CONTRIBUTED PAPER
Bi ds ha a e mo e commonly encoun e ed in he wild
a ac highe public in e es online
John C. Mi e meie
1
| U i Roll
2
| Thomas J. Ma hews
3,4
|
Rica do Co eia
5,6
| Rich G enye
1
1
School o Geog aphy and The
En i onmen , Uni e si y o Ox o d,
Ox o d, UK
2
Mi ani Depa men o Dese Ecology,
The Jacob Blaus ein Ins i u es o Dese
Resea ch, Ben-Gu ion Uni e si y o he
Nege , Bee sheba, Is ael
3
School o Geog aphy, Ea h and
En i onmen al Sciences, and Bi mingham
Ins i u e o Fo es Resea ch, Uni e si y o
Bi mingham, Edgbas on,
Bi mingham, UK
4
CE3C –Cen e o Ecology, E olu ion
and En i onmen al Changes/Azo ean
Biodi e si y G oup and Uni . dos Aço es –
Dep o de Ciências e Engenha ia do
Ambien e, Ang a do He oísmo, Po ugal
5
Helsinki Lab o In e disciplina y
Conse a ion Science (HELICS),
Depa men o Geosciences and
Geog aphy, Uni e si y o Helsinki,
Helsinki, Finland
6
Helsinki Ins i u e o Sus ainabili y
Science (HELSUS), Uni e si y o Helsinki,
Helsinki, Finland
Co espondence
Rich G enye , School o Geog aphy and
The En i onmen , Uni e si y o Ox o d,
Ox o d, OX1 3QY, UK.
Email: [email p o ec ed].ac.uk
Abs ac
La ge body size, he de ining cha ac e is ic o “cha isma ic mega auna,”is o en
iewed as he mos signi ican co ela e o highe public in e es in species. How-
e e , common, local species (many o which a e no la ge) can also gene a e
public in e es . We explo ed he ela i e impo ance o body size e sus local
occu ence in pa e ns o online in e es in bi ds using a la ge sample o digi al
human-wildli e in e ac ions (367 million Wikipedia page iews) ha included
mo e han 10,000 bi d species and a ange o cul u al and geog aphic con ex s
( ep esen ed by 25 Wikipedia language edi ions). We compa ed in e es in
Wikipedia, as measu ed by page iews, wi h a bi d's body size and i s egional
obse a ion equency (using da a om eBi d.o g). We ound ha local species
(i.e., hose ha occu in he wild in he coun y esponsible o he majo i y o a
Wikipedia language edi ion's page iews) a ac mo e page iews han global spe-
cies. Bo h body size and obse a ion equency had a posi i e co ela ion wi h
Wikipedia page iews ac oss languages, bu eBi d obse a ion equency
explained mo e o he a iance in page iews on a e age. In a model ha
included bo h obse a ion equency and body size, obse a ion equency was a
signi ican ly be e p edic o o page iews han body size in 24 o 25 languages.
Ou esul s demons a e ha he oppo uni y o encoun e bi ds in he wild is a
signi ican co ela e o inc eased online in e es in bi ds ac oss mul iple linguis-
ic and geog aphic con ex s. This ela ionship p o ides insigh in o why some
species a ac g ea e in e es han o he s and emphasizes he o e looked po en-
ial o common species in conse a ion ma ke ing.
KEYWORDS
bi d conse a ion, ci izen science, conse a ion cul u omics, eBi d, geog aphy, Wikipedia
1|INTRODUCTION
Unde s anding why some species command g ea e pub-
lic in e es han o he s ha e long been o in e es o
conse a ionis s (Kelle , 1982; Lo ime , 2007; Macdon-
ald e al., 2015). Iden i ying he biological and ecological
ai s o species ha co ela e wi h inc eased public in e -
es can bene i conse a ion ma ke ing and be used o
Recei ed: 4 July 2020 Re ised: 22 No embe 2020 Accep ed: 23 No embe 2020
DOI: 10.1111/csp2.340
This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided
he o iginal wo k is p ope ly ci ed.
© 2021 The Au ho s. Conse a ion Science and P ac ice published by Wiley Pe iodicals LLC. on behal o Socie y o Conse a ion Biology
Conse a ion Science and P ac ice. 2021;3:e340. wileyonlinelib a y.com/jou nal/csp2 1o 11
h ps://doi.o g/10.1111/csp2.340
iden i y po en ial lagship species (Smi h, Ve íssimo,
Isaac, & Jones, 2012; Ve issimo e al., 2013). P e ious
s udies ha e e alua ed he ole o biological ai s such as
colo a ion (Liško á & F yn a, 2013), enomousness (Roll
e al., 2016), and he pe cei ed a ac i eness o species
(Gunn ho sdo i , 2001) in de e mining public in e es .
While a a ie y o ai s ha e been ecognized as co ela -
ing wi h inc eased in e es , la ge body size is equen ly
iden i ied as he mos impo an (Be i, Monsa a ,
Munk, Ja ie, & S enning, 2020; Clucas, McHugh, &
Ca o, 2008; Macdonald e al., 2015). Indeed, he wide-
sp ead use o he e m “cha isma ic mega auna” o
desc ibe species ha a ac high public in e es sugges s
ha la ge body size is o en iewed as synonymous wi h
inc eased public in e es in he conse a ion communi y.
The impo ance o la ge body size sugges s ha di ec
in e ac ions wi h animals in he wild may be ela i ely
unimpo an o de e mining in e es in species. While
la ge species a e s ongly ep esen ed in media (Clucas
e al., 2008) and cap i e en i onmen s (Ma in, Lu biecki,
Joy, & Mooe s, 2014), hey a e o en a e and in e-
quen ly encoun e ed in he wild. In ex eme examples,
species such as ige s (Pan he a ig is) and gian pandas
(Ailu opoda melanoleuca) a ac subs an ial public in e -
es despi e only a iny mino i y o people e e obse ing
hem in he wild. Con e sely, di ec encoun e s wi h spe-
cies in he wild do seem o d i e public in e es in some
si ua ions. The commonness o bu e ly and bi d species
in he UK and Poland co ela es wi h he a en ion hese
species ecei e online (Zmiho ski, Dzia ska-Palac,
Spa ks, & T yjanowski, 2013). In B azil, species ha o e -
lap wi h a eas o highe human popula ion densi y
ecei e g ea e in e es online, as measu ed by Google
sea ch equency (Co eia, Jepson, Malhado, &
Ladle, 2016). This ela ionship be ween people's abili y o
di ec ly encoun e a species in he wild and hei in e es
in i has ele ance o conse a ion. I di ec encoun e s
a e ela i ely unimpo an , hen emphasizing la ge and
cha isma ic species as well as i ual encoun e s h ough
a ious o ms o media should be he op p io i y o
engaging g ea e public in e es . In his scena io, public
in e es in a species is unlikely o decline i he species
becomes a e o e en goes ex inc in he wild. Con-
e sely, i di ec encoun e s wi h wild animals a e impo -
an o gene a ing public in e es , hen app oaches ha
emphasize di ec encoun e s wi h species in he wild may
be mo e e ec i e o inc easing public engagemen in
conse a ion. Fu he mo e, i public in e es is closely
linked o wild encoun e a es hen he declining num-
be s o wild animals (Di zo e al., 2014) and ends o
inc easing u baniza ion and de achmen om na u e
(Mille , 2005; Soga & Gas on, 2016) could lead o
dec eased in e es in some species.
Conse a ion cul u omics is a new esea ch a ea ha
uses online digi al da a o in es iga e ques ions a ound
human-na u e in e ac ions ha a e ele an o conse a-
ion (Ladle e al., 2016). P e ious s udies ha e used con-
se a ion cul u omic me hods o assess people's in e es
in ep iles (Roll e al., 2016) and B azilian bi ds (Co eia
e al., 2016), and o compa e seasonal pa e ns in biodi-
e si y awa eness (Mi e meie , Roll, Ma hews, &
G enye , 2019). He e we used hese me hods o assess he
ela ionship be ween online in e es , body size and he
equency wi h which people ha e di ec encoun e s spe-
cies in he wild. Using bi ds as case s udy, we assessed
pa e ns o public in e es (as measu ed by Wikipedia
page iews) ac oss a la ge numbe o digi al in e ac ions
and a ange o cul u al con ex s (25 di e en Wikipedia
languages). We compa ed Wikipedia page iew da a o
in o ma ion on he equency wi h which bi ds a e
obse ed in he wild de i ed om eBi d, an online da a-
base o a ian dis ibu ions. Wi h mo e han 700 million
eco ds, eBi d is he wo ld's la ges biodi e si y- ela ed
communi y science p ojec and p o ides occu ence da a
o bi ds a ound he wo ld (Le a ich & Ligocki, 2020; Sul-
li an e al., 2009; Wood, Sulli an, Ili , Fink, &
Kelling, 2011). Mul iple s udies ha e used eBi d da a o
moni o pa e ns in a ian dis ibu ion and abundance a
a a ie y o scales (eBi d, 2020).
We es ed h ee hypo heses a ound he impo ance o
di ec encoun e s in gene a ing online in e es in bi ds.
Fi s , we assessed whe he he geog aphic dis ibu ion o
bi ds impac ed hei Wikipedia page iews. I wild
encoun e s a e unimpo an , we p edic ed ha he local
occu ence o bi d species would ha e a minimal impac
on hei page iews; i di ec encoun e s a e impo an ,
we p edic ed page iews would be skewed owa d local
species (i.e., hose ha occu ed in he wild in a gi en
egion). Second, we es ed o co ela ions be ween
page iews and body size/eBi d equency. We p edic ed a
posi i e co ela ion be ween page iews and body size i
indi ec encoun e s a e impo an and a posi i e co ela-
ion be ween page iews and eBi d equency i di ec
encoun e s wi h species a e impo an . Thi d, in
ins ances whe e we had bo h body size and eBi d e-
quency da a, we es ed he ela i e impo ance o bo h
a iables in p edic ing Wikipedia page iews in a com-
bined model. He e, we p edic ed ha body size would
explain mo e a ia ion in page iews i di ec encoun e s
a e less impo an and eBi d equency would explain
mo e a ia ion i di ec encoun e s a e mo e impo an .
P e iously, Mi e meie e al. (2019) iden i ied empo al
co espondence be ween eBi d sigh ings and Wikipedia
page iews o a sample o mig a o y bi ds in I aly, Ge -
many, Sweden, and he Uni ed S a es. Thei s udy demon-
s a ed ha , o some mig a o y bi ds, he physical p esence
2o 11 MITTERMEIER ET AL.
o he species in a egion co ela ed wi h inc eased public
in e es . This s udy builds on hese indings by e alua ing
he ela ionship be ween he obse a ion equency and
online in e es ac oss a much la ge ange o languages
(25 language edi ions), geog aphic egions (25 egions
including bo h empe a e and opical coun ies) and spe-
cies (mo e han 10,000 bi d species, nea ly he en i e global
di e si y). Thus, we p o ide a gene alized unde s anding o
he ela ionship be ween wild encoun e equency and
public in e es in bi ds i espec i e o he phenomenon o
mig a ion and es he impo ance o he ela ionship ac oss
a ange o linguis ic and geog aphic con ex s.
2|METHODS
2.1 |Da a selec ion and ex ac ion
We used he numbe o page iews ha a page ecei es in
Wikipedia as a measu e o online in e es (e.g., Mi e meie
e al., 2019; Roll e al., 2016). Wikipedia does no include
geog aphic in o ma ion wi h i s page iews, bu Wikipedia
edi ions a e cons uc ed in di e en languages and sum-
ma y da a a e a ailable ha lis he p opo ion o
page iews each language edi ion ecei es by coun y
(Zach e, 2020). Following p e ious s udies, we used hese
coun y-le el summa ies as a coa se p oxy o geog aphy
by pai ing Wikipedia language edi ions wi h he geo-
g aphic egion ha accoun s o he majo i y o a language
edi ion's page iews (Gene ous, Fai child, Deshpande, Del
Valle, & P iedho sky, 2014; Mi e meie e al., 2019).
We selec ed 25 Wikipedia edi ions o non-a i icial lan-
guages ha , as o June 22, 2019, had o e 100,000 a icles, a
Wikipedia edi ing dep h highe han 10 (a measu e o he
language edi ion's quali y; Wikimedia 2019), and mo e han
50% o he language's page iews o igina ing om a single
coun y (Appendix S1; Wikipedia, 2020; Zach e, 2020).
Selec ing language edi ions wi h mo e han 50% o hei
page iews o igina ing om a single coun y p o ided
highe con idence in he geog aphic o igin o he page iews
in ou da a se . I also esul ed in se e al widely spoken lan-
guages wi h la ge Wikipedia edi ions no being included in
ou da a se (e.g., Spanish, English).
We iden i ied pages o bi d species in Wikipedia
using he Wikida a Que y Se ice (h ps://que y.
wikida a.o g/) o ex ac a lis o en i ies agged wi h an
eBi d axon ID (Wikida a p ope y: P3444) on June
23, 2019. Wikida a is a seconda y da abase ha collec s
s uc u ed da a o Wikimedia p ojec s, including all
Wikipedia language edi ions (h ps://www.wikida a.o g).
We c oss- e e enced ou lis o Wikida a en i ies wi h he
eBi d/Clemen s global a ian axonomy (Clemen s
e al., 2018) o ensu e ha non-bi d pages we e no
included. We ob ained page links o Wikipedia pages in
languages ha me ou c i e ia and downloaded
page iews o each page o he pe iod be ween July
1, 2015 and June 22, 2019 (1,453 days) using “page iews”
in R (Keyes & Lewis, 2016).
To ob ain eBi d da a, we downloaded he eBi d Basic
Da a se ( e sion Ap il 2019) wi h eco ds o all species
and all yea s (eBi d, 2019) o he egions associa ed wi h
each o ou 25 Wikipedia language edi ions. In mos
cases, we de ined a egion as a single coun y, howe e ,
in a e ins ances whe e he dis ibu ion o a language
co esponded s ongly o a speci ic subna ional egion,
we downloaded da a o ha egion a he han o he
en i e coun y (e.g., Tamil Wikipedia was pai ed wi h
eBi d eco ds om Tamil Nadu a he han all o India).
We limi ed ou analyses o egions wi h a minimum o
10,000 unique sampling e en s in eBi d, and o species
ha appea ed in mo e han 10 sampling e en s in he
egion's eBi d da a se . These h esholds helped minimize
biases p esen in smalle eBi d da a se s and educe
ins ances whe e inco ec ly iden i ied species had no ye
been emo ed ia eBi d's e iew p ocess (Wood
e al., 2011). Bi d pages in each Wikipedia language edi-
ion we e assigned as ei he “local”i hey occu ed on
he eBi d lis o he associa ed egion o “global”i hey
did no (eBi d, 2019). Fo local species, we calcula ed he
obse a ion equency o a species as he o al epo s o
ha species in he egion di ided by he o al unique
sampling e en s in he egion (Sulli an e al., 2009). This
app oach p o ided a ough o e all measu e o obse a-
ion equency o all local species in egion. I did no
accoun o seasonal a ia ions (such as o mig a o y
species) o o how species we e dis ibu ed wi hin he
egion. Since global species by de ini ion had no sam-
pling e en s in a egion, hey did no ha e an eBi d e-
quency o ha egion in ou da a se . We ob ained bi d
body mass da a om Dunning (2008).
2.2 |Da a analysis
We explo ed he ela ionship be ween Wikipedia
page iews, eBi d sigh ings, and a ian body mass ac oss
each o he 25 language- egion pai s in ou da a se . Since
language edi ions a ied subs an ially in hei iews, we
assessed pa e ns in each language sepa a ely. Fu he -
mo e, since ou da a we e spa se (we only had eBi d e-
quency da a o local species, and we did no ha e body
mass da a o all species) we analyzed da a o each lan-
guage in ou sepa a e analyses. Fi s , o all o he bi d
Wikipedia pages in a language, we looked a he o e all
pa e n be ween page iews o global as opposed o local
species and compa ed he median page iews o local
MITTERMEIER ET AL.3o 11
and global species using a Wilcoxon ank-sum es . Sec-
ond, o all he bi ds wi h Wikipedia pages in a language
o which we had eBi d sigh ing equency (i.e., he local
species ha occu ed in he egion associa ed wi h he
language), we es ed he ela ionship be ween eBi d
sigh ing equency as a p edic o and Wikipedia
page iews as a esponse a iable using a simple linea
eg ession model. Thi d, o all he bi ds wi h Wikipedia
pages in a language o which we had body mass da a
( his included bo h local and global species), we modeled
he ela ionship be ween body mass (p edic o ) and
Wikipedia page iews ( esponse) using a simple linea
eg ession model. Since we had body mass da a o bo h
global and local species in each language, we es ed
whe he a species being classi ied as local o global
in luenced he ela ionship be ween body mass and
page iews using an analysis o co a iance (ANCOVA).
Fou h, o he bi ds wi h Wikipedia pages in a language
o which we had bo h eBi d equency da a and body
mass da a (i.e., local species wi h body mass da a) we
used a combined linea eg ession model wi h eBi d e-
quency and body mass as p edic o s o Wikipedia
page iews. In he combined model, we compa ed he el-
a i e in luence o eBi d equency and body mass as p e-
dic o s o Wikipedia page iews using a ia ion
pa i ioning (Oksanen e al., 2017). Fo all analyses, we
no malized a iables using a logi ans o ma ion o p o-
po ions (eBi d equency) and a log ans o ma ion o
coun s and con inuous a iables (Wikipedia page iews,
body mass). Final models we e es ed o assump ions o
no mali y and homoscedas ici y o esiduals.
3|RESULTS
3.1 |Da a selec ion and ex ac ion
Ou da a se con ained 78,415 Wikipedia pages o bi ds
ac oss 25 languages (Appendix S2). In o al, 10,174 bi d
species had a leas one Wikipedia page in ou da a se
(96.1% o he o al bi d species; Clemen s e al., 2018).
Numbe s o bi d pages pe language a ied om
103 (Hindi) o 10,103 in Du ch (mean pages pe language
3,137, SD 3,034). Wikipedia pages in ou da a se ecei ed
367 million iews o e he sampling pe iod ( iews pe
language 417,000–70.9 million; mean 14.7 million, SD
18.5 million). Wi h he excep ion o B azil (which
accoun ed o 79.7% o he page iews o Po uguese
Wikipedia), all o ou geog aphic egions we e in Eu ope
(15 egions) and Asia (nine egions).
Ou eBi d da a se con ained 2.3 million unique sam-
pling e en s wi h eco ds o 4,340 bi d species. Unique
sampling e en s pe egion anged om 10,600 (Sou h
Ko ea) o 817,000 (India; mean 91,900, SD 169,000). eBi d
species ichness pe egion anged om 313 species
(Czech Republic) o 1,716 (B azil; mean o e all 615, SD
374). No all species ha occu ed in a egion had a
Wikipedia page in he associa ed language. The p opo -
ion o eBi d species in a egion wi h Wikipedia pages
a ied om 7.62% o species (Hindi—India) o >98% o
species in ele en language– egion pai s (mean o e all
82.5%, SD 25.6%). We ob ained body mass da a o 9,406
species.
3.2 |Wikipedia page iews o local
s. global species
Ou da a se con ained many mo e Wikipedia pages ha
we classi ied as global (70,250 global s. 8,165 local).
Despi e his, pages o local species ecei ed mo e iews
han hose o global species (218 million s. 149 million
page iews; mean page iews o local species 26,700, SD
74,900; mean page iews o global species 2,120, SD
14,300). The mean page iews o local species was highe
han he mean page iews o global species in all 25 lan-
guages, wi h he di e ence be ween he means being s a-
is ically signi ican in all bu wo languages (Wilcoxon
ank-sum p< .01; mean page iews local 723–131,000;
mean page iews global 106–16,600; Figu e 1,
Appendix S3). The e ec size o he di e ence in mean
page iews o local as opposed o global species was
ei he la ge o mode a e in 20 languages (la ge de ined as
>0.5, mode a e as 0.3–0.5).
3.3 |eBi d equency as a p edic o o
Wikipedia page iews
Fo local bi ds in ou da a se (i.e., hose o which we
had eBi d equency da a) we es ed he ela ionship
be ween eBi d equency as a p edic o and Wikipedia
page iews as a esponse a iable using a linea eg ession
model. The numbe o species on a egion's eBi d lis
wi h Wikipedia pages anged om 82 o 1,138 (mean
399, SD 216; o al local pages ac oss all languages 8,165).
eBi d equency showed a signi ican posi i e co ela ion
wi h Wikipedia page iews in all 25 languages (p< .01).
Ac oss languages, a linea eg ession model wi h eBi d
equency as a p edic o explained be ween 6.9 and 49.0%
o he a iance in Wikipedia page iews o bi d pages
(adj. R
2
= .07–.49, mean 0.25, SD .13; Figu e 2, Table 1).
The mos equen ly epo ed bi ds in eBi d in each
language ecei ed a high p opo ion o he Wikipedia
page iews o bi ds. Among local species, he 20 mos e-
quen ly eco ded species in eBi d accoun ed o 20.2 o
4o 11 MITTERMEIER ET AL.
49.8% o he page iews o local bi d species, while mak-
ing up 3.14 o 20.1% o he local bi d species pages (mean
p opo ion o iews 40.0%, SD 8.24%; mean p opo ion o
species 13.1%, SD 5.09%). This pa e n was also e iden
when conside ing he o al bi d Wikipedia pages in a lan-
guage (i.e., bo h local and global). Ac oss languages, he
20 species wi h he highes eBi d equency accoun ed o
3.16 o 19.6% o he o al Wikipedia page iews o bi d
species, while consis ing o only 0.20 o 4.20% o he o al
bi d species pages (mean p opo ion o iews 13.36%, SD
4.66%; mean p opo ion o species 1.29%, SD 1.14%;
Appendix S4).
FIGURE 1 Bi d species ha occu in he egion esponsible o he majo i y o a Wikipedia language edi ion’s page iews (“local”
species) a ac mo e page iews han bi d species ha do no occu in he egion (“global”species). Poin s indica e ou lie s occu ing > 1.5x
he in e qua ile ange beyond he median. S a is ical signi icance o he di e ence be ween means calcula ed using a Wilcoxon ank-sum
es wi h p < 0.01; wo language edi ions whe e di e ence be ween means was no s a is ically signi ican a e ma ked wi h an as e isk
(Malay and Hindi)
MITTERMEIER ET AL.5o 11

FIGURE 2 eBi d obse a ion equency co ela es posi i ely wi h Wikipedia page iews o local bi d species ac oss mul iple
Wikipedia language edi ions. In a linea model wi h eBi d obse a ion equency (logi ans o med) as a p edic o o page iews (log
ans o med), obse a ion equency p edic ed be ween 49% and 6.9% o he a iance in page iews ( op h ee ows show 9 language
edi ions wi h he highes adj. R2 in he linea model; bo om ow shows 3 languages wi h he lowes adj. R2 in he linea model; see
Table 1 o ull esul s)
6o 11 MITTERMEIER ET AL.
3.4 |Body mass as a p edic o o
Wikipedia page iews
Fo speciesinou da ase o whichwehadbodymassda a,
we es ed he ela ionship be ween body mass (as a p edic-
o ) and Wikipedia page iews (as a esponse) using a linea
eg ession model. The numbe o species pe language o
which we had body mass da a a ied om 99 o 9,403 (mean
2,993, SD 2,875; 74,830 pages o al). Body mass showed a sig-
ni ican posi i e co ela ion wi h Wikipedia page iews in
21 o 25 languages (p< .01; Table 2). In languages whe e he
ela ionship was signi ican , body mass desc ibed be ween
2.1 and 24.5% o he a iance in page iews (p< .01, adj.
R
2
=.02–.25, mean 0.11, SD 0.08). The in e ac ion be ween
local–global and body mass was signi ican in 12 o he
25 languages (p< .01). In ins ances whe e he ela ionship
was signi ican , body mass had a s onge posi i e ela ion-
ship wi h page iews o global species han o local ones in
all bu one language (local mean coe icien 0.18, SD 0.11,
global mean coe icien 0.34, SD 0.05; Appendix S5).
3.5 |eBi d equency e sus body mass
as p edic o s o Wikipedia page iews o
local species
Fo all species o which we had bo h body mass and
eBi d equency da a (i.e., local species wi h body mass
da a; 2,563 species wi h 8,014 pages), we i ed a linea
eg ession model ha included bo h eBi d equency and
body mass as p edic o s o Wikipedia page iews. This
join model was signi ican in all 25 languages (p<.01)
and explained be ween 10.5 and 54.9% o he a iance in
TABLE 1 Model esul s o a linea model wi h eBi d
obse a ion equency (logi ans o med) as a p edic o o
page iews (log ans o med) o pages o local bi d species in 25
Wikipedia language edi ions
Language
Es ima e
(β)SE p- alue
Model
adj. R
2
Ge man 0.50 0.03 7.16E−51 .49
F ench 0.42 0.03 9.58E−40 .35
Du ch 0.42 0.03 1.61E−40 .44
Russian 0.50 0.05 3.57E−24 .20
I alian 0.40 0.04 5.06E−21 .25
Polish 0.53 0.04 6.82E−28 .37
Japanese 0.49 0.03 2.02E−39 .35
Chinese 0.23 0.02 1.83E−19 .16
Po uguese 0.38 0.02 1.71E−53 .19
Se bian 0.29 0.06 1.62E−06 .14
Ca alan 0.42 0.03 9.62E−38 .40
No wegian 0.43 0.04 5.53E−28 .34
Indonesian 0.35 0.04 5.34E−18 .14
Finnish 0.41 0.03 7.01E−35 .43
Ko ean 0.19 0.05 7.67E−04 .07
Hunga ian 0.45 0.04 1.89E−20 .28
Czech 0.48 0.04 1.51E−28 .40
Basque 0.24 0.02 8.19E−20 .25
Malay 0.19 0.03 2.68E−08 .07
Danish 0.50 0.04 2.55E−29 .38
Heb ew 0.35 0.05 1.26E−12 .16
G eek 0.26 0.06 8.94E−06 .11
Thai 0.18 0.04 2.57E−05 .10
Hindi 0.23 0.06 5.80E−04 .13
Tamil 0.24 0.03 8.68E−12 .16
TABLE 2 Model esul s o a linea model wi h body mass (log
ans o med) as a p edic o o page iews (log ans o med) o pages
o local and global bi d species in 25 Wikipedia language edi ions
Language
Es ima e
(β)SE p- alue
Model
adj. R
2
Ge man 0.36 0.01 5.1E−167 .17
F ench 0.39 0.01 .00 .20
Du ch 0.42 0.01 .00 .25
Russian 0.28 0.01 5.2E−89 .12
I alian 0.32 0.01 2.8E–104 .12
Polish 0.26 0.02 1.4E–63 .10
Japanese 0.18 0.02 4.2E–16 .04
Chinese 0.36 0.01 9.2E–130 .21
Po uguese 0.43 0.01 1.9E–289 .23
Se bian 1.0E-03 0.05 0.99 .00
Ca alan 0.13 0.01 3.80E–32 .03
No wegian 0.34 0.02 4.00E–62 .13
Indonesian −0.05 0.03 0.06 .00
Finnish 0.39 0.01 1.2E–291 .20
Ko ean 0.13 0.03 2.2E–05 .04
Hunga ian 0.41 0.01 1.5E–296 .21
Czech 0.17 0.03 9.6E–12 .04
Basque 0.07 0.01 4.5E−53 .03
Malay 0.09 0.03 2.3E−04 .02
Danish 0.13 0.03 8.9E−08 .03
Heb ew 0.12 0.02 1.1E−07 .02
G eek 0.00 0.04 0.94 .00
Thai 0.11 0.04 0.01 .02
Hindi 0.03 0.10 0.77 −.01
Tamil 0.18 0.03 4.5E−09 .07
MITTERMEIER ET AL.7o 11
Wikipedia page iews o bi ds ac oss languages (adj.
R
2
= .10–.55, mean 0.33, SD 0.13). We used pa i ion o
a iance using pa ial linea eg ession o assess he ela-
i e con ibu ions o body mass and eBi d equency as
explana o y a iables. Fo all languages excep one
(Po uguese), eBi d equency explained mo e o he a i-
ance in page iews han body mass (adj. R
2
equency j
mass 0.07–0.52, mean 0.28, SD 0.13; adj. R
2
mass j
equency −0.01–0.26, mean 0.08, SD 0.06; Table 3).
O en he di e ence be ween he wo was subs an ial; in
14 languages eBi d equency explained o e h ee imes
mo e a iance han body mass.
4|DISCUSSION
We ou lined h ee hypo heses o assessing he ela i e
impo ance o a physical ai (body size) e sus he
po en ial o encoun e a species in he wild in
de e mining public in e es in bi ds. Fo all h ee hypo h-
eses, ou esul s s ongly suppo ed he impo ance o
encoun e equency o e body size in de e mining public
in e es online. Fi s , we obse ed a clea geog aphical
pa e n in he dis ibu ion o Wikipedia page iews o
bi ds, wi h local species a ac ing mo e in e es han
pages o global species ac oss languages (Figu e 1). Sec-
ond, while bo h body size and eBi d equency (Figu e 2;
Tables 1 and 2) co ela ed posi i ely wi h inc eased
page iews; eBi d equency explained mo e o he a i-
ance in page iews ac oss languages (mean adj. R
2
eBi d
equency 0.25, SD 0.13; mean adj. R
2
body size 0.11, SD
0.08). Thi d, when eBi d equency and body size we e
compa ed di ec ly in a combined model, he o me
explained signi ican ly mo e a iance in page iews in
24 o 25 languages in ou da a se (mean adj. R
2
e-
quency jmass 0.28, SD 0.13; mean adj. R
2
mass j e-
quency mean 0.08, SD 0.06; Table 3). In se e al language
edi ions, he di e ence be ween eBi d equency and
TABLE 3 Model adjus ed R
2
esul s
o a linea model wi h bo h body mass
and eBi d equency as p edic o s o
page iews o pages o local bi d species
in 25 Wikipedia language edi ions
Language Model adj. R
2
F equency jmass Mass j equency Di e ence
Ge man .55 0.52 0.06 0.46
Du ch .46 0.44 0.02 0.42
Czech .45 0.45 0.05 0.39
Ca alan .43 0.42 0.04 0.39
Polish .41 0.4 0.05 0.35
Finnish .51 0.42 0.08 0.34
Japanese .38 0.36 0.03 0.33
Danish .47 0.39 0.1 0.29
F ench .47 0.39 0.12 0.28
No wegian .39 0.33 0.05 0.28
Hunga ian .32 0.29 0.05 0.24
Basque .29 0.27 0.05 0.22
I alian .36 0.3 0.11 0.19
Heb ew .22 0.2 0.06 0.13
Hindi .12 0.12 −0.01 0.13
Indonesian .13 0.13 0.0 0.13
Russian .32 0.23 0.12 0.11
G eek .18 0.15 0.08 0.07
Se bian .24 0.17 0.11 0.06
Thai .25 0.2 0.14 0.06
Chinese .33 0.21 0.17 0.04
Tamil .33 0.22 0.17 0.04
Malay .13 0.09 0.06 0.04
Ko ean .11 0.07 0.04 0.03
Po uguese .44 0.18 0.26 −0.08
No e: Adjus ed R
2
o F equency jMass and Mass jF equency calcula ed using a ia ion pa i ioning
(Oksanen e al., 2017).
8o 11 MITTERMEIER ET AL.
body size as p edic o s o Wikipedia page iews in his
combined model was subs an ial: in Ge man Wikipedia
eBi d equency explained 52% o he a iance in
Wikipedia page iews o local species when used as he
sole p edic o in he combined linea model, while mass
explained only 6% o he a iance.
Fi ing he es ablished wisdom in conse a ion
(e.g., Be i e al., 2020) ou esul s ound ha body size
co ela ed posi i ely wi h online in e es . No ably, some
o he mos - iewed pages in se e al languages we e a ely
encoun e ed, global species. In a s iking example, he
Dodo (Raphus cuculla us), a ela i ely la ge bodied,
ex inc species, ecei ed he mos page iews o any bi d
page in bo h F ench and Po uguese Wikipedias. Thus,
indi ec ep esen a ions such as appea ances in media
and popula cul u e clea ly d i e in e es in some species
and can e en de e mine he mos iewed species pages
o e all. Howe e , when assessed ac oss la ge numbe s
o species, ou esul s clea ly showed ha body size was
seconda y o he po en ial o di ec wild encoun e s in
p edic ing Wikipedia page iews o bi ds. While high
public in e es in locally common species has been
shown in p e ious s udies using online da a (e.g., Co eia
e al., 2016; Zmiho ski e al., 2013), we demons a ed ha
his pa e n is p esen ac oss a wide ange o cul u al and
geog aphic con ex s.
The posi i e co ela ion be ween wild encoun e a es
and online in e es could p o e ele an o conse a ion
policy in wo ways. Fi s , om a me hodological s and-
poin i demons a es he impo ance o aking encoun e
equency in o accoun when measu ing public in e es
in species. Second, i highligh s he impo ance o local
and equen ly encoun e ed species in a ac ing people's
a en ion o he na u al wo ld. Few i any o he 20 mos
equen ly obse ed bi ds in he egions in ou da a se
would quali y as “cha isma ic mega auna”unde mos
c i e ia, and ye hey accoun ed o be ween 10 and 20%
o he o al bi d page iews in mos languages. As a esul
o he in e es hey a ac , hese commonly obse ed spe-
cies could ac as en y poin s o people's in e ac ions
wi h biodi e si y. Conse a ion ini ia i es ha imp o e
oppo uni ies o people o in e ac wi h common, local
species (such as he de elopmen o public pa ks and g e-
enspaces) could help inc ease public suppo o
conse a ion.
The e a e se e al ca ea s o conside when in e -
p e ing ou esul s. Ou me hod o ma ching Wikipedia
languages o egions p o ided only ough geog aphic es-
olu ion. Fo each language, a pe cen age o page iews
came om ou side o he associa ed egion. Fu he mo e,
species ha a e equen ly encoun e ed in one pa o a
egion may be absen om ano he , pa icula ly in la ge
egions such as B azil. Using cul u omic esou ces wi h
mo e ine-scale geog aphic esolu ion could be an
app oach o in es iga ing hese pa e ns a ine scales
(e.g., Co eia e al., 2016). eBi d da a also ha e impo an
biases (Sulli an e al., 2009). eBi d use s may be mo e
likely o isi ce ain loca ions, epo some species o e
o he s, and he co e age and ypes o use s may di e
be ween egions. I is also impo an o conside he ways
in which people use Wikipedia as opposed o o he
online esou ces (Co eia e al., 2021). Wikipedia
page iews e lec people's a en ion and hei sea ches o
addi ional knowledge, bu no necessa ily hei p e e -
ences. High in e es in a species could equally esul om
people liking a species as i could om people conside -
ing i is a pes and wan ing o emo e i . Thus, addi ional
con ex is equi ed unde s and he d i e s behind high
page iews o species. Finally, i is possible ha in e ac-
ions wi h wild animals a e mo e impo an o bi ds
han o o he axonomic g oups. Bi ds a e mo e easily
obse ed and iden i ied in he wild han many o he
o ganisms and gene a e unique o ms o human-na u e
in e ac ion h ough ac i i ies such as bi dwa ching and
bi d- eeding (Cocke , Tipling, Elphick, &
Fanshawe, 2013). I is possible ha obse a ion equency
may ha e a s onge posi i e co ela ion wi h online
in e es o bi ds han i does o o he g oups o o gan-
isms. Fu u e s udies may be able o add ess hese ca ea s
and expand upon ou esul s o in es iga e he mecha-
nisms unde lying he high online in e es in local bi d
species.
Ou esul s highligh ed he no el insigh s ha a e
possible using he la ge analy ical scales enabled by con-
se a ion cul u omics me hods. The ela ionship be ween
sigh ing equency and inc eased public in e es may be
obscu ed when public in e es in species is assessed using
small o un ep esen a i e samples. In smalle samples,
ai s o ex eme ou lie s (such as he Dodo) may ou -
weigh b oadscale pa e ns in he da a. Thus, ou indings
highligh ed he alue o la ge-scale analyses ac oss many
species and o e b oad cul u al and geog aphic con ex s.
Access o la ge digi al sou ces such as Wikipedia,
oge he wi h powe ul new analy ical ools hold much
p omise o u he no el esea ch insigh s using conse -
a ion cul u omics (Co eia e al., 2021).
Fu u e s udies can build on ou esul s by using mo e
ocused s udies o in es iga e he causal ela ionships
be ween public in e es and people's wild encoun e s
wi h species. I he e is a causal ela ionship be ween
wild encoun e s wi h species and inc eased in e es , as
he co ela ions we iden i ied sugges , his ela ionship
may hin a an upcoming challenge o conse a ion:
popula ions o many species a e declining a ala ming
a es esul ing in ewe oppo uni ies o people o
encoun e species in he wild (Di zo e al., 2014; Inge
MITTERMEIER ET AL.9o 11