Conservation ecology of butterflies on Cyprus in the context of Natura 2000

Abstract

In this study we investigate the environmental factors influencing butterfly communities and evaluate the Natura 2000 network’s effectiveness in representing butterfly species richness and abundance, taking as a case study the island of Cyprus. We sampled butterflies and 11 environmental factors in 60 randomly selected sites across four 500-m elevation zones, representing seven habitat types. Rural mosaics and riverine vegetation were the habitats with the highest diversity of butterflies. Within habitats, the number of flower heads was the most important factor favouring butterfly species richness and abundance and endemic butterfly richness, while soil humidity had a positive effect on species richness and abundance. Although the Natura 2000 network succeeds in including the majority of butterfly species and all Cyprian endemics, the transects sampled within the network did not support more butterfly species than those outside it, and were significantly poorer in terms of butterfly abundance and endemic butterfly species richness and abundance. We found a similar pattern for the Habitats Directive priority habitats, which held poorer overall and endemic butterfly communities than the other habitats. The effectiveness of existing protected area networks may need to be reassessed in regions such as the South East Mediterranean, to ensure that regionally important components of biological diversity are adequately protected. To this aim, our results suggest that new European and national policies as well as further inclusion of rural mosaics and riverine habitats in protected area networks are needed for effective butterfly conservation in Cyprus.

Full text

Vol.:(0123456789)
Biodiversity and Conservation
https://doi.org/10.1007/s10531-019-01755-2
1 3
ORIGINAL PAPER
Conservation ecology ofbutteries onCyprus inthecontext
ofNatura 2000
ElliTzirkalli1,2 · CostasKadis2,3· JohnM.Halley1· IoannisVogiatzakis4 ·
RobertJ.Wilson5 · KonstantinaZografou6 · AndreasAntoniou7· TakisTsintides8·
ChristodoulosMakris9· VassilikiKati1
Received: 18 May 2018 / Revised: 25 March 2019 / Accepted: 29 March 2019
© Springer Nature B.V. 2019
Abstract
In this study we investigate the environmental factors influencing butterfly communities
and evaluate the Natura 2000 network’s effectiveness in representing butterfly species rich-
ness and abundance, taking as a case study the island of Cyprus. We sampled butterflies
and 11 environmental factors in 60 randomly selected sites across four 500-m elevation
zones, representing seven habitat types. Rural mosaics and riverine vegetation were the
habitats with the highest diversity of butterflies. Within habitats, the number of flower
heads was the most important factor favouring butterfly species richness and abundance
and endemic butterfly richness, while soil humidity had a positive effect on species rich-
ness and abundance. Although the Natura 2000 network succeeds in including the majority
of butterfly species and all Cyprian endemics, the transects sampled within the network
did not support more butterfly species than those outside it, and were significantly poorer
in terms of butterfly abundance and endemic butterfly species richness and abundance. We
found a similar pattern for the Habitats Directive priority habitats, which held poorer over-
all and endemic butterfly communities than the other habitats. The effectiveness of exist-
ing protected area networks may need to be reassessed in regions such as the South East
Mediterranean, to ensure that regionally important components of biological diversity are
adequately protected. To this aim, our results suggest that new European and national poli-
cies as well as further inclusion of rural mosaics and riverine habitats in protected area net-
works are needed for effective butterfly conservation in Cyprus.
Keywords Community ecology· Diversity patterns· Islands· Lepidoptera·
Mediterranean· Protected areas
Communicated by Kwek Yan Chong.
* Vassiliki Kati
vkati@uoi.gr
Extended author information available on the last page of the article

Biodiversity and Conservation
1 3
Introduction
Invertebrates are the most abundant and diverse organisms in the world, constituting 80%
of all described species, yet they have been largely understudied and hence neglected
in conservation biology (Cardoso etal. 2011). This lack of knowledge and popularity
of invertebrates is reflected in the Habitats Directive (92/43/EEC) which, together with
the Birds Directive (2009/147/EC), form the foundation of the Natura 2000 network in
Europe (Trochet and Schmeller 2013; Orlikowska etal. 2016). Although the Habitats
Directive has been recently evaluated as fit for its purpose (Milieu etal. 2016), namely
to guarantee the long-term survival of Europe’s most valuable and threatened species
and habitats, the list of arthropods is incomplete, and many important species are not-
annexed, due to gaps in knowledge. It includes a tiny fraction of the European inverte-
brates, accounting for less than 6% of the annexed species (EIONET 2018). Thus, it is
widely recognized that this important biodiversity component is largely neglected from
the process of conservation planning and management at the European level (D’Amen
etal. 2013). The spatial designation of the European network has been in general evalu-
ated as adequate by the scientific community (Kati etal. 2015), although no general
consensus has been reached so far for invertebrates (e.g. D’Amen etal. 2013; Pellissier
etal. 2014; Rosso et al. 2017). In order to be able to strengthen Natura 2000s effec-
tiveness towards conserving the complete spectrum of biological diversity in Europe, a
broader view should be adopted by including non-annexed biodiversity components in
the conservation management schemes of the network, such as non-annexed invertebrate
species.
Butterflies (Rhopalocera, Lepidoptera) are among the few invertebrate groups that
are well represented in the Directive annexes, also being among the best-known inverte-
brate taxa in terms of their distribution patterns and abundance (van Swaay and Warren
1999; Cardoso 2012). They are considered sensitive indicators of environmental change
induced by factors such as agricultural intensification, land use change, habitat fragmen-
tation, overgrazing and climate change (Warren etal. 2001; Zografou etal. 2014; Her-
rando etal. 2015). They are also one of the 26 indicators used to assess progress towards
the EU’s biodiversity strategy of halting the loss of biodiversity and the degradation of
ecosystem services by 2020 (EEA 2012), whilst the European grassland butterfly indi-
cator is an important policy tool for this aim (van Swaay etal. 2016). However, even for
this group, the Directive annexes cover only 34% of the butterfly species that are consid-
ered threatened in the EU (van Swaay etal. 2011; Maes etal. 2013) and only one butter-
fly species (Polyommatus golgus) that is threatened in the Mediterranean region (Numa
etal. 2016), despite the high butterfly richness and the high degree of endemism charac-
terizing the area (Munguira 1995). In spite of the good background knowledge acquired
for butterflies across Europe and enhanced by well-developed monitoring schemes
(Thomas 2005) our knowledge of butterfly distribution patterns, ecological preferences,
threat factors and population trends is still scarce for south-east Europe.
The island of Cyprus is located within the Mediterranean Basin biodiversity hotspot
(Hewitt 2011), and hosts more than 1700 plant taxa, of which 8.2% are endemic (Tsin-
tides etal. 2007). Its insect diversity is also considered high as more than 7000 species
are estimated to be present on the island and more than 10% are endemic (Kadis etal.
2012). Although no butterfly species holds a legally protected status under European
legislation (92/43/EEC) and there is no national red list, the island hosts 53 species,
including six endemic species and subspecies (Makris 2003; John and Skule 2016), and

Biodiversity and Conservation
1 3
two Near Threatened species under the international IUCN criteria (van Swaay etal.
2010), underlining its conservation importance.
In this light, the goal of the current study was to increase ecological knowledge of Cyprus
butterflies, so as to open new perspectives for their adequate conservation management in the
future. We particularly aimed: (a) To explore the diversity patterns of the butterfly community
in Cyprus among different habitat types, (b) To investigate the effects of a set of environmen-
tal factors on the composition and abundance of butterfly communities (c) To evaluate the
effectiveness of the Natura 2000 network and its priority habitat types in conserving Cyprus
butterflies, and (d) To crystallize the implications of our findings for conservation actions.
Methods
Study area
The study area is situated in Cyprus, extending from sea level up to 1952m, covering an area
of 1400km2. The climate is Mediterranean, with a mean annual temperature of 18 o C, a mean
annual precipitation of 480mm, and an arid season extending over 7months in the lowlands
and over 5months in higher altitudes (Tsintides etal. 2002, 2007). Pine forests (Pinus bru-
tia) are the dominant vegetation type, in particular at higher altitudes, followed by scrubland,
whilst at medium altitudes agricultural landscapes cover a great part of the area (see Table4 in
Appendix). The study area includes nine Sites of Community Importance (SCI) of the Natura
2000 network, covering 14.3% of its area (Fig.1).
Site selection
We selected a set of 60 representative sample sites of 25ha each from the study area, using
ArcGIS v.10 (ESRI 2010), as described herein. We divided the study area into four 500-m
elevation zones (< 500, 501–1000, 1001–1500, > 1500m), and we super-imposed a standard
grid of 500 × 500m2. Each grid cell was classified according to its elevational zone (> 70%
of cover within the zone) and its habitat type (> 60% cover within the grid). We considered
seven broad habitat types, namely forests, sclerophyllous vegetation, transitional woodland/
shrubland, agriculture areas, heterogeneous agriculture areas (rural mosaics) and grasslands,
according to the Corine Land Cover Classification (EEA 2010), as well as riparian habitats
(data provided by the Department of Forests of Cyprus). We then randomly selected 15 of the
500 × 500 m2 grid cells for every elevation category and located a sampling transect at each
of these sites, in order to represent the habitat types that were present in each elevation zone
with at least one transect each (see Table4 in Appendix), under the condition that transect
centres were at least 500m apart, to avoid duplicate counts. The 60 sites selected for transects
included 32 sites falling within the Natura 2000 network and 18 sites within priority habitats
of the Habitats Directive 92/43/EEC (Fig.1; see Table5 in Appendix).
Sampling
One standard transect of 300-m length was established within each of the 60 selected sites.
We counted all butterfly species occurring within a band of 2.5m on each side of the tran-
sect, as well as up to 5m ahead (Pollard and Yates 1993), and we repeated each transect

Biodiversity and Conservation
1 3
six times at approximately 20day intervals (March to July; years 2012–2013), using a hand
held net and field guides (Tolman and Lewington 1997; Makris 2003). Transects were
walked at a constant speed, between 8 am and 5pm, under appropriate weather condi-
tions by one observer (first author). Since no year effect was observed either on species
richness or abundance (Generalized Linear Models (GLMs): z-value = 0.946, p > 0.05 and
z-value = 0.612, p > 0.05 respectively), we consolidated all the repeated surveys at each
transect by using the sum of abundances recorded for each butterfly species at each transect
across the 2 years; transect was therefore the unit of analysis.
We also recorded 11 environmental parameters for all 60 sites, related to butterfly ecol-
ogy (Samways etal. 2010). We recorded transect elevation (m) using a GPS (transect cen-
tre), and we used an ordinal four level scale (1: 0–10%, 2: 11–20%, 3: 21–35%, 4: > 35%)
to estimate slope (100% slope corresponds to slope of 90°), bare ground and rocky sub-
strate cover at the start of each transect (300 × 5m2). We also measured the following seven
parameters within three standard quadrats (25m2) that were positioned at the beginning
(0m), middle (150m) and end (300m) of each transect, and we considered the transect
average values in the analyses: Soil humidity and temperature were measured at the cen-
tre of each quadrat using a Hobo H21-002 data logger; Quadrat vegetation was described
by recording once (in May) the average cover of low shrubs (< 0.5 m), high shrubs
(0.5–2.5m), and trees (> 2.5m), as the vertical projection of their crown area (%). To cover
the seasonal variability of herb and flower heads, we recorded them twice (April–May) and
considered their average values in the analysis. Herb cover was estimated as percentage
Fig. 1 The location of the 60 sampling sites for butterflies in the study area, showing the four elevational
zones, areas designated as Sites of Community Importance (SCI) in the Natura 2000 network, and sampling
sites falling in priority habitat types (PH sampling sites)

Biodiversity and Conservation
1 3
(%) and flower heads as the number of flower heads from individuals flowers and inflo-
rescences, using the following ordinal scale 1: < 10, 2: 11–50, 3: 51–100, 4: 101–200, 5:
201–400, 6: 401–600, 7: > 600), considering both the herb and low shrub vegetation layers.
Data analysis
We estimated the overall expected butterfly species richness in the study area, in each
habitat type, in sites either inside or outside the Natura 2000 network and in priority or
non-priority sites using the nonparametric species richness estimator Chao 1 (1000 permu-
tations) (Magurran 2004; Colwell 2013). We calculated species richness (S), endemic spe-
cies richness (SE), number of specialist species, abundance of all species (A), abundance of
endemic species (AE), as well as the Shannon–Wiener diversity index (H) for each transect.
We calculated overall diversity values, as well as average transect diversity values for each
habitat type, for sites within and outside the Natura 2000 network, and for priority and non-
priority habitats. To identify specialist species, we calculated their specialization index,
as the average of three variables related to larval food plant, habitat demands and spatial
distribution following the methodology of Habel etal. (2018).
To investigate the environmental parameters influencing the composition of the butterfly
community, we conducted a Redundancy Analysis (RDA), after running a Hellinger trans-
formation of the species dataset (many species absences) (Monte Carlo randomisation test:
1000 permutations; variables without collinearity VIF < 5).
To assess the effects of the environmental variables on butterfly species richness,
endemic species richness and butterfly abundance we used Generalized Linear Models
(GLM), with Poisson and Negative Binomial error distributions. In order to deal with the
many explanatory variables, we utilized the following approach to model selection. We
first tested the 11 explanatory variables for multicollinearity (Spearman rank coefficient
r > 0.7), and one variable (temperature) was excluded due to its correlation with altitude
(r = − 0.851, p < 0.01). Then, we performed a hierarchical partitioning analysis (Mac Nally
2002) to identify the independent contribution of each environmental variable to our meas-
ures of community richness and composition, and ranked them according to their explana-
tory strength (Z-score), using 100 randomizations (with the function ‘‘rand.hp’’). To avoid
bias from variable position within our dataset which may produce considerable incon-
sistency when more than nine variables are used for analysis (Olea etal. 2010), we first
reshuffled all variables 100 times by using the function ‘‘sample’’ in R (i.e. the position of
variables was randomly changed). This procedure was used to determine the importance of
each variable prior to a series of GLMs (Matteson and Langellotto 2010). We then adopted
a stepwise simplification approach to conclude on the final model structure. Starting from
a full model with all explanatory variables we gradually dropped the variable with the low-
est Z-score (see Fig.3 in Appendix), as identified from hierarchical partitioning, until the
final model comprised of only statistically significant variables (significance level set at
α = 0.05). We also calculated the percentage of deviance explained (D2) as a measure of the
explanatory power of the best models. In all GLMs, the explanatory variables were stand-
ardized to allow comparison of model parameter estimates.
In addition, we tested whether species diversity differed significantly (a) between
sites within or outside the Natura 2000 network, and (b) between priority or non-prior-
ity habitat types from the Directive 92/43/EEC. We performed Welch’s t test when the
condition of a Normal distribution was satisfied (Shapiro–Wilk test) or the non-para-
metric alternative (Mann–Whitney U test), using the average transect diversity values.

Biodiversity and Conservation
1 3
Finally, we performed Analysis of Similarities (ANOSIM) to detect whether there was
a significant difference in species composition for the above two cases (a and b) and
further similarity percentage analysis (SIMPER), using the package “vegan” (Oksanen
etal. 2017), to pinpoint those species contributing to the differences in communities
(Clarke 1993). All analyses were performed in R software (version R 3.3.1, R Core
Team 2016) using the packages “vegan”, “MASS”, “car” and “hier.part” (Venables
and Ripley 2002; Fox and Weisberg 2011; Oksanen etal. 2017; Walsh and Mac Nally
2013).
Results
A total of 38 butterfly species (4708 individuals) was recorded, belonging to 5 families
and 32 genera. Sampling efficiency was high (Chao 1 > 98% for each habitat type, 100% of
species sampled in the study area) (Table1). Nymphalidae was the most dominant family
with 14 species, followed by Lycaenidae (11), Pieridae (7), Hesperiidae (4) and Papillioni-
dae (2). We recorded all three endemic species of Cyprus, namely Glaucopsyche paphos,
Hipparchia cypriensis and Maniola cypricola, as well as the three endemic sub-species
Zerynthia cerisyi cypria, Hipparchia syriaca cypriaca and Chazara briseis larnacana.
None of the recorded species is listed under the IUCN threat categories in Cyprus, apart
from Thymelicus acteon, which is Near Threatened at European level, but Least Concern at
Mediterranean level (van Swaay etal. 2010; Numa etal. 2016). Only two species (Aphari-
tis acamas, Zerynthia cerisyi cypria) were identified as specialists, occurring in four sites
in the study area (Table2).
Diversity patterns
We found that rural mosaics and riverine vegetation were the most diverse sites in terms
of the Shannon-Weiner index while forests were the least diverse, but sites from each of
these three habitats included the six endemic (sub) species of Cyprus. Rural mosaics and
transitional woodland/shrubland were highest in average endemic species richness, while
agricultural areas had the greatest average endemic species abundance. The two specialist
species co-occurred only in riverine vegetation.
Environmental parameters
We found that four environmental parameters: the number of flower heads, altitude, herb
cover and rock cover, were significantly related to the occurrence of species in the butter-
fly community (Redundancy Analysis, RDA; 20% of variance explained, p < 0.05; Fig.2).
An increased number of flower heads positively affected several species, and in particular
Polyommatus icarus, Lampides boeticus and Colias crocea. Some other species preferred
higher altitudes, such as Pseudochazara anthelea and Celastrina argiolus, while other
species were limited to lower altitudes, such as the Near Threatened species Thymelicus
acteon. Several species were dependent on higher levels of herb cover, such Aricia ages-
tis and the endemic Hypparchia cypriensis, or of rock cover in the case of Lassiommata
megera.
The best models that resulted from stepwise model simplification is presented in
Table 3. The number of flower heads was the most important parameter, significantly

Biodiversity and Conservation
1 3
Table 1 Butterfly diversity across habitat types, Natura 2000 sites, and the priority habitats (after Directive 92/43/EEC) in Cyprus
Chao 1 Estimated species richness according to Chao 1 non-parametric estimator. S species richness (in parenthesis number of specialist species), SE number of endemic spe-
cies or subspecies, A abundance (total number of individuals recorded). Mean S mean species richness of transects, Mean SE mean endemic species richness of transecs. Mean
A Mean abundance of transects, Mean AE Mean endemic species abundance of transects, Mean H mean Shannon index of transects. Habitat types are ranked by Mean H
Habitat types
Habitat types Overall diversity Average transect diversity N sites
Chao 1 S SEA Mean S Mean SEMean A Mean AEMean H
Rural mosaics 28.50 (± 1.39) 28 (1) 6 1021 11.63 (± 0.86) 2.00 (± 0.42) 127.63 (± 16.80) 23.43 (± 9.24) 2.26 (± 0.05) 8
Riverine vegetation 29.33 (± 0.92) 29 (2) 6 722 9.75 (± 0.84) 1.50 (± 0.46) 90.25 (± 10.78) 12.14 (± 2.99) 2.02 (± 0.09) 8
Transitional woodland/scrubland 18 (± 0.71) 18 (0) 4 373 9.00 (± 0.71) 1.75 (± 0.25) 93.25 (± 8.77) 16.25 ((± 4.59) 2.00 (± 0.05) 4
Grasslands 24 (± 0.63) 24 (1) 4 278 8.75 (± 0.63) 1.25 (± 0.25) 69.50 (± 16.29) 8.25 (± 2.78) 1.98 (± 0.05) 4
Sclerophyllous vegetation 21 (± 1.12) 21 (0) 3 793 8.33 (± 1.12) 1.11 (± 0.11) 88.11 (± 11.07) 9.00 (± 1.08) 1.93 (± 0.12) 9
Agricultural areas 23 (± 1.20) 23 (1) 4 544 8.33 (± 1.20) 1.67 (± 0.56) 90.67 (± 12.19) 26.75 (± 4.96) 1.91 (± 0.15) 6
Forest 23 (± 0.46) 23 (0) 6 977 6.00 (± 0.46) 1.05 (± 0.16) 46.52 (± 6.80) 7.81 (± 1.64) 1.61 (± 0.08) 21
Natura 2000 network
Natura 2000 32.50 (± 0.50) 32 (1) 6 2008 7.28 (± 0.48) 1.09 (± 0.14) 63.77 (± 7.18) 8.86 (± 1.20) 1.79 (± 0.08) 32
Non-Natura 2000 37 (± 0.71) 36 (2) 6 2700 9.11 (± 0.59) 1.64 (± 0.20) 98.33 (± 7.08) 17.29 (± 3.20) 1.99 (± 0.06) 28
Priority habitats
Priority habitats 23 (± 0.80) 22 (0) 4 917 6.72 (± 0.75) 0.89 (± 0.11) 50.94 (± 9.55) 5.80 (± 0.76) 1.68 (± 0.11) 18
Non-priority habitats 40 (± 0.50) 38 (2) 6 3791 8.86 (± 0.42) 1.5 (± 0.16) 90.26 (± 5.73) 15.57 (± 2.21) 1.97 (± 0.05) 42
Study area
Total 38.33 (± 0.43) 38 (2) 6 4708 8.22 (± 0.39) 1.57 (± 0.12) 78.47 (± 5.42) 12.75 (± 1.70) 1.88 (± 0.05) 60

Biodiversity and Conservation
1 3
Table 2 Inventory of butterfly species in the study area
N sites: number of sites the species was recorded. Total abundance (%): Share of the species abundance as
percentage of overall butterfly abundance (4708 individuals, considering the maximum abundance recorded
in the six visits). E endemic. NT near threatened. F food plant (monophagous-1; oligophagous-2; polypha-
gous-3); H habitat demands (specialist-1; intermediate-2; ubiquitious-3); D spatial distribution, local-1;
continental-2; global-3); SI specialization index (Habel etal. 2018)
Code Species E/NT Specialization index N sites Total abun-
dance (%)
F H D SI
Acard Anthocharis cardamines 2 2 3 2.33 5 0.38
Aacam Apharitis acamas – 1 2 1.00 1 0.04
Acrat Aporia crataegi 2 2 3 2.33 5 0.98
Apand Argynnis pandora 1 1 3 1.67 2 0.13
Aages Aricia agestis 1 3 3 2.33 9 0.87
Calce Carcharodus alceae 2 3 3 2.67 4 0.83
Cargi Celastrina argiolus 3 3 3 3.00 10 1.61
Cbril Chazara briseis larnacana E 2 2 1 1.67 6 1.17
Ctroc Chilades trochylus 2 3 3 2.67 5 0.74
Ccroc Colias crocea 2 3 3 2.67 40 10.07
Gpumi Gegenes pumilio 2 2 2 2.00 1 0.02
Gpaph Glaucopsyche paphos E 1 3 1 1.67 13 5.18
Gcleo Gonepteryx cleopatra 1 3 2 2.00 34 7.98
Hcypr Hipparchia cypriensis E 2 3 1 2.00 32 4.03
Hsyrc Hipparchia syriaca cypriaca E 2 2 1 1.67 7 0.62
Hlupi Hyponephele lupina 2 3 2 2.33 6 0.72
Kroxe Kirinia roxelana 2 2 2 2.00 3 0.08
Lboet Lampides boeticus 3 3 3 3.00 19 3.63
Lmaer Lasiommata maera 2 2 3 2.33 2 0.30
Lmege Lasiommata megera 2 2 3 2.33 5 0.64
Lpiri Leptotes pirithous 3 3 3 3.00 1 0.04
Lredu Limenitis reducta 1 2 2 1.67 3 0.23
Lphla Lycaena phlaeas 1 2 3 2.00 15 2.59
Lther Lycaena thersamon 1 2 2 1.67 3 0.19
Mcypr Maniola cypricola E 2 3 1 2.00 20 2.68
Pmach Papilio machaon 3 3 3 3.00 16 1.97
Paege Pararge aegeria 2 3 3 2.67 14 2.87
Pthra Pelopidas thrax 2 2 3 2.33 3 0.17
Pbras Pieris brassicae 3 3 3 3.00 49 12.93
Prapa Pieris rapae 3 3 3 3.00 45 14.87
Picar Polyommatus icarus 2 3 3 2.67 29 8.22
Pdapl Pontia daplidice 3 3 3 3.00 29 5.39
Panth Pseudochazara anthelea 2 2 2 2.00 7 1.32
Tacte Thymelicus acteon ΝΤ 2 3 3 2.67 10 2.93
Vatal Vanessa atalanta 2 3 3 2.67 2 0.08
Vcard Vanessa cardui 3 3 3 3.00 28 2.97
Zcerc Zerynthia cerisyi cypria E 1 2 1 1.33 4 0.34
Zkars Zizeeria karsandra 3 2 3 2.67 2 0.17

Biodiversity and Conservation
1 3
predicting overall butterfly species richness together with the soil humidity, whilst the
cover of trees had a negative effect. The number of flower heads was the sole significant
predictor for endemic butterfly richness. Two environmental parameters, the number of
flower heads, and soil humidity were positive predictors of butterfly abundance, whilst alti-
tude and the cover of trees were negative ones.
Evaluation oftheNatura 2000 network
We found that sites sampled within the Natura 2000 network (32 out of 60 sites) included
82% of the overall butterfly species observed, including all the endemic species. Two spe-
cies were found exclusively in the network (Argynnis pandora, Gegenes pumilio) and six
species outside it (Apharitis acamas, Lycaena thersamon, Leptotes pirithous, Zizeeria
karsandra, Vanessa atalanta, Pelopidas thrax). The comparison between transects fall-
ing within and outside the network was valid based on estimated sample coverage in each
group of sites (Chao 1 > 95%). No significant difference was found for average butterfly
species richness between the transects falling within and outside the network, but aver-
age butterfly abundance was significantly lower within the network (Welch’s t test = − 1.66,
p > 0.05; U = 257, p < 0.01, for species richness and abundance respectively). Transects
in the network were also found to be poorer in terms of average endemic butterfly spe-
cies richness and abundance as compared with the transects sampled outside it (U = 304.5,
p < 0.05; U = 307.5, p < 0.05, for endemic species richness and abundance respectively).
Both specialist species occurred outside the network and one was found within it. Finally,
no difference regarding butterfly community composition was observed in transects falling
within and outside the Natura 2000 network (ANOSIM: R = 0.15, p > 0.01).
The priority habitats of the Habitats Directive 92/43/EEC (18 sites) showed signifi-
cantly lower butterfly species richness and abundance compared to the remaining habitats
(42 sites) (U = 248, p < 0.05; U = 187.5, p < 0.01, respectively), as well as lower endemic
butterfly species richness and abundance (U = 224.5, p < 0.01; U = 201, p < 0.0.01, respec-
tively). In addition, no specialist species were recorded in priority habitats. The commu-
nity composition of butterflies inhabiting priority habitats was significantly different from
Fig. 2 Redundancy analysis
diagram (RDA) presenting the
significant (p < 0.05) environ-
mental factors (arrows) for
butterfly species (abbreviations
according to Table2)

Biodiversity and Conservation
1 3
sites in the remaining study area (ANOSIM: R = 0.33, p < 0.01). This differentiation was
mainly attributed to eight common species (P. rapae, P. brassicae, C. crocea, P. icarus, G.
cleopatra, P. daplidice, V. cardui, L. boeticus) and three endemic species (G. paphos, H.
cypriensis, M. cypricola) (SIMPER analysis), all being more abundant outside the network
(see Table6 in Appendix).
Discussion
Diversity patterns
Our results clearly show that rural mosaics have a high ecological value for butterfly con-
servation, in line with previous studies in the Mediterranean area (Grill and Cleary 2003;
Zografou et al. 2009). Their structural heterogeneity offers a variety of microhabitats,
increases the availability of food resources and oviposition sites, providing a wider range
of ecological niches for butterfly species than uniform landscapes (Flick etal. 2012). How-
ever, Cyprus is gradually losing its rural mosaics to forest cover, due to the gradual aban-
donment of agricultural land leading to the loss of traditional elements (e.g. low intensity
vineyards and carob cultivations), as well as due to the rural depopulation stemming from a
shift from an agricultural to a tourist economy (DF and FAO 2005; Delipetrou etal. 2008;
Table 3 Results of the best
generalized linear models
(GLMs) characterizing butterfly
species richness, endemic
butterfly species richness
and butterfly abundance by
explanatory variables
D2: percentage of deviance explained (%). (↑)—positive statistically
significant relationship; (↓)—negative statistically significant relation-
ship
Model selection was based on stepwise model simplification until only
significant variables were included in the best models (statistical sig-
nificance level set at α = 0.05)
Variables Estimate SE z value P
Butterfly species richness
Intercept 2.06 0.05 43.89 < 0.001
Flower heads (↑) 0.20 0.04 4.62 < 0.001
Soil humidity (↑) 0.15 0.05 3.30 < 0.001
Tree cover (↓)− 0.11 0.05 − 2.04 0.04
D2 = 66
Endemic butterfly species richness
Intercept 0.29 0.11 2.58 0.01
Flower heads (↑) 0.20 0.10 1.97 0.05
D2 = 32
Butterfly abundance
Intercept 4.25 0.05 91.01 < 0.001
Flower heads (↑) 0.25 0.05 5.10 < 0.001
Altitude (↓)− 0.31 0.05 − 6.35 < 0.001
Soil humidity (↑) 0.17 0.05 3.54 < 0.001
Tree cover (↓)− 0.14 0.05 − 2.71 0.01
D2 = 62

Biodiversity and Conservation
1 3
Ieronymidou 2012). This is a well-known trend in remote and less accessible agricultural
areas all over southern Europe (MacDonald etal. 2000; Geri etal. 2010). Forest encroach-
ment that follows agricultural land abandonment reduces landscape heterogeneity and open
habitat availability, which is known to negatively affect butterflies (Slancarova etal. 2016;
Bonelli etal. 2018), as well as other open-land dwelling invertebrates (Plexida etal. 2012),
or vertebrates (Mammides etal. 2015; Zakkak etal. 2015). Agricultural areas on the other
hand were far less diverse, but they seemed to hold large species populations, including
endemic species populations. This could be attributed to the presence of vineyards in the
study area (see Table5 in Appendix), which are known to be important habitats for butter-
flies (Puig-Montserrata etal. 2017).
Our results underline the great value of riverine vegetation for butterfly conservation in
Cyprus, including specialist species. The riverine vegetation consists mainly of restricted
galleries (Platanus orientalis, Alnus orientalis, Nerium oleander, Tamarix spp.) along the
riverside of non-permanent streams and permanent rivers, from higher elevational forested
zones to lowland farmlands. Such galleries provide humid and shady microhabitats that
are indispensable for butterfly thermoregulation, especially when located within open low-
land landscapes. Increased humidity is known to be related to delayed summer drought and
desiccation of larval host plants and nectar sources in the Mediterranean, explaining its
positive impact on butterfly diversity (Stefanescu etal. 2004; Gutiérrez Illán etal. 2010).
Riparian habitats in particular can act as local refuges against heat and drought stress, lead-
ing to localised concentrations of butterfly species during the summer period in the Medi-
terranean basin (Galiano etal. 1985).
Although grasslands are considered the most important habitat for butterflies in Europe
(van Swaay etal. 2010; van Swaay etal. 2016), our study indicates otherwise. In Cyprus,
grasslands consist mainly of dry xerothermic plant communities, which presumably can
only provide adequate food resources to butterflies during the short springtime period, due
to extensive drought afterwards. As expected, forested areas were the poorest habitats for
butterflies, consisting mainly of semi-open pinewoods of Pinus brutia with understory veg-
etation, and Pinus nigra subsp. pallasiana at higher altitudes. However, we found that they
offer adequate habitats for several forest-dwelling species, such as the endemic H. syriaca
cypriaca, and the endemic H. cypriensis that migrates to the forest zone of higher altitudes
for the summer period of drought stress (John and Parker 2002). Their semi-open charac-
ter, which is more pronounced in mature forests, has been shown to support the endemic
butterfly fauna in other mountainous parts of Cyprus (Özden etal. 2008). We argue there-
fore that in spite of their poor species richness, special attention should be paid to forest
management for butterfly conservation, as semi-shaded habitats and forest edges can have
a positive effect on delaying host plant senescence (Weiss etal. 1988; Gutiérrez Illán etal.
2010), and on providing microhabitats for butterfly feeding and thermoregulation.
Environmental parameters
The number of flower heads was found to be the most important factor influencing the
composition of butterfly communities, positively affecting several species, and it was fur-
ther found to positively predict species richness and abundance, as well the species rich-
ness of endemic species. The number of flower heads is associated with the availability
of nectar resources for adult butterflies, enhancing butterfly species richness in a range
of ecosystems (Dempster 1997; Krauss etal. 2004; Kuussaari et al. 2007; Curtis etal.
2015), including in the Mediterranean Basin (Grill etal. 2005; Zografou etal. 2009; Serrat

Biodiversity and Conservation
1 3
etal. 2015). Nevertheless, not all flower heads are suitable nectar resources for butterflies
depending on the species preferences (Tudor etal. 2004). Since we did not sample plant
diversity, it is possible that butterfly diversity would be associated with plant diversity as
well (Blair and Launer, 1997; Ehrlich and Raven 1964) along with nectar plant species
abundance (Steffan-Dewenter and Tscharntke 2000), as the diversity of feeding resources
is known to support more pollinators, including butterflies (Kitahara 2015).
Soil humidity was also found to significantly shape butterfly community patterns, and
predict butterfly species richness and abundance. Soil humidity influences vegetation com-
position, which in its turn directly affects the distribution of individual butterfly species
(Gorbach etal. 2010) and hence butterfly community composition (van Halder etal. 2008).
Bare moist soils are known to enhance nectar production during the growing season, and
to offer further feeding resources through puddling (Murphy and Wilcox 1986; Real and
Rathcke 1991). Humid microhabitats are of particular importance for arid Mediterranean
ecosystems such as those in Cyprus, and the aridity stress for species is expected to further
increase due to climate change (Giannakopoulos etal. 2010; EEA 2017).
Butterfly abundance was negatively affected by increased tree cover, probably because
of the associated reduction in suitable open habitats containing resources such as nectar
plants, larval food plants and oviposition sites (Settele etal. 2009). Butterfly abundance
also showed a decreasing trend with increasing altitude, but different species had differ-
ent altitudinal preferences. For instance, the endemic species G. paphos has been sighted
from all altitudinal zones (John and Skule 2016), but in this study was mainly conducted at
lower altitudes. Finally, herb cover was not a significant predictor for butterfly species rich-
ness, abundance or endemic butterfly species richness, but it was positively associated with
specific species such as the endemic H. cypriensis (Fig.2) that selects herbs for ovipositing
(Makris 2003).
Natura 2000 forbuttery conservation
The Natura 2000 network in Cyprus has one of the most extensive national covers in the
EU (28.82%) (EC 2016) and was found to hold most butterfly species and all six endem-
ics. However, the network seems to under-perform for butterfly diversity conservation at
the local scale, when comparing localities within and outside it, as the transects sampled
within the network were poorer in terms of butterfly abundance and endemic butterfly spe-
cies richness and abundance. Contrary to Cyprus, in most European countries the network
seems to maintain greater butterfly species richness than the areas outside it, having also
a large share of endemic and annexed species (Verovnik etal. 2011; Pellissier etal. 2014;
van der Sluis etal. 2016). In line with our results, the network has been found to have a low
effectiveness for protecting vascular plants in Cyprus (Christodoulou etal. 2018). We also
found that the priority habitats, protected under the Habitats Directive, held poorer but-
terfly and endemic butterfly communities than the other habitats. In the case of the Med-
iterranean island ecosystems of Cyprus, priority habitat types do not appear to function
as surrogates for other components of biodiversity, such as butterflies. This was however
expected, as priority habitats in Cyprus included mainly forest habitats, which were found
to be particularly species-poor for butterflies.

Biodiversity and Conservation
1 3
Conservation management
The conservation value of the island of Cyprus in terms of butterflies is quite high in com-
parison with other Mediterranean islands (Özden etal. 2008), hosting three endemic spe-
cies and three endemic subspecies. Nevertheless, no butterfly species of Cyprus merits
legally protected status under European legislation (92/43/EEC), underlining the issue of
the incomplete inventory of invertebrates in the relevant Annexes (Cardoso 2012; D’Amen
etal. 2013). Given the underperformance of the Natura 2000 network for Cyprus butterfly
conservation at a local scale, we stress the need for new complementary European and
national policies targeting the conservation of endemics and red listed species. Our find-
ings indicate more precisely that such reassessment of protected areas for Cyprus or similar
Mediterranean environments would benefit from the inclusion of rural mosaics and further
inclusion of riverine habitats.
The lack of basic knowledge about their conservation status makes a red list of Cyprus
butterflies indispensable at the national scale. Red listing is needed to assess the impact of
major underlying threats such as climate change or land abandonment and intensification
on the Cyprus butterfly fauna (van Swaay etal. 2010; Bonelli etal. 2018), and to determine
the conservation status of endemics and several Least Concern species at a European level
that were found to be scarce in our study. Furthermore, Cyprus would benefit from estab-
lishing a coordinated butterfly monitoring scheme, to gain knowledge of butterfly popula-
tion trends, and thus to guide ongoing conservation and to contribute internationally in the
context of the European grassland butterfly indicator (EEA 2012; van Swaay etal. 2016).
In the applied management context, our results showed that rural mosaics, recognized as
the backbone of the High Nature Value (HNV) farmlands of the island (Zomeni etal. 2018)
are indeed a conservation priority for Cyprus butterflies. We argue that maintenance of
structural heterogeneity in agricultural lands should be a distinct target integrated into the
Common Agricultural Policy (Pe’er etal. 2014; Zakkak etal. 2014), for the conservation
of butterflies and other biodiversity components. In the absence of a coherent protected
area network, HNV farmlands can complement existing Natura 2000 sites (see Sigura etal.
2010; Klimek etal. 2014) in achieving a functional countryside with short but also long
term benefits for biodiversity (Vos etal. 2008).
Our results also underline the need to preserve riverine vegetation habitats and moist
microhabitats for butterfly species conservation, in particular in the context of climate
change scenarios for Cyprus (Hadjinicolaou etal. 2011). Sustainable water management
should therefore be a key priority of the Cyprus Strategy for Adaptation to Climate Change
(DE 2017) and is also relevant for the Cyprus Biodiversity Strategy (Kadis etal. 2012).
Finally, our results pinpoint the need for a butterfly-oriented forest management towards
opening forests (Arany 2013), given their importance for specific endemics.
Acknowledgements The authors would like to thank the Department of Forests and the Department of
Environment for providing information on the study area and enabling field work. Also we are thankful to
George Michaelides for assistance during fieldwork and Angeliki Martinou for providing useful comments
on statistical analysis. This research was funded by the Research Promotion Foundation of Cyprus (Protocol
No. PENEK/0609/34), with co-funding from the European Union’s Structural Funds.
Appendix
See Tables4, 5 and 6 and Fig.3.

Biodiversity and Conservation
1 3
Table 4 The number of transects (N) sampled per elevational zone and per habitat type
The cover (%) of each of the habitat types is presented according to the Corine land cover classes (2006)
comprising the overall study area shown in Fig.1
Zone 1 Zone 2 Zone 3 Zone 4 Total
< 500 m 501–100 m 1001–1500m > 1500 m
Habitat type N N N N Cover N
Agricultural areas 2 2 2 0 9.03 6
Rural mosaics 2 3 3 0 10.03 8
Forests 3 4 4 10 44.29 21
Grasslands 2 0 0 2 0.30 4
Sclerophyllous vegetation 2 2 4 1 16.47 9
Transitional woodland/shrubland 2 2 0 0 3.45 4
Riverine vegetation 2 2 2 2 n/a 8
Total N and area cover (km2) 15 15 15 15 1400 60

Biodiversity and Conservation
1 3
Table 5 Description of the 60 transects sampled
N Habitat type Corine code Natura Habitat code Environmental parameters
Alt (m) Sl Temp (oC) SΗum (%) T HSh LSh He F R B
1Agricultural areas
Comprised by non-irrigated arable land (241) with Ceratonia
siliqua, permanently irrigated land (212) with annual crops
and vineyards (221)
241 – – 271 1 25 28 3 0 2 3 1 1 1
2 212 – – 53 1 27 20 4 0 0 2 1 1 3
3 221 – – 860 1 24 33 0 4 0 2 1 1 3
4 221 – – 774 1 26 29 0 4 0 1 1 1 3
5 221 CY2000009 – 1152 2 26 34 4 0 1 1 1 1 3
6 221 – – 1007 1 24 32 0 4 0 1 2 1 3
7Rural Mosaics
Comprised by land principally occupied by agriculture, with
significant areas of natural vegetation (243) mainly includ-
ing vineyards and Mediterranean maquis, while complex
cultivation patterns (242) included orchards, vineyards and
abandoned cultivations
243 – – 68 1 27 28 1 2 3 2 2 1 1
8 243 – – 222 1 30 20 1 1 3 2 3 2 2
9 243 – – 612 1 27 34 0 3 2 2 2 1 2
10 243 – – 852 1 26 36 1 0 4 2 3 1 1
11 242 – – 518 1 25 49 2 1 3 2 4 2 1
12 242 – – 1187 1 27 33 0 4 0 1 1 1 3
13 243 – – 1246 2 22 28 1 3 1 2 3 1 2
14 242 – – 1390 1 24 39 0 1 4 2 4 2 0
15 Forests
Coniferous forest with Pinus brutia (<1500 m) (312/9540) and
Pinus nigra subsp. pallasiana forest (9536*). Also including
Juniperus foetidissima woods (9563*) at higher altitudes
311 – – 11 1 26 33 3 2 1 2 1 1 0
16 312 – – 455 2 26 26 3 0 2 2 2 1 2
17 312 – – 449 3 27 20 4 0 1 2 1 2 1
18 312 – – 823 2 28 25 3 2 3 1 2 2 1
19 312 – – 856 2 28 21 3 2 3 2 3 2 0
20 312 CY5000001 9540 643 2 25 31 0 1 2 3 2 2 1
21 312 CY5000001 9540 833 2 28 31 0 2 2 3 2 2 1
22 312 CY5000004 9536* 1298 2 23 23 4 1 2 2 2 1 1
23 312 – – 1141 2 25 23 3 2 2 2 1 2 1
24 312 CY5000004 9536* 1364 3 25 20 4 4 0 3 2 2 0
25 312 CY5000004 9536* 1570 2 20 22 2 1 3 2 2 2 2

Biodiversity and Conservation
1 3
Table 5 (continued)
N Habitat type Corine code Natura Habitat code Environmental parameters
Alt (m) Sl Temp (oC) SΗum (%) T HSh LSh He F R B
26 312 CY5000004 9536* 1696 1 18 23 3 0 1 3 1 2 0
27 312 CY5000004 9536* 1595 2 21 22 3 1 2 2 1 2 1
28 312 CY5000004 9536* 1834 2 18 23 3 1 2 2 1 2 1
29 312 CY5000004 9536* 1866 2 17 25 2 1 2 3 2 2 1
30 312 CY5000004 9536* 1621 1 18 25 2 1 3 2 1 2 1
31 312 CY5000004 9536* 1632 1 20 31 2 1 2 3 1 3 1
32 312 CY5000004 9563* 1737 1 17 24 0 0 2 3 1 2 2
33 312 CY5000004 9563* 1922 2 17 22 0 2 3 2 1 2 1
34 312 CY5000004 9563* 1743 3 17 25 0 3 2 2 1 2 1
35 3121CY5000004 9540 1163 2 24 20 3 2 2 2 2 2 1
36 Grasslands
Dry xerophytic grasslands with annual vegetation and grasses in
the lowlands (CY14) and wet grasslands (6460)
3212CY5000001 CY14598 2 26 33 0 0 2 4 2 1 0
37 3213CY5000005 CY145284 2 27 31 0 1 2 4 2 1 0
38 312 CY5000004 6460 1646 1 21 48 2 0 2 4 1 1 1
39 312 CY5000004 6460 1583 1 18 43 2 0 2 4 1 1 1
40 Sclerophyllous vegetation
Mainly comprised by the endemic Quercus alnifolia scrub com-
munities (9390*). Other areas include arborescent matorral
with Juniperus phoenica
323 – – 34 1 28 26 0 2 3 2 1 2 2
41 323 – – 109 1 29 28 1 2 3 2 2 2 1
42 323 CY2000005 9390* 876 2 24 23 0 3 2 3 2 1 1
43 323 – – 506 2 28 20 0 0 4 2 2 2 1
44 3231CY5000004 9390* 1027 1 25 35 2 3 2 2 1 2 0
45 3234CY2000005 9390* 1444 2 23 28 0 3 2 2 2 2 1
46 323 CY2000005 9390* 1324 2 23 28 0 3 2 2 2 2 1
47 3234CY2000005 9390* 1136 1 24 28 0 3 2 2 1 1 0
48 3234CY2000005 9390* 1584 1 21 29 0 2 3 2 3 2 2

Biodiversity and Conservation
1 3
Table 5 (continued)
N Habitat type Corine code Natura Habitat code Environmental parameters
Alt (m) Sl Temp (oC) SΗum (%) T HSh LSh He F R B
49 Transitional woodland/shrubland
Open forest areas with low shrub communities of Cistus spp.
and maquis with scattered Pinus trees
324 – – 411 2 26 32 2 2 3 2 2 1 1
50 324 – – 432 1 27 20 2 2 3 3 1 1 1
51 324 – – 766 2 24 27 3 0 3 2 2 1 1
52 324 – – 647 2 27 21 2 2 2 2 2 2 2
53 Riverine vegetation6
River valleys characterized by the presence of Platanus orienta-
lis and Alnus orientalis (92C0). Nerium oleander communi-
ties found in lower elevations along with Tamarix spp.
311 – – 364 1 25 44 2 1 3 2 3 2 0
54 311 – – 115 1 27 41 0 0 2 5 2 1 0
55 312 CY2000012 92C0 638 1 26 59 2 3 2 2 2 2 2
56 323 CY2000005 92C0 561 2 25 52 4 2 1 2 2 1 1
57 312 CY2000005 92C0 1024 2 22 53 2 2 2 2 1 2 1
58 312 CY2000005 92C0 1027 2 23 50 3 2 1 2 2 2 1
59 312 CY5000004 92C0 1568 1 17 58 1 1 3 2 1 2 1
60 312 CY5000004 92C0 1502 2 21 40 4 1 2 2 1 2 1
Corine code: code according to Corine Land Cover classes (2006). Habitat code: code according to Annex I of the Directive 92/43/EEC, *priority habitat type. Environ-
mental Parameters: Alt altitude, Sl slope (1: 0–10%, 2: 11–20%, 3: 21–35%, 4: > 35%), Temp air temperature, SHum soil humidity, T trees (> 2.5 m height), HSh high shrubs
(0.5–2.5m height), LSh: low shrubs (< 0.5 m height), He herb cover (1: 1–5%, 2: 6–25%, 3: 26–50%, 4: 51–75%, 5: > 75%), F flower heads (1: < 10, 2: 11–20, 3: 21–50, 4:
51–100, 5: 101–200, 6: 201–300, 7: 301–400, 8: 401–500, 9: 500–600, 10: > 600), R rocks, B bare ground
Priority habitat marked withasterisk (*): 9390: Scrub and low forest of Quercus alnifolia, 9536: Forests of Pinus nigra subsp. pallasiana, 9563: Forests of Juniperus foetidis-
sima. The following deficiencies were recorded in Corine and Natura mapping: 1mapped as 324, 2mapped as 242, 3mapped as 323, 4mapped as 312, 5mapped as 6220*, 6river-
ine vegetation is not mapped in Corine

Biodiversity and Conservation
1 3
Table 6 Results of SIMPER
analysis for priority and non-
priority habitat sites
Cont. (%) the average contribution of each species to the communities’
differentiation, Cum. (%) the cumulative contribution of each species
Code Species Cont (%) Cum (%)
Acard Anthocharis cardamines 0.34 98.15
Aacam Apharitis acamas 0.01 100.00
Acrat Aporia crataegi 0.68 93.78
Apand Argynnis pandora 0.18 99.08
Aages Aricia agestis 1.05 90.69
Calce Carcharodus alceae 0.73 92.84
Cargi Celastrina argiolus 2.26 78.54
Cbril Chazara briseis larnacana 0.82 91.83
Ctroc Chilades trochylus 0.58 95.54
Ccroc Colias crocea 5.97 30.04
Gpumi Gegenes pumilio 0.02 99.98
Gpaph Glaucopsyche paphos 3.46 55.71
Gcleo Gonopteryx cleopatra 5.01 45.11
Hcypr Hipparchia cypriensis 3.17 64.55
Hsyrc Hipparchia syriaca cypriaca 0.57 96.33
Hlupi Hyponephele lupina 0.68 94.73
Kroxe Kirinia roxelana 0.09 99.82
Lboet Lampides boeticus 2.87 68.53
Lmaer Lasiommata maera 0.22 98.82
Lmege Lasiommata megera 0.54 97.08
Lpiri Leptotes pirithous 0.04 99.96
Lredu Limenitis reducta 0.26 98.51
Lphla Lycaena phlaeas 2.10 84.41
Lther Lycaena thersamon 0.16 99.30
Mcypr Maniola cypricola 2.49 71.98
Pmach Papilio machaon 1.63 89.24
Paege Pararge aegeria 2.47 75.41
Pthra Pelopidas thrax 0.14 99.69
Pbras Pieris brassicae 6.51 21.76
Prapa Pieris rapae 9.18 12.73
Picar Polyommatus icarus 5.86 38.16
Pdapl Pontia daplidice 4.18 50.90
Panth Pseudochazara anthelea 1.85 86.98
Tacte Thymelicus acteon 2.13 81.50
Vatal Vanessa atalanta 0.06 99.91
Vcard Vanessa cardui 3.21 60.16
Zcerc Zerynthia cerisyi cypria 0.43 97.67
Zkars Zizeeria karsandra 0.15 99.50
Fig. 3 Plots of Z-scores for independent contributions, from 100 randomizations of the data vector, for
explanatory predictor variables of a Butterfly species richness, b Endemic butterfly species richness and c
Butterfly abundance. Z-scores above the reference line of 1.65 indicate significance at the 95% confidence
level (Mac Nally 2002)
▸

Biodiversity and Conservation
1 3
(a)
Butterfly species richness
(b) Endemic butterfly species richness
-2
2
6
10
14
FSHumT LShHSh AltRHe Sl B
Z-score
(c) Butterfly abundance
-1
0
1
2
3
FSHumT He BLSh HShR Sl Alt
Z-score
-2
0
2
4
6
8
10
12
14
16
18
FAlt Shum TLSh RHSh Sl BHe
Z-score

Biodiversity and Conservation
1 3
References
Arany I (2013) Four Highlights from Natura 2000 Forests, NGO recommendations on management. CEE-
web for Biodiversity, Budapest
Blair RB, Launer AE (1997) Butterfly diversity and human land use: species assemblages along an urban
grandient. Biol Conserv 80:113–125. https ://doi.org/10.1016/S0006 -3207(96)00056 -0
Bonelli S etal (2018) The first red list of Italian butterflies. Insect Conserv Divers 11:506–521. https ://doi.
org/10.1111/icad.12293
Cardoso P (2012) Habitats directive species lists: urgent need of revision. Insect Conserv Divers 5:169–174.
https ://doi.org/10.1111/j.1752-4598.2011.00140 .x
Cardoso P, Erwin TL, Borges PAV, New TR (2011) The seven impediments in invertebrate conservation and
how to overcome them. Biol Conserv 144:2647–2655. https ://doi.org/10.1016/j.bioco n.2011.07.024
Christodoulou SC, Grifffiths GH, Vogiatzakis IN (2018) Using threatened plant species to identify conser-
vation gaps and opportunities on the island of Cyprus. Biodivers Conserv 27:2837–2858. https ://doi.
org/10.1007/s1053 1-018-1572-4
Clarke K (1993) Non-parametric multivariate analysis of changes in community structure. Aust J Ecol
18:117–143. https ://doi.org/10.1111/j.1442-9993.1993.tb004 38.x
Colwell R (2013) EstimateS 9.1: Statistical estimation of species richness and shared species from samples
Curtis RJ, Brereton TM, Dennis RLH, Carbone C, Isaac NJB (2015) Butterfly abundance is deter-
mined by food availability and is mediated by species traits. J Appl Ecol 52:1676–1684. https ://doi.
org/10.1111/1365-2664.12523
D’Amen M, Bombi P, Campanaro A, Zapponi L, Bologna MA, Mason F (2013) Protected areas and insect
conservation: questioning the effectiveness of Natura 2000 network for saproxylic beetles in Italy.
Anim Conserv 16:370–378. https ://doi.org/10.1111/acv.12016
Delipetrou P, Makhzoumi J, Dimopoulos P, Georghiou K (2008) Cyprus. In: Vogiatzakis I, Pungetti G,
Mannion AM (eds) Mediterranean island landscapes: natural and cultural approaches. Springer,
Dordrecht, pp 170–203
Dempster JP (1997) The role of larval food resources and adult movement in the population dynamics of the
orange-tip butterfly (Anthocharis cardamines). Oecologia 111:549–556. https ://doi.org/10.1007/s0044
20050 270
Development and Environment (2017) Cyprus strategy for adaptation to climate change. Department of
Environment, Ministry of Agriculture, Rural Development and Environment, Nicosia
DF, FAO (2005) Forestry outlook study for Cyprus: opportunities and challenges towards 2020. Department
of Forest, Ministry of Agriculture, Rural Development and Environment, Nicosia
EC (2016) Natura 2000 Barometer. European Commission. http://ec.europ a.eu/envir onmen t/natur e/natur
a2000 /barom eter/index _en.htm. Accessed 12 Feb 2018
EEA (2010) Corine Land Cover 2006. European Environment Agency. https ://www.eea.europ a.eu/data-and-
maps/data/clc-2006-vecto r-data-versi on. Accessed 10 Jan 2011
EEA (2012) Streamlining European biodiversity indicators 2020: Building a future on lessons learnt
from the SEBI 2010 process. Technical report, No 11/2012. European Environment Agency,
Luxembourg
EEA (2017) Climate change, impacts and vulnerability in Europe 2016: An indicator-based report.
Report, No 1/2017. European Environment Agency, Luxembourg
Ehrlich PR, Raven PH (1964) Butterflies and plants: a study in coevolution. Evolution 18:586–608. https
://doi.org/10.1111/j.1558-5646.1964.tb016 74.x
EIONET (2018) Reference Portal for Natura 2000. Eionet Central Data Repository. https ://bd.eione
t.europ a.eu/activ ities /Natur a_2000/refer ence_porta l. Accessed 16 Feb 2018
ESRI (2010) ArcGIS 10. Environmental Systems Research Institute, Redlands
Flick T, Feagan S, Fahrig L (2012) Effects of landscape structure on butterfly species richness and abun-
dance in agricultural landscapes in eastern Ontario, Canada. Agric Ecosyst Environ 156:123–133.
https ://doi.org/10.1016/j.agee.2012.05.006
Fox J, Weisberg S (2011) An R companion to applied regression, 2nd edn. SAGE, California
Galiano EF, Sterling A, Viejo JL (1985) The role of riparian forests in the conservation of butterflies in
a Mediterranean Area. Environ Conserv 12:361–362. https ://doi.org/10.1017/S0376 89290 00345 12
Geri F, Amici V, Rocchini D (2010) Human activity impact on the heterogeneity of a Mediterranean
landscape. Appl Geogr 30:370–379. https ://doi.org/10.1016/j.apgeo g.2009.10.006
Giannakopoulos C, Hadjinicolaou P, Kostopoulou E, Varotsos KV, Zerefos C (2010) Precipitation and
temperature regime over Cyprus as a result of global climate change. Adv Geosci 23:17–24. https ://
doi.org/10.5194/adgeo -23-17-2010

Biodiversity and Conservation
1 3
Gorbach VV, Saarinen K, Reznichenko ES (2010) On the ecology of the poplar admiral (Limenitis pop-
uli, Lepidoptera, Nymphalidae) in Eastern Fennoscandia. Entomol Rev 90:989–998. https ://doi.
org/10.1134/S0013 87381 00800 38
Grill A, Cleary FRD (2003) Diversity patterns in butterfly communities of the Greek nature reserve
Dadia. Biol Conserv 114:427–436. https ://doi.org/10.1016/S0006 -3207(03)00070 -3
Grill A, Knoflach B, Cleary DFR, Kati V (2005) Butterfly, spider, and plant communities in different
land-use types in Sardinia, Italy. Biodivers Conserv 14:1281–1300. https ://doi.org/10.1007/s1053
1-004-1661-4
Gutiérrez Illán J, Gutiérrez D, Wilson RJ (2010) Fine-scale determinants of butterfly species rich-
ness and composition in a mountain region. J Biogeogr 37:1706–1720. https ://doi.org/10.111
1/j.1365-2699.2010.02314 .x
Habel JC, Seibold S, Ulrich W, Schmitt T (2018) Seasonality overrides differences in butterfly species
composition between natural and anthropogenic forest habitats. Anim Conserv 21:405–413. https ://
doi.org/10.1111/acv.12408
Hadjinicolaou P, Giannakopoulos C, Zerefos C, Lange MA, Pashiardis S, Lelieveld J (2011) Mid-21st
century climate and weather extremes in Cyprus as projected by six regional climate models. Reg
Environ Change 11:441–457. https ://doi.org/10.1007/s1011 3-010-0153-1
Herrando S etal (2015) Assessing impacts of land abandonment on Mediterranean biodiversity using
indicators based on bird and butterfly monitoring data. Environ Conserv 43:69–78. https ://doi.
org/10.1017/S0376 89291 50002 60
Hewitt GM (2011) Mediterranean Peninsulas: the evolution of hotspots. In: Zachos FE, Habel JC (eds)
Biodiversity hotspots: distribution and protection of conservation priority areas. Springer, Berlin,
pp 123–147
Ieronymidou C (2012) Avian land-use associations in the Eastern Mediterranean. Dissertation, Univer-
sity of East Anglia
John E, Parker R (2002) Dispersal of Hipparchia cypriensis (Holik, 1949) (Lepidoptera:Nymphalidae,
Satyridae) in Cyprus, with notes on its ecology and life-history. Entomol Gaz 53:3–18
John E, Skule B (2016) Lepidoptera. In: Sparrow DJ, John E (eds) An introduction to the wildlife of
Cyprus. Terra Cypria, Cyprus, pp 268–385
Kadis etal (2012) Strategy for biodiversity in Cyprus. Frederick University, Nicosia
Kati V, Hovardas T, Dieterich M, Ibisch PL, Mihok B, Selva N (2015) The challenge of implement-
ing the European network of protected areas Natura 2000. Conserv Biol 29:260–270. https ://doi.
org/10.1111/cobi.12366
Kitahara M (2015) Nectar plant species richness explains butterfly diversity in and around a primeval
woodland of Mount Fuji, central Japan. In: Peck RL (ed) Nectar: production, chemical composition
and benefits to animals and plants. Nova Science Publishers, New York, pp 77–98
Klimek S, Lohss G, Gabriel D (2014) Modelling the spatial distribution of species-rich farmland to
identify priority areas for conservation actions. Biol Conserv 174:65–74. https ://doi.org/10.1016/j.
bioco n.2014.03.019
Krauss J, Steffan-Dewenter I, Tscharntke T (2004) Landscape occupancy and local population size depends
on host plant distribution in the butterfly Cupido minimus. Biol Conserv 120:355–361. https ://doi.
org/10.1016/j.bioco n.2004.03.007
Kuussaari M, Heliölä J, Luoto M, Pöyry J (2007) Determinants of local species richness of diurnal Lepidop-
tera in boreal agricultural landscapes. Agric Ecosyst Environ 122:366–376. https ://doi.org/10.1016/j.
agee.2007.02.008
Mac Nally R (2002) Multiple regression and inference in ecology and conservation biology: further com-
ments on identifying important predictor variables. Biodivers Conserv 11:1397–1401. https ://doi.
org/10.1023/A:10162 50716 679
MacDonald D etal (2000) Agricultural abandonment in mountain areas of Europe: environmental conse-
quences and policy response. J Environ Manage 59:47–69. https ://doi.org/10.1006/jema.1999.0335
Maes D etal (2013) Not the right time to amend the annexes of the European habitats directive. Conser Lett
6:468–469. https ://doi.org/10.1111/conl.12030
Magurran AE (2004) Measuring biological diversity. Wiley-Blackwell, Oxford
Makris C (2003) Butterflies of Cyprus. Bank of Cyprus Cultural Foundation, Nicosia
Mammides C, Kadis C, Coulson T (2015) The effects of road networks and habitat heterogeneity on the spe-
cies richness of birds in Natura 2000 sites in Cyprus. Landsc Ecol 30:67–75. https ://doi.org/10.1007/
s1098 0-014-0100-5
Matteson KC, Langellotto GA (2010) Determinates of inner city butterfly and bee species richness. Urban
Ecosyst 13:333–347. https ://doi.org/10.1007/s1125 2-010-0122-y

Biodiversity and Conservation
1 3
Milieu, IEEP, ICF (2016) Evaluation study to support the fitness check of the birds and habitats directives.
European Commission, Brussels
Munguira ML (1995) Conservation of butterfly habitats and diversity in Meditteranean countries. In: Pullin
AS (ed) Ecology and conservation of butterflies. Chapman and Hall, London, pp 277–287
Murphy DD, Wilcox BA (1986) Butterfly diversity in natural habitat fragments: a test of the validity of
vertebrate-based management. In: Verner J, Morrison ML, Ralph CJ (eds) Wildlife 2000, modeling
habitat relationships of terrestrial vertebrates. University of Wisconsin Press, Madison, pp 287–292
Numa C etal (2016) The status and distribution of Mediterranean butterflies. IUCN, Malaga
Oksanen J, Blanchet FG, Friendly M etal (2017) Vegan: community ecology package. R package version
2:4–4
Olea PP, Mateo-Tomás P, de Frutos Á (2010) Estimating and modelling bias of the hierarchical partition-
ing public-domain software: implications in environmental management and conservation. PLoS ONE
5:e11698. https ://doi.org/10.1371/journ al.pone.00116 98
Orlikowska EH, Roberge J-M, Blicharska M, Mikusiński G (2016) Gaps in ecological research on the
world’s largest internationally coordinated network of protected areas: a review of Natura 2000. Biol
Conserv 200:216–227. https ://doi.org/10.1016/j.bioco n.2016.06.015
Özden Ö, Ciesla WM, Fuller WJ, Hodgson DJ (2008) Butterfly diversity in Mediterranean islands and
in Pentadaktylos Pinus brutia forests of Cyprus. Biodivers Conserv 17:2821–2832. https ://doi.
org/10.1007/s1053 1-008-9382-8
Pe’er G et al (2014) EU agricultural reform fails on biodiversity. Science 344:1090–1092. https ://doi.
org/10.1126/scien ce.12534 25
Pellissier V, Schmucki RF, Jiguet R, Julliard J, Touroult RD, Evans D (2014) The impact of Natura 2000
on non-target species, assessment using volunteer-based biodiversity monitoring. Technical paper No
4/2014. European Topic Centre on Biological Diversity, Paris
Plexida SG, Sfougaris AI, Papadopoulos NT (2012) Quantifying beetle and bird diversity in a Mediterra-
nean Mountain Agro-Ecosystem. Isr J Ecol Evol 58:1–25. https ://doi.org/10.1560/IJEE.58.1.1
Pollard E, Yates TJ (1993) Monitoring butterflies for ecology and conservation. Chapman & Hall, London
Puig-Montserrata X, Stefanescu C, Torre I, Palet J, Fàbregas E, Dantart J, Arrizabalaga A, Flaquer C (2017)
Effects of organic and conventional crop management on vineyard biodiversity. Agric Ecosyst Environ
243:19–26. https ://doi.org/10.1016/j.agee.2017.04.005
R Core Team (2016) R: A Language and Environment for Statistical Computing. R Foundation for Sta-
tistical Computing, Vienna, Austria
Real LA, Rathcke BJ (1991) Individual variation in nectar production and its effect on fitness in Kalmia
latifolia. Ecology 72:149–155. https ://doi.org/10.2307/19389 10
Rosso A etal (2017) Effectiveness of the Natura 2000 network in protecting Iberian endemic fauna. Anim
Conserv 21:262–271. https ://doi.org/10.1111/acv.12387
Samways MJ, McGeoch MA, New TR (2010) Insect conservation: a handbook of approaches and methods.
Oxford University Press, New York
Serrat A etal (2015) Environmental factors influencing butterfly abundance after a severe wildfire in Medi-
terranean vegetation. Anim Biodivers Conserv 38:207–220
Settele J, Shreeve T, Konvicka M, van Dyck H (2009) Ecology of butterflies in Europe. Cambridge Univer-
sity Press, Cambridge
Sigura M, Peccol E, Piani L (2010) High nature value farmland (HNVF) and ecological networks: their
role in the sustainability of trans-border regions. disP 46:60–68. https ://doi.org/10.1080/02513
625.2010.10557 112
Slancarova J etal (2016) Life history traits reflect changes in Mediterranean butterfly communities due
to forest encroachment. PLoS ONE 11:e0152026. https ://doi.org/10.1371/journ al.pone.01520 26
Stefanescu C, Herrando S, Páramo F (2004) Butterfly species richness in the north-west Mediterranean
Basin: the role of natural and human-induced factors. J Biogeogr 31:905–915. https ://doi.org/10.11
11/j.1365-2699.2004.01088 .x
Steffan-Dewenter I, Tscharntke T (2000) Butterfly community structure in fragmented habitats. Ecol
Lett 3:449–456. https ://doi.org/10.1111/j.1461-0248.2000.00175 .x
Thomas JA (2005) Monitoring change in the abundance and distribution of insects using butterflies and
other indicator groups. Philos Trans R Soc B 360:339–357. https ://doi.org/10.1098/rstb.2004.1585
Tolman T, Lewington R (1997) Butterflies of Britain and Europe. Princeton University Press, Princeton
Trochet A, Schmeller DS (2013) Effectiveness of the Natura 2000 network to cover threatened species.
Nat Conserv 4:35–53. https ://doi.org/10.3897/natur econs ervat ion.4.3626
Tsintides T, Hadjikyriakos G, Christodoulou CS (2002) Trees and Shrubs in Cyprus. Foundation A.G.
Leventis, Cyprus Forest Association, Nicosia

Biodiversity and Conservation
1 3
Tsintides T, Christodoulou CS, Delipetrou P, Georghiou K (2007) The red data book of the flora of
Cyprus. Cyprus Forestry Association, Nicosia
Tudor O, Dennis RLH, Greatorex-Davies JN, Sparks TH (2004) Flower preferences of woodland butter-
flies in the UK: nectaring specialists are species of conservation concern. Biol Conserv 119:397–
403. https ://doi.org/10.1016/j.bioco n.2004.01.002
van der Sluis T etal (2016) How much Biodiversity is in Natura 2000? The “Umbrella Effect” of the
European Natura 2000 protected area network. Alterra report 2730B. Alterra, Wageningen
van Halder I, Barbaro L, Corcket E, Jactel H (2008) Importance of semi-natural habitats for the conser-
vation of butterfly communities in landscapes dominated by pine plantations. Biodivers Conserv
17:1149–1169. https ://doi.org/10.1007/s1053 1-007-9264-5
van Swaay C etal (2016) The European Butterfly Indicator for Grassland species: 1990–2015. Report
VS2016.019, De Vlinderstichting, Wageningen
van Swaay C, Warren MS (1999) Red Data Book of European butterflies (Rhopalocera). Nature and
Environment, No. 99. Council of Europe Publishing, Strasbourg
van Swaay C etal (2010) European Red List of Butterfies. Publications Office of the European Union,
Luxembourg
van Swaay C etal (2011) Applying IUCN criteria to invertebrates: how red is the Red List of European
butterflies? Biol Conserv 144:470–478. https ://doi.org/10.1016/j.bioco n.2010.09.034
Venables WN, Ripley BD (2002) Modern Applied Statistics with S, 4th edn. Springer, New York
Verovnik R, Govedič M, Šalamun A (2011) Is the Natura 2000 network sufficient for conservation of
butterfly diversity? A case study in Slovenia. J Insect Conserv 15:345–350. https ://doi.org/10.1007/
s1084 1-010-9308-0
Vos CC et al (2008) Adapting landscapes to climate change: examples of climate-proof ecosys-
tem networks and priority adaptation zones. J Appl Ecol 45:1722–1731. https ://doi.org/10.111
1/j.1365-2664.2008.01569 .x
Walsh C, Mac Nally R (2013) hier.part: Hierarchical Partitioning. R package version 1.0-4
Warren MS etal (2001) Rapid responses of British butterflies to opposing forces of climate and habitat
change. Nature 414:65–69. https ://doi.org/10.1038/35102 054
Weiss SB, Murphy DD, White RR (1988) Sun, slope, and butterflies: topographic determinants of habi-
tat quality for Euphydryas Editha. Ecology 69:148–1496. https ://doi.org/10.2307/19416 46
Zakkak S, Chatzaki M, Karamalis N, Kati V (2014) Spiders in the context of agricultural land aban-
donment in Greek Mountains: species responses, community structure and the need to preserve
traditional agricultural landscapes. J Insect Conserv 18:599–611. https ://doi.org/10.1007/s1084
1-014-9663-3
Zakkak S, Radovic A, Nikolov SC, Shumka S, Kakalis L, Kati V (2015) Assessing the effect of agri-
cultural land abandonment on bird communities in southern-eastern Europe. J Environ Manage
164:171–179. https ://doi.org/10.1016/j.jenvm an.2015.09.005
Zografou K, Sfenthourakis S, Pullin A, Kati V (2009) On the surrogate value of red-listed butterflies for
butterflies and grasshoppers: a case study in Grammos site of Natura 2000, Greece. J Insect Con-
serv 13:505–514. https ://doi.org/10.1007/s1084 1-008-9198-6
Zografou K, Kati V, Grill A, Wilson RJ, Tzirkalli E, Pamperis LN, Halley JM (2014) Signals of climate
change in butterfly communities in a Mediterranean protected area. PLoS ONE 9:e87245. https ://doi.
org/10.1371/journ al.pone.00872 45
Zomeni M, Martinou A, Stavrinides M, Vogiatzakis IN (2018) High nature value farmlands in Cyprus:
identification, mapping and synergies with nature conservation. Nat Conserv
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.

Biodiversity and Conservation
1 3
Aliations
ElliTzirkalli1,2 · CostasKadis2,3· JohnM.Halley1· IoannisVogiatzakis4 ·
RobertJ.Wilson5 · KonstantinaZografou6 · AndreasAntoniou7· TakisTsintides8·
ChristodoulosMakris9· VassilikiKati1
1 Department ofBiological Applications andTechnology, University ofIoannina, 45110Ioannina,
Greece
2 Frederick Research Center, Filokyprou 7-9, Palouriotissa, 1036Nicosia, Cyprus
3 Nature Conservation Unit, Frederick University, P.O Box24729, 1303Nicosia, Cyprus
4 School ofPure andApplied Sciences, Open University ofCyprus, PO Box12794, 2252Nicosia,
Cyprus
5 Department ofBiogeography andGlobal Change, National Museum ofNatural Sciences (MNCN-
CSIC), 28006Madrid, Spain
6 Department ofBiology, Temple University, 1900 North 12th St., Philadelphia, PA19122, USA
7 Department ofEnvironment, Ministry ofAgriculture, Rural Development andEnvironment,
2414Nicosia, Cyprus
8 Department ofForest, Ministry ofAgriculture, Rural Development andEnvironment,
1414Nicosia, Cyprus
9 Lemesos, Cyprus