Decline o aqua ic plan s in an iconic Eu opean p o ec ed na u al a ea
Pablo Ga cia-Mu illo
a
, Ca men Díaz-Paniagua
b,*
, Rocío Fe n´
andez-Zamudio
b
a
Depa men o Plan Biology and Ecology, Uni e si y o Se illa, 41004, Se illa, Spain
b
Do˜
nana Biological S a ion-CSIC, A da Am´
e ico Vespucio 26, 41013 Se illa, Spain
ARTICLE INFO
Keywo ds:
Aqua ic mac ophy es
Do˜
nana
Th ea ened lo a
Ex inc ions
We lands
Medi e anean ponds
ABSTRACT
We examined occu ence pa e ns o e ime o aqua ic ascula plan species in a majo Eu opean na u al
p o ec ed a ea: Do˜
nana Na ional Pa k (sou hwes e n Spain). We used all a ailable eco ds (1965 o he p esen )
o aqua ic ascula plan s ound in he pa k’s pond ne wo k. In o ma ion was a ailable o 38 species ac oss 263
si es, including mo e han 100 ponds ha we e in ensi ely moni o ed be ween 2001 and 2023. Ou esul s show
ha aqua ic ascula plan species began o decline in he las hi d o he 20 h cen u y. Two phases we e
appa en : 1) an ea ly phase (la e 20 h cen u y), du ing which declines in aqua ic ascula plan popula ions we e
la gely p o oked by he a i al o an in asi e species, P ocamba us cla kii, and 2) a mo e ecen phase du ing
which he o e exploi a ion o he aqui e has esul ed in many ponds d ying up and in sho e looding pe iods in
he emaining ponds. A p esen , nine species ha e disappea ed, and 72% o he emaining species ha e smalle
anges. Ou indings sugges ha he d ama ic local ex inc ion and decline o Do˜
nana’s a ious aqua ic plan
popula ions is a consequence o he apid habi a loss ha has occu ed in ecen decades. His o ically, he pa k’s
pond ne wo k ac ed as a e uge o many endange ed and unique species acing majo h ea s o hei su i al.
Such no longe appea s o be he case.
1. In oduc ion
Habi a loss has been desc ibed as he g ea es challenge o ou imes
(Lambdon & C onk, 2020) and is leading o species ex inc ion. Aqua ic
habi a s ank among he wo ld’s mos h ea ened ecosys ems: app oxi-
ma ely 87 % o na u al we lands ha e been los since he ea ly 18 h
cen u y (Da idson, 2014; Dudgeon e al., 2006).
An eno mous amoun o wa e is cu en ly equi ed o mee ag i-
cul u al, u ban, and indus ial demands (Ac eman e al., 2022), which is
leading o he al e a ion and des uc ion o aqua ic habi a s. The esul is
a wo ldwide loss o na i e aqua ic species, especially in places well
sui ed o ag icul u al p oduc ion and u ban de elopmen (Kolp e al.,
2021). The ea lies eco ds signaling declines in aqua ic ege a ion da e
back o 1900, wi h he pa e n being seen in he No he n and Sou he n
Hemisphe es and a high and low la i udes; his decline has accele a ed
o e ecen decades (Zhang e al., 2017). In Eu ope o e he las cen u y,
aqua ic habi a loss and de e io a ion ha e educed he ichness o
eshwa e ascula plan species (Sand-Jensen e al., 2018). Addi ion-
ally, a ecen assessmen conduc ed as pa o Eu ope’s Wa e F ame-
wo k Di ec i e (WFD) e ealed ha a ound 60 % o wa e bodies in he
EU do no ha e a heal hy ecological s a us and poin ed o
hyd omo phological changes as a key p essu e on su ace wa e bodies
(K is ensen e al., 2018).
The mos e ec i e esponse o his challenge has been p o ided by
p o ec ed na u al a eas, in which he p ese a ion o op imal, unal e ed
habi a s os e s he pe sis ence o plan and animal popula ions, espe-
cially hose o endange ed species (Cazzolla e al., 2023; Pimm e al.,
2018; Se gio e al., 2022). In ecen yea s, he conce ns o conse a ion
biologis s abou managing h ea s o p e en species ex inc ion ha e
esul ed in conside able changes o conse a ion policies; no ably, he
numbe and quali y o ese es ha e inc eased, and managemen p ac-
ices ha e imp o ed. Much mo e a en ion is also being paid o species
losses and habi a de e io a ion, which se e as ea ly wa ning signs ha
p o ec ed a eas a e no longe se ing hei unc ional pu pose (Se gio
e al., 2022). Howe e , s akeholde s in Medi e anean ecosys ems a y
in how hey in e p e his in o ma ion (Mo ei a e al., 2019; Salme ´
on
e al., 2021).
Loca ed in sou hwes e n Spain, Do˜
nana Na ional Pa k is one o
Eu ope’s bes known na u al p o ec ed a eas, and i has long se ed as
an example o how o p ese e species and biological communi ies. Fo
example, he pa k has helped many species escape ex inc ion, including
animals such as he Ibe ian lynx (Pacín e al., 2024), he c es ed coo
* Co esponding au ho .
E-mail add ess: [email p o ec ed] (C. Díaz-Paniagua).
Con en s lis s a ailable a ScienceDi ec
Jou nal o Na u e Conse a ion
jou nal homepage: www.else ie .com/loca e/jnc
h ps://doi.o g/10.1016/j.jnc.2024.126814
Recei ed 1 Oc obe 2024; Recei ed in e ised o m 18 Decembe 2024; Accep ed 18 Decembe 2024
Jou nal o Na u e Conse a ion 84 (2025) 126814
A ailable online 19 Decembe 2024
1617-1381/© 2024 The Au ho (s). Published by Else ie GmbH. This is an open access a icle unde he CC BY license
(
h p://c ea i ecommons.o g/licenses/by/4.0/ ).
(Ho as e al., 2023), and he ma bled eal (O ego e al., 2024) and
plan s such as A me ia gadi ana, Lina ia u sica, and Vulpia on que ana,
among many o he s (Asensi and Diez-Ga e as, 2017; Fe n´
andez-
Zamudio e al., 2019; Ga cía-Mu illo e al., 2019). Do˜
nana is one o
Eu ope’s mos impo an we lands—i is a majo o e win e ing si e o
wa e owl and con ains he con inen ’s la ges empo a y pond ne wo k
(Díaz-Paniagua e al., 2015; G een e al. 2018; Sche e e al., 2015). I
has ecei ed se e al in e na ional designa ions o no e (Ramsa Si e,
Wo ld He i age Si e, Special P o ec ion A ea o Bi ds [Eu opean Bi ds
Di ec i e]) and is also a si e o communi y impo ance (SCI; Eu opean
Habi a s Di ec i e), all o which unde sco es he pa k’s impo an ole in
biodi e si y conse a ion in he wes e n Medi e anean Basin and
wes e n Eu ope.
Mos o Do˜
nana’s we lands a e empo a y, and hose loca ed on
sandy subs a es a e g oundwa e dependen (Cus odio e al., 2009;
Díaz–Paniagua e al., 2015), and hei inunda ion cycle equi es
adequa e p oximi y o he wa e able o ensu e ha , a e au umn o
win e ains, hey can lood and pe sis long enough o pe mi he suc-
cess ul de elopmen o he o ganisms hey hos . The na ional pa k has
been g an ed Spain’s highes possible le el o legal p o ec ion. Un o -
una ely, he aqui e unde Do˜
nana ex ends a beyond he pa k’s
bounda ies, and ag icul u al i iga ion and esiden ial usage (local il-
lages and ou is eso s) in neighbo ing a eas is esul ing in in ensi e
g oundwa e abs ac ion. Consequen ly, he g oundwa e held in he
Do˜
nana aqui e has been g adually declining o e ecen decades
(Cus odio e al., 2009; Dimi ou e al., 2017). These dynamics, combined
wi h he e ec s o clima e change (i.e., educed annual ain all o e he
las 10 yea s), ha e led o sho e inunda ion cycles in app oxima ely 80
% o ponds and he comple e absence o inunda ion in 59 % o ponds (de
Felipe e al., 2023; Díaz–Paniagua e al., 2023; 2024). Howe e , he
magni ude o he impac p oduced by he o e exploi a ion o he aqui e
on he aqua ic o ganisms o his p o ec ed a ea has no ye been docu-
men ed in de ail, which we u gen ly equi e i we wan o make
e idence-based decisions o be e p o ec he Do˜
nana ecosys ems.
Do˜
nana ep esen s an impo an ho spo o aqua ic plan s (Ga cía-
Mu illo e al., 2006; Ga cía-Mu illo & Fe n´
andez–Zamudio, 2015). The
lo a o a gi en a ea is a ec ed by in e nal and ex e nal o ces, whose
impac s a e pa icula ly p onounced in habi a s expe iencing an h o-
pogenic dis u bances. Mode n ecologis s a e using plan species eco ds
o be e unde s and his o ical condi ions, landscape ecology, and con-
se a ion (MacKenzie e al., 2019). Flo is ic s udies can p o ide a
e e ence poin o gauging he success o conse a ion e o s. To un-
de s and he deg ada ion o inland we lands o e ime, i is c ucial o
cha ac e ize mac ophy e occu ence and communi y composi ion since
al e a ions can lead o d ama ic dec eases in habi a s uc u al
complexi y and biodi e si y (Can ona i e al., 2020). Mac ophy es
s ongly in luence hei local physicochemical en i onmen s
(Fe n´
andez–Zamudio e al., 2016). Changes in aqua ic plan communi y
abundance o composi ion a e o en ob ious signals ha he ecological
condi ion o a we land has changed (Bi k e al., 2012) and a e good
indica o s o eu ophica ion o o ganic pollu ion and hyd ological o
mo phological changes (Bi k e al., 2012). By assessing which species
occu wi hin a gi en egion o e ime, i is possible o de e mine
whe he ce ain plan species ha e disappea ed o expe ienced shi s in
dis ibu ion o abundance, which p o ides insigh in o pa e ns o oc-
cupancy and agmen a ion and can yield indings ha in o m conse -
a ion ac ions (Bü ne e al., 2022). In his sense, he ex inc ion o plan
species and hus he loss o biodi e si y in a ge egions can be shown
by compa ing his o ical da a (he ba ium specimens and/o old lo is ic
publica ions, his o ical da ase s) wi h he p esen -day si ua ion. This
app oach has been used o examine coun y-scale ex inc ion pa e ns in
sou he n Eu ope (Bo nand e al., 2016; Kempel e al., 2020).
Since i s i s conside a ion as a p o ec ed a ea in 1964, se e al
bo anical sampling su eys we e conduc ed a ound ha ime, including
some ha collec ed aqua ic plan s (Cabezudo, 1974; Cas o iejo,
Vald´
es-Be mejo, Ri as-Ma ínez, & Cos a, 1980; Ri as-Ma ínez, Cos a,
Cas o iejo, & Valdes, 1980); hese e o s ha e p o ided ex ensi e
his o ical in o ma ion ha can be used o assess he in ensi y o ecen
pa k de e io a ion. In his manusc ip , we ha e compiled all a ailable
da a on he aqua ic ascula plan s o Do˜
nana’s pond ne wo k wi h he
aim o quan i ying he changes in his g oup’s biodi e si y. Indeed,
aqua ic ascula plan s a e pa icula ly sensi i e o he en i onmen al
ans o ma ions ha ha e been occu ing in he icini y o Do˜
nana
Na ional Pa k.
He e, we examine he loss o aqua ic ascula plan species o e ime
in he we lands o he pa k’s sandy a ea. Ou main goal was o quan i y
a ia ion in his g oup’s biodi e si y. Mo e p ecisely, he objec i es
we e o: 1) iden i y how many aqua ic ascula plan species ha we e
p esen 50 yea s ago we e no longe obse ed a e 2020; and 2) assess
he changes in aqua ic ascula plan occu ence o e ime, e idence
ha can in o m u gen decision-making aimed a p ese ing Do˜
nana’s
unique aqua ic habi a s.
2. Ma e ials and me hods
In his s udy, we used all a ailable eco ds— om 1965 o he p e-
sen — o aqua ic ascula plan s in he eshwa e we lands in he sandy
a ea o Do˜
nana Na ional Pa k. We also ga he ed da a on plan p esence
om gene al s udies, he ba ia, pe sonal collec ions, and ield no es
made by bo anis s and he au ho s.
2.1. S udy a ea
Do˜
nana Na ional Pa k (a ound 36◦59
′
N, 6◦27
′
W) is loca ed in
Fig. 1. Loca ion o he ponds su eyed o aqua ic plan s in he sandy a ea o
Do˜
nana Na ional Pa k.
P. Ga cia-Mu illo e al.
Jou nal o Na u e Conse a ion 84 (2025) 126814
2
sou hwes e n Spain (Fig. 1) and spans an a ea o app oxima ely 54,300
ha. I is on he igh bank o he Guadalqui i Ri e , in he es ua y zone
whe e he i e encoun e s he A lan ic Ocean. I expe iences a Medi-
e anean clima e—mild win e s (mean win e empe a u e =10.9 ◦C),
ho d y summe s (mean summe empe a u e =23.5 ◦C), and mean
annual ain all o a ound 550 L/m
2
(da a eco ded om 1978 o 2023 in
a me eo ological s a ion loca ed in he cen al a ea o he pa k). This
egion hos s a di e se ange o ecosys ems and is widely acknowledged
as one o Eu ope’s mos impo an we lands (G een e al., 2018). The
pa k con ains wo well-di e en ia ed geomo phological zones: an
ex ensi e ma sh on a clay subs a e and a sandy a ea. The ma sh o ms a
b oad loodplain ha is ed by wa e om a ious ibu a y s eams and
he accumula ed ain all om au umn o ea ly sp ing. The sandy a ea is
o med o mo ing and s able dunes, in e spe sed wi h a mul i ude o
i egula and mode a ely sized dep essions ha o m g oundwa-
e –dependen ponds a e hea y ains. The ansi ion zones be ween he
ma shes and he s able dune a ea and be ween he mo ing dunes and
s able dunes ( he pe idune a ea) con ain he la ges abundance o ponds
in gene al, as well as he ponds wi h he la ges hyd ope iods (i.e.,
inunda ion pe iod du a ion); consequen ly, he ansi ion zones con ain
he pa k’s highes le els o biodi e si y (Díaz–Paniagua e al., 2015). In
his s udy, we ocus on he aqua ic ascula plan s ound in he pond
ne wo k, which we ha e been in ensi ely su eying since 2003. The
ma shes we e no conside ed wi hin he scope o ou s udy.
2.2. Gene al desc ip ion o he pond ne wo k
Mos o Do˜
nana’s ponds ha e been classi ied as Medi e anean
empo a y ponds (EU p io i y habi a code 3170). They a e shallow
g oundwa e -dependen we lands ha lood a e ainy pe iods and
commonly d y ou in summe . Pond looding may occu in au umn,
win e , o sp ing (excep ionally), depending on he iming o quan i y o
annual ains, which a e highly unp edic able in he Medi e anean.
The e o e, pond hyd ope iods can a y widely om yea o yea . In a
we yea (2004), esea che s obse ed mo e han 3,000 empo a y
ponds in he pa k using emo e-sensing da a (G´
omez-Rod íguez e al.,
2011). In con as , na u al pe manen ponds a e uncommon. We
conside ha he e a e only ou : h ee big ponds in he sandy a ea and a
ou h along he bo de o he ma sh; hey e ain wa e yea ound,
excep du ing se e ely d y pe iods. Recen ly, he hyd ope iods o some
o hese pe manen ponds ha e changed. Since 2012, one pond has d ied
ou e e y summe , ano he has d ied ou in 3 o he las 12 yea s, and a
hi d, he bigges and deepes one, d ied ou du ing he summe s o 2022
and 2023. A p esen , only he pond loca ed along he bo de wi h he
ma sh can be said o pe manen ly e ain a shallow le el o wa e .
G oundwa e discha ge o igina ing benea h he mo ing dunes o ms
small pe manen s eams (called e ue as by locals) ha un be ween
he ma shes and he dunes. The e a e also in e mi en s eams ha , in
ainy pe iods, low h ough he a ea and in o he neighbo ing ma sh. In
p e ious decades, he s eams would o m ponds (ca˜
nos) a hei mou hs
ha could hold wa e all yea . Howe e , in he las decade, mos ha e
d ied ou in he summe . In addi ion, in 2005, wo la ge a i icial ponds
we e buil o sedimen deposi ion and he pa ial ea men o wa e s
a i ing om ag icul u al a eas a ound he pa k. Likewise, Do˜
nana
con ains a ound 200 ponds ha we e a i icially ans o med in o pe -
manen wa e bodies. Named zacallones by he locals, hey a e usually
small and deep (app oxima ely 15 x 4 m in leng h and 1.5–2 m in dep h)
and we e c ea ed by deepening a small a ea o na u al empo a y ponds
o make g oundwa e a ailable o di e en adi ional uses (e.g., wa-
e ing animals, i iga ing small a ms). They a e no longe used o hese
pu poses, bu hey emain impo an pe manen aqua ic habi a s o
many o ganisms, wi h aqua ic plan s nea he op o he lis .
The physicochemical cha ac e is ics o Do˜
nana ponds ha e been
desc ibed in p e ious s udies (Fe n´
andez-Zamudio e al., 2016; G´
omez-
Rod íguez e al. 2009; Se ano e al., 2006). They a e eshwa e ponds,
wi h pH commonly anging om 5.7 o 9.3, low elec ical conduc i i y
(0.24–4.8 mS/cm), high concen a ions o Fe
2+
, and poo ca bona e
con en . Only in excep ional ponds, we ha e eco ded high conduc i i y
alues, as eco ded in a pond loca ed in he sou he n pa k, in which we
egis e ed 12.87 mS/cm in 2021.
2.3. Species nomencla u e and selec ion c i e ia o aqua ic plan s
De ining an “aqua ic plan ” can be complica ed. He e, we applied he
c i e ia o Ci ujano e al. (2014), o whom aqua ic plan s a e plan s ha
comple e hei li e cycles as subme ged o loa ing o ganisms in aqua ic
en i onmen s. Unlike Ci ujano e al. (2014), bu in ag eemen wi h
Bolpagni e al. (2020), Fois e al. (2024), and Mu phy e al. (2019), we
de ined A ella a is ulosa, E yngium co nicula um, and Juncus he e o-
phyllus as aqua ic plan s because, o success ully de elop, hey mus li e
in wa e du ing a leas he ea ly s ages o de elopmen . E idence o
his equi emen can be seen in s em and lea ana omy o ju eniles o
hese species.
Fo plan nomencla u e and axonomy, we used Flo a Ibe ica
(Cas o iejo e al., 1986–2020) c i e ia, e en i some nomencla u al
and/o axonomic changes occu ed a e he publica ion o ha wo k.
Ranunculus pel a us is a p oblema ic axon because he e is a high deg ee
o plas ici y in ce ain diagnos ic ai s; we hus ocused on he species
le el, a oiding he di e en c i e ia used by di e en au ho s o iden i y
subspecies (Cook, 1986; Piza o, 1995; Velayos, 1988; Wiegleb e al.,
2017). The diminu i e Ela ine species (E. b ochonii, E. hexand a, and
E. mac opoda) display s ong pheno ypic plas ici y (Łysko e al., 2022;
Razi a d e al., 2017) and can only be easily dis inguished using ep o-
duc i e ai s. Because we equen ly obse ed hem in hei ege a i e
s a e, we e e o hem as he Ela ine minu e g oup (Ela ine sp.).
In his s udy, we only discuss au och honous species. We a e awa e
ha Azolla iliculoides was i s de ec ed in he pa k in 2001
(Ga cía–Mu illo e al., 2007) and was seen co e ing a la ge pe cen age
o he ma sh bo de s be ween 2001 and 2009, including some ponds in
close p oximi y o he sandy a ea (Díaz-Delgado e al., 2011). The spe-
cies’ p esence in he Do˜
nana pond ne wo k is now limi ed o a ew
isola ed poin s nea he ma shes (see he p esen -day map in Appendix
A).
2.4. Th ea ened species
Fo he species conside ed in his s udy, we ha e aken in o accoun
h ea le els and p o ec ion le els, based on egional, na ional, and
Eu opean egula ions, as well as he in o ma ion con ained in ed lis s
and ed books. These sou ces o e e ence a e indica ed in Table 1.
2.5. Sou ces o aqua ic plan da a
Because Do˜
nana is an a ea o g ea na u al alue, i has been ela-
i ely well p ospec ed om a lo is ic poin o iew. Howe e , su eys
la gely ocused on e es ial plan s, gene ally igno ing aqua ic plan s.
F om 1965 o 2000, ew in ensi e da a we e collec ed ac oss he pa k;
mos o he in o ma ion a ailable ela es o species occu ence. These
da a a e none heless impo an because hey indica e he p esence o
singula o h ea ened species ha we ha e no obse ed in mo e ecen
su eys. The i s da a on aqua ic ege a ion we e collec ed by Ma -
azano (1967), du ing one o he ini ial expedi ions o Do˜
nana and we e
a complemen a y pa o desc ip ions o he a ea’s biodi e si y,
including we land biodi e si y. The i s publica ions speci ically
ocused on he lo a in he p o ec ed a ea we e Cabezudo (1974; 1975;
1978) and Galiano & Cabezudo (1976), which we e ollowed by Cas-
o iejo e al. (1980) and Ri as-Ma ínez e al. (1980). This body o wo k
es ablished he basis o wha is known abou Do˜
nana’s lo a and
ege a ion. Howe e , i is impo an o ei e a e ha hese s udies we e
especially in e es ed in e es ial plan s and p o ide spa se in o ma ion
abou aqua ic plan s. Da a on he pa k’s aqua ic plan s gene ally come
om Ga cía-Mu illo e al. (2006; 2014), Ga cia–Mu illo &
P. Ga cia-Mu illo e al.
Jou nal o Na u e Conse a ion 84 (2025) 126814
3
Table 1
Lis o he aqua ic plan species his o ically ound in he sandy a ea o Do˜
nana Na ional Pa k, indica ing bio ype, habi a (pe manen o empo a y), yea o he las
obse a ion, and p esence in he pa k. To al species occupancy (%) was es ima ed by di iding he o al numbe o g id cells in which each species was de ec ed by he
o al numbe o g id cells su eyed his o ically (n =135). Species occupancy ac oss he pa k was also examined by looking a he numbe o g id cells occupied by he
species in he no he n pa k (N), cen al pe idune a ea (P), and sou he n pa k (S). Resul s a e shown o he GLMMs conduc ed using he numbe o g id cells occupied
by each species be ween 2001 and 2010 (P2000) and be ween 2021 and 2023 (P2020) (signi ican di e ences in bold), only o hose species o which we ha e da a in
hese wo pe iods. Also indica ed a e he ed-lis ca ego ies in which each species is classi ied (and he ela ed e e ences).
Species Red-lis ca ego y Bio ype Habi a Las yea
obse ed
%
o al
% N % S % P P2000 P2020 GLMM esul s
Apium inunda um DD
4
AM empo a y 2023 37.04 68.00 12.00 20.00 26 11
χ
2
¼7.7121, p ¼
0.0055
A ella a is ulosa EN
10, 11
CR
1, 2, 4, 7
AM empo a y 2023 13.33 90.00 5.00 5.00 10 4
χ
2
¼12.394, p ¼
0.0004
Calli iche b u ia LC
1
SA empo a y 2023 44.44 78.33 8.33 13.33 37 23
χ
2
¼6.6601, p ¼
0.010
Calli iche lusi anica NT
1
EN
2
DD
4
SA empo a y 2023 13.33 72.22 11.11 16.67 6 2
χ
2
=1.923, p =
0.1655
Calli iche ob usangula DD
4
SA empo a y 2023 28.89 66.67 12.82 20.51 20 7
χ
2
¼16.794, p <
0.0005
Calli iche s agnalis LC
1
AM empo a y 2023 32.59 70.45 11.36 18.18 25 10
χ
2
¼9.5807, p ¼
0.0020
Calli iche unca a subsp.
occiden alis
DD
1
SA empo a y 2023 17.78 58.33 12.50 29.17 4 9
χ
2
=2.7359, p =
0.0981
Ce a ophyllum deme sum DD
4
SP pe manen 1978 2.22 66.67 0.00 33.33 0 0
Ela ine alsinas um NT
1, 2
SA empo a y 2023 17.78 87.50 0.00 12.50 9 6
χ
2
=0.8637, p =
0.3527
Ela ine sp. NT
1, 2
* DD
4
SA empo a y 2023 16.30 54.55 9.09 36.36 7 7
χ
2
=0.8637, p =
0.3527
E yngium co nicula um VU
4
AM empo a y 2023 25.93 71.43 8.57 20.00 24 2
χ
2
¼10.445, p ¼
0.0012
Hyd ocha is mo sus- anae CR
2, 4, 7
EN
10, 11
VU
12
FP pe manen 2006 1.48 0.00 100.00 0.00 0 0
Illeceb um e icilla um LC
1
AM empo a y 2023 40.74 78.18 0.00 21.82 31 17
χ
2
¼11.103, p ¼
0.0009
Isoe es ela a NT
1
DD
4
SA empo a y 2021 8.89 83.33 0.00 16.67 10 1
χ
2
¼4.4664, p ¼
0.0346
Isolepis lui ans DD
4
FA empo a y 2021 16.30 77.27 0.00 22.73 16 2
χ
2
¼27.498, p <
0.0005
Juncus he e ophyllus NT
1
AM empo a y 2022 43.70 72.88 6.78 20.34 29 28
χ
2
=2.2013, p =
0.1379
Lemna gibba LC
1
FA pe manen 2023 34.81 48.94 44.68 6.38 27 8
χ
2
¼11.599, p ¼
0.0007
Lemna mino LC
1
FA pe manen 2023 51.85 51.43 34.29 14.29 35 20
χ
2
¼8.0529, p ¼
0.0045
Lemna isulca CR
2, 3
DD
4
FP pe manen 2022 7.41 0.00 100.00 0.00 5 4
χ
2
=0.4399, p =
0.5072
Ma silea s igosa VU
1, 2, 4, 5, 6, 8, 10,
12, 13
AM empo a y 2023 1.48 100.00 0.00 0.00 0 1
χ
2
¼6.2389, p ¼
0.0125
My iophyllum al e ni lo um LC
1
SA empo a y 2023 48.15 75.38 4.62 20.00 39 20
χ
2
¼9.9806, p ¼
0.0016
Najas ma ina LC
1
SA pe manen 2018 0.74 0 0 100 0 1
Nupha lu eum CR
4
VU
10
FP pe manen 1981 0.74 100.00 0.00 0.00 0 0
Nymphaea alba CR
4
VU
10
FP pe manen 2008 2.22 100.00 0.00 0.00 0 0
Po amoge on c ispus LC
1
SP pe manen 1978 2.22 66.67 0.00 33.33 0 0
Po amoge on lucens DD
4
SP pe manen 2023 10.37 7.14 64.29 28.57 8 7
χ
2
¼8.9623, p ¼
0.0028
Po amoge on na ans DD
4
SP pe manen 2023 15.56 90.48 0.00 9.52 6 6
χ
2
=0.3116, p =
0.5767
Po amoge on pec ina us LC
1
SP pe manen 2023 24.44 42.42 33.33 24.24 14 15
χ
2
=0, p =1
Po amoge on polygoni olius DD
4
AM pe manen 2010 5.93 87.50 12.50 0.00 5 0
χ
2
¼122706, p <
0.00001
Po amoge on ichoides LC
1
SA empo a y 2023 23.70 46.88 28.13 25.00 8 10
χ
2
=1.1898, p =
0.2754
Ranunculus pel a us LC
1
SA empo a y 2023 70.37 61.05 25.26 13.68 44 48
χ
2
=0.0308, p =
0.8606
Ranunculus ipa i us LC
1
SA empo a y 2021 20.00 81.48 3.70 14.81 13 2
χ
2
¼22.342, p ¼
0.00001
Ruppia d epanensis DD
4
SA empo a y 2022 5.19 0.00 57.14 42.86 3 1 Х
2
=2.9929, p =
0.0836
Spi odela poly hiza DD
4
FP pe manen 2013 2.22 33.33 66.67 0.00 0 0
U icula ia aus alis EX
4
, EN
10
FP pe manen 1977 4.44 83.33 16.67 0.00 0 0
U icula ia gibba NT
1
, CR
2, 4, 7
VU
10, 12
FP pe manen 1977 0.74 100.00 0.00 0.00 0 0
Wol ia a hiza EN
2,4,
VU
10, 12
FP pe manen 2023 11.11 6.67 93.33 0.00 9 1
χ
2
¼323323, p <
0.00001
Zannichellia ob usi olia NT
1
, VU
4
SA empo a y 2023 25.19 26.47 47.06 26.47 8 13 Х
2
=2.4203, p =
0.1198
P. Ga cia-Mu illo e al.
Jou nal o Na u e Conse a ion 84 (2025) 126814
4
Fe nandez–Zamudio (2015), and Fe n´
andez-Zamudio e al. (2016), in
addi ion o a ious monog aphs on hyd ophy e gene a (Ga cia–Mu illo
e al., 2000; Tala e a e al., 1986) and cho ological s udies
(Fe n´
andez–Zamudio e al., 2006; Ga cía–Mu illo e al., 1990).
Al hough cha ophy es a e impo an componen s o aqua ic ecosys ems,
he e is no abundan his o ical in o ma ion abou hese mac ophy es in
Do˜
nana’s ponds. The e o e, we ha e no included his g oup in his
s udy, al hough we a e cu en ly assessing hei conse a ion s a us o a
u u e s udy.
O e all, o he pe iod om 1965 o 2023, he e a e ege a ion da a
o 263 si es ac oss he pa k (Fig. 1), including 127 empo a y ponds (o
which 25 ha e d ied up in ecen yea s: 2021–2023); 21 ponds o med
by in e mi en s eams (ca˜
nos); 4 pe manen na u al ponds (o which 3
ha e d ied ou du ing a leas one summe o e he las decade); 2
a i icial ponds; 106 zacallones (o which 16 ha e disappea ed and 32
now d y ou in he summe ); and 3 e ue as (Table 2). In p e ious
s udies desc ibing he dynamics o animal and plan assemblages in he
Do˜
nana pond ne wo k (Fe n´
andez-Zamudio e al., 2016; Flo encio e al.,
2014), we in ensi ely moni o ed 167 si es om 2001 o 2010, no ing all
he aqua ic plan species occu ing a each si e. To e alua e changes in
he pa k’s plan assemblages, we conduc ed new su eys om 2021 o
2023, du ing which we e isi ed 114 o he abo e si es as well as 31
o he si es ha had hos ed ponds in p e ious decades (bu ha ha e
d ied up in mo e ecen yea s).
We also consul ed he ollowing he ba ia o his wo k: SEV (Uni-
e si y o Se ille, Se ille, Spain), MA (Royal Bo anical Ga dens-CSIC,
Mad id, Spain), UNEX (Uni e si y o Ex emadu a, Badajoz, Spain),
and HSS (Resea ch Cen e o La O den-Valdeseque a, M´
e ida, Spain).
2.6. Da a analyses
We cons uc ed a da a ma ix comp ising he species o which we
had eco ds and he yea s hey we e obse ed ( om 1965 o 2023). The
da a we e g ouped in o ou pe iods: a) be o e 2000, b) 2001–2010
(he ea e , P2000), c) 2011–2020, and d) 2021–2023 (he ea e ,
P2020). We di ided he en i e pa k in o Uni e sal T ans e se Me ca o
(UTM) g id cells (1 km x 1 km) on o which we mapped he pe iod-
speci ic p esence o each species using Qgis ( . 3.16). While we only
had spo adic obse a ions o he pe iod be o e 2000, he in o ma ion
a ailable o us was ex emely aluable in de ec ing he local ex inc ion
o ce ain species. In o al, we had in o ma ion o 135 g id cells. Da a
om 2003 onwa ds included pe iods o in ensi e and sys ema ic su eys
ha employed simila me hodologies; hey we e use ul o e alua ing
changes in species occupancy (i.e., he p esence o a species in he g id
cells) h oughou he pa k. Speci ically, we had in ensi ely su eyed
ponds (n =164) in 73 g id cells o e wo pe iods (P2000 and P2020),
allowing us o conduc s a is ical compa isons.
The pa k’s ponds a e he e ogeneously dis ibu ed. Pond densi y is
high in he no he n pa k, e en highe in he cen al pe idune a ea, and
lowe in he sou he n pa k (G´
omez-Rod íguez e al., 2011). Because
compa ing indi idual ponds could ha e gi en ise o pseudo eplica ion,
we used he g id cells as he uni s in which he p esence o aqua ic plan
species was quan i ied. We es ima ed he numbe o g id cells occupied
by each species and hen di ided his numbe by he o al numbe (n =
73) o cells su eyed du ing bo h pe iods (P2000 and P2020) o a i e a
ela i e species abundance. To conduc compa isons, we used gene al-
ized linea mixed models (GLMMs) wi h a binomial e o dis ibu ion;
pe iod was he p edic o a iable, and g id cell iden i y was a andom
a iable. We also es ima ed he numbe o species eco ded in each g id
cell du ing he wo pe iods and hen pe o med compa isons using a
GLMM wi h a Poisson e o dis ibu ion (g id cell iden i y = andom
ac o ).
Fo each species, we es ima ed he numbe o g id cells occupied
du ing each pe iod. We hen used he di e ence in g id cells occupied
be ween pe iods as he esponse a iable in a gene alized linea model
(GLM) wi h a Poisson e o dis ibu ion. In his model, he numbe o
g id cells occupied in P2000 was a co a iable, and he p edic o a i-
ables we e species a ea o dis ibu ion (no he n, cen al, and sou he n);
bio ype (AM: amphibious annual; SA: subme ged annual; FA: loa ing
annual; SP: subme ged pe ennial; FP: loa ing pe ennial) and habi a
ype (pe manen o empo a y in P2000). Pos -hoc Tukey es s we e used
o pai ed compa isons be ween di e en a eas o he pa k.
3. Resul s
3.1. Disappea ance and decline o Dona˜
na’s aqua ic ascula plan
species
We we e able o ob ain eco ds om 1965 o 2023 o 38 species o
au och honous aqua ic ascula plan s ha occu in he we lands o
Do˜
nana’s sandy a ea (Table 1). These species comp ised 8 amphibious
plan s, 11 loa ing plan s, and 19 subme ged plan s. In e ms o habi a ,
21 species we e ound in empo a y we lands, and 17 occu ed in pe -
manen we lands. All we e membe s o Spe ma ophy a, wi h he
excep ion o Ma silea s igosa, a membe o he aqua ic P e idophy a, and
Isoe es ela a as a Lycopodiophy a. We mapped species occu ence in
space and ime (Appendix A).
I was appa en om he eco ds we ga he ed ha nine aqua ic
ascula plan species—Ce a ophyllum deme sum, Po amoge on c ispus,
Nupha lu eum, Nymphaea alba, U icula ia gibba, U icula ia aus alis,
Po amoge on polygoni olius, Hyd ocha is mo sus- anae, and Spi odela pol-
y hiza—ha e disappea ed om he pa k. All nine o hese species ended
o occu in pe manen wa e bodies; six we e loa ing pe ennials, wo
we e subme ged pe ennials, and one was amphibious (Fig. 2).
Ce a ophyllum deme sum and P. c ispus we e ela i ely abundan in
he Ibe ian Peninsula un il he 1980 s. They a e species ypical o pe -
manen backwa e s and a e bo h easy o iden i y. Thei i s eco ds in
he s udy a ea a e ound in Ma azano (1967), who collec ed
C. deme sum in 1965 in a pe manen pond in he cen al pe idune a ea;
he ound P. c ispus in his same pond, as well as in a ca˜
no along he ma sh
bo de . Bo h species we e las obse ed in 1978, when C. deme sum was
collec ed om a di e en pe manen pond, and P. c ispus was collec ed
om a di e en ca˜
no a he ma sh bo de (Cas o iejo e al., 1980).
Do˜
nana used o con ain wo wa e lilies—N. lu eum and N. alba— ha
a e cu en ly p o ec ed by egional law (Dec e o 23/2012 in Table 1).
AM: amphibious; SP: subme ged pe ennial; SA: subme ged annual; FP: loa ing pe ennial; FA: loa ing annual.
1: IUCN (2024): (h ps://www.iucn edlis .o g/); 2: Mo eno (2010); 3: Ba˜
na es e al. (2006); 4: Cabezudo e al. (2005); 5: Mo eno e al. (2019); 6: Ba˜
na es e al. (2010);
7: Ba˜
na es e al. (2004); 8: Habi a s Di ec i e 92/43/EEC; 9: Real Dec e o 139/2011 (h ps://www.boe.es/eli/es/ d/2011/02/04/139); 10: Dec e o 23/2012 (h ps://
www.jun adeandalucia.es/boja/2012/60/6); 11: Spanish Ca alogue o Endange ed Species (h ps://www.mi eco.gob.es/es/biodi e sidad/ emas/conse acion-de-es
pecies/especies-p o eccion-especial/ce-p o eccion-lis ado.h ml); 12: Blanca e al. (1999), 13: Lis o Wildli e Species unde Special P o ec ion Regimen (h ps://www.
mi eco.gob.es/es/biodi e sidad/ emas/conse acion-de-.
Table 2
Numbe o di e en aqua ic habi a s in Do˜
nana su eyed o aqua ic plan s
be ween 1965 and 2023. Also shown a e he numbe o he pe manen and
empo a y ponds su eyed ha ha e d ied up and he numbe o o me ly
pe manen si es ha now d y ou in he summe .
N D ied up D y in summe
Tempo a y ponds 127 25 102
Pe manen na u al ponds 4 0 3
S eams (including ca˜
nos) 21 0 10
Pe manen a i icial ponds 2 0 0
Re ue as (pe manen dune s eams) 3 0 0
Zacallones (a i icially deepened ponds) 106 16 32
P. Ga cia-Mu illo e al.
Jou nal o Na u e Conse a ion 84 (2025) 126814
5
Nupha lu eum was i s eco ded in he pa k in 1981, mos equen ly in
he la ge, deep backwa e s o a pe manen s eam. Howe e , i was no
ound in la e yea s. Un il he 1980 s, i was also ela i ely common in
pe manen backwa e s in he Spanish egions o Andalusia, Ex em-
adu a, and Cas illa La Mancha (Ci ujano e al., 2014; An hos, 2024). I
hen began o disappea om many places and is cu en ly an ex emely
a e species in Andalusia, whe e i is lis ed in he Red Lis o Andalusian
Flo a, as c i ically endange ed (Cabezudo e al., 2005). To h i e, i
needs slow, medium-dep h wa e s in eu h ophic, bu sca cely pollu ed
en i onmen s (Ci ujano e al., 2014; Melendo e al., 2003). Nymphaea
alba also began o disappea om he sou he n Ibe ian Peninsula in he
1980 s and was obse ed in Do˜
nana na ional pa k un il 2008. A p esen ,
i is a a e species ound in he sou he n Ibe ian Peninsula. I s ill pe sis s
in he a ea a ound he na ional pa k, a one loca ion owa ds he
beginning o La Rocina s eam, conside ed o be nea ly he las popu-
la ion in wes e n Andalusia. To h i e, i needs condi ions simila o
hose o he p e ious species (Ci ujano e al., 2014; Melendo e al.,
2003).
U icula ia aus alis and U. gibba a e species ypically ound in s ill,
dys ophic wa e s ha a e poo in dissol ed nu ien s and ich in humic
acids esul ing om plan decomposi ion, which con ibu e o en i-
onmen al acidi ica ion (Ci ujano e al., 2014). Un il he 1970 s,
U. aus alis was ound in he s udy a ea in he pe manen pa s o wo
ca˜
nos along he ma sh bo de . U icula ia gibba was obse ed in Do˜
nana
un il he 1980 s, in pe manen backwa e s o in e mi en and pea y
wa e cou ses ha lowed in o he ma sh. Bo h species we e also p esen
in some o he pe manen s eams ha lowed in o he no he n pa k.
U icula ia aus alis is p o ec ed by egional law (Dec e o 23/2012 in
Table 1) bu is classi ied as ex inc in he Red Lis o Andalusian Flo a,
which has labelled U. gibba as c i ically endange ed (Cabezudo e al.,
2005).
Po amoge on polygoni olius is also a species ypically ound in i e s,
s eams, pea bogs, and pea y we lands, which ha e acidic wa e s ha
a e poo in sal s and nu ien s (Ci ujano e al., 2014). In he s udy a ea, i
occu ed in he pe manen wa e s o in e mi en s eams along he
ma sh bo de , whe e i coexis ed wi h U. aus alis and U. gibba. I was
s ill ound in a single pea y pond up un il 2010. Al hough no longe
occu ing in he pa k, i is p esen in he su ounding a ea.
His o ically, Hyd ocha is mo sus- anae occu ed in he e ue as and
one o he pe manen ponds along he ma sh bo de . The las he ba ium
specimen was collec ed in 1985 om a e ue a, he only place i
emained; his unique habi a is o med by g oundwa e discha ge
eme ging om benea h he mo ing dunes, on he edge o he ma sh.
This si e was he las loca ion whe e he species was ound on he Ibe ian
Peninsula (Ga cia-Mu illo e al., 2000). I is p o ec ed by Spanish na-
ional law (Real Dec e o 139/2011 in Table 1, amended Ap il 7, 2023)
and egional law (Dec e o 23/2012 in Table 1). Likewise, H. mo sus-
anae has been classi ied as c i ically endange ed by he 2010 Red Lis o
Spanish Vascula Flo a (Mo eno, 2010), he Red Lis o Andalusian
Vascula Flo a (Cabezudo e al., 2005), and he A las and Red Book o
Spain’s Th ea ened Vascula Flo a (Ba˜
na es e al., 2004). I has been he
ocus o a ious ein oduc ion and epopula ion p og ams in Do˜
nana,
whe e i s p esence has been in ensi ely moni o ed. The species has no
been obse ed na u ally in he wild since 2010, bu indi iduals o he
ein oduced popula ion we e p esen om 2011 o 2019 (Cobo, pe -
sonal communica ion). New ein oduc ion e o s a e unde way.
Spi odela poly hiza, o g ea e duckweed, was a e bu obse ed be-
ween 1988 and 2013 in a pe manen zacall´
on ha has now d ied up.
A e 2013, we ne e again ound i a ha si e.
O he species appea o be a e because hey we e p esen ac oss less
han 8 % o he s udy a ea in he pe iod om 2000 o 2010. These
species we e Po amoge on na ans, Lemna isulca, Calli iche lusi anica. C.
unca a subsp. occiden alis, and Ruppia d epanensis. Be ween 2021 and
2023, we obse ed low le els o occupancy (<4 %) o A ella a is ulosa,
Isolepis lui ans, Ranunculus ipa i us, and Isoe es ela a. Fu he mo e, in
ecen yea s, Ma silea s igosa has only been seen in an epheme al pond,
and Wol ia a hiza has only been obse ed in a e ue a.
O he 38 aqua ic plan species o which we had eco ds, all hem
appea in a IUCN ed-lis ca ego y, based on he a ious ed lis s and ed
books we used. Twel e species (31.6 %) a e classi ied in ca ego ies
conside ed o be h ea ened wi h ex inc ion ( ulne able, endange ed o
c i ically endange ed). One species is ex inc a he egional le el:
U icula ia aus alis; 6 a e c i ically endange ed (CR): A ella a is ulosa,
Hyd ocha is mo sus- anae, Lemna isulca, Nupha lu eum, Nymphaea alba,
and U icula ia gibba; 2 a e endange ed (EN): Calli iche lusi anica and
Wol ia a hiza; 3 a e ulne able (VU): E yngium co nicula um, Ma silea
s igosa and Zannichellia ob usi olia; 3 a e nea h ea ened (NT): Ela ine
alsinas um, Juncus he e ophyllus and Isoe es ela a; 10 a e da a de icien
(DD): Apium inunda um, Calli iche ob usangula, C. unca a, Ce a o-
phyllum deme sum, Isolepis lui ans, Po amoge on lucens, P. na ans, P. pol-
ygoni olius, Ruppia d epanensis and Spi odela poly hiza; and he emaining
ones a e classi ied as leas conce n (LC) (Table 1).
Eigh (21.2 %) a e p o ec ed unde Eu opean, Spanish, o Andalusian
law. No ably, M. s igosa is p o ec ed a all h ee le els; A. is ulosa and
H. mo sus- anae a e p o ec ed a na ional and egional le els; and
N. lu eum, N. alba, U. aus alis, U. gibba and W. a hiza a e p o ec ed a
he egional le el (Table 1).
3.2. Changes in species occupancy (2003–2007 o 2021–2023)
In addi ion o he ex inc species, 21 species displayed a dec ease in
occupancy o e ime (Fig. 3). Du ing P2000, 11 species we e e y
abundan and occupied mo e han 30 % o he g id cells. Du ing P2020,
only h ee species had high le els o occupancy (>30 %); o hese, only
Ranunculus pel a us, he mos widely dis ibu ed species (Fig. 4), had a
highe occupancy le el han p e iously, while Calli iche b u ia and
Juncus he e ophyllus had a lowe occupancy (Fig. 3). All o he equen
species in P2000 exhibi ed, in P2020, a educ ion o 20–30 %, as in he
case o Apium inunda um (Fig. 3 and Fig. 4). Two o he species (Zanni-
chellia ob usi olia and Po amoge on pec ina us) displayed small, non-
–signi ican inc eases in occupancy le els: hey wen om being among
he species wi h lowe occupancy in P2000 o he g oup o common
Fig. 2. Numbe o ascula aqua ic plan species (ex an s. locally ex inc )
obse ed in he we lands o he Do˜
nana sandy a ea, classi ied by bio ype and
habi a (pe manen s. empo a y).
P. Ga cia-Mu illo e al.
Jou nal o Na u e Conse a ion 84 (2025) 126814
6
species in P2020. O he 28 species o which we had da a o bo h P2000
and P2020, 15 unde wen a signi ican dec ease in occupancy be ween
pe iods, wi h an a e age educ ion o 17.8 % ( ange: 1.36–30.1 %,
Table 1, Fig. 3).
A ea o dis ibu ion (
χ
2
=19.378, d =2, p =0.006) and bio ype (
χ
2
=11.169, d =4, p =0.0245) signi ican ly in luenced species occupancy
be ween he wo pe iods. The species ound in he no he n pa k di e ed
signi ican ly om he species ound ac oss he en i e s udy a ea (Tukey
es : p =0.012), and he subme ged annuals di e ed om he loa ing
annuals (Tukey es : p <0.019). Occupancy dec eased he mos o he
subme ged pe ennials, a pa e n ha seemed a ibu able o he signi -
ican dec ease in Po amoge on lucens, which disappea ed om 4 o he 9
g id cells i had p e iously occupied. We obse ed ha bo h P. lucens and
P. na ans now only occu in ponds in a educed numbe o ege a i e
o ms ha ha e no lowe s o seeds (Fig. 5). In he case o P. na ans, he
species occupied a simila numbe o ponds in bo h pe iods, bu ha
p esence was limi ed o isola ed indi iduals in P2020, con as ing wi h
i s high deg ee o pond co e age and lowe p oduc ion in P2000
(Fig. 5).
The numbe o species pe g id cell di e ed among he h ee main
a eas o he pa k (
χ
2
=31.41, d =2, p <0.0005). The cen al pe idune
a ea had he g ea es species ichness du ing bo h pe iods (Table 3). The
numbe o species pe g id cell also di e ed be ween he wo pe iods (
χ
2
=61.00, d =1, p <0.0005). In P2000, he a e age numbe o species
pe g id cell was 7.25 (+0.61) and dec eased o 4.15 (+0.34) in P2020.
The dec ease was highes in he no he n pa k (49.5 %) and less p o-
nounced in he cen al pe idune a ea and sou he n pa k (29.5 and 29.7
%, espec i ely; Table 3 and Fig. 6).
4. Discussion
Ou wo k assembled exis ing eco ds o 38 species o ascula
Fig. 3. Rela i e abundance (%) o aqua ic plan s obse ed in he Do˜
nana pond ne wo k (numbe o g id cells in which a species was p esen di ided by he numbe o
g id cells su eyed) o 2001–2010 (P2000) and 2021–2023 (P2020).
Fig. 4. Maps o Ranunculus pel a us and Apium inunda um p esence in he pa k. G id cell colo indica es he yea o he las obse a ion: be o e 2000 ( ed); be ween
2001 and 2010 (blue); be ween 2011 and 2020 (g ay); and be ween 2021 and 2023 (black). (Fo in e p e a ion o he e e ences o colo in his igu e legend, he
eade is e e ed o he web e sion o his a icle.)
P. Ga cia-Mu illo e al.
Jou nal o Na u e Conse a ion 84 (2025) 126814
7
aqua ic plan s ound in he ponds o Do˜
nana Na ional Pa k. Ou esul s
e ealed he impo an ole played by his p o ec ed a ea in main aining
Ibe ian and Eu opean biodi e si y (Diaz–Paniagua e al., 2015; Flo -
encio e al., 2014; 2016). Indeed, mo e han 60 % o hese species a e
classi ied in o one o he IUCN’s ed lis ca ego ies, and nea ly 25 % a e
p o ec ed unde a ious laws. Thus, Do˜
nana has played a c ucial ole in
p o ec ing species ha a e highly sensi i e o he an h opogenic changes
ha ha e occu ed o e ecen decades in his egion. Fo mos o hese
species, exis ence ou side he pa k’s bo de s is impossible, gi en he
ways he en i onmen is being agg essi ely modi ied by an h opogenic
ac i i ies, which include in ensi e a ming, hea y ou ism, and pe o-
chemical ope a ions. In addi ion, e en Do˜
nana is no longe a sa e ha en,
as ag icul u e and ou ism ou side he pa k’s bo de s a e p o oundly
ans o ming he habi a s wi hin he p o ec ed a ea, as we discuss
below.
Aqua ic plan popula ions play a key ole in aqua ic habi a s
(Fe n´
andez-Zamudio e al., 2016; Ga cía Mu illo & Fe n´
andez-Zamudio,
2015) by p o iding i al s uc u al and unc ional ecosys em se ices
(O’Ha e e al., 2018). The e o e, he conse a ion o he aqua ic plan s
in Do˜
nana’s ponds would help gua an ee p ope ecosys em unc ioning.
Un o una ely, he pa k has no been spa ed om he widesp ead
de e io a ion o aqua ic ecosys ems occu ing elsewhe e. In ac , he
pa k is cu en ly acing se e al conse a ion challenges, which a e
consequences o bo h clima e change and an h opogenic dis u bances.
Based on all he eco ds we ga he ed, we we e able o iden i y wo
dis inc phases o species decline. The e was an ea ly phase, a ound he
end o he 20 h cen u y, which coincided wi h he in oduc ion and
sp ead o he ed swamp c ay ish (P ocamba us cla kii), an exo ic species
ha can ac as an ecosys em enginee because o i s d ama ic impac s on
he physical and chemical cha ac e is ics o aqua ic habi a s, leading o
biodi e si y losses (Geige e al., 2005; Ghe a di, 2007; Sou y-G osse
e al., 2016). The e was also a mo e ecen phase ela ed o he o e -
exploi a ion o he egional aqui e , one o he main ac o s, which
oge he wi h he inc ease in empe a u e and lowe ain all, has led
many ponds o d y up comple ely, wi h he emaining ponds
Fig. 5. Po amoge on na ans was commonly seen co e ing ponds wi h la ge numbe s o ep oduc i e plan s (2007 on le ) bu now only occu s as spa se ege a i e
o ms (2023 on igh ).
Table 3
Mean numbe o ascula aqua ic plan species pe g id cell in he no he n pa k,
he cen al pe idune a ea, and he sou he n pa k, as well as he o al numbe o
ponds su eyed. The ela i e numbe o species los (%) is also indica ed (mean
numbe o species p esen du ing he second ocal su ey pe iod [2021–2023]
e sus he i s ocal su ey pe iod [2001–2010]).
No he n
pa k
Pe idune
a ea
Sou he n
pa k
To al
2001–2010 8.33 (±0.79) 9.54 (±1.11) 3.70 (±0.95) 7.25
(±0.607)
2021–2023 4.21 (±0.39) 6.82 (±0.84) 2.60 (±0.65) 4.15 (±0.34)
% species loss 49.46 29.45 29.73 42.76
Fig. 6. Compa ison o mean species numbe pe g id cell be ween he wo ocal
su ey pe iods: 2001–2010 (P2000) and 2021–2023 (P2020). The g id colo s
indica e species loss ( ed) o gain (blue) o no di e ence be ween pe iods
(whi e). (Fo in e p e a ion o he e e ences o colo in his igu e legend, he
eade is e e ed o he web e sion o his a icle.)
P. Ga cia-Mu illo e al.
Jou nal o Na u e Conse a ion 84 (2025) 126814
8
expe iencing educed hyd ope iods, pa icula ly in he no hwes e n
pa k (de Felipe e al., 2023).
4.1. Loss o aqua ic ascula plan species
App oxima ely 24 % o he aqua ic ascula plan s e e eco ded in
he Do˜
nana pond ne wo k ha e disappea ed. Fi e species (U icula ia
aus alis, U. gibba, Nymphaea alba, Ce a ophyllum deme sum, Po amoge on
c ispus) we e las obse ed be o e 2000, and hei disappea ance can be
linked o he expansion o he exo ic ed swamp c ay ish (P. cla kii). This
in asi e species was in oduced in o he Do˜
nana ma sh in 1974 and
sp ead in o he ponds in he sandy a ea a ound 1983 (Díaz-Paniagua
e al., 2014). Resea ch obse ed he disappea ance o he abundan
biomass p oduced by subme ged and loa ing mac ophy es in he ma sh
and i s adjacen s eams be o e he 1990 s. This pa e n was a ibu ed o
he c ay ish (Dua e e al., 1990), which s ongly p e e s mac ophy es
o e o he po en ial ood sou ces (Ci ujano e al., 2004). The c ay ish’s
impac on pond s uc u e has been s udied expe imen ally (A ibas
e al., 2014)—i has been ound o inc ease u bidi y, boos nu ien
concen a ions, educe oxygen le els, and de ou all mac ophy e
biomass. All i e o he plan species men ioned abo e we e inhabi an s
o pe manen wa e s (e.g., na u al ponds o zacallones), which is also he
habi a ha allows he c ay ish’s popula ions o pe sis (Díaz-Paniagua
e al., 2014).
Addi ionally, he explosion in ag icul u al ac i i y a ound he pa k
has led o a e i ied in lux o nu ien s in Do˜
nana’s aqua ic habi a s,
which ha e a i ed ia en y wa e s and he inc ease in u ban was e-
wa e s om he inc easingly popula ed illages adjacen o he pa k
(Pa edes e al., 2019; 2021; Pe is e al., 2024). Li es ock a ming, high
li es ock densi ies, e ilize usage combined wi h i iga ion ag icul u e,
and u ban was ewa e s in a eas su ounding Do˜
nana ha e con ibu ed
o he eu ophica ion o aqua ic habi a s in he no he n pa k, whe e he
i e abo e species used o g ow. These ac o s, in e ac ed wi h he
sp ead o he exo ic c ay ish, had a signi ican impac on he popula ions
o U icula ia aus alis, U. gibba, and Po amoge on polygoni olius, which
need dys ophic wa e s poo in sal s and nu ien s (Ci ujano e al.,
2014). Mos o he o he subme ged species we e a ec ed as well.
Th ee species esis ed he c ay ish’s in asion somewha longe . They
pe sis ed in isola ed pe manen wa e s be o e inally disappea ing du -
ing he nex decade, when eu ophica ion and he b oad-scale educ ion
in pond hyd ope iods made he species’ pe manen habi a s empo a y.
Indeed, Po amoge on polygoni olius, and Spi odela poly hiza we e ound in
isola ed ponds and ca˜
nos up un il 2010 and 2013, espec i ely, and
Hyd ocha is mo sus- anae was obse ed in one pond in which i had
na u ally occu ed in p e ious decades.
The disappea ance o aqua ic plan species om Do˜
nana ep esen s
mo e han a local loss o biodi e si y; i has impac s a la ge scales,
a ec ing biodi e si y on he Ibe ian Peninsula and e en in Eu ope. The
pe sis ence o hese species has special impo ance, since mos a e pa
o pe iphe al popula ions, and hei disappea ance means ha he spe-
cies’ ex en o occu ence (EOO) is g ea ly educed (Joppa e al., 2016).
When EOO signi ican ly dec eases o e a sho pe iod o ime, i is a
ecognized indica o ha a species’ conse a ion s a us is de e io a ing
(Izco, 2015). Fo H. mo sus- anae, he closes e i ied popula ions a e
ound in F ance (Ga cía Mu illo e al., 2000; 2004); o U. gibba, i is
Mo occo. Alge ia and Tunisia, in no he n A ica (de B´
elai & Rhazi,
2010). In he case o U. aus alis, i is now no he n Spain (Laínz, 1979;
Mayo & Al a ez, 1978; Robinson e al., 2009), as he las ime he
species was seen in he sou he n hal o he coun y was in 1979, in a
s eam in he Guadamella o Valley in he p o ince o C´
o doba (A enas
e al., 1983); i is now conside ed o be ex inc in Andalusia (Cabezudo
e al., 2005). Fo N. lu eum, he nea es popula ions a e ound in he
p o ince o Badajoz in wes e n Spain (UNEX he ba ium). The si ua ion
is simila o N. alba: he nex closes popula ion is loca ed a ound 280
km o he no hwes , also in he p o ince o Badajoz (Bau is a e al.,
2012). The same is ue o S. poly hiza (specimens in HSS he ba ium),
P. polygoni olius (Ga cía Río, 2006), and P. c ispus (T´
ellez e al., 2008).
4.2. Recen dec ease in species anges
O he aqua ic ascula plan species s ill ound in he Do˜
nana pond
sys em, 78.9 % ha e expe ienced a decline o e en ex inc ion o hei
popula ions. This eali y is s a ling gi en ha Do˜
nana is one o he mos
impo an na ional pa ks in Eu ope and ha se e al o hese species a e
also p o ec ed by na ional o egional laws. The e ec s on aqua ic plan
communi ies may be ela ed o he gene al de e io a ion o aqua ic
habi a s ha has been occu ing in he egion o e he las ou decades.
The g ea es h ea o he conse a ion o Do˜
nana’s we lands is
an h opogenic p essu e on wa e esou ces: Do˜
nana ponds mus ha e
access o adequa e le els o g oundwa e o main ain hei inunda ion
and desicca ion cycles. Recen esea ch has desc ibed he de e io a ion
o he pa k’s pond ne wo k, cha ac e ized by a g adual educ ion in
pond hyd ope iod and su ace a ea, as well as he d ying up o a ound
60 % o ponds (de Felipe e al., 2023). The egion’s aqui e is being
o e exploi ed o i iga e ag icul u al ields a ound he pa k and supply
wa e o neighbo ing aca ion eso s and esidences. Local dep ession
cones a ound he pumping a ea a e causing declining wa e le els and
educing na u al discha ge om he aqui e o ce ain ponds (Cus odio
e al., 2009; Manzano e al., 2005; Se ano & Se ano, 1996). A simila
e ec is being caused by he inc eased empe a u es and lowe ain all
le els in ecen decades (de Felipe e al., 2023). Ou s udy also con i ms
ha aqua ic habi a s a e being los , gi en ha nea ly 20 % o he si es we
su eyed be ween he 2000 s and 2020 s ha e d ied up. These habi a s
include Medi e anean empo a y ponds, an EU p io i y habi a (code
3170), conside ed close o be a disappea ing ecosys em (Bagella, 2023;
Zacha ias and Zampa as, 2010). Addi ionally, la ge pe manen ponds
ha e been ans o med in o seasonal ponds. Unsu p isingly, he loss o
ponds is a ec ing aqua ic plan popula ions, as we ha e unde sco ed in
his s udy.
The g ea es loss o ascula aqua ic plan species has occu ed in he
pa o he pa k wi h he g ea es loss o ponds and la ges dec ease in
pond hyd ope iod (see Fig. 6 in de Felipe e al., 2023), namely he
no hwes e n pa k, which has been he mos a ec ed by aqui e o e -
exploi a ion. In ac , some species commonly ound h oughou he pa k
ha e been in equen in his a ea because he e a e no longe any ponds
o inhabi , and he species ha a e es ic ed o he no he n pa k o o
he cen al pe idune a ea a e hose whose anges ha e dec eased he
mos .
Some species ha ha e his o ically been a e in Do˜
nana s ill pe sis ,
e en i hei p esence is now e en mo e limi ed. Some imes, hei pop-
ula ions a e es ic ed o a ew isola ed loca ions, whe e only a small
numbe o indi iduals can de elop. Gi en hese ci cums ances, i seems
likely ha hey will e en ually disappea . This end is pa icula ly
ala ming o A. is ulosa, a monospeci ic genus whose las emaining
popula ions a e hose in Do˜
nana (O iz He e a e al., 2008). I inhabi s
we a eas co e ed by hyg o u bous g asslands in he eco one be ween
he sandy a ea and ma shland. The species has been se e ely a ec ed by
habi a loss (due o he lack o wa e ) and in ense he bi o y (Ga cia-
Mu illo & Fe n´
andez-Zamudio, 2015). Fo ce ain o he species, he
disappea ance o hei popula ions om Do˜
nana would conside ably
educe hei EOO. Such is he case o Po amoge on na ans (Ma ín-
Blanco & Ca asco, 2005; Melendo e al., 2003), P. lucens (Ma ín-Blanco
and Ca asco, 2005), and Lemna isulca (Ci ujano e al., 2006; Ga -
cía–Mu illo e al., 1990), o which hei nea es popula ions a e hun-
d eds o kilome es o he No h. Ou da a abou he p esence o hese
Po amoge on species do no s ill de ec a educ ion o hei ange in he
a ea, bu i is ala ming he ans o ma ion obse ed om dense co e o
ep oduc i e indi iduals in he 2000
′
s o he p esence o a sca ce numbe
o non- ep oduc i e indi iduals s ill pe sis ing a he same locali ies.
They a e subme ged pe ennial plan s ha s ill pe sis in some zacal-
lones, P. na ans om he no he n a ea, and P. lucens om he cen al
and sou he n a ea. Thei main h ea is he empo aliza ion o e en
P. Ga cia-Mu illo e al.
Jou nal o Na u e Conse a ion 84 (2025) 126814
9
