob Gutschlhofer
tion of the past two decades. We illustrated the types and amounts of and Mirko Lieber for their support in compiling the data, and two
commodities extracted, along with a detailed assessment of their anonymous reviewers, who helped to improve this paper. This work was
geographical location. Backed by the rich empirical evidence that funded by the European Commission under the ERC Consolidator Grant
mining activities induce hazardous changes to the environment, we FINEPRINT (Grant Number 725525). Anke Schaffartzik acknowledges
considered areas around mining sites to be more strongly at risk and the financial support of the Austrian Science Fund (FWF) through Her
reflected how severe mining pressures may impact ecosystems, in tha Firnberg project T949.
particular those already recognisable as vulnerable. It is remarkable
that, compared to the total global surface, relatively few and small areas Appendix A. Supplementary data
supply the metallic basis for the entire industrialised world. However,
based on what we know so far about the impacts of minerals extraction Supplementary data associated with this article can be found in the
on the environment and their spatial transmission, it seems highly likely online version at https://doi.org/10.1016/j.gloenvcha.2021.102303.
12
References Martinez-Alier, J., 2001. Mining conflicts, environmental justice, and valuation.
J. Hazard. Mater. 86 (1), 153–170. https://doi.org/10.1016/S0304-3894(01)00252-
7.
Asner, G.P., Llactayo, W., Tupayachi, R., Ráez Luna, E., 2013. Elevated rates of gold
Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well-being:
mining in the Amazon revealed through high resolution monitoring. Proc. Natl.
Biodiversity Synthesis. World Resources Institute, Washington, DC. https://www.
Acad. Sci. U.S.A. 110 (46), 18454–18459. https://doi.org/10.1073/
millenniumassessment.org/documents/document.354.aspx.pdf.
pnas.1318271110.
Maus, V., Giljum, S., Gutschlhofer, J., et al., 2020. A global-scale data set of mining areas.
Asner, G.P., Tupayachi, R., 2016. Accelerated losses of protected forests from gold
Sci. Data 7 (289). https://doi.org/10.1038/s41597-020-00624-w.
mining in the Peruvian Amazon. Environ. Res. Lett. 12 (9), 094004 https://doi.org/
Mudd, G.M., Jowitt, S.M., 2018. Growing Global Copper Resources, Reserves and
10.1088/1748-9326/aa7dab.
Production: Discovery Is Not the Only Control on Supply. Econ. Geol. 113 (6),
Ali, S.H., Giurco, D., Arndt, N., et al., 2017. Mineral supply for sustainable development
1235–1267. https://doi.org/10.5382/econgeo.2018.4590.
requires resource governance. Nature 543 (7645), 367–372. https://doi.org/
Murguía, D.I., Bringezu, S., Schaldach, R., 2016. Global direct pressures on biodiversity
10.1038/nature21359.
by large-scale metal mining: Spatial distribution and implications for conservation.
Bebbington, A., Chicchon, A., Cuba, N., et al., 2020. Opinion: Priorities for governing
J. Environ. Manage. 180, 409–420. https://doi.org/10.1016/j.
large-scale infrastructure in the tropics. Proc. Nat. Acad. Sci. 117 (36),
jenvman.2016.05.040.
21829–21833. https://doi.org/10.1073/pnas.2015636117.
Mwitwa, J., German, L., Muimba-Kankolongo, A., Puntodewo, A., 2012. Governance and
Bebbington, A., Hinojosa, L., Humphreys Bebbington, D., Burneo, M.L., Warnaars, X.,
sustainability challenges in landscapes shaped by mining: Mining-forestry linkages
2008. Contention and Ambiguity: Mining and the Possibilities of Development. Dev.
and impacts in the Copper Belt of Zambia and the DR Congo. For. Policy Econ. 25,
Change 39 (6), 887–914. https://doi.org/10.1111/j.1467-7660.2008.00517.x.
19–30. https://doi.org/10.1016/j.forpol.2012.08.001.
Bebbington, A., Williams, M., 2008. Water and Mining Conflicts in Peru. Mt. Res. Dev. 28
Northey, S.A., Mudd, G.M., Werner, T.T., et al., 2017. The exposure of global base metal
(3), 190–195. https://doi.org/10.1659/mrd.1039.
resources to water criticality, scarcity and climate change. Global Environ. Change
Birkmann, J., Wisner, B., 2006. Measuring the unmeasurable: the challenge of
44, 109–124. https://doi.org/10.1016/j.gloenvcha.2017.04.004.
vulnerability. UNU Institute for Environment and Human Security, Bonn.
OECD, 2019. Global Material Resources Outlook to 2060. OECD, Paris. https://doi.org/
Bridge, G., 2004. Contested Terrain: Mining and the Environment. Annu. Rev. Environ.
10.1787/9789264307452-en.
Resour. 29 (1), 205–259. https://doi.org/10.1146/annurev.
OECD, 2021. Environmental Policy Stringency Index. Accessed: 23 Feb 2021. Paris:
energy.28.011503.163434.
OECD. https://stats.oecd.org/Index.aspx?DataSetCode=EPS.
Brunsdon, C., Fotheringham, A.S., Charlton, M.E., 1996. Geographically Weighted
Peša, Iva, 2020. Mining, Waste and Environmental Thought on the Central African
Regression: A Method for Exploring Spatial Nonstationarity. Geog. Anal. 28 (4),
Copperbelt, 1950–2000. Environment and History (pre-print). https://doi.org/
281–298. https://doi.org/10.1111/j.1538-4632.1996.tb00936.x.
10.3197/096734019X15755402985703.
Caballero Espejo, J., Messinger, M., Román-Dañobeytia, F. et al., 2018. Deforestation and
Resolve. http://ecoregions2017.appspot.com.
Forest Degradation Due to Gold Mining in the Peruvian Amazon: A 34-Year
Rockström, Johan, Steffen, Will, Noone, Kevin, et al., 2009. A safe operating space for
Perspective. In: Remote Sensing 10.12, p. 1903. https://doi.org/10.3390/rs
humanity. Nature 461, 472–475. https://doi.org/10.1038/461472a.
10121903.
Salem, Jorge, Amonkar, Yash, Maennling, Nicolas et al., 2018. An analysis of Peru: Is
Conde, M., 2017. Resistance to Mining. A Review. Ecol. Econ. 132, 80–90. https://doi.
water driving mining conflicts? Resources Policy. In press. https://doi.org/10.1016/
org/10.1016/j.ecolecon.2016.08.025.
j.resourpol.2018.09.010.
Dinerstein, E., Olson, D., Joshi, A., et al., 2017. An Ecoregion-Based Approach to
SNL, 2020. Metals and Mining Database. S&P Global Market Intelligence, New York.
Protecting Half the Terrestrial Realm. Bioscience 67 (6), 534–545. https://doi.org/
Schaffartzik, A., Mayer, A., Eisenmenger, N., Krausman, F., 2016. Global patterns of
10.1093/biosci/bix014.
metal extractivism, 1950–2010: Providing the bones for the industrial society’s
Dorninger, C., Hornborg, A., Abson, D.J., et al., 2021. Global patterns of ecologically
skeleton. Ecol. Econ. 122, 101–110. https://doi.org/10.1016/j.
unequal exchange: Implications for sustainability in the 21st century. Ecol. Econ.
ecolecon.2015.12.007.
179, 106824 https://doi.org/10.1016/j.ecolecon.2020.106824.
Schaffartzik, A., Mayer, A., Gingrich, S., et al., 2014. The global metabolic transition:
Durán, A.P., Rauch, J., Gaston, K.J., 2013. Global spatial coincidence between protected
Regional patterns and trends of global material flows, 1950–2010. Global Environ.
areas and metal mining activities. Biol. Conserv. 160, 272–278. https://doi.org/
Change 26, 87–97. https://doi.org/10.1016/j.gloenvcha.2014.03.013.
10.1016/j.biocon.2013.02.003.
Schandl, H., Fischer-Kowalski, M., West, J., et al., 2017. Global Material Flows and
EJatlas, 2020. Rio Tinto/BHP’s iron mining destroys sacred Aboriginal sites, Western
Resource Productivity: Forty Years of Evidence. J. Ind. Ecol. 22, 827–838. https://
Australia. In: Atlas of Environmental Justice. Accessed: 3 Dec 2020. https://ejatlas.
doi.org/10.1111/jiec.12626.
org/conflict/rio-tintos-iron-mine-and-destruction-of-sacred-aboriginal-sites-australi
Scheidel, A., Del Bene, D., Liu, J., et al., 2020. Environmental conflicts and defenders: A
a.
global overview. Global Environ. Change 63, 102104. https://doi.org/10.1016/j.
Elshkaki, A., Shen, L., 2019. Energy-material nexus: The impacts of national and
gloenvcha.2020.102104.
international energy scenarios on critical metals use in China up to 2050 and their
Sonter, L., Ali, S.H., Watson, J.E.M., 2018. Mining and biodiversity: key issues and
global implications. Energy 180, 903–917. https://doi.org/10.1016/j.
research needs in conservation science. Proc. R. Soc. B 285, 285. https://doi.org/
energy.2019.05.156.
10.1098/rspb.2018.1926.
Elshkaki, A., Graedel, T.E., Ciacci, L., Reck, B.K., 2018. Resource Demand Scenarios for
Sonter, L.J., Dade, M.C., Watson, J.E.M., et al., 2020. Renewable energy production will
the Major Metals. Environ. Sci. Technol. 52 (5), 2491–2497. https://doi.org/
exacerbate mining threats to biodiversity. Nat. Commun. 11 (1), 1–6. https://doi.
10.1021/acs.est.7b05154.
org/10.1038/s41467-020-17928-5.
Freitas, R. and Almeida, F., 2020. Um ano após tragédia da Vale, dor e luta por
Sonter, L.J., Herrera, D., Barrett, D.J., et al., 2017. Mining drives extensive deforestation
justiçaunem famílias de 259 mortos e 11 desaparecidos. In: Globo. Accessed: 3 Dec
in the Brazilian Amazon. Nat. Commun. 8 (1), 1013. https://doi.org/10.1038/
2020. https://g1.globo.com/mg/minas-gerais/noticia/2020/01/25/um-ano-apos-
s41467-017-00557-w.
tragedia-da-vale-dor-e-luta-por-justica-unem-familias-de-259-mortos-e-11-desapare
Steffen, W., Broadgate, W., Deutsch, L., et al., 2015. The trajectory of the Anthropocene:
cidos.ghtml.
The Great Acceleration. The Anthropocene Review 2 (1), 81–98. https://doi.org/
Fonseca do Carmo, F., Kamino, L.H.Y., Tobias Junior, R., et al., 2017. Fundão tailings
10.1177/2053019614564785.
dam failures: the environment tragedy of the largest technological disaster of
Supriatna, J., Shekelle, M., Fuad, H.A.H., et al., 2020. Deforestation on the Indonesian
Brazilian mining in global context. Perspect. Ecol. Conserv. 15 (3), 145–151. https://
island of Sulawesi and the loss of primate habitat. Global Ecol. Conserv. 24, e01205
doi.org/10.1016/j.pecon.2017.06.002.
https://doi.org/10.1016/j.gecco.2020.e01205.
Gibson, C.C., Ostrom, E., Ahn, T.K., 2000. The concept of scale and the human
Temper, L., Del Bene, D., Martinez-Alier, J., 2015. Mapping the frontiers and front lines
dimensions of global change: a survey. Ecol. Econ. 32 (2), 217–239. https://doi.org/
of global environmental justice: the EJAtlas. J. Political Ecol. 22 (1), 255–278.
10.1016/S0921-8009(99)00092-0.
https://doi.org/10.2458/v22i1.21108.
Görg, C., Plank, C., Wiedenhofer, D., et al., 2020. Scrutinizing the Great Acceleration: the
Tost, M., Murguia, D., Hitch, M., et al., 2020. Ecosystem services costs of metal mining
Anthropocene and its analytic challenges for social-ecological transformations.
and pressures on biomes. The Extractive Industries and Society 7 (1), 79–86. https://
Anthropocene Rev. 7 (1), 42–61. https://doi.org/10.1177/2053019619895034.
doi.org/10.1016/j.exis.2019.11.013.
Gudynas, E., 2010. The new extractivism of the 21st century: Ten urgent theses about
Turner, B.L., Kasperson, R.E., Matson, P.A., et al., 2003. A framework for vulnerability
extractivism in relation to current South American progressivism. Americas Program
analysis in sustainability science. Proc. Nat. Acad. Sci. 100 (14), 8074–8079. https://
Report 21, 1–14.
doi.org/10.1073/pnas.1231335100.
ICMM, 2020. International Council on Mining and Metals: Mining Principles. Accessed: 3
UN IRP, 2017. Global Material Flows Database: Version 2017. United Nations
Dec 2020. http://www.icmm.com/mining-principles.
Environment Programme, Paris. http://www.resourcepanel.org/global-mate
IRMA, 2020. Initiative for Responsible Mining Assurance. Accessed: 3 Dec 2020. https
rial-flows-database.
://responsiblemining.net/.
UN IRP, 2019. Global resources outlook 2019: Natural resources for the future we want.
Krausmann, F., Wiedenhofer, D., Haberl, H., 2020. Growing stocks of buildings,
United Nations Environment Programme, Nairobi. https://www.resourcepanel.org/r
infrastructures and machinery as key challenge for compliance with climate targets.
eports/global-resources833outlook.
Global Environ. Change 61, 102034. https://doi.org/10.1016/j.
UN IRP, 2020. Mineral Resource Governance in the 21st Century: Gearing extractive
gloenvcha.2020.102034.
industries towards sustainable development. United Nations Environment
Krausmann, F., Wiedenhofer, D., Lauk, C., et al., 2017. Global socioeconomic material
Programme, Nairobi. https://www.resourcepanel.org/reports/mineral-resource
stocks rise 23-fold over the 20th century and require half of annual resource use.
-governance-21st-century.
Proc. Natl. Acad. Sci. 114 (8), 1880–1885. https://doi.org/10.1073/
UNCTAD, 2019. State of Commodity Dependence 2019. New York: United Nations
pnas.1613773114.
Conference on Trade and Development. https://unctad.org/en/pages/Publication
Webflyer.aspx?publicationid=2439.
13
UNEP-WCMC, 2020. Protected Planet: The World Database on Protected Areas 03/2020. Watari, T., Nansai, K., Nakajima, K., 2020. Review of critical metal dynamics to 2050 for
UNEP-WCMC and IUCN, Cambridge. https://www.protectedplanet.net. 48 elements. Resour. Conserv. Recycl. 155, 104669 https://doi.org/10.1016/j.
UNESCO, 2020. Lorentz National Park. Accessed: 4 Dec 2020. United Nations resconrec.2019.104669.
Educational, Scientific and Cultural Organization, Paris. https://whc.unesco.org/en/ Weißhuhn, P., Müller, F., Wiggering, H., 2018. Ecosystem Vulnerability Review: Proposal
list/955/. of an Interdisciplinary Ecosystem Assessment Approach. Environ. Manage. 61 (6),
Vale (2016). Vale inaugurates the largest project in the history of the mining industry. 904–915. https://doi.org/10.1007/s00267-018-1023-8.
Accessed: 4 Dec 2020. http://saladeimprensa.vale.com/en/Paginas/Articles.aspx? Werner, T.T., Mudd, G.M., Schipper, A.M., et al., 2020. Global-scale remote sensing of
r=Vale_inaugurates_the_largest_project_in_the_history__of_the_mining_industry_ mine areas and analysis of factors explaining their extent. Global Environ. Change
&s=Mining&rID=979&sID=6. 60, 102007. https://doi.org/10.1016/j.gloenvcha.2019.102007.
Government, WA, 2009. Pilbara water in mining guideline. Accessed: 4 Dec 2020. Wikramanayake, E., Dinerstein, E., Loucks, C.J., et al., 2002. Terrestrial Ecoregions of the
Government of Western Australia, Department of Water, Perth. https://www.water. Indo-Pacific: a Conservation Assessment. Island Press, Washington, DC.
wa.gov.au/__data/assets/pdf_file/0013/1570/88526.pdf. Xu, Z., Li, Y., Chau, S.N., et al., 2020. Impacts of international trade on global sustainable
WULCA. http://www.wulca-waterlca.org. development. Nature Sustain. https://doi.org/10.1038/s41893-020-0572-z.
14
View publication stats

Luckeneder Et Al 2021

Uploaded by

Copyright:

Available Formats

Luckeneder Et Al 2021

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Luckeneder Et Al 2021

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Surge in global metal mining threatens vulnerable ecosystems

Article in Global Environmental Change · July 2021

Sebastian Luckeneder Stefan Giljum

SEE PROFILE SEE PROFILE

Anke Schaffartzik Victor Maus

SEE PROFILE SEE PROFILE

Prosperous Pathways in the EU Bioeconomy (BIOWAY) View project

The user has requested enhancement of the downloaded file.

Contents lists available at ScienceDirect

Global Environmental Change

Surge in global metal mining threatens vulnerable ecosystems

2.1. Mining data and environmental spatial layers

View publication stats

You might also like