RESEARCH ARTICLE
Re ealing Be a-Di e si y Pa e ns o B eeding
Bi d and Liza d Communi ies on Inunda ed
Land-B idge Islands by Sepa a ing he
Tu no e and Nes edness Componen s
Xing eng Si
1
, And és Baselga
2
, Ping Ding
1
*
1College o Li e Sciences, Zhejiang Uni e si y, Hangzhou, Zhejiang, 310058, China, 2Depa amen o de
Zoología, Facul ad de Biología, Uni e sidad de San iago de Compos ela, c/Lope Gómez de Ma zoa s/n,
15782, San iago de Compos ela, Spain
*[email p o ec ed]n
Abs ac
Be a di e si y desc ibes changes in species composi ion among si es in a egion and has
pa icula ele ance o explaining ecological pa e ns in agmen ed habi a s. Howe e , i is
di icul o e eal he mechanisms i b oad sense be a-di e si y indices (i.e. yielding iden ical
alues unde nes edness and species eplacemen ) a e used. Pa i ioning be a di e si y
in o u no e (caused by species eplacemen om si e o si e) and nes edness- esul an
componen s (caused by nes ed species losses) could p o ide a unique way o unde s and
he a ia ion o species composi ion in agmen ed habi a s. He e, we collec ed occupancy
da a o b eeding bi ds and liza ds on land-b idge islands in an inunda ed lake in eas e n
China. We decomposed be a di e si y o b eeding bi d and liza d communi ies in o spa ial
u no e and nes edness- esul an componen s o assess hei ela i e con ibu ions and e-
spec i e ela ionships o di e ences in island a ea, isola ion, and habi a ichness. Ou e-
sul s showed ha spa ial u no e con ibu ed mo e o be a di e si y han he nes edness-
esul an componen . The deg ee o isola ion had no signi ican e ec on o e all be a di e -
si y o i s componen s, nei he o b eeding bi ds no o liza ds. In u n, in bo h g oups he
nes edness- esul an componen inc eased wi h la ge di e ences in island a ea and habi-
a ichness, espec i ely, while u no e componen dec eased wi h hem. The majo di e -
ence among bi ds and liza ds was a highe ele ance o nes edness- esul an dissimila i y
in liza ds, sugges ing ha hey a e mo e p one o local ex inc ions de i ed om habi a ag-
men a ion. The dominance o he spa ial u no e componen o be a di e si y sugges s ha
all islands ha e po en ial conse a ion alue o b eeding bi d and liza d communi ies.
In oduc ion
Be a di e si y, o he amoun o change in species composi ion among si es in a egion [1], has
pa icula ele ance o explaining ecological pa e ns in egional biodi e si y [2–4]. Nume ous
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 1/19
a11111
OPEN ACCESS
Ci a ion: Si X, Baselga A, Ding P (2015) Re ealing
Be a-Di e si y Pa e ns o B eeding Bi d and Liza d
Communi ies on Inunda ed Land-B idge Islands by
Sepa a ing he Tu no e and Nes edness
Componen s. PLoS ONE 10(5): e0127692.
doi:10.1371/jou nal.pone.0127692
Academic Edi o : Rica do Bom im Machado,
Uni e si y o B asilia, BRAZIL
Recei ed: Decembe 4, 2014
Accep ed: Ap il 17, 2015
Published: May 18, 2015
Copy igh : © 2015 Si e al. This is an open access
a icle dis ibu ed unde he e ms o he C ea i e
Commons A ibu ion License, which pe mi s
un es ic ed use, dis ibu ion, and ep oduc ion in any
medium, p o ided he o iginal au ho and sou ce a e
c edi ed.
Da a A ailabili y S a emen : All ele an da a a e
wi hin he pape and i s Suppo ing In o ma ion iles.
Funding: The Na ional Na u al Science Founda ion
o China (Nos. 31170397, 31210103908) (h p://www.
ns c.go .cn) o P.D. and he Fundamen al Resea ch
Funds o he Cen al Uni e si ies (h p://www.dos .
moe.edu.cn) unded his s udy. The unde s had no
ole in s udy design, da a collec ion and analysis,
decision o publish, o p epa a ion o he manusc ip .
Compe ing In e es s: The au ho s ha e decla ed
ha no compe ing in e es s exis .
dissimila i y indices ha e been p oposed in he li e a u e [5–7] o quan i y he a ia ion in spe-
cies composi ion among si es. All hese indices aim o in e he mechanisms behind a ia ion
in species composi ion [8]. Howe e , i has been shown ha b oad sense dissimila i y indices
yield iden ical esul s unde di e en pa e ns de i ing om di e en ecological p ocesses [9–
11]. In pa icula , wo an i he ic p ocesses bo h con ibu e o be a di e si y [2,12]: species e-
placemen (spa ial u no e ) and species ichness di e ences, which may be be ween nes ed o
no nes ed assemblages [9,13–15]. I is hus c i ical he need o pa i ion be a di e si y in o
sepa a e componen s accoun ing o hese pa e ns, because con using wo an i he ic phenom-
ena as being a single pa e n could p e en unde s anding he ecological p ocesses behind he
obse ed pa e ns [9,16]. As a consequence, pa i ioning be a di e si y will p o ide a unique
way o unde s and he a ia ion o species composi ion among si es, wi h in e es o basic bio-
geog aphy and ecological applica ions [17–19].
Recen ly, Baselga [9] p oposed ha Sø ensen dissimila i y [20], a mono onic ans o ma-
ion o s ic sense be a di e si y (i.e. gamma/alpha), could be decomposed in o wo addi i e
componen s accoun ing o spa ial u no e and nes edness- esul an dissimila i ies, espec-
i ely. Spa ial u no e is caused by he eplacemen o species om one si e o ano he , which
may be he esul o niche and dispe sal p ocesses, ei he con empo a y o his o ical [11,16,
21,22]. Con a y o u no e , he nes edness- esul an componen is de e mined by species
loss o gain in nes ed subse s, which may be due o con empo a y o his o ical p ocesses as se-
lec i e ex inc ion, selec i e coloniza ion, habi a nes edness [4,16,23–25]. Thus, sepa a ing
bo h componen s o dissimila i y can help o un eil he ecological p ocesses [26,27]. Fo exam-
ple, Baselga e al. [28] ound no clea la i udinal g adien in la ge-scale be a-di e si y pa e ns
o wo ld amphibians. Howe e , when bo h componen s we e sepa a ed, clea la i udinal g adi-
en s we e obse ed, wi h spa ial u no e domina ing low la i udes and nes edness- esul an
dissimila i y being mo e ele an a high la i udes, poin ing o ma ked di e ences in he p o-
cesses d i ing be a di e si y a low and high la i udes (see also [24]). Besides i s in e es o
basic biogeog aphical and ecological ques ions, quan i ying he p opo ion o each componen
o be a di e si y is also c ucial o planning conse a ion s a egies [22]. I nes edness con ib-
u es mo e han u no e in o o e all be a di e si y among si es, i migh sugges ha si es wi h
iche in species should be p io i ized o p o ec ion. On he o he hand, i he spa ial u no e
componen is he dominan phenomenon, all si es should be po en ial a ge s o conse a ion
[11]. As a esul , many s udies applied me hods o pa i ioning be a di e si y o delinea e con-
se a ion s a egies o a ious axonomic g oups [4,29,30].
Habi a agmen a ion caused by human ac i i ies comp omises he conse a ion o bio-
logical communi ies. Biodi e si y alls apidly unde habi a agmen a ion because species
a e p one o local ex inc ion due o habi a loss, and hus animal communi ies in small habi-
a s ha e highe ex inc ion ulne abili y [31,32]. In consequence, small habi a pa ches a e
gene ally conside ed o low conse a ion p io i y and hey a e gi en li le p o ec ion [33–35].
Meanwhile, agmen a ion inc eases be a di e si y by c ea ing pa chiness in species dis ibu-
ions because o di e en ial local ex inc ion o species among agmen s [36]. The e o e, ag-
men a ion ep esen s a challenge o conse a ion, because he emaining small agmen s
may no be su icien o suppo iable popula ions, bu s ill make a subs an ial con ibu ion
o egional di e si y [3,37]. In his way, small emnan s o agmen habi a s may ha e po en-
ial conse a ion alue ha should no be o e looked [37]. I is hus ele an o know whe he
composi ional di e ences be ween agmen s (i.e. be a di e si y) a e ela ed o eplacemen
and/o nes ed pa e ns. Al hough nume ous s udies ha e add essed he e ec s o agmen a-
ion on species communi ies (e.g. [38–41]), ew s udies ha e used be a-di e si y pa i ioning
o species communi ies in agmen ed habi a s, especially compa ing da a on mul iple axo-
nomic g oups [42,43].
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 2/19
He e, we s udy be a-di e si y pa e ns o b eeding bi d and liza d communi ies on islands
in an inunda ed lake in eas e n China by pa i ioning o e all dissimila i y among islands in o
u no e and nes edness- esul an componen s. We collec ed occupancy da a on 37 islands o
bi ds ( om 2007 o 2012) and o liza ds ( om 2007 o 2008). We used hese da a o assess he
spa ial pa e ns o be a di e si y and i s componen s, and add ess he ollowing ques ions: (1)
A e be a di e si y and i s componen s di e en be ween b eeding bi d and liza d communi ies?
(2) Which componen , u no e o nes edness- esul an dissimila i y, con ibu e mo e o o e -
all be a di e si y o bo h g oups? (3) Wha a e he ela ionships be ween be a di e si y compo-
nen s and di e ences in island a ibu es? (4) Based on he ela i e con ibu ions o u no e
and nes edness- esul an componen s, wha a e he po en ial conse a ion s a egies o biodi-
e si y managemen on land-b idge islands, such as he sys ems inunda ed ecen ly? Ou ini ial
expec a ions we e ha bi ds gene ally ha e lowe o e all be a di e si y, especially he nes ed-
ness- esul an componen , and hus a lowe con ibu ion o nes edness- esul an componen
in o o e all be a di e si y because o i s highe agili y [4,24,44].
Ma e ials and Me hods
E hics S a emen
Ou esea ch on b eeding bi d and liza d communi ies in he Thousand Island Lake was ap-
p o ed by he Chinese Wildli e Managemen Au ho i y and conduc ed unde Law o he Peoples
Republic o China on he P o ec ion o Wildli e (Augus 28, 2004). Chun’an Fo es y Bu eau and
he Thousand Island Lake Na ional Fo es Pa k g an ed he pe mi s o conduc he esea ches.
S udy A ea
Thousand Island Lake lies a 29°22´ o 29°50´ N, 118°34´ o 119°15´ E and is a la ge a i icial
ese oi in wes e n Zhejiang P o ince, eas e n China (Fig 1). The lake was c ea ed in 1959 by
he cons uc ion o he Xin’anjiang Dam o hyd oelec ici y. Flooding app oxima ely 580 km
2
,
i o med 1078 islands wi h a eas >0.25 ha when he wa e eached i s highes le el (108 m).
The dominan ege a ion on he islands is na u al seconda y o es , mainly o Pinus massoni-
ana, wi h many b oad-lea ed ees and sh ub species, such as Cas anopsis scle ophylla,Liquid-
amba o mosana,Rhododend on a e ae, and Lo ope alum chinense. The egion has a ypical
sub opical monsoon zone, wi h ma ked seasonali y. The a e age annual empe a u e is 17.0°C.
Daily empe a u e anges om—7.6°C in Janua y o 41.8°C in July. The annual p ecipi a ion o
he egion is 1430 mm, mainly concen a ed in ainy season be ween Ap il and June [43].
Land-b idge islands c ea ed by cons uc ion o dams, such as he islands o he Thousand Is-
land Lake, can be iewed as na u al expe imen s o assessing he e ec s o agmen a ion on com-
muni y composi ion a ia ion [45]. Ou s udy sys em is pa icula ly well sui ed o wo easons.
Fi s , he islands we e c ea ed essen ially simul aneously as a esul o dam cons uc ion and he
quick subsequen inunda ion, so ha all islands ha e he same ecological backg ound and clea
geological bounda ies [46,47]. Mo eo e , gi en he sho his o y o he lake (55 yea s), all he is-
lands s ill sha e an ances al pool o species [48], and he e ec s o long- e m his o ical p ocesses
(e.g. glacial cycles and specia ion) can be excluded [16,24,25]. Second, he islands a e ela i ely
small and accessible, so we we e able o su ey he animal communi ies on islands ho oughly se -
e al imes in each yea , p o iding a ela i ely comple e animal in en o y in ou esea ch sys em.
Sampling P o ocols
Island A ibu es. We selec ed 37 islands (numbe ed om la ges o smalles acco ding o
island a ea in Fig 1) o encompass as much a ia ion in a ea and isola ion as possible. These
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 3/19
islands ange om 0.57 ha o app oxima ely 1300 ha, and om abou 20 m sepa a ion om
he nea es sho e o he mainland o o e 3.71 km. We cha ac e ized islands in e ms o a ea,
isola ion and habi a ichness, as hese a iables a e ecognized as he key de e minan s o he
p obabili ies o coloniza ion and ex inc ion [49,50], hus being po en ially ele an ac o s be-
hind be a di e si y and i s componen s [51].
Fo each island, we measu ed a ea in hec a es (A). Fo isola ion measu es, we used a bu e -
based measu e o es ima e he isola ion, which is gene ally conside ed be e han dis ance-
based measu es [52]. We es ima ed isola ion (I) as he ac ion o bu e a ea ha is wa e wi h-
in a 2-km bu e egion a ound a ocal island [48]. Be ween Ap il and No embe in 2007, we
in ensi ely su eyed s udy islands, and classi ied he habi a s on each island in o se en ca ego-
ies: coni e ous o es , b oad-lea ed o es , mixed coni e ous b oad-lea o es s, bamboo
g o es, sh ubs, g asses and a mland [43]. We hen de ined habi a ichness (H ) as he numbe
o habi a ypes on each island (Table 1).
Bi d Sampling. We su eyed bi d communi ies on 37 islands du ing he b eeding season
(Ap il–June) annually om 2007 o 2012. The sampling e o o each island was oughly p o-
po ional o he loga i hm o a ea [53]. As a esul , eigh ansec ails we e sampled on Island
1 ( he la ges s udy island, a ea >1000 ha), ou on Islands 2 and 3 (island a ea >100 ha), wo
on ou islands (10 ha <island a ea <100 ha), and one on each o he emaining small islands
(island a ea 1 ha o mos islands) (Table 1). We used a global posi ioning sys em (GPS) o
eco d he o al leng h o ansec s on each island.
T ansec s we e gene ally placed along idge-lines, and we clea ed na ow census ails
(abou 20 cm wide) o acili a e su eys [54]. Whe e islands had mo e han one habi a ype, a
Fig 1. The 37 s udy islands in he Thousand Island Lake in Zhejiang P o ince, China. S udy islands
we e numbe ed in o de o dec easing a ea.
doi:10.1371/jou nal.pone.0127692.g001
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 4/19
Table 1. Cha ac e is ics o he 37 s udy islands in he Thousand Island Lake, China.
Island
code
La i ude Longi ude Island
a ea (ha)
Isola ion Species
ichness o
bi ds (n)
Species
ichness o
liza ds (n)
Habi a
ichness (n)
Numbe o
ansec s (n)
To al leng h o
ansec s (m)
1 29°31
´11.44´´N
118°52
´25.87´´E
1289.23 0.78 43 5 7 8 3200
2 29°30
´30.15´´N
118°49
´09.31´´E
143.19 0.88 34 4 6 4 1600
3 29°31
´09.85´´N
118°55
´15.92´´E
109.03 0.73 35 4 6 4 1600
4 29°31
´51.52´´N
118°56
´24.47´´E
55.08 0.79 32 4 5 2 800
5 29°29
´40.04´´N
118°53
´39.07´´E
46.37 0.68 31 3 5 2 800
6 29°30
´52.90´´N
118°50
´57.48´´E
35.64 0.63 27 4 5 2 800
7 29°32
´06.77´´N
118°56
´13.82´´E
32.29 0.88 30 3 5 2 800
8 29°29
´45.04´´N
118°55
´42.22´´E
5.69 0.57 30 2 3 1 375
9 29°37
´07.60´´N
118°58
´03.22´´E
3.42 0.69 24 0 4 1 300
10 29°30
´01.91´´N
118°53
´08.99´´E
2.90 0.67 20 1 3 1 275
11 29°29
´54.89´´N
118°54
´13.92´´E
2.83 0.77 25 2 4 1 150
12 29°29
´45.77´´N
118°54
´22.48´´E
2.29 0.77 26 3 4 1 300
13 29°32
´58.32´´N
118°54
´37.70´´E
2.23 0.98 23 0 3 1 400
14 29°34
´03.33´´N
118°53
´42.12´´E
2.00 0.74 23 2 3 1 300
15 29°36
´45.05´´N
118°57
´00.39´´E
1.93 0.73 24 1 4 1 250
16 29°30
´12.53´´N
118°53
´31.07´´E
1.74 0.66 20 2 3 1 300
17 29°29
´43.37´´N
118°54
´33.45´´E
1.54 0.76 25 1 3 1 375
18 29°30
´28.07´´N
118°49
´24.48´´E
1.52 0.71 24 1 3 1 250
19 29°31
´14.60´´N
118°49
´38.74´´E
1.40 0.68 23 1 3 1 375
20 29°28
´55.79´´N
118°54
´55.49´´E
1.26 0.42 25 1 3 1 200
21 29°29
´24.45´´N
118°54
´01.52´´E
1.20 0.67 22 1 3 1 225
22 29°30
´11.27´´N
118°53
´25.38´´E
1.20 0.67 23 2 3 1 225
23 29°30
´19.16´´N
118°53
´38.69´´E
1.17 0.63 20 2 3 1 250
24 29°34
´47.56´´N
118°54
´43.16´´E
1.15 0.67 23 0 3 1 275
25 29°34
´36.77´´N
118°55
´38.49´´E
1.03 0.80 18 1 3 1 250
26 29°31
´45.45´´N
118°55
´21.56´´E
1.01 0.72 24 1 3 1 250
(Con inued)
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 5/19
s a i ied andom placemen was used o cap u e all he ypes. We collec ed bi d occupancy
da a along hese ansec s [55] du ing b eeding seasons om 2007 o 2012. In each su ey, ob-
se e s walked each ansec a a cons an speed (c. 2.0 km × h
−1
). We eco ded all he bi ds
seen o hea d on he su ey island, bu excluded high- lying species passing o e he islands
du ing su eys. We su eyed each ansec on hese islands 78 imes o e he cou se o he en-
i e s udy [48]. Su eys an om a e hal an hou a e dawn o 11:00 h in he mo nings and
om 15:00 o hal an hou be o e sunse in he a e noons. We did no conduc bi d su eys i
he e was hea y ain, high wind, o high empe a u e. We al e na ed he di ec ion obse e
walked on each ansec andomly o elimina e he po en ial su ey bias [43,48,56].
We assessed he comple eness o ou su ey o he la ges , and p opo ionally leas sampled
island (Island 1; see Fig 1) by c ea ing species accumula ion cu es o each o he six yea s. We
ound all cu es clea ly le eled o be o e he comple ion o he su eys, indica ing sampling e -
o s we e su icien on his and smalle islands o cap u e he ull b eeding bi d communi ies
[48]. In ou s udy, we only conside ed e es ial b eeding bi ds, excluding di ing bi ds, ducks,
gulls, sho ebi ds, he ons and king ishe s whose habi a s ex ensi ely elies on wa e . Du ing he
cou se o b eeding seasons om 2007 o 2012, we eco ded da a o al o 60 e es ial b eeding
bi ds (S1 Da ase ).
Liza d Sampling. We used he line- ansec me hod [57] o su ey he liza d communi ies
along ansec s on 37 islands ha a e he same as bi d su eys du ing wo b eeding seasons in
2007 and 2008. Obse e s walked each ansec a a cons an speed (c. 10 m × min
−1
) o sea ch
he g ound and ee boles wi h binocula s [43]. We only included he con iden iden i ica ions
Table 1. (Con inued)
Island
code
La i ude Longi ude Island
a ea (ha)
Isola ion Species
ichness o
bi ds (n)
Species
ichness o
liza ds (n)
Habi a
ichness (n)
Numbe o
ansec s (n)
To al leng h o
ansec s (m)
27 29°33
´06.81´´N
118°54
´26.16´´E
0.96 0.98 19 0 3 1 250
28 29°36
´51.21´´N
118°57
´43.00´´E
0.91 0.74 23 0 4 1 275
29 29°31
´48.11´´N
118°55
´18.05´´E
0.86 0.70 22 1 3 1 225
30 29°30
´54.55´´N
118°49
´21.12´´E
0.83 0.74 22 1 3 1 275
31 29°36
´48.30´´N
118°57
´12.92´´E
0.83 0.75 20 0 4 1 250
32 29°31
´47.12´´N
118°55
´28.34´´E
0.80 0.74 25 1 2 1 300
33 29°33
´58.20´´N
118°53
´43.87´´E
0.73 0.80 23 2 3 1 300
34 29°34
´38.59´´N
118°54
´57.84´´E
0.67 0.75 18 1 3 1 325
35 29°34
´40.16´´N
118°54
´34.20´´E
0.59 0.68 25 0 3 1 225
36 29°29
´42.71´´N
118°54
´18.86´´E
0.59 0.78 21 1 3 1 250
37 29°32
´52.20´´N
118°54
´42.82´´E
0.57 0.95 19 0 3 1 200
Each island is numbe ed as in Fig 1. Isola ion was es ima ed as he ac ion o bu e a ea ha is wa e wi hin a 2-km bu e egion a ound a ocal island.
See mo e de ails o hese island a ibu es in he Ma e ials and Me hods.
doi:10.1371/jou nal.pone.0127692. 001
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 6/19
eco ded on he su ey island in o ou analyses [58]. Su eys an om be ween 1 h a e sun ise
un il 5 h a e sun ise in good wea he condi ion. The di ec ion obse e s walked on each an-
sec and he su eyed islands we e also andomly al e na ed.
In Zhejiang P o ince, he e a e only six liza d species [59] ha a e e y common and easy
o iden i y wi h con idence. We ound all he species on he su ey islands excep Takyd omus
sexlinea us, whose dis ibu ion on he nea by mainland is s ill con o e sial [43]. Based on he
high su ey equencies (20 imes pe each ansec ), we conside ha he species lis s in ou
da ase a e comple e and eliable (S2 Da ase ).
Pa i ioning Be a Di e si y
We pa i ioned be a di e si y in o wo sepa a e componen s o species u no e and nes ed-
ness- esul an dissimila i ies [9]. Speci ically, his me hod pa i ions he pai wise Sø ensen dis-
simila i y be ween wo communi ies (β
so
)(Eq 1) in o wo addi i e componen s accoun ing o
species spa ial u no e (β
sim
)(Eq 2) and nes edness- esul an dissimila i ies (β
sne
)(Eq 3). The
Simpson dissimila i y index (β
sim
) desc ibes species u no e wi hou he in luence o ichness
g adien s [5,30,60]. Since β
so
and β
sim
a e equal in he absence o nes edness, hei di e ence
is a ne measu e o he nes edness- esul an componen o be a di e si y, so ha β
sne
=β
so
−
β
sim
[9]. The pai wise dissimila i y indices a e o mula ed as:
bso ¼bþc
2aþbþcð1Þ
bsim ¼minðb;cÞ
aþminðb;cÞð2Þ
bsne ¼bso bsim ¼jbcj
2aþbþca
aþminðb;cÞð3Þ
whe e ais he numbe o species p esen a bo h si es, bis he numbe o species p esen a he
fi s si e bu no a he second, and cis he numbe o species p esen a he second si e bu no
a he fi s . The fi s ac ion o Eq 3,jbcj
2aþbþc, is simila as β
gl
index, 2max b;cðÞ2min ðb;cÞ
2aþmax b;cðÞþmin ðb;cÞ[5,13]—bo h
o hem es ima e species ichness di e ences. The second ac ion o β
sne
,a
aþmin ðb;cÞ, is he Simp-
son simila i y, 1 −β
sim
, which is a measu e o nes edness [61]. As a esul , β
sne
measu es he
ac ion o dissimila i y caused by ichness di e ences be ween nes ed subse s [9,15,19].
To es ima e he o e all be a di e si y o b eeding bi d communi ies among all islands, we
used he mul iple-si e dissimila i y [62]. O e all mul iple-si e dissimila i y was measu ed using
mul iple-si e Sø ensen dissimila i y (β
SOR
;Eq 4), which was decomposed in o spa ial u no e
(β
SIM
;Eq 5) and nes edness- esul an componen s (β
SNE
;Eq 6)[9]:
bSOR ¼
½Xi<jminðbij;bjiÞ þ ½Xi<jmaxðbij;bjiÞ
2½XiSiSTþ½
Xi<jminðbij;bjiÞ þ ½Xi<jmaxðbij;bjiÞ
ð4Þ
bSIM ¼
½Xi<jminðbij;bjiÞ
½XiSiSTþ½
Xi<jminðbij;bjiÞ
ð5Þ
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 7/19
bSNE ¼bSOR bSIM
¼
½Xi<jmaxðbij;bjiÞ ½Xi<jminðbij;bjiÞ
2½XiSiSTþ½
Xi<jminðbij;bjiÞ þ ½Xi<jmaxðbij;bjiÞ
½XiSiST
½XiSiSTþ½
Xi<jminðbij;bjiÞ
ð6Þ
whe e S
i
is he species ichness o island i,S
T
is he species ichness on all s udy islands (γdi e -
si y), and b
ij
and b
ji
a e he species ichness exclusi e o island iand j, espec i ely. The ac-
ions [S
i<
jmin (b
ij
,b
ji
)] and [S
i<
jmax (b
ij
,b
ji
)] o he mul iple-si e dissimila i y a e analogous
o he componen s band co pai wise dissimila i y, espec i ely. The ac ion [S
i
S
i
−S
T
] anal-
ogous o he componen ao pai wise dissimila i y (i.e. he species sha ed be ween bo h si es).
We also calcula ed he Jacca d pai wise dissimila i y indices p oposed by Baselga [19] and
Ca alho e al. [14] o compa ison o assess he obus ness o ou esul s. We ound o e all
be a di e si y and i s componen s had almos iden ical esul s o all h ee pa i ioning me h-
ods. A comp ehensi e e iew o hese me hods is ou o he scope o his pape , bu see [15,63].
Da a Analyses
Nes edness Analysis. We es ima ed nes edness o b eeding bi d and liza d communi ies
using a nes edness me ic: NODF (Nes edness me ic based on O e lap and Dec easing Filling)
[64]. NODF is gene ally conside ed o ha e be e s a is ical p ope ies, compa ed o ma ix
empe a u e [65] o disc epancy me ic [66]. Fo example, NODF a oids o e es ima ing nes -
edness ( ype I e o s), and allows decons uc ing o al nes edness (NODF) in o he indepen-
den con ibu ions o columns (NODFc) and ows (NODF ) (i.e. si es and species) o he
nes ed pa e ns [64,67,68]. We calcula ed he NODF indices wi h he p og am NODF e sion
2.0, and gene a ed null communi ies using 1000 andom simula ed ma ices based on p opo -
ional- ow and p opo ional-column (PP) algo i hms [64,69,70]. PP algo i hm is conside ed
as he p e e ed model when esea ch sys ems con ain ela i ely small islands and he scale o
analysis is small (e.g. Thousand Island Lake) [68,69].
In ou s udy, we aimed o in es iga e whe he si es wi h poo e in species a e subse s o si es
wi h iche in species, so we epo ed he indices o NODF o si es in ou s udy. In addi ion,
we ound no liza d species on eigh islands a e in ensi e su ey, and hus excluded hese is-
lands’communi ies om ou analyses (Table 1).
Species Richness Modelling. We es ed he ela ionships among island a iables using
pai wise Pea son co ela ion coe icien s ( ). Species ichness on islands was eg essed agains
island a ea, isola ion, and habi a ichness o b eeding bi ds and liza ds using linea eg es-
sion. We used backwa d s epwise eg ession o choose he bes - i ed model wi h a iable selec-
ion and model e alua ion based on he Akaike in o ma ion c i e ia (AIC) [71]. All island
a iables we e log- ans o med o no malize model esiduals.
Be a Di e si y Pa e ns. To make compa able dissimila i ies compu ed o b eeding bi d
and liza d communi ies wi h di e en numbe s o islands (bi d 37 islands s. liza d 29 islands),
we compu ed dissimila i y alues o b eeding bi d communi ies using a esampling p ocedu e,
aking 100 andom samples o 29 in en o ies and compu ing he a e age dissimila i y alues
[19]. We hen ob ained he p opo ion o nes edness- esul an componen o o e all mul iple-
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 8/19
si e dissimila i y o ep esen he ela i e con ibu ion o o e all be a di e si y: β
a io
=β
SNE
/
β
SOR
. Thus, β
a io
<0.5 indica es ha be a di e si y is de e mined dominan ly by species u n-
o e , and β
a io
>0.5 indica es nes edness is he dominan componen [24].
We es ed whe he di e ences in island a ibu es (a ea, isola ion and habi a ichness) had
signi ican ela ionships wi h bi d and liza d be a di e si y and hei componen s. The pai wise
dissimila i ies be ween islands a e no independen because he species composi ion o one is-
land a ec s he dissimila i y o his island wi h all o he islands. We hus used mul iple eg es-
sion models o dis ance ma ices (MRM) o examine he ela ionships be ween he ma ices o
o e all be a di e si y, u no e and nes edness- esul an dissimila i ies and he Euclidean dis-
ance ma ices o en i onmen al a iables [22,72,73]. We hen ob ained he eg ession slopes
(a) and in e cep s (b) by MRM. Because non-independen obse a ions o pai wise dis ances
will in la e he signi icance o s a is ical es s, and ou analyses could be a ec ed by spa ial au o-
co ela ion, we used pa ial Man el es s (9999 pe mu a ions) including spa ial dis ance be-
ween islands as a co a ia e o es ima e he p- alues and he Pea son co ela ion coe icien s ( )
[72]. In ou s udy, he eg ession slopes (a) and in e cep s (b)o β
so
equa ed he summa ions
o β
sim
and β
sne
, espec i ely because o he p ope y o he addi i e pa i ioning me hod.
We pe o med s a is ical analyses in R [74] using packages ape [75], be apa [76], ecodis
[77] and egan [78], and NODF p og am ( e sion 2.0) [70].
Resul s
Nes edness S uc u e
The obse ed NODF o si es o b eeding bi ds (N
obs
= 81.66) was signi ican ly lowe han ex-
pec ed om he null model (N
exp
= 85.38, Z- alue = −2.49, p= 0.01), whe eas he obse ed
NODF o si es o liza ds (N
obs
= 62.73) was no signi ican ly di e en (N
exp
= 61.70, Z-
alue = 0.39, p= 0.35) (Table 2). I indica ed b eeding bi d communi ies we e signi ican ly
an i-nes ed (i.e. obse ed communi y less nes ed han expec ed by null ma ices), and liza d
communi ies we e no signi ican ly mo e nes ed han andom pa e ns. S1 and S2 Figs showed
he maximally packed ma ix o b eeding bi ds and liza ds, espec i ely.
Pa e ns o Species Richness
Island a ea was signi ican ly co ela ed wi h habi a ichness ( = 0.89, p<0.05), indica ing
ha la ge islands gene ally suppo mo e habi a ypes. The e we e no signi ican co ela ions
be ween isola ion and island a ea o habi a ichness.
Species ichness pa e ns o b eeding bi ds and liza ds a e simila as shown by he linea e-
g ession models and s epwise analyses: only island a ea was e ained in he bes - i ed models
and was posi i ely ela ed o species ichness o bo h g oups (Table 3).
Table 2. Resul s o nes edness analysis using NODF p og am o species by si e ma ix o b eeding bi d and liza d communi ies on 37 s udy is-
lands in he Thousand Island Lake, China.
Speicies Numbe o species Numbe o island N
obs
N
exp
(SD) Filling Z- alue p
B eeding bi ds 60 37 81.66 85.38 (1.49) 41.0% –2.49 0.01
Liza ds 5 29 62.73 61.70 (2.66) 40.0% 0.39 0.35
Islands wi h no liza d species (N= 8) we e excluded om he analysis. Null model was based on p opo ional- ow and p opo ional-column cons ains wi h
1000 andomiza ions. Abb e ia ions: obse ed NODF o si es, N
obs
; expec ed NODF o si es wi h s anda d de ia ion, N
exp
(SD); Z- alue; Mon e Ca lo-
de i ed p obabili ies, p.
doi:10.1371/jou nal.pone.0127692. 002
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 9/19
Au ho Con ibu ions
Concei ed and designed he expe imen s: XS PD. Pe o med he expe imen s: XS PD. Ana-
lyzed he da a: XS AB. Con ibu ed eagen s/ma e ials/analysis ools: XS PD. W o e he pape :
XS AB PD.
Re e ences
1. Whi ake RH. E olu ion and measu emen o species di e si y. Taxon. 1972; 21:213–51.
2. Ha ison S, Ross SJ, Law on JH. Be a di e si y on geog aphic g adien s in B i ain. J Anim Ecol. 1992;
61:151–8.
3. Ka an GH, F anco P, Saa ed a-Rod íguez CA, Valde ama C, Rojas V, Oso io D, e al. Spa ial compo-
nen s o bi d di e si y in he Andes o Colombia: implica ions o designing a egional ese e sys em.
Conse Biol. 2006; 20:1203–11. PMID: 16922236
4. Calde ón-Pa ón JM, Mo eno CE, Pineda-López R, Sánchez-Rojas G, Zu ia I. Ve eb a e dissimila i y
due o u no e and ichness di e ences in a highly be a-di e se egion: he ole o spa ial g ain size,
dispe sal abili y and dis ance. PLoS ONE. 2013; 8:e82905. doi: 10.1371/jou nal.pone.0082905 PMID:
24324840
5. Kole P, Gas on KJ, Lennon JJ. Measu ing be a di e si y o p esence—absence da a. J Anim Ecol.
2003; 72:367–82.
6. Tuomis o H. A di e si y o be a di e si ies: s aigh ening up a concep gone aw y. Pa 2. Quan i ying
be a di e si y and ela ed phenomena. Ecog aphy. 2010; 33:23–45.
7. Legend e P, De Cáce es M. Be a di e si y as he a iance o communi y da a: dissimila i y coe icien s
and pa i ioning. Ecol Le . 2013; 16:951–63. doi: 10.1111/ele.12141 PMID: 23809147
8. Ande son MJ, C is TO, Chase JM, Vellend M, Inouye BD, F ees one AL, e al. Na iga ing he mul iple
meanings o βdi e si y: a oadmap o he p ac icing ecologis . Ecol Le . 2011; 14:19–28. doi: 10.1111/
j.1461-0248.2010.01552.x PMID: 21070562
9. Baselga A. Pa i ioning he u no e and nes edness componen s o be a di e si y. Global Ecol Bio-
geog . 2010; 19:134–43.
10. Fa o ini S. The in luence o geog aphical and ecological ac o s on island be a di e si y pa e ns. J Bio-
geog . 2010; 37:1061–70.
11. Angele DG. Re ealing a conse a ion challenge h ough pa i ioned long- e m be a di e si y: inc eas-
ing u no e and dec easing nes edness o bo eal lake me acommuni ies. Di e s Dis ib. 2013;
19:772–81.
12. Gas on KJ, Blackbu n TM. Pa e n and P ocess in Mac oecology. Ox o d: Blackwell Science; 2000.
13. Lennon JJ, Kole P, G eenwood J, Gas on KJ. The geog aphical s uc u e o B i ish bi d dis ibu ions:
di e si y, spa ial u no e and scale. J Anim Ecol. 2001; 70:966–79.
14. Ca alho JC, Ca doso P, Gomes P. De e mining he ela i e oles o species eplacemen and species
ichness di e ences in gene a ing be a-di e si y pa e ns. Global Ecol Biogeog . 2012; 21:760–71.
15. Legend e P. In e p e ing he eplacemen and ichness di e ence componen s o be a di e si y. Global
Ecol Biogeog . 2014; 23:1324–34.
16. Dappo o L, Fa o ini S, VodăR, DincăV, Vila R. Biogeog aphy o wes e n Medi e anean bu e lies:
combining u no e and nes edness componen s o aunal dissimila i y. J Biogeog . 2014; 41:1639–50.
17. Williams PH. Mapping a ia ions in he s eng h and b ead h o biogeog aphic ansi ion zones using
species u no e . P oc R Soc B. 1996; 263:579–88.
18. Baselga A. Disen angling dis ance decay o simila i y om ichness g adien s: esponse o Soininen
e al. 2007. Ecog aphy. 2007; 30:838–41.
19. Baselga A. The ela ionship be ween species eplacemen , dissimila i y de i ed om nes edness, and
nes edness. Global Ecol Biogeog . 2012; 21:1223–32.
20. Sø ensen T. A me hod o es ablishing g oups o equal ampli ude in plan sociology based on simila i y
o species and i s applica ion o analyses o he ege a ion on Danish commons. Biol Sk . 1948; 5:1–
34.
21. Lep ieu F, Tedesco PA, Hugueny B, Beaucha d O, Dü HH, B osse S, e al. Pa i ioning global pa -
e ns o eshwa e ish be a di e si y e eals con as ing signa u es o pas clima e changes. Ecol Le .
2011; 14:325–34. doi: 10.1111/j.1461-0248.2011.01589.x PMID: 21303436
22. Gu ié ez-Cáno as C, Millán A, Velasco J, Vaughan IP, O me od SJ. Con as ing e ec s o na u al and
an h opogenic s esso s on be a di e si y in i e o ganisms. Global Ecol Biogeog . 2013; 22:796–805.
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 16 / 19
23. Ul ich W, Almeida-Ne o M, Go elli NJ. A consume 's guide o nes edness analysis. Oikos. 2009; 118:3–
17.
24. Dob o olski R, Melo AS, Cassemi o FAS, Diniz-Filho JAF. Clima ic his o y and dispe sal abili y explain
he ela i e impo ance o u no e and nes edness componen s o be a di e si y. Global Ecol Biogeog .
2012; 21:191–7.
25. Liggins L, Boo h DJ, Figuei a WF, T eml EA, Tonk L, Ridgway T, e al. La i ude-wide gene ic pa e ns
e eal his o ical e ec s and con as ing pa e ns o u no e and nes edness a he ange pe iphe ies o
a opical ma ine ish. Ecog aphy. 2015; 38: doi: 10.1111/ecog.01398
26. K e H, Je z W. A amewo k o delinea ing biogeog aphical egions based on species dis ibu ions. J
Biogeog . 2010; 37:2029–53.
27. Hol BG, Lessa d JP, Bo egaa d MK, F i z SA, A aújo MB, Dimi o D, e al. An upda e o Wallace’s
zoogeog aphic egions o he wo ld. Science. 2013; 339:74–8. doi: 10.1126/science.1228282 PMID:
23258408
28. Baselga A, Gómez-Rod íguez C, Lobo JM. His o ical legacies in wo ld amphibian di e si y e ealed by
he u no e and nes edness componen s o be a di e si y. PLoS ONE. 2012; 7:e32341. doi: 10.1371/
jou nal.pone.0032341 PMID: 22384222
29. Aze ia ET, Iba zabal J, Bouche J, Hébe C. Towa ds a be e unde s anding o be a di e si y: Decon-
s uc ing composi ion pa e ns o sap oxylic bee les b eeding in ecen ly bu n bo eal o es s. In: Pa i-
no IY, edi o . Resea ch in Biodi e si y: Models and Applica ions: InTech Open Access Publishe ;
2011. p. 75–94.
30. Mouillo D, De Bo oli J, Lep ieu F, Pa a icini V, Kulbicki M, Bellwood DR. The challenge o delinea ing
biogeog aphical egions: nes edness ma e s o Indo-Paci ic co al ee ishes. J Biogeog . 2013;
40:2228–37.
31. Pimm SL. The Balance o Na u e?: Ecological Issues in he Conse a ion o Species and Communi ies.
Chicago, IL: Uni e si y o Chicago P ess; 1991.
32. K auss J, Bomma co R, Gua diola M, Heikkinen RK, Helm A, Kuussaa i M, e al. Habi a agmen a ion
causes immedia e and ime-delayed biodi e si y loss a di e en ophic le els. Ecol Le . 2010;
13:597–605. doi: 10.1111/j.1461-0248.2010.01457.x PMID: 20337698
33. Pimm SL, Diamond J, Reed TM, Russell GJ, Ve ne J. Times o ex inc ion o small popula ions o la ge
bi ds. P oc Na l Acad Sci USA. 1993; 90:10871–5. PMID: 11607439
34. Fe az G, Russell GJ, S ou e PC, Bie egaa d RO, Pimm SL, Lo ejoy TE. Ra es o species loss om
Amazonian o es agmen s. P oc Na l Acad Sci USA. 2003; 100:14069–73. PMID: 14614134
35. Gibson L, Lynam AJ, B adshaw CJA, He F, Bick o d DP, Wood u DS, e al. Nea -comple e ex inc ion
o na i e small mammal auna 25 yea s a e o es agmen a ion. Science. 2013; 341:1508–10. doi:
10.1126/science.1240495 PMID: 24072921
36. Guadagnin DL, Pe e ÂS, Pe ello LFC, Mal chik L. Spa ial and empo al pa e ns o wa e bi d assem-
blages in agmen ed we lands o sou he n B azil. Wa e bi ds. 2005; 28:261–72.
37. Benedick S, Hill JK, Mus a a N, Chey VK, Ma ya i M, Sea le JB, e al. Impac s o ain o es agmen a-
ion on bu e lies in no he n Bo neo: species ichness, u no e and he alue o small agmen s. J
Appl Ecol. 2006; 43:967–77.
38. Ka an GH, Al a ez-López H, Gi aldo M. Fo es agmen a ion and bi d ex inc ions: San An onio eigh y
yea s la e . Conse Biol. 1994; 8:138–46.
39. Lens L, Van Dongen S, No is K, Gi hi u M, Ma hysen E. A ian pe sis ence in agmen ed ain o es .
Science. 2002; 298:1236–8. PMID: 12424379
40. Ko an a NM, Newma k WD, Kau man MJ. Long- e m demog aphic consequences o habi a agmen-
a ion o a opical unde s o y bi d communi y. Ecology. 2012; 93:2548–59. PMID: 23431586
41. B egman TP, Seke cioglu CH, Tobias JA. Global pa e ns and p edic o s o bi d species esponses o
o es agmen a ion: implica ions o ecosys em unc ion and conse a ion. Biol Conse . 2014;
169:372–83.
42. W igh DH, Pa e son BD, Mikkelson GM, Cu le A, A ma W. A compa a i e analysis o nes ed subse
pa e ns o species composi ion. Oecologia. 1997; 113:1–20.
43. Wang Y, Bao Y, Yu M, Xu G, Ding P. Nes edness o di e en easons: he dis ibu ions o bi ds, liza ds
and small mammals on islands o an inunda ed lake. Di e s Dis ib. 2010; 16:862–73.
44. Qian H. Global compa isons o be a di e si y among mammals, bi ds, ep iles, and amphibians ac oss
spa ial scales and axonomic anks. J Sys E ol. 2009; 47:509–14.
45. Te bo gh J, Feeley K. Ecosys em decay in closed o es agmen s. In: Ca son WP, Schni ze SA, edi-
o s. T opical Fo es Communi y Ecology: Wiley-Blackwell, Hoboken, NJ; 2008. p. 308–21.
46. Williamson M. Island Popula ions. Ox o d: Ox o d Uni e si y P ess; 1981.
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 17 / 19
47. Whi ake RJ, Fe nández-Palacios JM. Island Biogeog aphy: Ecology, E olu ion, and Conse a ion.
2nd ed. Ox o d: Ox o d Uni e si y P ess; 2007.
48. Si X, Pimm SL, Russell GJ, Ding P. Tu no e o b eeding bi d communi ies on islands in an inunda ed
lake. J Biogeog . 2014; 41:2283–92.
49. MacA hu RH, Wilson EO. The Theo y o Island Biogeog aphy. P ince on, NJ: P ince on Uni e si y
P ess; 1967.
50. Losos JB, Rickle s RE. The Theo y o Island Biogeog aphy Re isi ed. P ince on, NJ: P ince on Uni e -
si y P ess; 2010.
51. Gas on KJ, Da ies RG, O me CDL, Olson VA, Thomas GH, Ding TS, e al. Spa ial u no e in he global
a i auna. P oc R Soc B. 2007; 274:1567–74. PMID: 17472910
52. Moilanen A, Nieminen M. Simple connec i i y measu es in spa ial ecology. Ecology. 2002; 83:1131–
45.
53. Schoe ede JH, Galbia i C, Ribas CR, Sob inho TG, Spe be CF, DeSouza O, e al. Should we use p o-
po ional sampling o species—a ea s udies? J Biogeog . 2004; 31:1219–26.
54. Te bo gh J, Lopez L, Tello SJ. Bi d communi ies in ansi ion: he Lago Gu i islands. Ecology. 1997;
78:1494–501.
55. Bibby CJ, Bu gess ND, Hill DA, Mus oe S. Bi d Census Techniques. 2nd ed. London: Academic
P ess; 2000.
56. Mac Nally R, Ho ocks G, Benne AF. Nes edness in agmen ed landscapes: bi ds o he box-i onba k
o es s o sou h-eas e n Aus alia. Ecog aphy. 2002; 25:651–60.
57. Jaege R. T ansec sampling. In: Heye WR, Donnelly MA, McDia min RW, Hayek LC, Fos e MS, edi-
o s. Measu ing and Moni o ing Biological Di e si y: S anda d Me hods o Amphibians. Washing on,
DC: Smi hsonian Ins i u ion P ess; 1994. p. 103–7.
58. Ge maine S, Wakeling B. Liza d species dis ibu ions and habi a occupa ion along an u ban g adien
in Tucson, A izona, USA. Biol Conse . 2001; 97:229–37.
59. Huang M. Fauna o Zhejiang: Amphibia, Rep ilia. Hangzhou: Zhejiang Science and Technology Pub-
lishing House; 1990.
60. McKnigh MW, Whi e PS, McDonald RI, Lamo eux JF, Sech es W, Ridgely RS, e al. Pu ing be a-di-
e si y on he map: b oad-scale cong uence and coincidence in he ex emes. PLoS Biol. 2007; 5:
e272. PMID: 17927449
61. Almeida-Ne o M, F ensel D, Ul ich W. Re hinking he ela ionship be ween nes edness and be a di e -
si y: a commen on Baselga (2010). Global Ecol Biogeog . 2012; 21:772–7.
62. Baselga A. Mul iple si e dissimila i y quan i ies composi ional he e ogenei y among se e al si es, while
a e age pai wise dissimila i y may be misleading. Ecog aphy. 2013; 36:124–8.
63. Baselga A, Lep ieu F. Compa ing me hods o sepa a e componen s o be a di e si y. Me hods Ecol
E ol. 2015; 6: doi: 10.1111/2041-210X.12388 PMID: 25954500
64. Almeida-Ne o M, Guima aes P, Guima ães PR, Loyola RD, Ul ich W. A consis en me ic o nes ed-
ness analysis in ecological sys ems: econciling concep and measu emen . Oikos. 2008; 117:1227–
39.
65. A ma W, Pa e son BD. The measu e o o de and diso de in he dis ibu ion o species in agmen ed
habi a . Oecologia. 1993; 96:373–82.
66. B ualdi RA, Sande son JG. Nes ed species subse s, gaps, and disc epancy. Oecologia. 1999;
119:256–64.
67. Hu G, Feeley K, Wu J, Xu G, Yu M. De e minan s o plan species ichness and pa e ns o nes edness
in agmen ed landscapes: e idence om land-b idge islands. Landscape Ecol. 2011; 26:1405–17.
68. Ma hews TJ, Co ee-Jones HEW, Whi ake RJ. Quan i ying and in e p e ing nes edness in habi a is-
lands: a syn he ic analysis o mul iple da ase s. Di e s Dis ib. 2015; 21:392–404.
69. Ul ich W, Go elli NJ. A null model algo i hm o p esence—absence ma ices based on p opo ional e-
sampling. Ecol Model. 2012; 244:20–7.
70. Almeida-Ne o M, Ul ich W. A s aigh o wa d compu a ional app oach o measu ing nes edness using
quan i a i e ma ices. En i on Model So wa e. 2011; 26:173–8.
71. Ande son DR. Model Based In e ence in he Li e Sciences: A P ime on E idence. New Yo k, NY:
Sp inge ; 2008.
72. Legend e P, Legend e LF. Nume ical Ecology. Ams e dan: Else ie ; 2012.
73. Lichs ein JW. Mul iple eg ession on dis ance ma ices: a mul i a ia e spa ial analysis ool. Plan Ecol.
2007; 188:117–31.
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 18 / 19
74. R De elopmen Co e Team. R: A Language and En i onmen o S a is ical Compu ing Vienna, Aus-
ia: R Founda ion o S a is ical Compu ing; 2013. A ailable om: h p://www. -p ojec .o g. doi: 10.
3758/s13428-013-0330-5 PMID: 23519455
75. Pa adis E, Claude J, S imme K. APE: analyses o phylogene ics and e olu ion in R language. Bioin-
o ma ics. 2004; 20:289–90. PMID: 14734327
76. Baselga A, O me CDL. be apa : an R package o he s udy o be a di e si y. Me hods Ecol E ol. 2012;
3:808–12.
77. Goslee SC, U ban DL. The ecodis package o dissimila i y-based analysis o ecological da a. J S a
So w. 2007; 22:1–19.
78. Oksanen J, Blanche F, Kind R, Legend e P, O'Ha a R, Simpson G, e al. egan: Communi y Ecology
Package. R package e sion 2.0–7 2013. A ailable om: h p://CRAN.R-p ojec .o g/package = egan.
79. Diamond JM. Assembly o species communi ies. In: Cody ML, Diamond JM, edi o s. Ecologyand E o-
lu ion o Communi ies. Camb idge, MA: Ha a d Uni e si y P ess; 1975. p. 342–444.
80. Weihe E, Keddy P. Ecological Assembly Rules: Pe spec i es, Ad ances, Re ea s: Camb idge Uni-
e si y P ess; 2001.
81. Nekola JC, Whi e PS. The dis ance decay o simila i y in biogeog aphy and ecology. J Biogeog . 1999;
26:867–78. PMID: 10370629
82. Qian H. Be a di e si y in ela ion o dispe sal abili y o ascula plan s in No h Ame ica. Global Ecol
Biogeog . 2009; 18:327–32.
83. Lomolino MV, B own JH, Sax DF. Island biogeog aphy heo y. Re icula ions and ein eg a ion o "A bio-
geog aphy o he species". In: Losos J, Rickle s RE, edi o s. The Theo y o Island Biogeog aphy Re is-
i ed. P ince ion, NJ: P ince ion Uni e si y P ess; 2010. p. 13–51.
84. Dennis RL, Dappo o L, Fa o ini S, Cook LM. The gene alism-specialism deba e: he ole o gene alis s
in he li e and dea h o species. Biol J Linn Soc. 2011; 104:725–37.
85. Dennis RL, Ha dy PB, Dappo o L. Nes edness in island aunas: no el insigh s in o island biogeog a-
phy h ough bu e ly communi y p o iles o coloniza ion abili y and mig a ion capaci y. J Biogeog .
2012; 39:1412–26.
86. Dennis RL, Dappo o L, Spa ks TH, Williams SR, G ea o ex-Da ies JN, Ashe J, e al. Tu no e and
ends in bu e ly communi ies on wo B i ish idal islands: s ochas ic in luences and de e minis ic ac-
o s. J Biogeog . 2010; 37:2291–304.
87. Janzen DH. Why moun ain passes a e highe in he opics. Am Na . 1967:233–49.
88. Cu le A. Nes ed aunas and ex inc ion in agmen ed habi a s. Conse Biol. 1991; 5:496–504.
89. W igh DH, Ree es JH. On he meaning and measu emen o nes edness o species assemblages.
Oecologia. 1992; 92:416–28.
Ve eb a e Dissimila i ies on Land-B idge Islands
PLOS ONE | DOI:10.1371/jou nal.pone.0127692 May 18, 2015 19 / 19
