Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 51
na u e ecology & e olu ion
h ps://doi.o g/10.1038/s41559-022-01934-5A icle
Mapping he plane ’s c i ical na u al asse s
Rebecca Chaplin-K ame 1,2,3 , Rachel A. Neuga en 4,5, Richa d P. Sha p1,
Pamela M. Collins5, S ephen Polasky 6, Da id Hole 5, Richa d Schus e 7,8,
Ma hew S imas-Mackey 9, Ma k Mulligan10, Ca e B andon11,
Sand a Diaz 12,13, E ienne Flue -Chouina d 14, L. J. Go en lo 15,
Jus in A. Johnson6, Ch is ina M. Kennedy 16, Pa ick W. Keys 17,
Ka e Longley-Wood 18, Pe e B. McIn y e4, Monica Noon5, Unai Pascual 19,20,21,
Ca he ine Reidy Lie mann22, Pa ick R. Roeh danz 5, Guido Schmid -T aub23,
M. Rebecca Shaw24, Ma k Spalding18,25, Will R. Tu ne 5,
A nou an Soesbe gen 10,26 & Reg A. Wa son 27
Sus aining he o ganisms, ecosys ems and p ocesses ha unde pin human
wellbeing is necessa y o achie e sus ainable de elopmen . He e we
de ine c i ical na u al asse s as he na u al and semi-na u al ecosys ems
ha p o ide 90% o he o al cu en magni ude o 14 ypes o na u e’s
con ibu ions o people (NCP), and we map he global loca ions o
hese c i ical na u al asse s a 2 km esolu ion. C i ical na u al asse s o
main aining local-scale NCP (12 o he 14 NCP) accoun o 30% o o al
global land a ea and 24% o na ional e i o ial wa e s, while 44% o land
a ea is equi ed o also main ain wo global-scale NCP (ca bon s o age and
mois u e ecycling). These a eas o e lap subs an ially wi h cul u al di e si y
(a eas con aining 96% o global languages) and biodi e si y (co e ing a ea
equi emen s o 73% o bi ds and 66% o mammals). A leas 87% o he
wo ld’s popula ion li e in he a eas bene i ing om c i ical na u al asse s
o local-scale NCP, while only 16% li e on he lands con aining hese asse s.
Many o he NCP mapped he e a e le ou o in e na ional ag eemen s
ocused on conse ing species o mi iga ing clima e change, ye his
analysis shows ha explici ly p io i izing c i ical na u al asse s and he
NCP hey p o ide could simul aneously ad ance de elopmen , clima e and
conse a ion goals.
Human ac ions a e apidly ans o ming he plane , d i ing losses o
na u e a an unp eceden ed a e ha nega i ely impac s socie ies and
economies, om accele a ing clima e change o inc easing zoono ic
pandemic isk
1,2
. Recognizing he accele a ing se e i y o he en i on-
men al c isis, he global communi y commi ed o Sus ainable De el-
opmen Goals and he Pa is Ag eemen on clima e change in 2015. In
2022, he UN Con en ion on Biological Di e si y (CBD) will adop new
a ge s o conse ing, es o ing and sus ainably managing mul iple
dimensions o biodi e si y, including na u e’s con ibu ions o peo
-
ple (NCP)3. Collec i ely, hese h ee policy amewo ks will shape he
sus ainable de elopmen agenda o he nex decade. All h ee depend
hea ily on sa egua ding na u al asse s
1,4
, he li ing componen s o ou
lands and wa e s. Fo ins ance, es o ing and ending con e sion and
deg ada ion o o es s, we lands and pea lands could seques e 9 G
CO2 pe yea by 2050 ( e . 5). While ambi ious new a ge s o p o ec
species and ecosys ems ha e been p oposed, including ‘hal Ea h’ (con-
se ing hal he Ea h’s a ea o na u e)
6
and ‘30 by 30’ (30% p o ec ed
by 2030) ( e .
7
), hese a ge s ha e been c i icized o insu icien ly
accoun ing o he needs o people, including many Indigenous and
local communi ies
8
. I is he e o e essen ial o demons a e how na u e
conse a ion con ibu es o human wellbeing. Ye despi e he u gency
o sa egua ding na u al asse s a ound he wo ld, we s ill ha e limi ed
Recei ed: 18 Ap il 2022
Accep ed: 13 Oc obe 2022
Published online: 28 No embe 2022
Check o upda es
A ull lis o a ilia ions appea s a he end o he pape . e-mail: [email p o ec ed]
Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 52
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
as Canada and Russia, equi e a less a ea o achie e NCP a ge s) o
la ge ecosys em he e ogenei y (i g ea e ecosys em di e si y yields
highe le els o di e se NCP in a smalle p opo ion o a ea, which may
explain pa e ns in Chile and Aus alia).
The highes - alue c i ical na u al asse s ( he loca ions deli e ing
he highes magni udes o NCP in he smalles a ea, deno ed by he
da kes blue o g een shades in Fig. 1c) o en coincide wi h di e se,
ela i ely in ac na u al a eas nea o ups eam om la ge numbe s o
people. Many o hese high- alue a eas coincide wi h a eas o g ea es
spa ial cong uence among mul iple NCP (Ex ended Da a Fig. 3). Spa-
ially co ela ed pai s o local NCP (Supplemen a y Table 4) include
hose ela ed o wa e ( lood isk educ ion wi h ni ogen e en ion
and ni ogen wi h sedimen e en ion); o es p oduc s ( imbe and
uelwood); and hose occu ing close o human-modi ied habi a s
(pollina ion wi h na u e access and wi h ni ogen e en ion). Coas al
isk educ ion, o age p oduc ion o g azing, and i e ine ish ha es
a e he mos spa ially dis inc om o he local NCP. In he ma ine ealm,
he e is subs an ial o e lap o ishe ies wi h coas al isk educ ion and
ee ou ism ( hough no be ween he la e wo, which each ha e much
smalle c i ical a eas han exis o ishe ies).
Numbe o people bene i ing om c i ical na u al asse s
We es ima e ha ~87% o he wo ld’s cu en popula ion, 6.4 billion
people, bene i di ec ly om a leas one o he 12 local NCP p o ided by
c i ical na u al asse s, while only 16% li e on he lands p o iding hese
bene i s (and hey may also bene i ; Fig. 2a). To quan i y he numbe
o bene icia ies o c i ical na u al asse s, we spa ially delinea e hei
bene i ing a eas (which a ies on he basis o NCP: o example, a eas
downs eam, wi hin he loodplain, in low-lying a eas nea he coas ,
o accessible by a sho a el). While ou op imiza ion selec s o he
p o ision o 90% o he cu en alue o each NCP, i is no gua an eed
ha 90% o he wo ld’s popula ion would bene i (since i does no
include conside a ions o edundancy in adjacen pixels and he e o e
many o he a eas selec ed bene i he same popula ions), so i is no able
ha an es ima ed 87% do. This es ima e o ‘local’ bene icia ies p ob-
ably unde es ima es he o al numbe o people bene i ing because
i includes only NCP o which bene icia ies can be spa ially delinea ed
o a oid double-coun ing, ye i is s iking ha he as majo i y, 6.1
billion people, li e wi hin 1 h a el (by oad, ail, boa o oo , aking
he as es pa h17) o c i ical na u al asse s, and mo e han hal o he
wo ld’s popula ion li es downs eam o hese a eas (Fig. 2b). Ma e-
ial NCP a e o en deli e ed locally, bu many also en e global supply
chains, making i di icul o delinea e bene icia ies spa ially o hese
NCP. Howe e , pas s udies ha e calcula ed ha globally mo e han 54
million people bene i di ec ly om he imbe indus y18, 157 million
om i e ine ishe ies
19
, 565 million om ma ine ishe ies
20
and 1.3
billion om li es ock g azing
21
, and ac oss he opics alone 2.7 billion
a e es ima ed o be dependen on na u e o one o mo e basic needs
22
.
Nea ly all coun ies ha e a la ge pe cen age (>80%) o hei popu-
la ions bene i ing om c i ical na u al asse s, bu small coun ies
ha e much la ge p opo ions o hei popula ions li ing wi hin he
oo p in o c i ical na u al asse s han do la ge coun ies (Fig. 2a and
Supplemen a y Da a 2). When people li e in hese a eas, and espe-
cially when cu en le els o use o na u al asse s a e no sus ainable,
egula ions o incen i es may be needed o main ain he bene i s hese
asse s p o ide. While p o ec ed a eas a e an impo an conse a ion
s a egy, hey ep esen only 15% o he c i ical na u al asse s o local
NCP (Supplemen a y Table 5); addi ional a eas should no necessa ily
be p o ec ed using designa ions ha es ic human access and use, o
hey could cease o p o ide some o he di e se alues ha make hem
so c i ical
23
. O he a ea-based conse a ion measu es, such as hose
based on Indigenous and local communi ies’ go e nance sys ems,
Paymen s o Ecosys em Se ices p og ammes, and sus ainable use
o land- and seascapes, can all con ibu e o main aining c i ical lows
o NCP in na u al and semi-na u al ecosys ems24.
unde s anding o he spa ial ex en o ecosys ems p o iding essen ial
bene i s o humani y9.
In his A icle, le e aging ecen ad ances in scien i ic unde s and-
ing, da a a ailabili y (Supplemen a y Tables 1 and 2) and compu a ional
powe , we unde ake a global analysis o 14 NCP (Ex ended Da a Fig.
1), he mos comp ehensi e se mapped globally so a 10,11. Mos o
hese NCP a e conside ed ecosys em se ices, which, acco ding o
he IPBES Concep ual F amewo k3,12, a e embedded wi hin he la ge ,
mo e inclusi e concep o NCP, and he e o e we use he la e e m
h oughou o be e align wi h in e na ional amewo ks and policy
13
.
Twel e o he NCP included in ou analysis deli e p ima ily local ben-
e i s ( hough some subsequen ly en e global supply chains), including
con ibu ions o he p o ision o ood, ene gy and aw ma e ials; he
egula ion o wa e quali y and disas e isk; and ec ea ional ac i i-
ies (Fig. 1a). We p io i ize hese 12 ‘local’ NCP a he coun y le el o
iden i y a eas needed o main ain hei consis en p o ision wi hin
each coun y. In con as , we p io i ize a he global scale o wo NCP
ela ed o clima e egula ion ( e es ial ecosys em ca bon s o age
and ege a ion- egula ed a mosphe ic mois u e ecycling), whose
bene i s acc ue a con inen al scales o globally.
Th ough mul i-objec i e spa ial op imiza ion a a esolu ion o
2 km, we map he loca ion o he plane ’s c i ical na u al asse s, de ined
as he na u al and semi-na u al ecosys ems p o iding 90% o cu en
le els o each NCP ( ha is, loca ions equi ed o all NCP o mee o
exceed a 90% a ge ). Beyond his a ge he e a e diminishing e u ns,
wi h disp opo iona ely mo e na u al a ea equi ed o each inc e-
men ally highe magni udes o p o ision o NCP (Fig. 1b). We use he
e m ‘c i ical’ o con ey ha losing hese na u al asse s would lead o
disp opo iona ely la ge losses in NCP. This usage is ela ed o, hough
no he same as, he e m ‘c i ical na u al capi al’, which has been o -
mally de ined in he ecological economics li e a u e as ecosys ems
p o iding impo an unc ions ha canno be subs i u ed
14
. Howe e ,
as we do no explo e subs i u es o na u al capi al ound in al e na i e
o ms o capi al ( ha is, manu ac u ed capi al), we use he e m ‘na u al
asse s’ ins ead o a oid con usion. The asse s we ocus on a e na u al
and semi-na u al ecosys ems (Supplemen a y Table 3), excluding
de eloped lands (c oplands and u ban a eas) and un ege a ed a eas, o
p o ide insigh s o conse a ion p io i ies ele an o he CBD; global
p io i ies o es o a ion
15
o o managemen o de eloped a eas
16
a e
beyond he scope o his i s e o o map c i ical na u al asse s. Ou
analysis e eals h ee key indings abou c i ical na u al asse s: (1) hei
ex en and loca ion; (2) he numbe o people bene i ing om, and
li ing wi hin, hese c i ical a eas; and (3) he deg ee o o e lap be ween
c i ical na u al asse s deli e ing local bene i s and hose deli e ing
global bene i s and be ween c i ical na u al asse s and o he global
p io i ies o he CBD (biodi e si y and cul u al di e si y).
Resul s and discussion
Ex en and loca ion o c i ical na u al asse s
C i ical na u al asse s p o iding he 12 local NCP (Fig. 1a) occupy only
30% (41 million km2) o o al land a ea (excluding An a c ica) and 24%
(34 million km
2
) o ma ine Exclusi e Economic Zones (EEZs), e lec -
ing he s eep slope o he agg ega e NCP accumula ion cu e (Fig.
1b). Despi e his modes p opo ion o global land a ea, he sha es
o coun ies’ land a eas ha a e designa ed as c i ical can a y sub-
s an ially. The 20 la ges coun ies equi e only 24% o hei land a ea,
on a e age, o main ain 90% o cu en le els o NCP, while smalle
coun ies (10,000 o 1.5 million km
2
) equi e on a e age 40% o hei
land a ea (Supplemen a y Da a 1). This high a iabili y in he NCP–
a ea ela ionship is p ima ily d i en by he p opo ion o coun ies’
land a eas made up by na u al asse s ( ha is, excluding ba en, ice
and snow, and de eloped lands), bu e en when his is accoun ed o ,
he e a e ou lie s (Ex ended Da a Fig. 2). Ou lie s may be due o spa ial
pa e ns in human popula ion densi y ( o example, coun ies wi h
dense popula ion cen es and as expanses wi h ew people, such
Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 53
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
O e laps be ween local and global p io i ies
Unlike he 12 local NCP p io i ized he e a he na ional scale, ce ain
bene i s o na u al asse s acc ue con inen ally o e en globally. We
he e o e op imize wo addi ional NCP a a global scale: ulne able
e es ial ecosys em ca bon s o age ( ha is, he amoun o o al
ecosys em ca bon los in a ypical dis u bance e en 25, he ea e ‘eco-
sys em ca bon’) and ege a ion- egula ed a mosphe ic mois u e ecy-
cling ( he supply o a mosphe ic mois u e and p ecipi a ion sus ained
by plan li e26, he ea e ‘mois u e ecycling’). O e 80% o he na u al
asse loca ions iden i ied as c i ical o he 12 local NCP a e also c i ical
a
Lis o local NCP modelled
Ni ogen e en ion o wa e quali y egula ion
Sedimen e en ion o wa e quali y egula ion
Pollina o habi a su iciency o pollina ion-dependen c ops
Fodde o li es ock
Timbe p oduc ion
Fuelwood p oduc ion
Flood egula ion
Ri e ine ish ha es
Access o e es ial na u e ( o local ec ea ion and ga he ing)
Coas al isk educ ion ( e es ial and ma ine)
Ma ine ish ha es
Ma ine ec ea ion (co al- ee ou ism and associa ed li elihoods)
0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100
Pe cen age o o al local NCP
Pe cen age a ea o all coun ies
Ocean (wi hin EEZ)
Land
24% 30%
b
c
Mos impo an
(si es p o iding
op 5% o NCP)
Agg ega e impo ance o local NCP (coun y-le el op imiza ion ac oss 12 ypes o NCP)
C i ical na u al asse s
Si es p o iding
90% o NCP
Na u al asse s
(100% o NCP)
On land
In he ocean
Fig. 1 | C i ical na u al asse s, de ined as he na u al and semi-na u al
e es ial and aqua ic ecosys ems equi ed o main ain 12 o na u e’s
‘local’ con ibu ions o people (local NCP) on land (g een) and in he ocean
(blue). a, The 12 local NCP analysed ( ha is, no including global NCP, shown
in Supplemen a y Fig. 4). b, The NCP accumula ion cu e, e lec ing he o al
a ea equi ed o main ain a ge le els o all NCP in e e y coun y, wi h do ed
lines deno ing he a ea o c i ical na u al asse s (90% o NCP in 30% o land a ea
and 24% o EEZ a ea). A eas selec ed by op imiza ion wi hin each coun y a e
agg ega ed ac oss all coun ies o c ea e a single global accumula ion cu e; o
a ea equi emen s in indi idual coun ies, see Supplemen a y Da a 1. c, Map o
c i ical na u al asse s, wi h da ke shades conno ing c i ical na u al asse s ha
a e associa ed wi h highe le els o agg ega ed NCP. G ey a eas show he ex en
o emaining na u al asse s no designa ed ‘c i ical’ bu included in his analysis;
whi e a eas (c opland, u ban and ba e a eas, ice and snow, and ocean a eas
ou side he EEZ) we e excluded om he op imiza ion.
Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 54
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
o he wo global NCP (Fig. 3). The spa ial o e lap be ween c i ical
na u al asse s o local and global NCP accoun s o 24% o land a ea,
wi h an addi ional 14% o land a ea c i ical o global NCP ha is no
conside ed c i ical o local NCP (Ex ended Da a Fig. 4). Toge he , c i i-
cal na u al asse s o secu ing bo h local and global NCP equi e 44%
o o al global land a ea. When each NCP is op imized indi idually
(ca bon and mois u e NCP a he global scale; he o he 12 a he coun y
scale), he o e lap be ween ca bon o mois u e NCP and he o he NCP
exceeds 50% o all e es ial (and eshwa e ) NCP excep coas al isk
educ ion (which o e laps only 36% wi h ecosys em ca bon, 5% wi h
mois u e ecycling; Supplemen a y Table 4).
Syne gies can also be ound be ween NCP and biodi e si y and
cul u al di e si y. C i ical na u al asse s o local NCP a na ional le els
o e lap wi h pa o all o he a ea o habi a (AOH, mapped on he basis
o species ange maps, habi a p e e ences and ele a ion27) o 60% o
28,177 e es ial e eb a es (Supplemen a y Da a 3). Bi ds (73%) and
mammals (66%) a e be e ep esen ed han ep iles and amphibians
(44%). Howe e , hese c i ical na u al asse s ep esen only 34% o he
a ea o endemic e eb a e species (wi h 100% o hei AOH loca ed
wi hin a gi en coun y; Supplemen a y Da a 3) and 16% o he a ea o
all e eb a es i using a mo e conse a i e ep esen a ion a ge ame
-
wo k based on he IUCN Red Lis c i e ia ( hough, no ably, achie ing
Red Lis ep esen a ion a ge s is impossible o 24% o species wi hou
es o a ion o o he expansion o exis ing AOH; Supplemen a y Da a
4). Cul u al di e si y (p oxied by linguis ic di e si y) has a highe
o e laps wi h c i ical na u al asse s han does biodi e si y; hese a eas
in e sec 96% o global Indigenous and non-mig an languages
28
(Sup-
plemen a y Da a 5). The deg ee o which languages a e ep esen ed
in associa ion wi h c i ical na u al asse s is consis en ac oss mos
coun ies, e en a he high end o language di e si y (coun ies con ain-
ing >100 Indigenous and non-mig an languages, such as Indonesia,
Nige ia and India). This high co espondence p o ides u he suppo
o he impo ance o sa egua ding igh s o access c i ical na u al
asse s, especially o Indigenous cul u es ha bene i om and help
main ain hem. Despi e he la ge land a ea equi ed o main aining
he global NCP compa ed wi h local NCP, global NCP p io i y a eas
o e lap wi h sligh ly ewe languages (92%) and wi h only 2% mo e spe-
cies (60% o species AOH), al hough a subs an ially g ea e o e lap is
seen wi h global NCP i Red Lis c i e ia a e conside ed (36% compa ed
wi h 16% o local NCP; Supplemen a y Da a 4). These esul s p o ide
di e en insigh s han p e ious e o s a smalle scales, pa icula ly
a simila exe cise in Eu ope ha ound less o e lap wi h p io i y a eas
o biodi e si y and NCP29. Howe e , he 40% o all e eb a e species
whose habi a s did no o e lap wi h c i ical na u al asse s could d i e
e y di e en pa e ns i biodi e si y we e included in he op imiza ion.
Al hough hese 14 NCP a e no comp ehensi e o he my iad ways
ha na u e bene i s and is alued by people23, hey cap u e, spa ially
and hema ically, many elemen s explici ly men ioned in he Fi s
D a o he CBD’s pos -2020 Global Biodi e si y F amewo k
13
: ood
secu i y, wa e secu i y, p o ec ion om haza ds and ex eme e en s,
li elihoods and access o g een and blue spaces. Ou emphasis he e is o
highligh he con ibu ions o na u al and semi-na u al ecosys ems o
human wellbeing, speci ically con ibu ions ha a e o en o e looked
in mains eam conse a ion and de elopmen policies a ound he
wo ld. Fo example, conside a ions o global ood secu i y o en
include only c op p oduc ion a he han na u e’s con ibu ions o
i ia pollina ion o ege a ion-media ed p ecipi a ion, o li es ock
p oduc ion wi hou pa i ioning ou he con ibu ion o g asslands
om mo e in ensi ied eed p oduc ion.
Gaps and nex s eps
Ou syn hesis o hese 14 NCP ep esen s a subs an ial ad ance beyond
o he global p io i iza ions ha include NCP limi ed o ecosys em
ca bon s ocks, esh wa e and ma ine ishe ies
30–32
, hough s ill alls
sho o including all impo an con ibu ions o na u e such as i s
ela ional alues
33
. Despi e he omission o many NCP ha we e no able
o be mapped, u he analyses indica e ha esul s a e ai ly obus
o inclusion o addi ional NCP. D opping one o he 12 local NCP a a
ime esul s in <1–3% change in he o al global land a ea equi ed o
main ain 90% o cu en le els o hese NCP (Supplemen a y Table 6)
and a high deg ee o spa ial ag eemen . In ac , 62% o he o al a ea
on land is sha ed by all op imiza ion solu ions, and 97% o he a ea is
included in 11 o he 12 solu ions; simila alues a e ound ac oss mos
coun ies (Supplemen a y Da a 1). Ne e heless, his same mul i-NCP
op imiza ion app oach could accommoda e addi ional NCP as spa ial
da a become a ailable a su icien esolu ion and app op ia e scales.
The e also is unce ain y in he iden i ica ion o c i ical na u al
asse s ela ed o model e o in he indi idual NCP ha we we e able o
include. We acknowledge ha NCP models, like all models, ha e e o s,
and ha consis en global-scale modelling will miss de ails impo an
o ce ain speci ic loca ions. Valida ion o NCP is pa icula ly di icul
gi en ha he e a e no di ec measu emen s o many NCP wi h assess-
men elian on emo ely sensed p oxies. We u ilize he bes a ail-
able global modelling app oaches and da a, mos o which ha e been
alida ed in a leas some loca ions
19,25,34–41
. Whe e unce ain y exis ed
abou wha dis ance was mos app op ia e o model he deli e y o
NCP ( o example, how a o model people downs eam o how a
people migh a el o na u al asse s), we pe o med u he sensi i i y
analysis and con i med ha he es ima ed land a ea o c i ical na u al
asse s is obus o he dis ance chosen (impac ing esul s by <5%; Sup-
plemen a y Table 6). As a ailabili y o global models o many o hese
NCP inc eases, u u e wo k should mo e owa ds ensemble modelling
app oaches, which ha e been shown o inc ease accu acy and educe
unce ain y compa ed wi h indi idual model ou pu s42,43.
1,169
656
0.89
6,389
3,827
1.01
0 20 40 60 80 100
Global
La ges 30 coun ies
Smalles 30 coun ies
Pe cen age o o al popula ion
Popula ion
li ing on CNA
Popula ion
bene i ing om
CNA
6,102
3,652
1,545
1,267
193
0 20 40 60 80
Popula ion wi hin 1 h
Downs eam popula ion
Floodplains popula ion
Popula ion bene i ing
Coas al popula ion
Pe cen age o ele an popula ion bene i ing
Rele an
popula ion
(millions)
7,342
7,342
3,717
954
1,839
a
b
Fig. 2 | People bene i ing om and li ing on c i ical na u al asse s (CNA). a,b,
‘Local’ bene icia ies we e calcula ed h ough he in e sec ion o a eas bene i ing
om di e en NCP, o a oid double-coun ing people in a eas o o e lap; only
hose NCP o which bene icia ies could be spa ially delinea ed we e included
( ha is, no ma e ial NCP ha en e global supply chains: ishe ies, imbe ,
li es ock o c op pollina ion). Ba s show pe cen ages o o al popula ion globally
and o la ge and small coun ies (a) o he pe cen age o ele an popula ion
globally (b). Numbe s inse in ba s show millions o people making up ha
pe cen age. Numbe s o he igh o ba s in b show o al ele an popula ion (in
millions o people, equi alen o o al global popula ion om Landscan 2017 o
popula ion wi hin 1 h a el o downs eam, bu limi ed o he o al popula ion
li ing wi hin 10 km o loodplains o along coas lines <10 m abo e mean sea le el
o loodplain and coas al popula ions p o ec ed, espec i ely, and o u al poo
popula ions o uelwood).
Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 55
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
Da a and modelling gaps p e en ed a b oade explo a ion o
issues ele an o he ecological supply side o NCP. Al hough esul s
p esen ed he e sugges ha na ionally p io i ized a eas o local NCP
can deli e on global p io i ies in many egions, hey also highligh
a need o in eg a ed modelling o ep esen in e ac ions be ween
di e en NCP. Fo example, a mosphe ic mois u e e apo anspi ed
by Amazonian o es s alls as ain in o he pa s o Sou h Ame ica,
suppo ing ecosys ems ha p o ide ood, uel and o he bene i s26.
Fu he wo k is needed o mo e beyond he spa ial o e laps explo ed
he e owa ds unde s anding unc ional in e -dependencies be ween
NCP. We also acknowledge ha u ban and c opland sys ems we e
omi ed om his analysis owing o da a and modelling limi a ions
ha would ail o adequa ely cap u e he NCP suppo ed by di e en
land use ypes and land managemen p ac ices wi hin hose sys ems.
Likewise, a c ic and dese ecosys ems, owing o spa se ege a ion and
low human popula ion densi ies, a e no well ep esen ed in ou NCP
models and ye a e e y impo an o he people who li e in and depend
on hem. As da a and modelling gaps a e illed, u u e assessmen o
c i ical na u al asse s should expand o ecognize unique con ibu ions
o cu en ly unde alued ecosys ems and should include possible
gains om es o ing and sus ainably managing human-domina ed
sys ems
15,16
o conside how hese di e en conse a ion s a egies
can complemen one ano he .
The e a e also, pe haps e en mo e p onounced, da a and model-
ling gaps o ill on he social side o NCP. In he NCP modelled he e, we
ep esen ealized bene i s o na u al asse s—weigh ed by bene icia y
popula ion when easible—bu his unde s a es he ange o ways in
which na u al asse s di ec ly and indi ec ly con ibu e o people’s
wellbeing. Limi ed socio-economic da a and lack o eliable models
linking NCP o wellbeing indica o s p eclude mo e p ecise alua ion o
mos NCP a he global scale. Addi ional insigh could be gained om
mapping c i ical na u al asse s ha suppo he mos ulne able o
dependen 22 people, including Indigenous peoples whose li elihoods
and cul u al iden i ies di ec ly depend on na u e (and indeed o e lap
subs an ially wi h c i ical na u al asse s, on he basis o ou es ima es
o Indigenous language di e si y on hese lands; Supplemen a y Da a
5), and he poo who may lack access o an h opogenic subs i u es o
NCP (see also philosophical conside a ions in Supplemen a y No e
1). Recen p og ess in linking ecological modelling wi h in eg a ed
assessmen modelling and gene al equilib ium economic modelling
44
shows g ea p omise o assessing he bene i s o c i ical na u al asse s
o socie y and he global economy. Such e o s could also e eal el-
ecoupling o c i ical na u al asse s a ising om ansbounda y lows
be ween coun ies such as ia in e na ional ade45.
Finally, u he wo k is needed o in es iga e whe he c i ical
na u al asse s a e necessa y o su icien o mee ing humani y’s needs,
by conside ing he a ailabili y o subs i u es and wha cons i u es
‘enough’. Though mo i a ed by he ecological economic concep o
‘c i ical na u al capi al’, we we e no able o cap u e he deg ee o
which an h opogenic asse s could eplace na u al asse s. Howe e ,
gi en he high co espondence o many NCP (Supplemen a y Table
4), i seems unlikely ha an h opogenic asse s could subs i u e o all
bene i s p o ided by na u al asse s in a pa icula a ea ( o example,
g een in as uc u e deli e s many co-bene i s beyond he single ben-
e i buil in as uc u e is designed o deli e
46,47
). I seems mo e likely
ha mo e han 44% o he plane may be equi ed o secu e he 14 NCP
mapped he e (le alone he o he di e se alues o na u e we we e no
able o map), mos impo an ly because i is unclea how much o he
cu en need o na u e is al eady unme . In many pa s o he wo ld,
na u al ecosys ems a e al eady deg aded o con e ed, so main aining
90% o cu en le els o NCP may be a oo li le ( o example, places
p one o ca as ophic looding due o habi a con e sion48, whe e
g azing lands ha e been dese i ied
49
and whe e ish popula ions ha e
c ashed
50
). Fu he mo e, wha is c i ical now may change in he u u e
owing o clima e change, popula ion g ow h o change in echnology
o consump ion pa e ns1,11. O he a eas may no di ec ly bene i peo-
ple bu may be c i ical o a oid ecological ipping poin s and collapse
o NCP in a ec ed ecosys ems51,52. We he e o e p opose he analysis
conduc ed he e as a i s a emp o de ine a minimum se o c i ical
na u al asse s, and sugges ha he o e all app oach p o ides a use ul
amewo k o explo ing such issues wi h u u e scena io modelling o
de elop mo e esilien conse a ion o na u e and people.
Fo bo h
Fo global (clima e) NCP
Fo local NCP
C i ical na u al asse s:
Fig. 3 | Spa ial o e laps be ween c i ical na u al asse s o local and global
NCP. Red and eal deno e whe e c i ical na u al asse s o global NCP (p o iding
90% o ecosys em ca bon and mois u e ecycling globally) o o local NCP
(p o iding 90% o he 12 NCP lis ed in Fig. 1), espec i ely, bu no bo h, occu ;
gold shows a eas whe e he wo o e lap (24% o he o al a ea). Toge he , local
and global c i ical na u al asse s accoun o 44% o o al global land a ea
(excluding An a c ica). G ey a eas show na u al asse s no de ined as ‘c i ical’ by
his analysis, hough s ill p o iding some alues o ce ain popula ions. Whi e
a eas we e excluded om he op imiza ion.
Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 56
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
We acknowledge ha ou app oach is s ongly an h opocen ic
(NCP, and all he concep s included in hem, a e by de ini ion an h o-
pocen ic3). As such, i is no in ended o cap u e he in insic alues
o na u e, o he alue o he ecosys ems o species p o iding he NCP
highligh ed he e o o he non-human o ganisms. Ou ocus and ind-
ings should no be in e p e ed as dismissing hose alues, and u he
wo k could explo e na u al asse s p o iding impo an con ibu ions o
biodi e si y ha a e no cap u ed by species maps alone. Fo example,
egula ing con ibu ions, including wa e quali y egula ion, na u al
haza ds esilience, pollina ion and a mosphe ic mois u e ecycling,
main ain he condi ions unde which cu en biodi e si y h i es.
Delinea ing species o high-biodi e si y a eas as he ‘bene icia ies’
o many o hese con ibu ions may be an impo an s ep owa ds
e lec ing na u e’s con ibu ions o na u e.
Conclusion
Iden i ying c i ical na u al asse s could enable na ional and global
leade s o p io i ize he conse a ion o a wide ange o NCP. We ind
i encou aging ha secu ing 90% o he NCP mapped he e is easible
wi h an a ea compa able o o he p oposed conse a ion a ge s
6,7,32,53
.
Global analyses such as his can se a b oade con ex o local decisions
including unde s anding o dis an connec ions ha ex end ou side
a coun y’s bo de s, p o ide apid in o ma ion o global ac o s such
as CBD signa o ies and non-go e nmen al o ganiza ions wo king on
conse a ion p io i ies ac oss many coun ies, and supplemen gaps
in local in o ma ion while i is s ill being gene a ed
54
. Na ional- and
local-le el use o spa ial da a is imely and ele an o coun ies seeking
o access sca ce in e na ional Sus ainable De elopmen Goal- ela ed
inance, as i can imp o e he ecological, social and economic aspec s o
p ojec eadiness
55,56
. Howe e , we emphasize he alue o he app oach
de eloped he e mo e han he maps o da a. This app oach can be
adap ed and e ined a he scales a which policy implemen a ion
occu s, wi h he bes a ailable da a and complemen ed by inpu om
local expe s and di e se s akeholde s, o imp o e accu acy and pub-
lic legi imacy57,58 and o ensu e ha human igh s and di e se human
ela ionships wi h na u e a e sa egua ded. Mo eo e , c ea ion and
use o spa ially explici in o ma ion allows o a ocus on ecosys em
quali y o e quan i y, helping o a oid po en ially pe e se ou comes
o a ea a ge s o conse a ion
59
. This app oach o iden i ying c i ical
na u al asse s is a i al s ep o wa d in empowe ing ac o s a all le els
o make decisions ha bene i bo h na u e and people.
Me hods
Modelling NCP
The 14 NCP in his analysis (Ex ended Da a Fig. 1) we e chosen o span
de elopmen and clima e goals, and o be mappable wi h spa ially
explici da a ep esen ing he pe iod 2000–2020. We use Eu opean
Space Agency (ESA) 2015, o land co e a 10 a csec (~300 m a he
equa o ) esolu ion, Landscan 2017 o popula ion60 a 30 a csec
(~1 km) esolu ion (which we e he mos cu en da a a ailable a he
ime we began ou analysis). We ocus on ‘na u e’s con ibu ions’ o
key bene i s o in e es ( o example, secu i y in ood, wa e , haza ds,
ma e ials and cul u e), meaning we pa i ion ou he ole o na u al and
semi-na u al ecosys ems in p oducing hose bene i s. Fo ood secu-
i y, we include he con ibu ions o pollina o habi a su iciency o
pollina ion-dependen c op p oduc ion, ege a ion-media ed a mos-
phe ic mois u e ecycling o c op and li es ock p oduc ion (included
as a global NCP ha also links o clima e secu i y), g assland odde
p oduc ion o li es ock p oduc ion, and wild i e ine and ma ine ish-
e ies. Fo wa e secu i y, we include he con ibu ions o wa e quali y
egula ion, ia sedimen e en ion and nu ien e en ion, bu no wa e
yield since he ole o ecosys ems in de e mining he quan i y o wa e
is minimal (o he han by e apo anspi a ion, which is al eady cap u ed
in he ege a ion-media ed mois u e ecycling, and egula ion o im-
ing o lows which is cap u ed in lood isk educ ion). Fo secu i y o
p o ec ion om na u al haza ds, we include lood isk educ ion and
coas al isk educ ion. Fo ma e ials, we include imbe p oduc ion,
uelwood p oduc ion and access o na u e (which could be used o
ga he ing, and also links o cul u e). Fo cul u al bene i s we include
co al ee ou ism (as he only globally mapped o m o ma ine-based
ou ism) and access o na u e again (which in addi ion o ga he ing also
cap u es ec ea ion o o he uses o nea by g eenspace). Finally, o
clima e secu i y, in addi ion o mois u e ecycling men ioned abo e,
we include o al ecosys em ca bon s o age (as a global NCP). Below we
b ie ly summa ize he models used o map hese local NCP (Supplemen-
a y Table 1) and global NCP (Supplemen a y Table 2); ull in o ma ion
on each model is a ailable in Supplemen a y No e 1.
Local NCP.
1. Ni ogen e en ion o egula e wa e quali y o downs eam
popula ions is modelled a na i e ESA esolu ion (a 10 a csec
o 300 m) using he InVEST61 Nu ien Deli e y Ra io model,
which is based on e ilize applica ion, p ecipi a ion, opog-
aphy and he e en ion capaci y o ege a ion, and has been
p e iously used in global applica ions11. The people bene i ing
om ni ogen e en ion a e hose who would o he wise be
exposed o ni ogen con amina ion in hei d inking wa e . In
his analysis, he numbe o people downs eam was calcula ed
o e e y pixel o habi a , o p o ide a sense o which habi a
po en ially bene i s he mos people. Ideally, o map ealized
ni ogen e en ion, we would be able o con e biophysical se -
ice p oduc ion in o a measu e o change in wellbeing, whe he
mone a y, in heal h e ms o o he wise. Howe e , he s a e o
he science and da a a ailable globally p ecludes his o mos
se ices, so ou p oxy was he numbe o people downs eam
who could po en ially bene i om ha e en ion. NCP o
ni ogen e en ion is exp essed as ni ogen e en ion on na u al
and semi-na u al pixels mul iplied by he numbe o people
downs eam o hose pixels ( o mo e de ails, see Supplemen-
a y No e 1 Sec ion 1).
2. Sedimen e en ion o egula e wa e quali y o downs eam
popula ions is modelled a na i e ESA esolu ion (10 a csec o
300 m) by adap ing he InVEST Sedimen Deli e y Ra io model,
which maps o e land sedimen gene a ion and deli e y o he
s eam using he Re ised Uni e sal Soil Loss Equa ion (RUSLE)
based on clima e, soil ex u e, opog aphy and land co e , and
a conduc i i y index based on he upslope and downslope a eas
o each pixel. Ideally sedimen e en ion would be delinea ed
o ese oi s, i iga ion canals o o he wa e deli e y in a-
s uc u e ha is mos impac ed by sedimen a ion, bu lacking
a comp ehensi e global da ase iden i ying all such in as uc-
u e, we again use he p oxy o numbe o people downs eam
(as desc ibed o ni ogen e en ion, abo e). NCP o sedimen
e en ion is exp essed as sedimen e en ion on na u al and
semi-na u al pixels mul iplied by he numbe o people down-
s eam o hose pixels (Supplemen a y No e 1 Sec ion 2).
3. Pollina o habi a su iciency o pollina ion-dependen c op
p oduc ion is modelled a na i e ESA esolu ion (10 a csec o
300 m) wi h a simpli ied e sion o InVEST, imp o ing upon
p e ious global mapping o he po en ial con ibu ion o wild
pollina o s o nu i ion p oduc ion by mapping he alue om
he a m ield back o habi a 11. This con ibu ion is calcula ed
as he su iciency o habi a su ounding a mland mul iplied by
he pollina ion dependency o c ops ( he p opo ion o yields
ha a e pollina ion dependen , based on he c op mixes g own
he e). While his does no cap u e o he impo an ea u es
ha a ec pollina o s, such as p esence o speci ic species
impo an o nes ing o lo al esou ces o he in ensi y o
equency o pes icide applica ions, i p o ides a p oxy o he
Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 57
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
con ibu ion o nea by na u al habi a o pollina ion. This NCP
is exp essed in e ms o he a e age equi alen numbe o peo-
ple ed by pollina ion-dependen c op p oduc ion, a ibu ed
o nea by ecosys ems based on he a ea o pollina o habi a
wi hin pollina o ligh dis ance o c ops (Supplemen a y No e 1
Sec ion 3).
4. Fodde p oduc ion o li es ock is modelled using e sion 3 o
Co$ ing Na u e62 a a esolu ion o 5 a cmin (~10 km), and hen
masked o ele an ESA na u al and semi-na u al habi a s (a
10 a csec o 300 m esolu ion; Supplemen a y Table 3). Supply
o odde is calcula ed om emo ely sensed d y ma e p o-
duc i i y o he non-c opland co e ac ion, and demand is
es ima ed by he headcoun o li es ock in a g id cell mul iplied
by he a e age biomass in ake equi emen s pe animal. NCP
o odde p oduc ion o li es ock is exp essed in e ms o
an index (0–1), e-scaled om he ealized se ice, which is e-
po ed as he smalle o he supply o demand (i consump ion
exceeds p oduc i i y, he gap is assumed o be me wi h eed).
The bes a ailable global inpu s o d y ma e p oduc i i y,
li es ock headcoun , c opland and land co e a e used as inpu
(Supplemen a y No e 1 Sec ion 4).
5. Timbe p oduc ion includes comme cial ( o example, o
ade o expo ) and domes ic ( o example, o local use)
imbe , modelled using e sion 3 o Co$ ing Na u e as wo
spa ially mu ually exclusi e laye s, because hey ep esen wo
di e en se s o bene icia ies. This model ou pu s da a a a
esolu ion o 5 a cmin (~10 km), and is hen masked o ele an
ESA na u al and semi-na u al habi a s (a 10 a csec o 300 m
esolu ion; Supplemen a y Table 3). NCP o imbe p oduc ion
is exp essed as an index (0–1) based on emo ely sensed o es
p oduc i i y and accessibili y o ha es . To al po en ial sus-
ainable supply o imbe is es ima ed om he bes a ailable
global abo eg ound ca bon s ock map mul iplied by ac ional
ee co e o u al a eas only. The sus ainable ha es is calcu-
la ed as he ecip ocal o he numbe o yea s aken o de elop
he s ock a he annual seques a ion a e, acco ding o d y
ma e p oduc i i y da a. Demand is calcula ed di e en ly o
comme cial e sus domes ic imbe based on di e en assump-
ions o accessibili y. Comme cial accessibili y is de ined as
wi hin 6 h a el ime o a popula ion cen e o >50,000 people
and on slope g adien s <70%. Domes ic accessibili y is de ined
as a eas inaccessible o comme cial ha es , and ha es a es
a e based on pe capi a consump ion mul iplied by popula ion
wi hin 10 km (Supplemen a y No e 1 Sec ion 5).
6. Fuelwood p oduc ion is calcula ed as a byp oduc o he imbe
model om e sion 3 o Co$ ing Na u e. NCP o uelwood
p oduc ion, like imbe , is ep esen ed as an index (0–1) based
on o es p oduc i i y and accessibili y o ha es , bu in his
case speci ically by u al people. Fuelwood can o e lap spa ially
wi h domes ic and comme cial imbe use, gi en ha domes ic
and comme cial imbe ha es will no consume all sus ainably
a ailable woody biomass in all places, owing o he slope g adi-
en limi and/o in places whe e demand is less han supply, and
uelwood is o en a byp oduc o imbe ha es . This model
ou pu s da a a a esolu ion o 5 a cmin (~10 km), and is hen
masked o ele an ESA na u al and semi-na u al habi a s (a
10 a csec o 300 m esolu ion; Supplemen a y Table 3) (Sup-
plemen a y No e 1 Sec ion 6).
7. Flood egula ion is modelled using e sion 2 o he hyd ologic
model Wa e Wo ld35. To map na u e’s in luence on lood isk
educ ion, we iden i y he ups eam places whe e canopies,
we lands and soils (g een s o age) e ain and slowly elease
ain all, o he bene i o downs eam communi ies. NCP o
lood egula ion is exp essed as an index (0–1) o ‘g een’ wa e
s o age (based on we land s o age, canopy s o age and soil
s o age) mul iplied by he numbe o people downs eam. This
model ou pu s da a a a esolu ion o 5 a cmin (~10 km), and is
hen masked o ESA na u al and semi-na u al habi a s (a 10 a c-
sec o 300 m esolu ion) (Supplemen a y No e 1 Sec ion 7).
8. Access o na u e is used as a p oxy o nume ous di ec and
indi ec bene i s o people, such as ec ea ion, hun ing and
ga he ing, aes he ics, men al and physical heal h, and o he
cul u al alues ha depend on he abili y o people o access
na u e. This p oxy NCP is exp essed as he numbe o u ban and
u al63 people wi hin 1 h (o 6 h, o sensi i i y analysis) a el o
na u al and semi-na u al habi a , aking he leas -cos pa h (by
oo , oad, ail o boa ) ac oss a ic ion laye 17 a a esolu ion o
2 km (equal-a ea p ojec ion) (Supplemen a y No e 1 Sec ion 8).
9. Ri e ine ish ca ch is based on spa ial disagg ega ion o na-
ionally epo ed ca ch o 2007–2014 ( e . 19) and upda ed o
include ca ch es ima ed by household consump ion su eys in
32 coun ies wi h se e e unde epo ing64. Ca ches om la ge
lakes we e excluded. To spa ially disagg ega e he global ca ch
o 13.3 × 106 onnes wi hin coun y bo de s, a mul iple linea e-
g ession model o o al ish ca ch in i e basins compiled om
he li e a u e was i ed wi h h ee p edic o a iables: popula-
ion densi y, i e discha ge and pe cen age we land co e
(n = 40, R2adj = 0.69). NCP o i e ine ish ha es is ep esen ed
as me ic onnes o ish caugh pe km2 o land a ea pe yea ,
spa ially alloca ed o he loca ions o he ha es . This model
ou pu s da a a a esolu ion o 5 a cmin (~10 km), and is hen
masked o ESA na u al and semi-na u al habi a s (a 10 a csec o
300 m esolu ion) (Supplemen a y No e 1 Sec ion 9).
10. Ma ine ish ca ch is based on da a om he Sea A ound Us,
Global Fishing Wa ch, and o he sou ces o map ish ca ch
o 2010–2014 wi hin 30 min (~55 km) g id cells ac oss he
ocean65,66. NCP o ma ine ish ha es is ep esen ed as me ic
onnes o ish caugh pe km2 pe yea , spa ially disagg ega ed
o he loca ions o he ca ch (Supplemen a y No e 1 Sec ion 10).
11. Co al ee ou ism is aken om he Mapping Ocean Weal h
da ase 34, which epo s NCP o co al- ee -associa ed ou ism
as dolla s o ou ism expendi u e (in deciles 1–10) a 15 a csec
(~500 m) esolu ion. Na ional expendi u e da a a e spa ially
dis ibu ed on he basis o h ee independen sou ces: ho el
ooms om he comme cial Global Accommoda ion Re e -
ence Da abase, di e shops and di e si es om Di eboa d, and
use -gene a ed pho os om he image-sha ing websi e Flick
(Supplemen a y No e 1 Sec ion 11).
12. Coas al isk educ ion is modelled wi h InVEST o e es ial
and coas al/o -sho e habi a s67–70 a na i e ESA esolu ion
(10 a csec o 300 m) upda ing p e ious global modelling11
h ough he inclusion o new da a and p ojec ing he alue
om he sho eline back o he p o ec i e habi a (which may
be o -sho e). Coas al isk educ ion depends on he physical
exposu e o coas al haza ds (based on wind, wa es, sea le el
ise, geomo phology and ba hyme y), wi h and wi hou na u-
al habi a o a enua e s o m su ge, and he people exposed.
NCP o coas al isk educ ion is exp essed as a uni less index
o he coas al isk educed by habi a mul iplied by he numbe
o people wi hin he p o ec i e dis ance o he habi a (Supple-
men a y No e 1 Sec ion 12).
Global NCP.
1. Vulne able e es ial ecosys em ca bon s o age is mapped
a a esolu ion o 1 a csec (~30 m), as he abo eg ound and
belowg ound ecosys em ca bon los in a ‘ ypical’ dis u bance
e en , a he han he o al s ock25. This includes e es ial and
coas al (mang o e, sal ma sh and seag ass) ecosys em ca bon
pools (abo eg ound, belowg ound and soils), based on wha
Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 58
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
ca bon is likely o be eleased i he ecosys em we e con e ed
(Supplemen a y No e 1 Sec ion 13).
2. A mosphe ic mois u e ecycling is he p ocess o wa e a ising
om he su ace o he Ea h as e apo a ion, lowing h ough
he a mosphe e as wa e apou , and e u ning o he su ace
o he Ea h as p ecipi a ion. Sou ces o e apo a ion include
canopy in e cep ion, soil in e cep ion, soil e apo a ion,
ege a ion anspi a ion and open wa e e apo a ion. We
employ an Eule ian mois u e acking model, WAM-2 laye s26,
o quan i y he sou ce o mois u e, whe e i a els h ough he
a mosphe e and whe e i alls ou downwind ac oss a 1.5° g id
(~166 km). NCP o mois u e ecycling, which is o say he mois-
u e associa ed wi h ege a ion, is exp essed as he olume o
wa e e apo anspi ed ha alls on all ain ed p oduc i e land
(c opland, angelands and wo king o es s71) (Supplemen a y
No e 1 Sec ion 14).
A ibu ion o alue o na u al asse s
The i s s ep in iden i ying c i ical na u al asse s, be o e he op imiza-
ion can selec he highes alue a eas ac oss all NCP, is o a ibu e he
magni ude o bene i s and, whe e possible, he numbe o bene icia -
ies, o he ecosys ems p o iding hose bene i s o each indi idual
NCP on he basis o i s unique cha ac e is ics ( o example, a ibu ing
he alue o pollina ion occu ing on c oplands o he nea by habi a
supplying he pollina o s, o he coas al isk educed and numbe o
people p o ec ed along he coas line o o -sho e as well as on-sho e
habi a s). We de ine na u al asse s as na u al and semi-na u al e
-
es ial ecosys ems (including semi-na u al lands like angelands and
p oduc ion o es s, bu excluding c opland, u ban a eas, ba e a eas
and pe manen snow and ice; Supplemen a y Table 3) and inland and
ma ine wa e s. Model ou pu s o pollina ion and coas al isk educ ion
a e mapped back o habi a on he basis o pollina o ligh dis ance
(Supplemen a y No e 1 Sec ion 3) and p o ec i e dis ance o coas al
habi a (Supplemen a y No e 1 Sec ion 12), espec i ely. Fo sedimen
and ni ogen e en ion, he coun o people downs eam o each habi a
pixel was summed acco ding o a hyd ologic low accumula ion (Sup-
plemen a y No e 1 Sec ion 1), and o na u e access, he coun o people
was calcula ed o each pixels by summing he popula ion pixels wi hin
1 h a el ime acco ding o a ic ion su ace (Supplemen a y No e 1
Sec ion 8). All o he model ou pu s a e coa se han ESA esolu ion and
a e masked o he land co e ypes de ined as na u al asse s ele an
o ha NCP ( o example, only o es s o imbe bu excluding o es s
o g azing; Supplemen a y Table 3).
Op imiza ion o NCP
Using in ege linea p og amming (p io i iz , Supplemen a y No e 1
Sec ion 22), we iden i y minimum a eas equi ed (1) wi hin each coun-
y’s land bo de s and ma ine EEZs o he local NCP and (2) wi hin all
global land a ea (excluding An a c ica) and all coun ies’ combined
EEZ a ea o he global NCP, o each a ge le els ( anging om 5% o
100%) o e e y NCP. Be o e op imiza ion he da a we e e-sampled o
a 2 km equal-a ea p ojec ion o consis ency (Supplemen a y No e
1 Sec ion 24). The op imiza ion selec s o he highes alues ac oss
all NCP, p o iding he mos bene i and/o o he g ea es numbe o
people, bu no accoun ing o complemen a i y o edundancies o
adjacen pixels ( ha is, no dynamically op imizing a e each pixel’s
selec ion). Land and ma ine bo de s a e based on Flande s Ma ine
Ins i u e (2020; Supplemen a y Table 2), and o e lapping claims a e
excluded om he na ional analyses. The 12 local NCP a e op imized
o each coun y, hen agg ega ed globally, while he wo global NCP
a e op imized globally. In addi ion o hese wo main op imiza ions,
we assess he sensi i i y o scale by op imizing he 12 local NCP glob-
ally (ins ead o by each coun y; Ex ended Da a Fig. 5), bo h wi h and
wi hou he wo global NCP, and by subs i u ing di e en scales o
bene icia ies mapping o people downs eam and o access o na u e
(Supplemen a y Table 6). We also assess he sensi i i y o he a ea and
loca ion o c i ical na u al asse s ( he op imiza ion solu ion o he
90% a ge ) o di e en NCP combina ions. These a ia ions include
op imizing o each NCP indi idually, and op imizing o all NCP bu
d opping each local NCP om he se o 12 o e alua e i s e ec on he
o e all op imiza ion (Supplemen a y Table 6). We also examine he
co espondence be ween NCP and he obus ness o hese di e en
solu ions, by calcula ing he pe cen age o a ea sha ed by di e en
pai s o se ices (Supplemen a y Table 4) o he pe cen age o a ea
sha ed by all solu ions (Supplemen a y Da a 1). We summa ize he
land and ocean a eas equi ed by coun y in Supplemen a y Da a 1.
Numbe o people bene i ing om c i ical na u al asse s
We map he a eas bene i ing om c i ical na u al asse s o calcula e
he numbe o di ec bene icia ies o hese asse s, and o compa e he
numbe o bene icia ies o he numbe o people li ing on he lands
comp ising hese c i ical na u al asse s. Fo his analysis we a e only
able o include NCP o which he low o he bene i can be spa ially
delinea ed: downs eam wa e quali y egula ion (sedimen e en ion
and ni ogen e en ion), lood mi iga ion, na u e access, uelwood
p o ision and coas al isk educ ion. The bene i ing a eas o some o
he ma e ial NCP ha a e aded ( ish, imbe , li es ock and c ops ha
a e pollina ed) o he loca ion o people who buy hose aded goods
a e no easily mapped, so people bene i ing om hese NCP a e no
included in his analysis o bene icia ies. Howe e , people wi hin 1 h o
c i ical na u al asse s may p o ide a su oga e o many o he ma e ial
NCP ha a e locally consumed. Fo wa e quali y egula ion, we ake
he popula ion wi hin he a eas downs eam (Supplemen a y No e 1
Sec ion 1) o c i ical na u al asse s. Fo na u e access, we ake he popu-
la ion wi hin 1 h a el (by oo , ca , boa o ail; Supplemen a y No e 1
Sec ion 8) o c i ical na u al asse s. Likewise, o he o he NCP we ake
he ele an popula ion downs eam, wi hin he p o ec i e dis ance, o
a ga he ing dis ance o c i ical na u al asse s. The ele an popula ion
o each NCP is conside ed o be he o al global popula ion o na u e
access and wa e quali y egula ion, bu is limi ed o he o al popula-
ion li ing wi hin 10 km o loodplains o lood mi iga ion, popula ion
along coas lines (in exposed a eas: <10 m abo e mean sea le el) o
coas al isk educ ion, and u al poo popula ions o uelwood. To al
‘local’ bene icia ies a e calcula ed h ough he in e sec ion o a eas
bene i ing om di e en NCP, o a oid double-coun ing people in
a eas o o e lap. We calcula e he numbe o people and pe cen o
ele an popula ion bene i ing globally o each NCP (Fig. 2b) and
he o al ‘local’ bene icia ies globally (Fig. 2a) and by coun y (Sup-
plemen a y Da a 2).
O e lap analysis
We e alua e how well local and global c i ical na u al asse s align spa-
ially wi h each o he , and wi h biodi e si y ( e es ial e eb a e
species AOH
27
; Supplemen a y No e 1 Sec ion 15) and cul u al di e si y
(p oxied by he numbe o Indigenous and non-mig an languages28;
Supplemen a y No e 1 Sec ion 16), o iden i y syne gies be ween hese
di e en po en ial p io i ies. To examine he le el o o e lap be ween
a eas iden i ied as c i ical o he 12 local NCP e sus he 2 global NCP,
we calcula e he a ea (globally and by coun y; Supplemen a y Da a 1)
whe e local NCP a e selec ed by he op imiza ion and global NCP a e
no , whe e global NCP a e selec ed by he op imiza ion and local NCP
a e no , and whe e bo h a e selec ed by hei espec i e op imiza ions
( he o e lap). To calcula e he species and languages ep esen ed by
c i ical na u al asse s, we coun he numbe o species whose AOH a ea
a ge s o e laps hese a eas, acco ding o bo h log-linea ep esen a-
ion a ge s (Supplemen a y Da a 3) and a ge s based on IUCN Red Lis
c i e ia (Supplemen a y Da a 4), and he numbe o languages pa ially
in e sec ing hese a eas (Supplemen a y Da a 5) globally and wi hin
each coun y ( o mo e de ail, see Supplemen a y No e 1 Sec ion 23).
Na u e Ecology & E olu ion | Volume 7 | Janua y 2023 | 51–61 59
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
Repo ing summa y
Fu he in o ma ion on esea ch design is a ailable in he Na u e Po -
olio Repo ing Summa y linked o his a icle.
Da a a ailabili y
All inal da a ou pu s a e a ailable on Open Science F amewo k72
(h ps://os .io/ 5xz7/).
Code a ailabili y
Code o gene a e op imiza ion esul s is also a ailable on Open Science
F amewo k72 (h ps://os .io/ 5xz7/).
Re e ences
1. Summa y o Policymake s o he Global Assessmen Repo on
Biodi e si y and Ecosys em Se ices o he In e go e nmen al
Science-Policy Pla o m on Biodi e si y and Ecosys em Se ices
(IPBES, 2019).
2. Dobson, A. P. e al. Ecology and economics o pandemic
p e en ion. Science 369, 379–381 (2020).
3. Diaz, S. e al. Assessing na u e’s con ibu ions o people. Science
359, 270–272 (2018).
4. Hole, D. G. e al. Make na u e’s ole isible o achie e he SDGs.
Glob. Sus ain. 5, e8 (2021).
5. Roe, S. e al. Con ibu ion o he land sec o o a 1.5 °C wo ld. Na .
Clim. Chang. 9, 817–828 (2019).
6. Wilson, E. O. Hal -Ea h: Ou Plane ’s Figh o Li e (WW No on &
Company, 2016).
7. Baillie, J. & Zhang, Y.-P. Space o na u e. Science 361,
1051 (2018).
8. Büsche , B. e al. Hal -Ea h o whole Ea h? Radical
ideas o conse a ion, and hei implica ions. O yx 51,
407–410 (2017).
9. Schmid -T aub, G. e al. In eg a ing clima e, biodi e si y, and
sus ainable land-use s a egies: inno a ions om China. Na l Sci.
Re . h ps://doi.o g/10.1093/ns /nwaa139 (2020).
10. Mye s, S. S. e al. Human heal h impac s o ecosys em al e a ion.
P oc. Na l Acad. Sci. USA 110, 201218656 (2013).
11. Chaplin-K ame , R. e al. Global modeling o na u e’s
con ibu ions o people. Science 366, 255–258 (2019).
12. Diaz, S. e al. The IPBES Concep ual F amewo k—connec ing
na u e and people. Cu . Opin. En i on. Sus ain. 14, 1–16 (2015).
13. Con en ion on Biodi e si y. Fi s D a o he Pos -2020 Global
Biodi e si y F amewo k. CBD/WG2020/3/5 (UN, 2021).
14. Ekins, P. The Kuzne s Cu e o he en i onmen and
economic g ow h: examining he e idence. En i on. Plan. A 29,
805–830 (1997).
15. S assbu g, B. B. N. e al. Global p io i y a eas o ecosys em
es o a ion. Na u e 586, 724–729 (2020).
16. Kennedy, C. M., Oaklea , J. R., Theobald, D. M., Ba uch-Mo do,
S. & Kiesecke , J. Managing he middle: a shi in conse a ion
p io i ies based on he global human modi ica ion g adien . Glob.
Chang. Biol. 25, 811–826 (2019).
17. Weiss, D. J. e al. A global map o a el ime o ci ies o assess
inequali ies in accessibili y in 2015. Na u e 553, 333–336 (2018).
18. Lebedys, A. & Yanshu, L. Fo es Finance. Con ibu ion o he
Fo es y Sec o o Na ional Economies, 1990–2011. FAO (2014).
19. McIn y e, P. B., Lie mann, C. A. R. & Re enga, C. Linking
eshwa e ishe y managemen o global ood secu i y
and biodi e si y conse a ion. P oc. Na l Acad. Sci. USA 113,
12880–12885 (2016).
20. Selig, E. R. e al. Mapping global human dependence on ma ine
ecosys ems. Conse . Le . 12, e12617 (2018).
21. Wo ld li es ock: ans o ming he li es ock sec o h ough he
sus ainable de elopmen goals. FAO h p://bea econ.com/
po olio-da a/ ao-ldg/ca1201en.pd (2018).
22. Fedele, G., Dona i, C. I., Bo nacelly, I. & Hole, D. G.
Na u e-dependen people: mapping human di ec use o na u e
o basic needs ac oss he opics. Glob. En i on. Change 71,
102368 (2021).
23. Summa y o Policymake s o he Me hodological Assessmen
o he Di e se Values and Valua ion o Na u e o he
In e go e nmen al Science-Policy Pla o m on Biodi e si y and
Ecosys em Se ices (IPBES) (IPBES, 2022)
24. Ba on, D. N. e al. Chap e 4. Value exp ession in
decision-making. Zenodo h ps://doi.o g/10.5281/
zenodo.6633784 (2022).
25. Noon, M. L. e al. Mapping he i eco e able ca bon in Ea h’s
ecosys ems. Na . Sus ain. 5, 37–46 (2021).
26. Keys, P. W., Wang-E landsson, L. & Go don, L. J. Re ealing in isible
wa e : mois u e ecycling as an ecosys em se ice. PLoS ONE 11,
e0151993 (2016).
27. B ooks, T. M. e al. Measu ing e es ial a ea o habi a (AOH)
and i s u ili y o he IUCN Red Lis . T ends Ecol. E ol. 34,
977–986 (2019).
28. Go en lo, L. J., Romaine, S., Mi e meie , R. A. & Walke -Painemilla,
K. Co-occu ence o linguis ic and biological di e si y in
biodi e si y ho spo s and high biodi e si y wilde ness a eas. P oc.
Na l Acad. Sci. USA 109, 8032–8037 (2012).
29. O’Conno , L. M. J. e al. Balancing conse a ion p io i ies o
na u e and o people in Eu ope. Science 372, 856–860 (2021).
30. Jung, M. e al. A eas o global impo ance o conse ing
e es ial biodi e si y, ca bon and wa e . Na . Ecol. E ol.
h ps://doi.o g/10.1101/2020.04.16.021444 (2021).
31. Sala, E. e al. P o ec ing he global ocean o biodi e si y, ood
and clima e. Na u e 592, 397–402 (2021).
32. Dine s ein, E. e al. A ‘Global Sa e y Ne ’ o e e se biodi e si y
loss and s abilize Ea h’s clima e. Sci. Ad . 6, eabb2824 (2020).
33. Chan, K. M. A., Gould, R. K. & Pascual, U. Edi o ial o e iew:
ela ional alues: wha a e hey, and wha ’s he uss abou ? Cu .
Opin. En i on. Sus ain. 35, A1–A7 (2018).
34. Spalding, M. e al. Mapping he global alue and dis ibu ion o
co al ee ou ism. Ma . Policy 82, 104–113 (2017).
35. Gunnell, K., Mulligan, M., F ancis, R. A. & Hole, D. G. E alua ing
na u al in as uc u e o lood managemen wi hin he
wa e sheds o selec ed global ci ies. Sci. To al En i on. 670,
411–424 (2019).
36. Bags ad, K. J., Semmens, D. J., Waage, S. & Win h op, R. A
compa a i e assessmen o decision-suppo ools o
ecosys em se ices quan i ica ion and alua ion. Ecosys . Se . 5,
27–39 (2013).
37. an de En , R. J., Tuinenbu g, O. A., Knoche, H.-R., Kuns mann, H.
& Sa enije, H. H. G. Should we use a simple o complex model o
mois u e ecycling and a mosphe ic mois u e acking? Hyd ol.
Ea h Sys . Sci. 17, 4869–4884 (2013).
38. Redhead, J. W. e al. Na ional scale e alua ion o he InVEST
nu ien e en ion model in he Uni ed Kingdom. Sci. To al En i on.
610–611, 666–677 (2018).
39. A kema, K. K. e al. Coas al habi a s shield people and
p ope y om sea-le el ise and s o ms. Na . Clim. Chang. 3,
913–918 (2013).
40. Cab al, P. e al. Assessing Mozambique’s exposu e o coas al
clima e haza ds and e osion. In . J. Disas e Risk Reduc . 23,
45–52 (2017).
41. Benez-Secanho, F. J. & Dwi edi, P. Does quan i ica ion o
ecosys em se ices depend upon scale ( esolu ion and ex en )?
A case s udy using he InVEST nu ien deli e y a io model in
Geo gia, Uni ed S a es. En i onmen s 6, 52 (2019).
42. Hoo man, D. A. P. e al. Reducing unce ain y in ecosys em
se ice modelling h ough weigh ed ensembles. Ecosys . Se .
53, 101398 (2022).
Na u e Ecology & E olu ion
A icle h ps://doi.o g/10.1038/s41559-022-01934-5
Ex ended Da a Fig. 5 | C i ical na u al asse s iden i ied h ough op imiza ion
a he global le el o 12 ‘local’ NCP. As in Fig. 1, he NCP accumula ion cu e
e lec s he o al a ea equi ed o main ain a ge le els o all 12 local NCP (bu in
his case op imized globally, no wi hin each coun y), wi h do ed lines deno ing
he a ea o c i ical na u al asse s (90% o he 12 NCP lis ed in Fig. 1a in 22% o land
a ea and 13% o EEZ a eas). The map shows c i ical na u al asse s o local NCP,
wi h da ke shades conno ing g ea e con ibu ion o agg ega e NCP.
