Content uploaded by Rúben M. Correia Rego
Author content
All content in this area was uploaded by Rúben M. Correia Rego on Aug 21, 2024
Content may be subject to copyright.
RESEARCH Open Access
© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Rego et al. BMC Ecology and Evolution (2024) 24:111
https://doi.org/10.1186/s12862-024-02300-8 BMC Ecology and Evolution
*Correspondence:
Rúben M. Correia Rego
ruben.mc.rego@uac.pt
1CIBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos,
Faculdade de Ciências e Tecnologias, InBIO Laboratório Associado,
Universidade dos Açores, R. Mãe de Deus 13A, Ponta Delgada
9500-321, Portugal
2BIOPOLIS Program in Genomics, Biodiversity and Land Planning, CIBIO de
Vairão, Vairão, Portugal
3Faculdade de Ciências e Tecnologias, UNESCO Chair – Land Within Sea:
Biodiversity & Sustainability in Atlantic Islands, Universidade dos Açores, R.
Mãe de Deus 13A, Ponta Delgada 9500-321, Portugal
4Banco Germoplasma & Dpto. Biología Reproductiva, Jardín Botánico
Canario “Viera y Clavijo” – u.a. CSIC, Las Palmas de Gran Canaria, Spain
Abstract
Background Anthropogenic threats are causing alteration of coastal areas worldwide. Most of the coastal
biodiversity is endangered, taking a particular toll on island ecosystems, like the Azores. To better understand the
biotic and abiotic factors constraining the distribution and conservation status of two endemic plants, Azorina vidalii
(Campanulaceae) and Lotus azoricus (Fabaceae), we performed a global survey of coastal plant communities in the
archipelago, also covering environmental descriptors, natural and anthropogenic threats. Moreover, we revised their
IUCN conservation status and estimated the population fractions within protected areas.
Results Non-indigenous plants were commonly found in plots with or without the target endemics, contributing
to the absence of well-dened coastal plant communities. Nonetheless, indigenous taxa commonly occurred at the
plots with L. azoricus. With a larger area of occurrence, A. vidalii ecological niche diered from that of L. azoricus, the
latter being restricted to dry and rocky sea clis, mostly in Santa Maria Island. Besides the presence of invasive plants,
signs of habitat destruction, trampling and grazing, and of natural threats, such as coastal erosion, were commonly
observed.
Conclusions Occurrence data indicated an endangered status for both species, although this would change to
critically endangered for L. azoricus when using smaller-sized occurrence cells. Both species are threatened since their
habitat is restricted to a very narrow vegetation belt, strongly limited by sea inuence and human pressure, and with
the frequent presence of invasive plants. While focusing on two endemic plants, our study allowed a broader view of
the impact of anthropogenic disturbance on Azorean coastal plant communities.
Keywords Anthropogenic disturbance, Coastal degradation, Invasive species, Conservation, Coastal plant
communities, Campanula Vidalii, Lotus Azoricus
Anthropogenic disturbance has altered the
habitat of two Azorean endemic coastal plants
Rúben M. CorreiaRego1,2,3*, MónicaMoura1,2,3, MariaOlangua-Corral4, GuilhermeRoxo1,2,3, RobertoResendes1,2,3 and
LuísSilva1,2,3
Page 2 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
Introduction
Coastal areas function as the connection zone between
land, sea, and atmosphere, and are subjected to high lev-
els of oscillation, making them particularly sensitive to
natural and anthropogenic disturbance [1].
In oceanic islands, intensification of traditional land
uses [2, 3] and the growing expansion of human activi-
ties into the coastal areas originate habitat loss, changes
in vegetation structure and the fragmentation of endemic
plant populations [4], leading to biodiversity loss, plant
extinctions [5, 6] and to decreases in abundance and
diversity [7]. Among these threats, the proliferation of
non-indigenous taxa [8], biological invasions [4, 7], and
climate change are paramount. e latter is expected
to increase the frequency of extreme weather events,
to cause sea level rise, and accelerate coastal erosion [9,
10] further altering coastal plant populations in oceanic
islands [1, 9]. Moreover, many populations of indigenous
taxa are often found outside the circumscription of pro-
tected areas, making their conservation more difficult,
due to lower land use restrictions and monitoring [11].
e Azores Archipelago integrates the Macaronesian
biodiversity hotspot, being characterised by a 7% rate
of endemic vascular plants [12], translating to approxi-
mately 97 taxa [13], but with more than 3000 introduced
plant taxa [14]. Most species have been introduced after
the Portuguese settlement, as described in historic litera-
ture [15]. Anthropogenic change included the installation
of agricultural crops, orchards, vineyards and hedgerows,
and the introduction of farm animals and other terres-
trial vertebrates that became established in the wild.
More recently, other threats gained relevance, such as
the expansion of pasturelands into the coast, the con-
struction of infrastructure, the occasional deposition
of solid waste at coastal areas, and the cultivation and
subsequent spread of alien species, becoming large-
scale invaders [15, 16]. Furthermore, marine erosion of
coastal areas is a well-known phenomenon in different
types of sand and boulder beaches in the Azores [17].
ese events narrowed the occurrence of endemic veg-
etation to relatively inaccessible areas, such as mountain
slopes, craters, or coastal cliffs [16, 18]. erefore, there
is an urgent need to preserve the natural heritage of the
Azores [19], that is, designing holistic recovery plans for
endangered plants, focussing on monitoring, ecological
modelling, habitat restoration and genetics [20].
Among Azorean threatened plant taxa, coastal endemic
withstands considerable levels of natural and anthro-
pogenic disturbance [19]. Recent projects addressed
reproductive and morphological traits of several Macaro-
nesian endemics (MacFlor: INTERREG MAC 2014–2020
MAC/4.6d/190; MacFlor 2: MAC2/4.6d/386). Project Life
Vidalia specifically aimed to improve the conservation
status of Azorina vidalii H.C.Watson (Campanulaceae),
and Lotus azoricus P.W.Ball (Fabaceae), through popula-
tion reinforcement, and habitat restoration, in the islands
of Faial, São Jorge and Pico (see https://www.lifevidalia.
eu/). Both endemic taxa have been considered as top pri-
orities for conservation in the Azores, being protected
by Azorean legislation (Decreto Legislativo Regional
n.o. 15/2012/A, Anexo II), by the Natura 2000 Network,
under the EU Habitats Directive (Council Directive
92/43/CEE of 21 May 1992), and by Berne Convention
[19]. Previous studies suggest that some populations may
be declining or even disappearing [21, 22].
According to literature [23–27], coastal plant commu-
nities include: (i) Coastal scrubland; (ii) Chamaephyte
plant communities from rocky coasts (e.g., rolled pebbled
beaches); (iii) Halophyte and halohydrophyte meadows;
(iv) Vegetation typical from sandy beaches or dunes;
(v) Coastal wetlands (e.g., halophyte reeds); (vi) Soggy
meadows, and coastal brackish water ponds. A global
quantitative assessment of the coastal plant communi-
ties is currently pertinent, given the emergence of several
anthropogenic threats, the conservation projects in place,
and the present network of protected areas. us, herein
we have focused on two different taxa, one with a broad
and another with a narrow occurrence area, also consid-
ering the factors potentially conditioning their ecological
niches.
Within this framework we formulated two sets of
research questions:
• e rst regarding the denition of their present
habitat - are there signicant dierences in terms
of: (i) plant community and (ii) environmental
descriptors between areas with or without the two
taxa?
• e second regarding habitat change and
conservation status: (i) Are coastal plant
communities still dominated by indigenous taxa?
(ii) Are there any relevant threats present at the
occurrence areas? (iii) Are the occurrences mainly
found within protected areas? and (iv) Has their
conservation status improved in recent years?
Based on the framework described above, as our start-
ing hypotheses, (i) we expect that non-indigenous plant
taxa presently correspond to a relevant component of
the herbaceous coastal plant communities; (ii) we expect
that the coastal habitat is under natural and anthropo-
genic threats; and (iii) we don’t expect a deterioration
of their conservation status, given the areas designated
for conservation and the restoration measures being
implemented.
We performed a thorough ecological survey of the
herbaceous coastal plant communities in the nine
Azores islands: (i) we included a comparison of the
Page 3 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
plant communities where the target taxa were present or
absent, considering the effect of anthropogenic threats
like invasive species and changes in the vegetation struc-
ture; (ii) we analysed environmental descriptors (climate,
altitude, substrate) to better define their habitats; (iii)
we identified potential threats; and (iv) we applied the
IUCN criteria to revise their conservation status, while
also evaluating the effectiveness of protected areas in
covering the respective populations. Although our main
targets were the two endemic taxa − Azorina vidalii and
Lotus azoricus−, we consider the global analysis of the
herbaceous coastal communities as a baseline require-
ment for a holistic understanding of their habitats and
conservation status.
Methods
Study site
e Azores archipelago (37º-40º N, 25–31W; Fig.1) is
situated in the north Atlantic Ocean, and it’s composed
of nine volcanic islands and several islets, divided into
three groups: Western (Flores and Corvo), Central (Ter-
ceira, Graciosa, São Jorge, Pico, and Faial) and Eastern
(São Miguel and Santa Maria). e archipelago is in a
warm temperate zone with high relative humidity, low
thermal amplitude, and rainfall throughout the year. e
average temperature at the coastal areas ranges between
14º and 17ºC [28].
Target species
Azorina vidalii is a synonym of Campanula vidalii
(H.C.Watson) Feer. Its taxonomy is currently under revi-
sion, due to clustering within other Campanula species
in a published phylogeny [29], is a glabrous chamaephyte
common below 50m a.s.l., found in all islands [12], rarely
occurring at altitudes above 100m [19]. e phytosocio-
logical alliance Euphorbio azoricae-Festucion petraeae
[27], included the association Azorinetum vidalii [23],
characteristic of sea cliff communities. However, this spe-
cies displays a very diverse ecology, being associated with
different coastal plant communities, on different islands,
from common rocky chamaephyte communities to halo-
hydrophyte meadows, typical of Corvo and Terceira
islands [24].
e endemic legume, Lotus azoricus, is a semi-herba-
ceous hemicryptophyte, whose distribution has been
referred to the islands of Santa Maria, São Miguel, Pico,
São Jorge and Flores [12], however, its presence in some
islands (e.g. São Miguel and Flores) lacks confirmation. It
grows on rocky shores, coastal cliffs and lava flows from
5 up to 95m a.s.l., on sand or rubble soils from incipi-
ent to a few centimetres thick [19, 21]. Lotus azoricus
has been associated with the phytosociological alliance
of Tolpido succulentae-Agrostion congestiorae [25].
Although L. azoricus has been considered as a synonym
of L. argyrodes R.P.Murray [30], a divergence time of 2.5
Fig. 1 Geographical location of the Azores archipelago, the islands composing each subarchipelago of the Azores and target species: (a) Azorina vidalii;
(b) Lotus azoricus
Page 4 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
Mya between the two species [31] has been estimated,
and available DNA sequences show differences at several
positions. erefore, as is often the case with plant taxa
isolated in different archipelagos [32], we consider the
two taxa as separate evolutionary, ecological and taxo-
nomic units.
Distribution of sampling stations
Using Quantum GIS 3.28.2 [33], the coastal area of each
island was divided using a grid of 500 × 500m, allowing to
overview which areas would be accessible for sampling,
as well as to make sampling distributed homogeneously
across the coast. All locations were georeferenced using a
GPS device (Garmin Montana, 680). In total, we selected
148 sampling stations (500 × 500m cells) across the nine
islands of the archipelago that were accessible by walk-
ing trails. ese sampling stations were selected before
the start of field work, independently of the distribu-
tion of the two target plants, and previously to any field
observation.
Plant community sampling
Field work was carried out from June to November of
2022. Following previous work [34], we used 5 × 5m plots
to prospect and describe coastal plant communities. We
used an average of three plots at each sampling station,
with a minimum of one plot at most sampling stations
without the target species, and a maximum of eleven
plots at a sampling station in Santa Maria Island with a
relatively large extension and elevation span.
In total 231 5 × 5m plots were sampled, including plots
with and without the target plants. erefore, four plot
types were defined: (i) A – plant communities with Azo-
rina vidalli only; (ii) L – plant communities with Lotus
azoricus only; (iii) B – plant communities including both
target species; and C – plant communities where both
species were absent.
In many locations, the area available for the target her-
baceous coastal plant communities was constrained by
the sea level, below, and by dense scrubland with indig-
enous or/and non-indigenous taxa, or humanised areas
(housing, crops, pastures), above. We focused our sam-
pling effort on that intermediate belt. In some cases,
like coastal cliffs with smoother slopes, the herbaceous
coastal vegetation extended to relatively high elevations
(100m).
At each 5 × 5m plot we recorded the percent cover for
each vascular plant taxa, which was visually estimated by
a vertical projection of above ground plant parts at each
of four equal sized subplots. Plant species were identi-
fied in situ or sampled and later identified in the labora-
tory with the help of field guides and floras [18, 35–37]
(Plant material identification undertaken by Luís Silva,
Guilherme Roxo and Rúben M. C. Rego). e sampled
voucher specimens were preserved in collection at the
AZB herbarium (Voucher ID’s AZB 4311 to AZB 4381;
Azores University, Ponta Delgada, Portugal). In total, 197
taxa were recorded in the 231 plots made (Supplemen-
tary Table S1).
Colonisation status
To evaluate the level of anthropogenic alteration in the
composition of the plant communities, we categorised
the plant taxa according to their colonisation status, fol-
lowing species lists [12, 38] and local legislation (DLR n.º
15/2012/A, 2 de Abril), into (see Supplementary Table
S2): indigenous (i.e., taxa that arrived or evolved on the
islands in the absence of human intervention), with two
subgroups – “native” and “endemic”; (ii) non-indigenous
(i.e., taxa that were intentionally or accidentally intro-
duced as the result of human activities), with two sub-
groups – “naturalised”, and “invasive”.
Life-forms
To evaluate possible differences in vegetation structure
between plots with or without the target species, we used
the Raunkiaer [39] life-form system, revised by Braun-
Blanquet [40], to categorise plant taxa, based on the posi-
tion of the resting buds, considering the following types
(see Supplementary Table S2): therophyte, hemicrypto-
phyte, chamaephyte, geophyte, and phanerophyte.
Habitat ecology
We used published floras [35, 36, 41] to categorise spe-
cies according to the ecology of their habitats, into: xero-
phyte, halophyte, mesophyte, hygrophyte and generalist
(see Supplementary Table S2).
Environmental descriptors evaluated in situ
We used the following environmental descriptors to
describe the habitat of the target species: elevation, slope,
exposure, type of substrate and threats to the habitat.
Exposure and elevation were recorded using a portable
GPS (Garmin Montana, 680).
Substrate
e following substrates were considered and defined,
following literature [42] (see Supplementary Table S2):
sand, clay, lapilli, lava flow, boulders, rolled pebbles, soil
and rocky soil.
Soil parameters
Soil was collected whenever it was present, since in most
plots we only found a rocky substrate, without a soil layer.
e 50 samples were sent to the Soil and Plant Labora-
tory of the University of Trás-os-Montes e Alto Douro
(Vila Real, Portugal) for the following analyses: pH (H2O
and KCl), organic matter (OM), extractable content
Page 5 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
of the main nutrients (phosphorus and potassium, by
the Eletrolite Replacement and Olsen method), cations
(Ca++, Mg++, K+, Na+, Al3++), effective cation exchange
capacity (CEC), electrical conductivity (1:5 soil to water
ratio), total nitrogen and texture (granulometry).
Identied threats
To define the most common potential threats to the
survival of the target species, we recorded signs of the
effect of natural or anthropogenic threats in situ. Natural
threats included (see Supplementary Table S2): (i) coastal
erosion; (ii) direct sea submersion; and (iii) storms.
Anthropogenic threats included: (i) agriculture; (ii) ani-
mal husbandry; (iii) construction work; (iv) waste dis-
posal; (v) habitat destruction; (vi) human infrastructure;
(vii) trampling; and (viii) invasive species.
Climatic data
We used climatic data to determine if there were relevant
variations or differences between the different plot types,
regarding the main bioclimatic descriptors. Climate data
was obtained through the CHELSA (Climatologies at
high resolution for the earth’s land surface) V.2.1 data-
base [43]. We used 19 bioclimatic variables (11 related
with temperature and 8 with precipitation), consisting of
data retrieved between 1981 and 2010. A full list of vari-
ables is provided in Supplementary Table S3.
Statistical treatment
Plant diversity
To compare diversity levels between plot types, we calcu-
lated species richness (number of taxa per plot), Shannon
entropy (total diversity per plot), maximum entropy [44]
(maximum theoretical diversity) and evenness [45].
Quantitative variables
For quantitative variables, i.e., diversity indexes, quantita-
tive environmental descriptors, comparison of plot types
was performed using boxplots and the Kruskal-Wallis
test, followed by non-parametric multiple comparison
tests (R “pgirmess” package [46]). We opted for a con-
servative approach, using non-parametric tests, since
we could not ensure normality and homoscedasticity,
required for parametric tests.
Categorical variables
For categorical variables, i.e., frequency of threats and
substrate, we used Pearson’s chi-square test followed by a
test for comparison of proportions, included in the pack-
age “gmodels” [47], and bar charts for graphical repre-
sentation. For soil texture, due to the small sample size,
we used Fisher’s exact test and the chi-square test with
the option of bootstrap, to confirm possible significant
differences.
Statistical applications and output
All the statistical analyses were performed using Rx64
4.2.3 [48]. Overall results from the Kruskal-Wallis and
Pearson’s chi-square tests are given in Supplementary
Tables S4, S5, S6, S7, S8, S10 and S12, and the significant
differences between plot types are indicated using differ-
ent letters in the respective figures. We should note that
the Kruskal-Wallis test statistic, H, is provided as a chi-
squared approximation in the R output, as is common in
many statistical applications.
Species frequency, cover, and importance
In order to determine the plant taxa that were physi-
ognomically dominant, we calculated and plotted the
frequency, abundance, and importance of each taxon
as follows [49]: (i) frequency as the percentage of plots
with the taxon; (ii) abundance (based on percent cover)
as the total abundance of the taxon, divided by the total
abundance of all taxa; and (iii) importance as frequency
(%) + abundance (%) divided by two.
Clustering and ordination
To detect possible differences between plot types, we
used hierarchical cluster analysis with the “vegan” pack-
age [50], based on species cover. Several combinations of
distance and agglomeration methods [51] were consid-
ered and ultimately, we concluded that Bray-Curtis dis-
similarity combined with Unweighted Pair Group Mean
Average (UPGMA) provided the highest cophenetic cor-
relation coefficient. e optimal number of plot groups in
the dendrogram was determined using two algorithms: (i)
according to silhouette widths (Rousseeuw quality index)
[52], and (ii) according to the Mantel statistic (Pearson)
[53]. We also represented the Bray-Curtis dissimilar-
ity matrix using Non-Metric Multidimensional Scaling
(NMDS), as commonly used in numerical ecology.
Indicator species analysis
To detect possible differences between plot types, we
used indicator species analysis with the R package “Indic-
species” [54], consisting in an improvement of the Ind-
Val method initially established by Dufréne and Legendre
[55]. As an abundance metric, species percent cover was
used. e algorithm calculates fidelity (limitation to a sin-
gle site or set of sites) and consistency (consistent species
occurrence among sites within site groups) and returns a
statistic (IndVal) and the corresponding p-value.
Binary logistic regression
Following previous work [56], we used the “glm()” func-
tion in R to calculate binomial generalised linear models,
to determine which factors (i.e., climate, substrate, per-
cent of endemic, native, naturalised and invasive taxa, or
Page 6 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
a mix of several factors) potentially affect the occurrence
of the target species in the coastal habitats.
For both species, a null model, not including explana-
tory variables, but only an intercept or model constant,
was considered as a benchmark, for comparison with
the explanatory models described below. Models under-
performing when compared with the null model were
excluded.
We tested several models, including different com-
binations of the explanatory variables: (i) a bioclimatic
model including the principal components extracted
from a principal component analysis, applied to the 19
bioclimatic variables; (ii) a model assessing the effect of
substrate types; (iii) a model regarding the contribution
of plants with different colonisation status, and diversity
measures; (iv) a saturated model including the effect of
all previous factors; and (v) several models resulting from
the simplification of the saturated model, converging to a
simplified final model.
e maximum likelihood approach was used for model
selection and simplification, based on Akaike’s Infor-
mation Criterion (AIC) - the lower the better. All mod-
els were compared with a null model only including an
intercept. Given the large number of samples (more than
200 observations) and the fact that the possible outcome
is binary (the species is present or absent), we consider
the application of binomial GLMs as appropriate, follow-
ing previous work [56]. We also performed a final model
selection, by keeping only those variables that would
exhibit significant regression coefficients.
Populations included in Natura 2000 or Island Natural
Parks
To access whether the populations of the target species
were located within the areas covered by the Natura
2000 network or by Island Natural Parks in the Azores,
the georeferenced populations of A. vidalii and L. azori-
cus were mapped in QGiS [33] and intersected with the
shapefiles representing the protected areas (Source:
Azorean Government).
IUCN Red List assessment of the target species
We evaluated the conservation status of the target species
following the guidelines of the IUCN Red List, v.15.1 [57].
We performed calculations of the extent of occurrence
(EOO) and of the area of occupancy (AOO) using Geo-
CAT [58]. We calculated AOO using 2 × 2km grid cells
(area of 4km [2]) and based the estimates of the number
of mature individuals on counts made during field work.
Results
Plant community sampling
A total of 197 taxa were recorded, with an average of 9.3
taxa per plot. e highest relative cover and frequency
was obtained for the endemic Festuca petraea (C: 14.0%,
F: 119 occurrences, 51.5%). Azorina vidalii (C: 6.5%, F:
99 occurrences, 42.9%) was clearly more abundant and
frequent than Lotus azoricus (C: 1.4%, F: 17 occurrences,
7.4%). Some native (e.g., Crithmum maritimum) and sev-
eral invasive taxa appeared frequently in the sampled
plots (e.g., Carpobrotus edulis, Tetragonia tetragonoides,
Paspalum dilatatum, Arundo donax, Cynodon dactylon),
which contributed with high percentages of cover and
frequency (Supplementary Figures S1 and S2).
Colonisation status
e 197 records corresponded to 108 naturalised, 40
invasive, 27 native and 22 endemic taxa. For endemic
plant cover, significant differences (p < 0.005) were found
between plot types A, B and L towards C plots, with
lower values for the latter. Endemic plant frequency
also showed significant differences between plot types
A towards B and C, and between plots L and B, towards
plots C, with the lowest values for the latter (Fig.2; see
Supplementary Table S4). No significant differences
between plot types were found for the remaining coloni-
sation status, in cover or frequency, with median values
around 20% for native and naturalised taxa, and below
20% for invasive taxa.
Life-forms
We detected significant differences between plots A and
plot types L and C for the cover of chamaephytes, with
higher values for A and B (Fig.2). Similarly, we found sig-
nificant differences in chamaephyte frequency, between
plots A and plots L and C, and between plots B and L,
with higher values for A and B (Fig. 2, Supplementary
Table S5). No significant differences between plot types
were found for the cover and frequency of the remaining
life forms, with median values ranging 1% for geophytes,
below 20% for phanerophytes, and ranging from 20 to
50% for therophytes and hemicryptophytes (data not
shown).
Habitat ecology
Across the ecology types considered, only halophyte,
generalist and mesophyte species were relevant, as the
cover and frequency of hygrophytes and xerophytes was
always below 5% and sometimes null, among plot types.
e cover and frequency of mesophytes was around 20%,
among plot types. e cover of halophyte species ranged
between 70% in L plots and 90% in A and B plots, while
the frequency was around 60–70%, among plot types.
Generalist species dominated the cover and frequency in
C plots (> 40%), appearing also with high cover and fre-
quency in A plots (40%), but much lower in plots with L.
azoricus (around 25%). No significant differences were
detected for the cover and frequency of the ecology
Page 7 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
types, among plot types, after the Kruskal-Wallis analysis
(Supplementary Figure S3; see Supplementary Table S6).
Plant community clustering
e highest cophenetic correlation was obtained for
Bray-Curtis’s dissimilarity Index and UPGMA (0.707).
e best value for the number of plot clusters was 32. We
found significant differences for Bray-Curtis dissimilar-
ity (Supplementary Table S7) between plot types A and
L and C plots, the latter showing the highest values (Sup-
plementary Figure S4).
e results of the NMDS plot (Fig.3) showed no clus-
tering of the four plot types. Type C plots were mostly
found on the periphery of the plot, while most A plots
were concentrated at the middle, and L plots were mostly
scattered at the top.
Fig. 2 Cover and frequency results for top) endemic plant taxa and bottom) chamaephyte taxa found in 231 plots sampled in coastal areas in the nine
Azores islands. Plot types: A – including Azorina vidalii, L – including Lotus azoricus, B – including both species, C – controls without both species. Dierent
letters indicate signicant dierences (p < 0.05); Results of a non-parametric multiple comparison test applied after Kruskal-Wallis test
Page 8 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
Plant diversity
We detected significant differences between plot types A
and C for species richness and Shannon diversity (Sup-
plementary Table S7), with higher values for the former
(Fig.4), but no significant differences between plot types
for evenness which was concentrated around 0.8 (data
not shown).
Indicator species
Considering the four types of plots, we found 14 taxa
with significant indicator value (p < 0.05), six taxa being
associated with one plot type, 7 taxa associated with two
plot types, and only one taxon associated with three plot
types (Table1).
Environmental variables
Elevation and exposed soil
Significant differences were found between plot types for
elevation, with the highest median value for plot type L
(Supplementary Figure S5; Supplementary Table S8). No
significant differences were found between plot types, for
Fig. 3 Results of a Non-Metric Multidimensional Scaling (NMDS) applied to the Bray-Curtis dissimilarity matrix, based on 231 plots sampled in coastal
areas in the nine Azores islands. Dierent colours represent the four plot types
Page 9 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
Fig. 4 Boxplots of Species richness and Shannon diversity, based on 231 plots sampled in coastal areas in the nine Azores islands. Plot types: A – including
Azorina vidalii, L – including Lotus azoricus, B – including both species, C – controls without both species. Dierent letters indicate signicant dierences
(p < 0.05); Results of a multiple comparison test applied after the Kruskal-Wallis test
Page 10 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
the proportion of exposed soil, which only had relevance
in plots B (≤ 20%, data not shown).
Type of substrate
Regarding substrate type, six out of nine substrates had
a general frequency of at least 10% (lapilli, lava flow,
boulders, rolled pebbles, sand and soil), but only soil
and sand retrieved significant differences in the multiple
comparison test, between plot types A and C (Fig.5; Sup-
plementary Table S8), with a higher frequency of sand
in type C plots and a higher frequency of soil in A type
plots.
No significant differences were found between plot
types, for the remaining substrates, with overall frequen-
cies of 4.8% for clay substrate, 9.5% for rocky soil, 10.8%
for rolled pebbles, 22.1% for sand, 24.7% for lapilli, 42.8%
for lava flow, 48.0% for soil, and the highest frequency,
which was of 76.6%, for boulders.
Soil parameters
Most soil parameters did not show any significant dif-
ferences between plot types (Supplementary Table S8).
Significant differences were only observed between plots
B and C, for the extractable content of phosphorus, with
higher levels in the former (Supplementary Figure S6).
Although the Kruskal-Wallis test indicated significant
differences between the plot types for electric conduc-
tivity, the subsequent multiple comparison test failed to
confirm those differences.
No significant differences were found between plot
types for field texture (Supplementary Table S8). Sand,
sandy clay loam and silt loam soil textures obtained over-
all frequencies below 5%, the frequency of clay loam was
8%, loamy sand obtained 18%, loam had a frequency of
Table 1 Indicator species associated with four plot types, from
197 taxa retrieved from 231 coastal plots
Plot type Taxa Indicator value p-value
LLolium rigidum 0.587 0.005
Agave americana 0.485 0.020
BSpergularia azorica 0.788 0.005
Calendula suruticosa 0.707 0.005
Gaudinia coarctata 0.626 0.005
Calluna vulgaris 0.408 0.020
A + B Azorina vidalii 1.000 0.005
L + B Lotus azoricus 1.000 0.005
Euphorbia azorica 0.631 0.035
Reichardia picroides 0.608 0.005
Limonium diasii 0.553 0.015
Lysimachia arvensis 0.473 0.040
Plantago lanceolata 0.427 0.045
A + L + B Sonchus tenerrimus 0.710 0.010
Plot typ e abbreviations are as f ollows: A: Azorina vidalii only; L: Lo tus azoricus only;
B: both taxa; C : control (without A. vidalii or L. a zoricus)
Fig. 5 Frequency of sand and soil substrates, based on 231 plots sampled in coastal areas in the nine Azores islands. Plot types: A – including Azorina
vidalii, L – including Lotus azoricus, B – including both species, C – controls without both species. Dierent letters indicate signicant dierences (p < 0.05);
Results of a multiple comparison test applied after chi-square test
Page 11 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
20%, and the highest percentage was observed for sandy
loam soils with 46% (Supplementary Figure S7).
Climate
e 19 bioclimatic variables were reduced to three prin-
cipal components, explaining 94% of the variance (Sup-
plementary Figure S9). e first principal component
was strongly supported by precipitation variables, while
the second principal component showed to be positively
influenced by precipitation, and negatively influenced by
temperature variables. Finally, the most important vari-
ables for the third principal component were related with
the mean temperatures in the different quarters of the
year, whilst having some mixed influence of both temper-
ature and precipitation variables (Supplementary Table
S9). We only found significant differences among plot
types for the first two components (Supplementary Table
S10), indicating larger precipitation for A type plots, and
lower precipitation for plot types L and B.
Binary logistic regression
In total, 24 models with different explanatory variable
combinations were tested, separately, for the two target
species (further details on the variables selected for each
model are available in Supplementary Table S11).
Based on our best simplified model, with an AICc
much lower than the null model, the presence of A.
vidalii was positively affected by higher species rich-
ness and the presence of soil, coupled with higher levels
of endemic and naturalised plant cover (Table 2). e
occurrence of this species also appears strongly corre-
lated with higher precipitation values, as defined by the
first principal component of the climate, and higher tem-
perature and thermal amplitude, as defined by the second
principal component.
Based on our best simplified model, with an AICc
much lower than the null model, higher values for the
frequency of endemic taxa and Shannon diversity appear
to be beneficial for the occurrence of Lotus azoricus
(Table2). However, it appears negatively affected by the
first principal component of the climate, associated with
higher precipitation values.
Threats
Nine of the 11 threats considered in this study displayed
an overall frequency of at least 10%, including three of
natural origin (sea exposure, storms, and coastal ero-
sion), and six anthropogenic (invasive species, trash dis-
posal, human presence, habitat destruction, trampling,
and animal husbandry). Figure6 shows that significant
differences were found for direct sea submersion (overall
frequency: 40.7%), between plot types L and B, and plots
of type C, with a higher frequency in the latter; as well as
for animal husbandry (overall frequency: 11.2%), between
plots L and C (Supplementary Table S12).
No significant differences were detected, among plot
types, for the remaining threats (please see Supplemen-
tary Figure S9), obtaining the following overall frequen-
cies: 1.7% for construction work, 9.5% for agriculture,
18.2% for trampling and trails, 38.1% for habitat destruc-
tion, 53.7% for coastal erosion, 54.5% for human infra-
structure, 58.4% for waste disposals, 71.0% for storms
and 82.7% for invasive alien plants.
Intersection with protected areas
For A. vidalii, most plots were located within Protected
Areas for Management of Species and Habitats (PAMSH)
and Special Conservation Zones (SCZ) (Supplementary
Table S13). Most of the INP and Natura 2000 network
areas overlapped, except for one site at Pico, which was
only covered by an SCZ. e islands of Santa Maria, São
Jorge and Flores had the highest percentages of plots
within any of the protected frameworks, while in Corvo
or Graciosa there was none. Globally, 41.41% of the plots
for this species were found within protected areas.
Table 2 A Summary of the simplied binary logistic regression models obtained for Azorina Vidalii and Lotus Azoricus. For the
complete set of calculated models see Supplementary Table S10. Explanatory variables dened in the text. Regression coecients and
the respective standard error and signicance. AICc for the models and the respective standardised pR2 value are also given
Taxon Explanatory
variables
Regression coecients Std. Error Sig. Model AICc pR2Null model AICc
Azorina vidalii PCA1 climate 1.839 0.468 *** 256.95 0.222 317.50
PCA2 climate 1.690 0.645 **
Species richness 0.182 0.047 ***
Endemic cover 0.035 0.009 ***
Naturalised cover 0.020 0.010 *
Soil substrate 0.789 0.320 *
Lotus azoricus PCA1 climate -3.907 1.038 *** 81.55 0.396 132.48
Shannon Index 2.374 0.854 **
Endemic frequency 0.105 0.026 ***
Signican ce levels are as follows: <0.001 ‘***’ 0.001 ‘**’ 0.01 ‘*’
Page 12 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
For L. azoricus, the highest percentages of plots were
also found in PAMSH and SCZ (Supplementary Table
S13). All the plots containing this taxon were under
protected land in Santa Maria and São Jorge islands,
although within different typologies, while only 40% of
the plots were within protected areas in Pico. Overall,
82.35% of the plots for this species were found within
protected areas.
Although some of the larger islands, such as São
Miguel or Pico, contained the greatest amounts of area
of occupancy for the two target taxa, it appears that the
percentage of populations from these populations under
protected areas is lower than in some of the smaller
islands (e.g., Santa Maria, Faial, or Flores).
IUCN evaluation
Table3 shows the conservation assessments made for the
two target species, which resulted in the attribution of
the Endangered category to both species.
According to the data available (Supplementary Tables
S14 and S15), the evaluation for both species followed
conditions B2 and a) since we have observed that popula-
tions appear severely fragmented, and b), (iii), due to the
observed decrease in the area and quality of the habitat.
For Lotus azoricus, criteria C2 was also selected, since the
populations are small and decline in population size has
been observed at some locations, by following a) and the
conditions (i) and (ii). However, based on the number of
mature individuals, which was 11250, the conservation
status of Azorina vidalii would be Least Concern, while
Lotus azoricus would still be in the Endangered category,
with only ca. 1520 mature individuals (see Supplemen-
tary Table S15). We have also calculated the area of occu-
pancy using 1 × 1km cells, instead of the IUCN standard
of 2 × 2km, which resulted in the attribution of the Criti-
cally Endangered (CR) category to L. azoricus.
Discussion
Sampling constraints
We have sampled in a systematic way along the coast,
wherever access was possible by walking along coastal
trails, including habitats such as: coastal cliffs, sand
beaches, dunes, and rolled pebbled beaches, streams, lava
flows and volcanic gravel.
e coast includes a narrow fringe of coastal vegeta-
tion, as described by Tutin [26]. Functioning as an eco-
tone between marine and terrestrial ecosystems, the
zonation from sea level to the upper vegetation belts is
determined by the exposure to salinity. Starting from
sea level, upwards, we would first find the typical coastal
(halophyte) species, such as Spergularia azorica, Crith-
mum maritimum and Euphorbia azorica, progress-
ing to Festuca meadows and, above, to Erica-Morella
scrublands or Picconia-Morella lowland forest [59]. We
focused our sampling effort in the first herbaceous veg-
etation belts, where the target species generally appear.
However, the presence of inaccessible vertical coastal
cliffs, extensive humanised areas (housing, agricultural
land, pastures), the occurrence of dense stands of inva-
sive species (e.g., Arundo donax, Pittosporum undu-
latum) limited the sampling area available for coastal
herbaceous plants, such as our target species.
Fig. 6 Frequency of animal husbandry and direct sea submersion found in 231 plots made in coastal areas on nine islands of the Azores. Plot types: A
– including Azorina vidalii, L – including Lotus azoricus, B – including both species, C – controls without both species. Dierent letters indicate signicant
dierences (p < 0.05); Results of the multiple comparison test applied after Chi-square test
Page 13 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
Nonetheless, we are confident that our sampling
scheme was unbiased. We visited all the sites where the
target species were previously reported (including recent
surveys by LIFE projects), but still, we could not confirm
previous records of L. azoricus for São Miguel and Flores
islands [12].
However, we found new occurrences, such as small
populations of L. azoricus in Santa Maria, and small
restored populations in Pico. New locations were also
found or rediscovered for A. vidalii. Since its occurrence
area is much larger than that of L. azoricus, the total
sampled areas for both species differ, but only due to the
rarity of L. azoricus, not to sampling bias. As planned, we
managed to sample plots with or without these two taxa,
a clear example being the huge number of samples per-
formed in Graciosa, where only one plot with A. vidalii
was found. Nonetheless, it is possible that we have not
found all the target taxa populations, since we aimed to
sample and not to completely survey the coastline.
Missing areas would correspond mainly to vertical
cliffs covered by native or invasive coastal scrubland,
with large boulder beaches at sea level. Further sampling
was out of our project capabilities, due to logistic, fund-
ing and time constraints, and could only be undertaken
Table 3 Conservation status assessment of Azorina Vidalii and Lotus Azoricus, following IUCN guidelines
Criteria Sub-criteria Condition Azorina vidalii Lotus
azoricus
Conservation category
CR EN VU
A Pop. size
reduction
A1 Population reduction (measured over the longer
of 10 years or 3 generations) based on any of A1
to A4
No data from the previous 10
years or 3 generations, to either
observe, estimate, infer or suspect
a decrease in population sizes.
≥ 90% ≥ 70% ≥ 50%
A2, A3, A4 ≥ 80% ≥ 50% ≥ 30%
B Geo-
graphic
range
B1 (Based on
EOO)
- 43816.38 km2 (LC) 5487.01 km2
(VU)
< 100
km2< 5000
km2< 20,000
km2
B2 (Based on
AOO)
- 276 km2 (EN); 78
km2 (EN)*
32 km2 (EN);
9 km2 (CR)*
< 10 km2< 500
km2< 2000
km2
AND at least 2 of the following 3 conditions
(a) Severely fragmented or number of populations (a) (a) 1 ≤ 5 ≤ 10
(b) (iii) Observed decline in: area, extension and/or
quality of the habitats
(b) (iii) (b) (iii)
(c) Extreme uctuations
C Small
population
size and
decline
Number of ma-
ture individuals
- 11,250 (LC) 1520 (EN) < 250 < 2500 < 10,000
AND at least one of C1 or C2
C1 An observed, estimated or projected continuing
decline of at least (up to a max. of 100 years in
future)
25% in
3 yrs./1
gen.
20% in
5 yrs./2
gen.
10% in
10 yrs./3
gen.
C2 An observed, estimated, project-
ed or inferred continuing decline
AND at least 1 of the following 3
conditions:
(a) (i) Number
of mat. ind.
in each
subpopulation
(a) (i) (VU) (a) (i) (VU) ≤ 50 ≤ 250 ≤ 1000
(a) (ii) % of mat.
ind. in one
subpopula-
tion =
(a) (ii) (CR) 90–100% 95–
100%
100%
(b) Extreme
uctuations in
the number of
mat. ind.
D Very
small or
restricted
population
D1 Number of mature individuals 11,250 (LC) 1520 (LC) < 50 < 250 < 1000
D2 Only applies to the VU category. Restricted area of occupancy or
number of locations with a plausible future threat that could drive
the taxon to CR or EX in a very short time.
276 km2 (LC); 78
km2 (LC)*; Nº loc.
>10
32 km2 (LC);
9 km2 (VU)*;
Nº loc. 8
(LC)
AOO < 20
km2/nº
loc. ≤5
E Quantitative Analysis - No data available ≥ 50%
in 10
years/3
gen.
≥ 20%
in 20
years/5
gen.
≥ 10%
in 100
years
Conservation status and
codes
- EN B2ab(iii) EN B2ab(iii); C2a(ii) CR B2ab(iii)*
*For both ta xa, we have also calculated th e area of occupancy (AOO) usi ng 1 × 1km (1km [2]) cells, in a ddition to the IUCN stand ardised 2 × 2km cells
Page 14 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
if regional entities would support boat access to coastal
areas inaccessible by land.
It can be argued that the herbaceous vegetation at
coastal areas is mainly dominated by generalist taxa,
and that L. azoricus and A. vidalli are not truly coastal
plants. We used a pragmatic approach that considered
all herbaceous species present in the coast, above sea
level, and below coastal scrubland (either natural or
partially anthropogenic) [59] or anthromes - although
this strip could be wider or narrower depending on site
conditions (e.g., contour, human activities surround-
ing the area). However, our target species showed to be
clearly restricted to this coastal vegetation belt, on highly
exposed sea cliffs or rocky substrates, with incipient
nutrients [60] and often together with F. petraea and C.
maritimum [25, 26, 61], even when found at higher loca-
tions (100m) such as at sea cliffs with gentler slopes (e.g.,
Corvo island).
Recent changes in the coastal ora
Festuca petraea was the most frequent coastal taxon in
our study, forming coastal meadows [59], but also found
in dunes and chamaephyte rocky shore communities [24].
e present taxonomic circumscription differentiates it
from F. francoi Fern.Prieto, C.Aguiar, E.Días & M.I.Gut,
an endemic taxon found at higher elevations [32].
Overall, naturalised and invasive taxa, when consid-
ered together, were clearly the most frequent across all
plot types, as previously reported [16]. is is linked to
the occurrence of many generalist taxa adapted to natural
or anthropogenic disturbance (e.g., Portulaca oleracea,
Sonchus spp.), exhibiting subcosmopolitan distributions
[16]. Although ocean-related disturbance leads to harsh
coastal environments where non-indigenous taxa could
have more difficulties to establish and thrive [62], our
results show a tendency for an expansion of these taxa.
is was also confirmed through the indicator species
analysis, where alien taxa showed a relevant indicator
value.
e occurrence of a high number of generalist, non-
indigenous species on the coastal habitats, is leading to
alterations in plant composition, cover, species richness,
diversity, and evenness, due to disassembly processes
[63], explaining the lack of support for the existence of
well-defined coastal plant communities in the cluster
analysis. While indigenous species still thrive in coastal
plant communities of halophytic or lithophytic character
[61], these are becoming scarce, given the expansion of
non-indigenous taxa, with many invasive species (e.g.,
A. donax, C. edulis, A. americana, P. undulatum) [38]
becoming dominant and invading large areas near the
coast, where only a few endemic plant taxa survive, often
at marginal habitats.
Life forms and ecological adaptations
Many coastal endemic taxa (e.g., Azorina vidalii, Euphor-
bia azorica, Limonium diasii, Spergularia azorica, Tol-
pis spp.) were often observed [24], contributing to the
considerable frequence of chamaephytes. e presence
of endemic chamaephytes in island floras could be the
result of secondary woodiness, prompting island taxa to
longer life cycles and sturdy woody habits, in detriment
of herbaceous habits [64]. However, hemicryptophytes
and therophytes dominated the studied areas, due to the
frequent presence of invasive, generalist taxa, as seen
elsewhere [16].
As expected, our results showed that halophytes are
still relevant in Azorean coastal plant communities [24,
61]. However, we also found many generalist taxa in con-
trol plots, but also in communities with A. vidalii. is
resulted from their wider geographical and ecological
niche, in their ability to thrive in different types of habi-
tats, which could be exacerbated by climate change [65].
Azorina vidalii
We found that A. vidalii still commonly occurs in the
Azorean coasts, in varied conditions, particularly at the
base of sea cliffs [24], although not being restricted to the
Euphorbio azoricae-Festucion petraeae alliance [27], sup-
porting a broad ecology [24]. At sea level it is frequently
found at Festuca meadows and other halophytic chamae-
phyte coastal communities from cliffs and rolled pebbled
beaches [24, 26]. Less frequently, it was found at the mar-
gins of coastal scrubland or lowland juniper stands in
Pico island [61], or more rarely, at higher elevation and
inland areas, with reduced salinity and lower tempera-
tures (e.g., Corvo and Faial islands) [24, 66]. Nonethe-
less, at A. vidalii plots we confirmed a high prevalence
of non-indigenous taxa [16, 19] since it was often found
at the margins of disturbed habitats (e.g., dense stands of
Arundo donax or of other invasive species). Finally, we
found some distinction between plots with A. vidalli and
without both target plants, namely, a larger heterogene-
ity of species composition in the latter, as evident in the
numerical ecology analyses (NMDS and Bray-Curtis).
Lotus azoricus
We confirmed the rarity of L. azoricus in the Azorean
coasts, being restricted to highly exposed plant commu-
nities in difficult access areas (e.g., high elevation coastal
cliffs in Santa Maria island), where it likely escapes the
intense human disturbance [19, 21] found at flatter areas.
It also occurs in volcanic substrates, highlighting the role
of endemic taxa in the primary succession [67]. It was,
in some cases, found in well-preserved halo-xerophytic
communities (e.g., in Pico island), with F. petreae and
Plantago coronopus, in salty-slime or clay deposits [61],
with high levels of endemic species, reduced disturbance,
Page 15 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
and some biotic resistance towards invaders [68]. How-
ever, the largest population of this species, located in
Ponta do Castelo, Santa Maria, was found among sev-
eral invasive plants (see below). Plots with L. azoricus or
with both target species were rare, the latter only occur-
ring at Ponta do Castelo, in Santa Maria. is island
presents relatively higher temperatures and less rainfall
[69], favouring the occurrence of halo-xerophytic com-
munities where L. azoricus and A vidalii appeared with
F. petreae but also with Carpobrotus edulis and Agave
americana [61].
Environmental descriptors
Our results showed some level of nutrient enrichment of
the soils, namely, high levels of extractable phosphorus
in plots with both target endemics, at Santa Maria island
(5 out of 6 plots). is suggests that agriculture, animal
husbandry, and the presence of seabird colonies [61, 70]
might be causing soil eutrophication. is, together with
favourable climate, likely allows the expansion of eutro-
phication adapted taxa, potentially affecting island plant
assemblages [67], and threatening the endemic halo-
phytes present at the transition between marine and ter-
restrial ecosystems.
Climate appears as a relevant factor for the establish-
ment and thrive of L. azoricus. According to the PCA and
binary regression analyses, it was negatively associated
with higher precipitation levels, justifying its common
occurrence in warmer and drier habitats, contributing to
its restricted and fragmented occurrence, mainly in Santa
Maria Island [69]. is might be, however, related with its
phenology, namely the onset of flowering [21].
In contrast, A. vidalii was found at places with wider
climatic variation (i.e., thermal amplitude), from drier
habitats, such as those of L. azoricus, to areas with higher
precipitation and humidity levels [69]. But also, in a vari-
ety of substrates (from almost vertical cliffs, to rolled
pebble beaches and soil filled rock crevices), appearing
also on shallow soil deposits [61].
A large proportion of control plots were observed in
sand substrates, probably linked with the human expan-
sion in these areas, placing sandy shores among the most
invaded terrestrial environments in Europe [71].
Threats
Invasive species were the most frequent threat observed,
with many known coastal invaders, such as Carpobro-
tus edulis and Arundo donax [38], accompanying indig-
enous taxa. While a worldwide concern [4, 7], their
occurrence is intrinsically linked to human disturbance
[72]. In the Azores, the proliferation of invasive species
in coastal areas is related with traditional land use and,
more recently, with the expansion of human infrastruc-
tures and economic activity [16, 73]. Abandonment
of agricultural land allowed the expansion of deliber-
ately introduced species in the coast, previously used as
hedgerows (e.g., Arundo donax, Metrosideros robusta or
Tamarix africana) [16, 38].
We found that another important threat is the expan-
sion of pastureland to low elevation, further constrain-
ing habitat availability for coastal plants, often already
reduced to a very narrow belt above sea level or restricted
to coastal cliffs, as observed in Santa Maria Island, for
both target species. Additional threats arise from free
roaming animals that can graze or trample on the pop-
ulations of rare endemic coastal plants, reducing native
plant cover and opening space for non-indigenous oppor-
tunists [34]. Lotus azoricus is affected by cattle grazing,
rabbits and rats [61], which negatively impact fitness and
seed production [74].
We observed the occurrence of construction work
near several populations of the target species, including
threatened populations of L. azoricus. Increased eco-
nomic activity, expansion of human activities and con-
struction of infrastructures in coastal areas often result
in habitat destruction and population fragmentation
[4–6], raising new challenges to the survival of coastal
endemic plants worldwide [3, 8]. e occurrence of ille-
gal waste disposal in coastal cliffs, beaches or close to
water streams, potentially affects marine species [75],
degrades indigenous plants habitat [73], and facilitates
the spread of invasive plant taxa. Conservative evolution
on island plants often resulted in increased susceptibility
to anthropic disturbance, and decreased defences against
herbivory [76].
Several sea level populations of A. vidalii are threat-
ened due to climate change and warming, which can
raise sea level and intensify the occurrence of extreme
weather events, leading to coastal flooding [1, 10] and to
the potential loss of unique genetic characteristics of this
species [29]. Previous work suggests that, due to climate
change, the suitable climate space of A. vidalii could
decline [77].
Direct sea exposure, through hydric stress, and
anthropic disturbance can promote the erosion of plant
fixing substrates [19, 21], triggering the occurrence of
landslides and exacerbating natural erosion [78], thereby
facilitating the establishment of naturalised taxa. Besides
coastal disturbance, the mild Azorean climate might
have also facilitated the proliferation of non-indigenous,
generalist plants [69], which could still be aggravated by
global warming [9, 77].
Conservation status and prospects
Despite 24.1% of the Azorean territory is within pro-
tected areas, our results showed that less than half of the
occurrences of A. vidalii were covered, while almost all L.
azoricus populations in Santa Maria and São Jorge were
Page 16 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
covered, but not most of its populations in Pico. Despite
the important conservation areas found in this island,
nearly one third of natural habitat patches were found
outside protected areas, under considerable degradation
[11]. is is worrying, given the important presence of L.
azoricus and A. vidalii in Pico.
Reassessing those areas for conservation will be vital
for the preservation of both taxa in the larger islands,
whose populations are more susceptible to threats, such
as invasive species and others mentioned above, to avoid
further losses due to anthropogenic disturbance [79].
Additionally, the frequent occurrence of both taxa within
protected landscapes does not represent effective con-
servation, due to the low level of restrictions. Monitor-
ing of these species inside and outside of protected areas
is fundamental [79], since many impacting activities like
agriculture or animal husbandry expanded almost to
sea level. Outside protected areas, endemic plants are
even more susceptible due to the lack of monitoring and
impact assessments [79].
Previous evaluations considered A. vidalii as an Endan-
gered species [22], while the conservation status of L.
azoricus in Santa Maria island was Vulnerable [21].
Our evaluation of the conservation status of both spe-
cies resulted in an Endangered (EN) status, showing a
stable trend in the conservation status of these species.
Given the rarity of L. azoricus [19], we argue that the
standardised 2 × 2km cells recommended by the IUCN
[57] for determining the Area of Occupancy are likely
not suitable to access its conservation status, since when
using 1 × 1 km cells, we obtained Critically Endangered
(CR) for L. azoricus.
e results provided by this research reinforce the
need for active conservation measures for both species,
but mostly for L. azoricus. e control of invasive taxa,
training of municipality and environmental workers, and
the restriction of cattle access should be undertaken [21,
80]. e involvement of local communities should also
be a priority in monitoring and cleaning of trash dispos-
als and trampling [75]. Citizen science initiatives that
aim to instil local populations with conservation behav-
iours have generated positive outcomes elsewhere [81].
Finally, the role of botanic gardens in providing back-up
materials for eventual ex situ conservation actions (e.g.,
Life Vidalia project) and the use of molecular genetic
approaches to assess extinction risk and detect reduced
genetic variation and inbreeding among populations will
be particularly important for the conservation of these
endemic species.
Conclusions
Our work raises significant ecological questions regard-
ing the current definition of the coastal herbaceous com-
munities in the Azores. e communities previously
described in the literature [23–27] are becoming rare,
with indigenous plant taxa being restricted to a narrow
vegetation belt constrained by sea level below and by
coastal scrubland or anthromes above.
While endemic chamaephytes and halophytes still are
an important component of these communities (e.g., in
Festuca meadows, rocky chamaephyte communities and
halo-xerophytic communities) [24, 61], we are observing
an increased presence of generalist non-indigenous taxa,
contributing for the homogenisation of the coastal plant
communities. At several locations, the halophytic plant
communities are at stake due to expansion of non-indige-
nous plants, despite the harsh coastal conditions.
Although environmental factors like dry climate, high
salinity, poor nutrient availability and rocky substrates
are still important ecological factors shaping Azorean
coastal herbaceous plant communities [60, 69], increased
anthropogenic disturbance and the expansion of highly
competitive invasive taxa [16, 38] has gained importance.
Increased anthropogenic disturbance derived from the
expansion of economic activities [73] into the coastal
areas is expected to continue in future years. us, natu-
ral coastal habitats are becoming scarce, due to the nar-
rowing and replacement of the respective vegetation belt.
is makes conservation and monitoring activities both
inside and outside protected areas a priority, and sug-
gests the need to periodically reevaluate the design of
coastal protected areas.
e conservation status of these two species have not
deteriorated, remaining as Endangered, suggesting that
restoration initiatives were useful to avoid their further
decline. erefore, it is essential to continue to raise
awareness for the conservation of the Azorean coastal
plant communities [19].
Supplementary Information
The online version contains supplementary material available at https://doi.
org/10.1186/s12862-024-02300-8.
Supplementary Material 1
Acknowledgements
We would like to thank Direção Regional do Ambiente e Alterações Climáticas
and Direção Regional da Ciência e Transição Digital for the licenses used to
collect the samples and conduct the scientic experiments (Lic. Nº 13/2022/
DRAAC; CCIR: 07/2022/DRCTD; SAI-SRAAC/2022/1841 and Lic. Nº 32/2023/
DRAAC; CCIR-RAA/2023/14; SAI/DRCT/2023/289). We would also like to
thank to the Azorean Island Natural Parks, and to Dr. Pedro Casimiro, Dr.
Mafalda Sousa, and Dr. João Bettencourt, from the Life Vidalia project, for the
support given during eld work. We acknowledge the important revisions
and comments made by two anonymous reviewers, that contributed for the
improvement of the manuscript.
Author contributions
RMCR, MM, MOC and LS proposed the topic of study and made the study
design. RMCR, GR, RR and LS performed the sampling, data collection, and
prepared materials. Plant specimens were identied by RMCR, GR and LS.
RMCR and LS performed statistical data analysis and prepared gures. The rst
Page 17 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
draft of the manuscript was written by RMCR, and all authors have contributed
with comments and edits. All authors read, reviewed and approved the nal
manuscript. Funding was acquired by RMCR, MM and LS. MM, MOC and LS
supervised the study.
Funding
This work is part of a PhD scholarship funded by Fundo Regional para a
Ciência e Tecnologia (CONCURSO PRO-SCIENTIA/FRCT/2021/M3.1.a/001,
ref. of application: M3.1.a/F/014/2021). This work was funded by Direção
Regional da Ciência e Transição Digital through 1.1.c – Apoio ao
desenvolvimento de projetos exploratórios de investigação – 2022 (M1.1.C/
PROJ. EXPLORATÓRIOS/011/2022). Additionally, it was nanced by MACFLOR2
(MAC2/4.6d/386) Project under the Cooperation Programme INTERREG MAC
2014–2020. This work is funded by National Funds through FCT - Foundation
for Science and Technology under the project UIDB/50027/2020.
Data availability
The datasets used and/or analysed during the current study are available from
the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The appropriate permits and licences were obtained for the collection of plant
materials during this research (Lic. Nº 13/2022/DRAAC; CCIR: 07/2022/DRCTD;
SAI-SRAAC/2022/1841).
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Received: 30 May 2024 / Accepted: 14 August 2024
References
1. IPCC. Climate Change 2007: Synthesis Report. Contribution of Working
Groups I, II and III to the Fourth Assessment Report of the Intergovernmen-
tal Panel on Climate Change. (2007). [Core Writing Team, Pachauri, R.K and
Reisinger, A, editors]. IPCC, Geneva, Switzerland, 104 pp.
2. Zhao H, Wu R, Hu J, Yang F, Wang J, Guo Y, et al. The contrasting east–west
pattern of vegetation restoration under the large-scale ecological restora-
tion programmes in southwest China. Land Degrad Dev. 2020;31:1688–98.
https://doi.org/10.1002/ldr.3520.
3. Ellis EC. Land use and ecological change: a 12,000-year history.
Annu Rev Environ Resour. 2021;46:1–33. https://doi.org/10.1146/
annurev-environ-012220-010822.
4. Lenzner B, Latombe G, Capinha C, Bellard C, Courchamp F, Diagne C, et al.
What will the future bring for biological invasions on islands? An expert-
based assessment. Front Ecol Evol. 2020;8:280. https://doi.org/10.3389/
fevo.2020.00280.
5. Sax DF, Gaines SD. Species invasions and extinction: the future of native
biodiversity on islands. Proc Natl Acad Sci USA. 2008;105:11490–7. https://doi.
org/10.1073/pnas.0802290105.
6. Whittaker RJ, Fernández–Palacios JM. Island Biogeography: Ecology, Evolution
and Conservation. Oxford: University; 2007.
7. Walentowitz A, Lenzner B, Essl F, Strandberg N, Castilla-Beltrán A, Fernández‐
Palacios JM, et al. Long‐term trajectories of non‐native vegetation on islands
globally. Ecol Lett. 2023;26:729–41. https://doi.org/10.1111/ele.14196.
8. Aguilera MA, Pacheco S, Manzur T. Human-derived eects and failure in
management drive coastal urban foredune degradation and novel vegeta-
tion structure. J Environ Manag. 2022;311:114843. https://doi.org/10.1016/j.
jenvman.2022.114843.
9. Harter DE, Irl SD, Seo B, Steinbauer MJ, Gillespie R, Triantis KA, et al. Impacts
of global climate change on the oras of oceanic islands–Projections,
implications and current knowledge. PPEES. 2015;17:160–83. https://doi.
org/10.1016/j.ppees.2015.01.003.
10. National Academies of Sciences, Division on Earth. Life studies, Board on
Atmospheric Sciences, Committee on Extreme Weather events, Climate
Change Attribution. Attribution of extreme weather events in the context of
climate change. National Academies; 2016.
11. Moreira M, Fonseca C, Vergílio M, Calado H, Gil A. Spatial assessment of habi-
tat conservation status in a macaronesian island based on the InVEST model:
a case study of Pico Island (Azores, Portugal). Land use Policy. 2018;78:637–
49. https://doi.org/10.1016/j.landusepol.2018.07.015.
12. Silva L, Moura M, Schaefer H, Rumsey F, Dias EF. List of vascular plants (Tra-
cheobionta). In: Borges PAV, Costa A, Cunha R, Gabriel R, Gonçalves V, Martins
AF, et al. editors. A list of the terrestrial and marine biota from the Azores.
Portugal: Princípia; 2010. pp. 117–46.
13. eAZora. Creative Solutions, Portugal. https://eazora.org/. Accessed 05 Octo
2023.
14. Borges Silva L, Moura M, Elias RB, Roxo G, Silva L. Caracterização global Da
ora introduzida nos Açores. Abstract retrieved from Spring Seminar 5ª
Edição: Book of abstracts. Azores, Portugal: University of the Azores, Ponta
Delgada; 2022. p. 3.
15. Frutuoso G. Saudades Da terra: livro IV (1589). Ed. Ponta Delgada: Instituto
Cultural de Ponta Delgada, Portugal; 1998.
16. Silva L, Smith CW. A quantitative approach to the study of non-indigenous
plants: an example from the Azores Archipelago. Biodivers Conserv.
2006;15:1661–79. https://doi.org/10.1007/s10531-004-5015-z.
17. Ávila SP, Johnson ME, Rebelo AC, Baptista L, Melo CS. Comparison of Modern
and Pleistocene (MIS 5e) Coastal Boulder deposits from Santa Maria Island
(Azores Archipelago, NE Atlantic Ocean). J Mar Sci Eng. 2020;8:386. https://
doi.org/10.3390/jmse8060386.
18. Schaefer H. Flora of the Azores. A eld guide. 2nd ed. Weikersheim: Margraf;
2005. p. 346.
19. Silva L, Martins MC, Maciel GB, Moura M. Azorean vascular ora. Priorities in
Conservation. Ponta Delgada: Amigos dos Açores and CCPA; 2009. p. 116.
20. Silva L, Dias EF, Sardos J, Azevedo EB, Schaefer H, Moura M. Towards a more
holistic research approach to plant conservation: the case of rare plants on
oceanic islands. AoB Plants. 2015;7:plv066. https://doi.org/10.1093/aobpla/
plv066.
21. Pereira MJ, Trota AN, Bairos J. Conservação activa de Lotus azoricus PW Bell na
ilha de Santa Maria. Rel Com Dep Biol. 2010;36:121–9.
22. Bilz M. Azorina Vidalii. IUCN Red List Threatened Species.
2011;eT161816A5499367. https://doi.org/10.2305/IUCN.UK.2011-1.RLTS.
T161816A5499367.en. Accessed 28 April 2023.
23. Lüpnitz D. Geobotanische Studien Zur natürlichen Vegetation Der Azoren
Unter Berücksichtigung Der Chorologie Innerhalb Makaronesiens. Beitr Biol
Panzen. 1976;51:149–319.
24. Dias E. Vegetação natural dos Açores. Ecologia e sintaxonomia das orestas
naturais. PhD Dissertation. University of the Azores, Angra do Heroísmo,
Portugal. 1996, 302 pp.
25. Aguiar C, Fernández Prieto JA, Dias E. Plantas vasculares endémicas do Arqui-
pélago Dos Acores. In: Dias E, Fernández Prieto JA, Aguiar C, editors. Guia Da
excursão Geobotanica: a paisagem vegetal Da Ilha Terceira (Açores). Portugal:
Universidade dos Açores; 2006. pp. 71–8.
26. Tutin TG. The vegetation of the Azores. J Ecol. 1953;41:53–61. https://doi.
org/10.2307/2257099.
27. Lüpnitz D. Subalpine und alpine Panzengesellschaften auf Der Insel Pico
(Azoren). Bot. Jahrb. 1975;95:149–73.
28. Raposeiro PM, Ritter C, Abbott M, Hernandez A, Pimentel A, Lasher E, et al.
Late holocene climate dynamics in the Azores archipelago. Quat Sci Rev.
2024;331:108617. https://doi.org/10.1016/j.quascirev.2024.108617.
29. Menezes T, Romeiras MM, de Sequeira MM, Moura M. Phylogenetic
relationships and phylogeography of relevant lineages within the complex
Campanulaceae family in Macaronesia. Ecol Evol. 2018;8:88–108. https://doi.
org/10.1002/ece3.3640.
30. Sandral G, Remizowa MV, Sokolo DD. A taxonomic survey of Lotus section
Pedrosia (Leguminosae, Loteae). Wulfenia. 2006;13:97–192.
31. Ojeda DI, Santos-Guerra A, Oliva-Tejera F, Jaen-Molina R, Caujapé-Castells J,
Marrero-Rodríguez Á, et al. DNA barcodes successfully identied Maca-
ronesian Lotus (Leguminosae) species within early diverged lineages of
Cape Verde and mainland Africa. AoB Plants. 2014;6:plu050. https://doi.
org/10.1093/aobpla/plu050.
32. Fernandez Prieto JA, Aguiar C, Dias E, Gutiérrez Villarías MI. On the identity
of Festuca Jubata Lowe (Poaceae) and the description of a new Festuca
species in the Azores Islands. Bot J Linn Soc. 2008;157:493–9. https://doi.
org/10.1111/j.1095-8339.2008.00837.x.
Page 18 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
33. QGIS Development Team. QGIS Geographic Information System. Open
Source Geospatial Foundation Project. http://qgis.osgeo.org (2024) Accessed
24 Jan 2024.
34. Queiroz RE, Ventura MA, Silva L. Plant diversity in hiking trails crossing
Natura 2000 areas in the Azores: implications for tourism and nature
conservation. Biodivers Conserv. 2014;23:1347–65. https://doi.org/10.1007/
s10531-014-0669-7.
35. Franco JA. Nova Flora De Portugal (Continente E Açores), volume I, Lycopo-
diaceae – Umbelliferae. Lisboa: Author’s edition; 1971.
36. Franco JA. Nova Flora De Portugal (Continente E Açores), volume II, Clethra-
ceae – Compositae. Lisboa: Author’s Editions; 1984.
37. ABP. (2024). A zorean Biodiversity Portal https://azoresbioportal.uac.pt/.
Accessed 19 April 2024.
38. Silva L, Ojeda Land E, Rodríguez Luengo JL. Invasive terrestrial ora and fauna
of macaronésia. Top 100 in Azores, Madeira and canaries. Ponta Delgada:
ARENA; 2008.
39. Raunkiaer C. The life form of plants and statistical plant geography. Oxford:
Clarendon; 1934. p. 632.
40. Braun-Blanquet J. Fitossociologia. Bases para El estúdio De las comunidades
vegetales. Madrid: Ed. H. Blume; 1979.
41. Flora-On. Flora de Portugal Interactiva. (2024). Sociedade Portuguesa de
Botânica. www.ora-on.pt. Assessed 3 July 2024.
42. França Z, Cruz JV, Nunes JC, Forjaz VH. Geologia dos Açores: uma perspectiva
actual. Açoreana. 2003;10:11–140.
43. Karger DN, Conrad O, Böhner J, Kawohl T, Kreft H, Soria-Auza RW, et al. Data
from: climatologies at high resolution for the earth’s land surface areas.
EnviDat. 2018. https://doi.org/10.16904/envidat.228.v2.1.
44. Shannon C. A mathematical theory of communication. Bell Syst Tech J.
1948;27:379–423. https://doi.org/10.1002/j.1538-7305.1948.tb01338.x.
45. Pielou EC. The measurement of diversity in diferent types of bio-
logical collections. J Theor Biol. 1966;13:131–44. https://doi.
org/10.1016/0022-5193(66)90013-0.
46. Giraudoux P, Antonietti J-P, Beale C, Groemping U, Lancelot R, Pleydell D et al.
pgirmess: Spatial Analysis and Data Mining for Field Ecologists. 2023; R pack-
age version 2.0.2. https://CRAN.R-project.org/package=pgirmess
47. Warnes GR, Bolker B, Lumley T, Johnson RC. gmodels: Various R Programming
Tools for Model Fitting. 2005; R package version 2.18.1.1 https://CRAN.R-
project.org/package=gmodels
48. R Core Team. R: A language and environment for statistical computing, R
Foundation for Statistical Computing, Vienna, Austria. 2023. https://www.R-
project.org/
49. Barbour MG, Burk JH, Pitts WD. Terrestrial plant Ecology. Califórnia: The Benja-
min/Cummings Publishing Company; 1987. p. 634.
50. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al.
Vegan: community ecology package. R Package Version. 2018;2:4–2. https://
CRAN.R-project.org/package=vegan.
51. Borcard D, Gillet F, Legendre P. Numerical Ecology with R. 1st ed. Springer-;
2011.
52. Rousseeuw PJ, Silhouettes. A graphical aid to the interpretation and valida-
tion of Cluster Analysis. J Comput Appl Math. 1987;20:53–65. https://doi.
org/10.1016/0377-0427(87)90125-7.
53. Legendre P, Legendre L. Numerical ecology. 2nd ed. Amsterdam: Elsevier;
1998.
54. De Cáceres M, Legendre P, Moretti M. Improving indicator species analysis
by combining groups of sites. Oikos. 2010;119:1674–84. https://doi.
org/10.1111/j.1600-0706.2010.18334.x.
55. Dufréne M, Legendre P. Species assemblages and indicator species: the need
for a exible asymmetrical approach. Ecol Monogr. 1997;67:345–66. https://
doi.org/10.1890/0012-9615(1997)067[0345:SAAIST]2.0.CO;2.
56. Roxo G, Brilhante M, Moura M, de Sequeira MM, Silva L, Costa JC, et al.
Genome size variation within Crithmum maritimum: clues on the colonization
of insular environments. Ecol Evol. 2023;13:e10009. https://doi.org/10.1002/
ece3.10009.
57. IUCN Standards and Petitions Subcommittee. Guidelines for Using the IUCN
Red List Categories and Criteria. Version 15.1. (2022). Prepared by the Stan-
dards and Petitions Subcommittee. http://www.iucnredlist.org/documents/
RedListGuidelines.pdf
58. Bachman S, Moat J, Hill AW, de la Torre J, Scott B. Supporting Red List threat
assessments with GeoCAT: geospatial conservation assessment tool. In:
Smith V, Penev L. editors. e-Infrastructures for data publishing in biodiversity
science. ZooKeys. 2011;150:117–126. http://geocat.kew.org/. Accessed 21
June 2023.
59. Elias RB, Gil A, Silva L, Fernández-Palacios JM, Azevedo EB, Reis F. Natural zonal
vegetation of the Azores Islands: characterization and potential distribution.
Phytocoenologia. 2016;46:107–23. https://doi.org/10.1127/phyto/2016/0132.
60. Coelho RML. Plano de gestão e conservação de Azorina vidalii (Wats.) Feer.
Master Dissertation. University of the Azores, Portugal; 2015.
61. Dias E, Mendes C, Aguiar C. Vegetação dos Açores. In: Capelo J, Aguiar
C. editors. Vegetação de Portugal. Lisboa, INCM. 2021;155–179. ISBN
978-972-27-2879-9.
62. Mologni F, Bellingham PJ, Cameron EK, Dinh K, Wright AE, Burns KC. Func-
tional traits explain non-native plant species richness and occupancy on
northern New Zealand islands. Biol Invasions. 2022;24:2135–54. https://doi.
org/10.1007/s10530-022-02762-1.
63. Eppinga MB, Haber EA, Sweeney L, Santos MJ, Rietkerk M, Wassen MJ.
Antigonon leptopus invasion is associated with plant community disassembly
in a Caribbean island ecosystem. Biol Invasions. 2022;24:353–71. https://doi.
org/10.1007/s10530-021-02646-w.
64. Nürk NM, Atchison GW, Hughes CE. Island woodiness underpins accelerated
disparication in plant radiations. New Phytol. 2019;224:518–31. https://doi.
org/10.1111/nph.15797.
65. Ainsworth A, Drake DR. Classifying hawaiian plant species along a habitat
generalist-specialist continuum: implications for species conservation under
climate change. PLoS ONE. 2020;15:e0228573. https://doi.org/10.1371/jour-
nal.pone.0228573.
66. Schaefer H. An inland population of Azorina Vidalii (Watson) Feer at Faial
Island. Arquipélago Life Mar Sci. 1999;17A:93–5.
67. Cabrera Rodriguez F, Otto R, Fernández-Palacios JM. Changes in soil chemical
properties and plant species composition during primary succession on an
oceanic island. J Veg Sci. 2022;33:e13137. https://doi.org/10.1111/jvs.13137.
68. Fernández-Palacios JM, Schrader J, de Nascimento L, Irl SD, Sánchez‐Pinto L,
Otto R. Are plant communities on the Canary Islands resistant to plant inva-
sion? Divers Distrib. 2023;29:51–60. https://doi.org/10.1111/ddi.13650.
69. Azevedo EB, Rodrigues MC, Fernandes JF. O clima dos Açores. In: Forjaz VH.
editor, Atlas Básico dos Açores. Azores, Ponta Delgada: Observatório Vulca-
nológico e Geotérmico dos Açores; 2004. pp. 25–48.
70. Otero XL, De La Peña-Lastra S, Pérez-Alberti A, Ferreira TO, Huerta-Diaz
MA. Seabird colonies as important global drivers in the nitrogen and
phosphorus cycles. Nat Commun. 2018;9:246. https://doi.org/10.1038/
s41467-017-02446-8.
71. Chytrý M, Maskell LC, Pino J, Pyšek P, Vilà M, Font X, et al. Habitat invasions by
alien plants: a quantitative comparison among Mediterranean, subcontinen-
tal and oceanic regions of Europe. J Appl Ecol. 2008;45:448–58. https://doi.
org/10.1111/j.1365-2664.2007.01398.x.
72. Macdonald DW, Chiaverini L, Bothwell HM, Kaszta Ż, Ash E, Bolongon G, et al.
Predicting biodiversity richness in rapidly changing landscapes: climate, low
human pressure or protection as salvation? Biodivers Conserv. 2020;29:4035–
57. https://doi.org/10.1007/s10531-020-02062-x.
73. Calado H, Borges P, Phillips M, Ng K, Alves F. The Azores archipelago,
Portugal: improved understanding of small island coastal hazards and
mitigation measures. Nat Hazards. 2011;58:427–44. https://doi.org/10.1007/
s11069-010-9676-5.
74. Parada-Díaz J, Cubas J, del Arco Aguilar MJ, López ÁF, González-Mancebo
JM. Changing the landscape: Short-term disruption of recruitment, height
growth and owering of endemic plants on macaronesian heathlands
by invasive alien rabbits. Biol Conservation. 2022;272:109596. https://doi.
org10.1016/j.biocon.2022.109596.
75. Rodríguez Y, Ressurreição A, Pham CK. Socio-economic impacts of marine
litter for remote oceanic islands: the case of the Azores. Mar Pollut Bull.
2020;160:111631. https://doi.org/10.1016/j.marpolbul.2020.111631.
76. Barton KE, Fortunel C. Island plant functional syndromes and competition
with invasive species. J Biogeogr. 2023;50:641–53. https://doi.org/10.1111/
jbi.14568.
77. Ferreira MT, Cardoso P, Borges PA, Gabriel R, Azevedo EB, Reis F, et al. Eects
of climate change on the distribution of indigenous species in oceanic
islands (Azores). Clim Change. 2016;138:603–15. https://doi.org/10.1007/
s10584-016-1754-6.
78. Vergílio M, Fjøsne K, Nistora A, Calado H. Carbon stocks and biodiversity
conservation on a small island: Pico (the Azores, Portugal). Land Use Policy.
2016;58:196–207. https://doi.org/10.1016/j.landusepol.2016.07.020.
79. Mulongoy K, Chape S. Protected areas and Biodiversity: an overview of Key
issues. Cambridge, UK: CBD Secretariat, Montreal Canada and UNEP-WCMC;
2004.
Page 19 of 19Rego et al. BMC Ecology and Evolution (2024) 24:111
80. Russell JC, Kueer C. Island biodiversity in the Anthropocene.
Ann Rev Environ Resour. 2019;44:31–60. https://doi.org/10.1146/
annurev-environ-101718-033245.
81. Fletcher M-S, Hamilton R, Dressler W, Palmer L. Indigenous knowledge and
the shackles of wilderness. Proc Natl Acad Sci USA. 2021;118:e2022218118.
https://doi.org/10.1073/pnas.2022218118.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional aliations.
