Quantifying the conservation value of Sacred Natural Sites

Full text

1
Quantifying the conservation value of Sacred Natural Sites
1
2
Avtzis DN
1
, Stara K
2
, Sgardeli V
3
, Betsis A
4
, Diamandis S
1
, Healey JR
5
, Kapsalis E
6
, Kati
3
V
2,6
, Korakis G
7
, Marini Govigli V
2,8
, Monokrousos N
2,9
, Muggia L
10
, Nitsiakos V
11
,
4
Papadatou E
12
, Papaioannou H
6
, Rohrer A
13
, Τsiakiris R
14
, van Houtan KS
15,16
, Vokou D
17
,
5
Wong J
18
, Halley, JM
2*
6
7
1
Forest Research Institute, Hellenic Agricultural Organization Demeter, 57006 Vassilika, Thessaloniki, Greece
8
2
Department of Biological Applications and Technology, University of Ioannina, 45110 Ioannina, Greece
9
3
Institute of Marine Biological Resources and Inland Waters (IMBRIW), Hellenic Centre for Marine Research,
10
Elliniko, P.C. 16604, Agios Κosmas, Attiki, Greece
11
4
Dodonis 13, 45221, Ioannina, Greece
12
5
School of Environment, Natural Resources and Geography, Bangor University, Bangor, Gwynedd, LL57
13
2UW, UK
14
6
Department of Environmental & Natural Resources Management, University of Patras, 30100 Agrinio, Greece
15
7
Department of Forestry and Management of the Environment and Natural Resources, Democritus University
16
of Thrace, 68200, Orestiada, Greece
17
8
European Forest Institute - Mediterranean Regional Office (EFIMED), St. Pau Art Nouveau Site – St. Leopold
18
Pavilion, St. Antoni Maria Claret 167, 08025 Barcelona, Spain
19
9
Institute of Soil and Water Resources, Hellenic Agricultural Organization Demeter, 14123 Athens, Greece
20
10
Institute of Life Sciences, University of Trieste, via Giorgieri 10, 34127 Trieste, Italy
21
11
University of Ioannina, Department of History and Archeology, Section of Folklore University campus 45110
22
Ioannina
23
12
Vernardou 14, 15235 Athens, Greece
24
13
Karl-Franzens University of Graz, Institute of Plant Science, Holteigasse 6, 8010 Graz, Austria
25
14
Department of Forest Administration and Management, Forestry Service of Ioannina, Marikas Kotopouli 62,
26
45445, Ioannina, Greece
27
15
Monterey Bay Aquarium, 886 Cannery Row, Monterey, CA 93940, United States
28
16
Nicholas School of the Environment, Duke University, Durham, North Carolina 27708, United States
29
17
Department of Ecology, School of Biology, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece
30
18
Wild Resources Ltd. Ynys Uchaf, Mynydd Llandygai, Bangor, Gwynedd LL57 4BZ, UK
31
32
*Correspondence jhalley@cc.uoi.gr, Telephone: +30-26510-07337
33
34
35
36
37

2
Abstract
37
Many have asserted that Sacred Natural Sites (SNS) play an important role in nature
38
protection but few have assessed their conservation effectiveness for different taxa. We
39
studied sacred groves in Epirus, NW Greece, where a large number of such SNS have been
40
identified. Based on historical, ethnographic and ecological criteria, we selected eight of
41
these groves and matching control sites and in them we studied fungi, lichens, herbaceous
42
plants, woody plants, nematodes, insects, bats and passerine birds. Our results reveal that the
43
contribution of SNS to species conservation is nuanced by taxon, vegetation type and
44
management history. We found that the sacred groves have a small conservation advantage
45
over the corresponding control sites. More specifically, there are more distinct sets of
46
organisms among sacred groves than among control sites, and overall biodiversity, diversity
47
per taxonomic group, and numbers of species from the European SCI list (Species of
48
Community Interest) are all marginally higher in them. Conservationists regard the often
49
small size of SNS as a factor limiting their conservation value. The sizes of SNS around the
50
globe vary greatly, from a few square meters to millions of hectares. Given that those
51
surveyed by us (ranging from 5 to 116 ha) are at the lower end of this spectrum, the small
52
conservation advantage that we testified becomes important. Our results provide clear
53
evidence that even small-size SNS have considerable conservation relevance; they would
54
contribute most to species conservation if incorporated in networks.
55
56
Keywords
57
Sacred Natural Sites; Conservation value; Biodiversity; Extinction Debt; Beta diversity
58
59
Research highlights
60
 Sacred Natural Sites (SNS) are thought to play an important role in conservation but
61
quantitative analyses are rare.
62
 We studied the conservation capacity of SNS at multiple sites for multiple taxonomic
63
groups.
64
 The SNS studied deliver a small but important conservation benefit compared with
65
corresponding control areas.
66
 The contribution of SNS to species conservation is nuanced by taxon, vegetation type and
67
management history.
68
 The best conservation strategy for small SNS is to join them as parts of networks within
69
conventional conservation schemes.
70
Abbreviations
71
Sacred Natural Sites: SNS
72
Species Abundance Relationships: SAR
73

3
1. Introduction
74
Conservation is closely aligned with modern ecological thinking and over the last two
75
centuries has become a major factor in policy decisions (Klein et al., 2009; Keppel et al.,
76
2015). Before the arrival of the modern ecology-motivated concept, conservation has been
77
practiced for many centuries in a variety of more traditional, community-based forms
78
(Malhotra et al., 2007). One such form was through social taboos and religious beliefs that
79
prescribed management regimes in sacred areas, often imposing limitations on certain
80
activities, so as to secure important resources and services for the whole community (Berkes
81
et al., 2000; Colding et al., 2001, Klepeis et al. 2016). These are the so-called sacred natural
82
sites (SNS) that not only reflect the religious and social needs of the community but at the
83
same time contribute important ecosystem services, from inspiration to air regulation, water
84
and micro-climate quality, or conservation of biological diversity (Jim, 2003; Soury et al.,
85
2007; Yuan and Liu, 2009; Wassie et al., 2010).
86
Sacred natural sites have been found in all inhabited continents (Hughes and Chandran, 1998)
87
and woodland sacred groves can be traced back to the time when human society was still in a
88
pre-agricultural state (Gadgil and Vartak, 1976). They have been associated with a wide
89
range of faiths and beliefs, socio-cultural systems, institutions and ritual practices, and may
90
be subject to changing conditions (Verschuuren et al., 2010). Around the Mediterranean
91
basin, forests have long been recognized as a resource with a multifunctional role that needs
92
particular care and protection. Groves or specific tree species, related mainly to sacrifice and
93
burial, were considered as sacred and thus gained a special protection status (Blondel and
94
Aronson, 1999). This was normally achieved through restrictions imposed by a local
95
authority, usually a religious authority, threatening transgressors with supernatural
96
consequences (Byers et al., 2001; Virtanen, 2002). At the same time, extended sacred forests
97
served as a protective levee for the local community against natural disasters, such as
98
landslides and floods (Stara et al., 2016). Sacred groves had flourished in Greece, since the
99
Ottoman period, mainly in the mountainous regions, where the above-mentioned natural
100
threats to local communities were much more severe and where historical circumstances
101
allowed the involvement of the Church in their management.
102
Epirus is a mountainous region in northwestern Greece, in which sacred groves are a
103
prominent component of the landscape; they form habitats dominated by mature trees that are
104
unique within the historically intensively used landscapes (Stara et al., 2015; Stara et al.,
105

4
2016). These groves were established through a range of ritual praxes. Some were dedicated
106
to specific saints, some were little more than community agreements, while others were
107
protected by the threat of excommunication. Different management regimes prevailed
108
through time with some groves being strictly protected, some subjected to controlled
109
management, whereas for others only the protection of mature trees is reported. The groves
110
appear either in the form of protective forests above or close to villages or as groups of
111
veteran trees that accompany outlying churches or icon stands (Stewart, 1993; Nixon, 2006;
112
See also Appendix G). Nonetheless, they served in many cases as multifunctional forests for
113
local communities providing among others shaded grazing areas for livestock. Especially in
114
deciduous sacred forests, grazing could be intensive (Papanastasis et al., 2008).
115
Different cultural groups coexisted in Epirus contributing to the variability of the landscape,
116
but they were all associated with sacred groves. Long-term ethnographic research has
117
revealed that of the 80 villages in the mountainous municipalities of Zagori and Konitsa
118
almost all had at least one sacred grove; these groves mostly lie within a narrow range of
119
elevation, typically from 800 to 1200 m (Stara et al., 2016). This is also the zone where most
120
mountain settlements, characterized by a mixed system of agriculture-animal husbandry, have
121
developed historically (Nitsiakos, 2016).
122
Even though the role of SNS in the conservation of biodiversity has long been recognized
123
(Kosambi, 1962; Gadgil and Vartak, 1976; Haridasan and Rao, 1985), they have recently
124
gained more attention amongst conservation biologists because of the many threats to
125
biodiversity due to anthropogenic activities (Pimm et al., 1995; Gao et al., 2013). It has been
126
suggested that incorporating these SNS into existing protected area networks might increase
127
their effectiveness in achieving conservation objectives (Bhagwat and Rutte, 2006; Soury et
128
al., 2007; Corrigan et al., 2013; Ormsby, 2013).
129
Despite the increasing interest in SNS as biodiversity refugia (Dudley et al., 2009), few
130
studies have assessed their effectiveness across taxa, whilst most have focused on specific
131
groups of organisms, such as plants (Boraiah et al., 2003; Khumbongmayum et al., 2006;
132
Frascaroli et al., 2016), small mammals (Decher, 1997; Reed and Carol, 2004) or butterflies
133
(Nganso et al., 2012). Most of these studies have been carried out in Asia, particularly India
134
and China (Nganso et al., 2012; Gao et al., 2013; Karthikeyan and Dhamatharan, 2015), or
135
Africa (Daye and Healey, 2015), with very little work in Europe (e.g. Frascaroli et al., 2016).
136
It could be argued that, as most SNS tend to be small, their relevance to conservation, though
137

5
tangible, is limited compared to large reserves (Bossart et al 2006, Aerts et al 2006). Area is
138
expected to affect the conservation effectiveness of SNS in several important ways. Firstly,
139
the species-area relationship indicates that smaller areas cannot support as many species as
140
larger ones. If a habitat shrinks, the level of biodiversity that it can sustain in the long term
141
also shrinks, but, in the short term, the habitat retains more species than it can support. This
142
surplus is called “extinction debt” (Diamond, 1972) and it must eventually be paid. The
143
process takes time, with the magnitude of the delay being greater in larger fragments (Halley
144
et al., 2016). Both the extinction debt and the time to the new equilibrium are also affected by
145
the degree of isolation and the habitability of the “matrix” (i.e. the area between fragments;
146
Koh and Ghazoul, 2010).
147
Focusing on a group of sacred groves in Epirus, the goal of this study is to investigate the
148
conservation effectiveness of SNS. We do this by assessing their biodiversity and comparing
149
them with matched control sites. For each sacred grove, a nearby woodland area without any
150
sacred status but with similar characteristics was chosen to serve as a control site. To achieve
151
a substantial breadth of studied organisms, eight different taxonomic groups were
152
investigated simultaneously. Estimates of diversity were assessed per taxonomic group and
153
per site. The importance of the size of the groves was also explicitly considered. In addition,
154
extensive ethnographic research highlighted the impact of different management practices on
155
the conservation status of these groves. The specific hypotheses that we are testing are as
156
follows: (I) sacred groves have a higher alpha-diversity than their control sites because they
157
enjoyed greater protection; (II) alpha-diversity differences will be accentuated for taxa, such
158
as fungi or lichens, that benefit from the presence of trees of great age; and (III) sacred groves
159
have higher beta-diversity than their control sites, since each sacred grove is expected to have
160
its own distinctive land-use history (and therefore forest structure).
161
2. Materials and Methods
162
2.1. Study Areas and Sampling
163
Numerous sacred groves have been identified in a wide area of north-western Greece (Fig. 1),
164
of which 22 were mapped. Of these, eight (1S-8S) were selected for the current study, based
165
on an integrated set of historical, ethnographic, management and ecological criteria
166
(Appendices A and G). Each of the selected sacred groves is situated in the mountainous
167
region of Zagori and Konitsa (Fig. 1). Since our main hypotheses are that sacred-grove status
168
involves higher biodiversity, for each grove we chose a single non-sacred site attempting an
169

6
assessment of biodiversity differences as practiced in other similar studies (Wortley et al.
170
2013, Derhé et al. 2016). We selected control sites (1C-8C) in close proximity; these
171
matched each sacred grove in terms of substrate, topographic position and type of vegetation.
172
In this study, we identified three types of groves in terms of vegetation: those dominated by
173
(i) coniferous, (ii) evergreen broadleaved or (iii) deciduous broadleaved trees. We sampled in
174
these eight pairs of sites over two consecutive years (2013 and 2014) following a sampling
175
protocol that was adapted to the unique characteristics of each taxonomic group (Appendix
176
B). The sampling effort was the same across all sites for any given taxonomic group, so that
177
estimates of biodiversity are comparable.
178
2.2. Dataset
179
In total, eight taxonomic groups (fungi, lichens, herbaceous plants, woody plants, nematodes,
180
insects, bats and passerine birds) were sampled in each sacred grove and the corresponding
181
control site. All observed organisms of these groups were identified to species level, except
182
for nematodes, which were identified to genus level. The data consist of abundance records
183
per species, except for lichens, herbaceous plants (including ferns) and woody plants, for
184
which only species presence was recorded.
185
2.3. Biodiversity analysis
186
The biodiversity we assess here is the total number of species recorded in each site, which we
187
call the species richness of the site.
188
2.3.1. Ordination
189
To visualize the difference in composition between sites, multidimensional scaling analysis
190
based on Bray-Curtis dissimilarity was conducted for each taxon, separately, and for all taxa
191
combined. This index is widely used as a measure of multidimensional “distance” between
192
samples for abundance data (e.g. Clarke et al., 2007; Birtel et al., 2015; Nicol et al., 2017); it
193
has the advantage, over some other ordination techniques, that differences in abundance are
194
scaled proportionally . The analysis was implemented in R 3.2.3 (R Core Team, 2015) using
195
function isoMDS of the MASS package (Venables and Ripley, 2002) and function vegdist of
196
the VEGAN package (Oksanen et al., 2016).
197
2.3.2. Species richness
198

7
Sacred groves and control sites were compared in terms of their species richness per site
199
(across all taxa), total species richness per taxon (across all sacred and all control sites) and
200
species richness per site per taxon.
201
Apart from their type (sacred or control), sites are characterized by their location within the
202
region of Epirus (Fig. 1), their vegetation (three forest types) and the area of the site (being
203
the area of the convex hull containing the sample plots within each site) (Table 1).
204
To investigate the effect of the different site characteristics on species richness, a generalized
205
linear regression model S ~ area + type + vegetation type + area:type with Poisson response
206
and a logarithmic link function was used. The model is applied to the total species richness
207
per site and to the species richness of each taxonomic group per site. In addition, we carried
208
out a number of tests (regression and paired t-test) comparing species richness in sacred sites
209
and control areas with and without conifer groves.
210
We also recorded the numbers of European SCI, Species of Community Interest (Official
211
Journal of the European Union, 2009; Council Directive, 1992), for all sacred groves and
212
corresponding control sites (Table E.1). We assessed the significance of the differences
213
between them using a paired Students t-test.
214
2.3.3. Beta diversity
215
Apart from the species richness per site (alpha diversity) and the species richness across sites
216
(gamma diversity), the sacred and control site communities were compared in terms of their
217
beta diversity or species turnover (Magurran, 2004). Beta diversity between the local scale
218
(sites) and the global scale (union of sites) was measured using Whittaker index and N* index
219
(Lazarina et al., 2013). Both indices give a measure of species turnover in space, which in
220
this case measures the difference in species composition between the local scale (site) and
221
global scale (the union of all sacred or all control sites). N* is roughly defined as the
222
sampling effort (number of samples) above which the samples accumulated will mostly
223
contain species that have already been found. The advantage of the N* index, as opposed to
224
other indices, is that it is independent of the sampling effort, provided that there are enough
225
samples for the index to be calculated (Lazarina et al 2013). The N* index was computed
226
using the R function provided by Lazarina et al (2013). We tested the significance of
227
differences between sacred groves and control sites at the 5% level.
228
All statistical tests and analyses were performed in R 3.2.3 (R Core Team, 2015).
229

8
2.3.4. Conservation capacity of SNS
230
By the term “conservation capacity” we refer to the ability of a protected area to conserve
231
biodiversity, assuming that management measures to protect the site are implemented.
232
Conservation capacity involves two components: the number of species that an area of a
233
given size can support at equilibrium, based on the species-area relationship (SAR, see for
234
example Halley et al., 2013), and the duration for which the area can retain species (if fully
235
protected). This is based on an estimation of the species relaxation curve for extinction debt
236
(Halley et al., 2016), a prominent factor in extinction ecology and conservation (Newmark et
237
al., 2017). Extinction debt becomes important when a fragment of habitat within a larger
238
habitat network connected by dispersal gets isolated, with no further dispersal possible.
239
Thereafter, the viability of each species is dependent on its population size within the
240
fragment so that current species richness may be a relic of earlier biodiversity levels rather
241
than true conservation capacity. The conservation capacity of the sacred groves was
242
estimated for each taxonomic group, separately, using the Arrhenius SAR:
243
…(1)
244
The constant z is typically between 0.2 and 0.3 for islands, while for continental areas it falls
245
within the range of 0.1 to 0.15 (Halley et al., 2013). Calibration of the SAR was achieved by
246
assuming a continental area with exponent 0.15; then c was determined by using the number
247
of species found in the control sites through the formula c = S/A
z
.
248
The first time-constant of relaxation is the expected time for half the extinction debt to be
249
paid off, which actually is the half-life of extinction debt in a habitat remnant. In the absence
250
of speciation and colonization, the half-life of extinction debt is equal to the time for species
251
richness to fall to half its original value. Based on the models developed in Halley et al.
252
(2016), this is approximately (in years):
253
50
0
2.77 A
tS
α
ρ
τ
 
≈
 
 
…(2) 254
Here, A is the area of the remnant forest,
ρ
is the typical total density of individuals of the
255
relevant taxonomic group,
τ
is the average generation time and S
0
is the initial number of
256
species in the area A at the time of area reduction or isolation. The factor
ρ
A/S
0
is important,
257
being the number of individuals per species. If the initial number of species, S
0
,
is not known,
258
the alternative is to use the SAR and substitute Eq. (1) for species number:
259
z
S cA
=

9
1
50
2.77
z
A
tc
α
ρ
τ
−
 
=
 
 
… (3) 260
In order to get
ρ
and
τ
, we assume a single average for each taxonomic group (Halley et al.,
261
2016). For passerine birds, herbaceous and woody plants,
ρ
and
τ
values are as in Halley et
262
al. (2016). For nematodes, our measurements indicated typical densities of 7.5
×
10
9
263
individuals per ha and we used a generation time of 19 days (Lee, 2002), while for bats we
264
used
ρ
=0.105 individuals per ha and for the generation time we used
τ
=8 years, which is half
265
the average longevity (Austad and Fischer, 1991). For insects, the value of
τ
=1 year was
266
typical of the species in our study, while
ρ
=7.83
×
10
4
individuals per ha that we used is
267
clearly a conservative number as it refers to ground-dwelling beetles (Didham et al., 1998).
268
We did not compute curves for lichens or fungi owing to known complications of defining
269
individuals and generation times for these groups.
270
2.3.5. SNS and National Parks (NP) size worldwide
271
To see how the size of the sacred groves that we studied fits into the global picture, using a
272
literature search, we assembled a database of SNS from various countries, for which we could
273
find the area (Table F.1) as well of National Parks in three countries: Greece, the United
274
Kingdom and the United States (Table F.2).
275
3. Results
276
In total, across all taxonomic groups studied, 816 species were observed and identified within
277
the eight pairs of sacred groves and control sites (Table C.1). There was great variability in
278
the species richness of the sacred sites relative to their respective control sites for different
279
taxonomic groups: in five of them, the total number of species observed was higher in the
280
sacred groves, and in three groups, it was higher in the control sites (Fig. 2a), but these
281
differences were not statistically significant except for fungi (p=0.001, see Table C.2), for
282
which richness was higher in sacred groves. Combining species across the taxonomic groups,
283
all except two localities had higher species richness in the sacred grove than the
284
corresponding control site (Fig. 2b). The two exceptions are localities 4 and 7 (Fig. 1) that are
285
associated with steeper slopes and are dominated by conifers. The other six pairs are
286
associated with the lowland or southern-aspect slopes and are dominated by broadleaved
287
trees. There is a strong correlation (Fig. 2b) between the species richness of the sacred groves
288
(x) and control sites (y) in each locality for the six pairs dominated by broadleaved trees,
289

10
reflecting the success of their matching in the sample design (y=0.727x+30.56, R
2
=0.912,
290
p=0.003). For these localities, there is also a significant difference between overall species
291
richness in the sacred groves and control sites (t-test, p=0.0085). These tests show a
292
consistent trend for greater overall species richness in the sacred groves than the control sites.
293
Ordination shows that the patterns of species composition amongst the three vegetation types
294
(Fig. D.1) varied by taxonomic group. However, with species of all groups combined, there
295
was a clear distinction between the vegetation types. Regarding the site type, there were no
296
consistent differences in composition between sacred groves and control sites for the
297
individual groups of species or for all species combined (Figs D.1 and D.2). The generalized
298
linear regression analysis shows (Table C.2) that the site area and type do not affect
299
significantly the total species richness per site (at a 5% significance level). However, their
300
interaction is significant meaning that the relationship between species richness and area
301
differs depending on the type of the site (sacred or control). As sacred sites are mostly
302
smaller in area than control sites (Table 1). The total species richness is also significantly
303
affected by vegetation type. On a taxonomic group level, the locality is not significant for any
304
group. The type of the site (sacred or control) is significant only for fungi, whereas vegetation
305
type is significant for lichens, herbaceous plants, and woody plants; none of these predictors
306
is significant for nematodes, insects, passerine birds or bats. The interaction between site
307
locality and type is also significant for herbaceous plants and lichens, as was also the case for
308
total species richness.
309
Of the 13 European SCI species that were encountered in the study area, more were found in
310
the sacred groves (eleven) than in their control sites (nine) especially for passerine birds (8
311
versus 4). However, overall the difference was not significant (paired t-test; p=0.30).
312
The Whittaker and N* indices of species turnover reveal significantly greater beta diversity
313
amongst the sacred groves than amongst the control sites (at the 5% level for both indices)
314
(Fig. 3). More specifically, beta diversity is greater in the sacred groves for five taxonomic
315
groups (lichens, herbaceous plants, woody plants, passerine birds and bats); it is slightly less
316
for insects, and very similar between the two site types for nematodes and fungi. Notably,
317
beta diversity is much lower for the nematodes than for all the other taxonomic groups of
318
species, presumably because nematodes were identified only to genus level and, hence, the
319
majority of nematode genera are found in all samples.
320

11
The area of the sacred groves was small, ranging from 4.9 ha to 115.7 ha with a median size
321
of 18.4 ha. Both the area and the taxonomic group are expected to affect the half-life of
322
species loss following habitat isolation (Fig. 4a) and, hence, their conservation capacity. The
323
predicted half-life varied greatly amongst taxonomic groups being low for bats and passerine
324
birds, under 100 years for most of the sacred groves, but very high, above 1000 years, for
325
nematodes and herbaceous plants (because of their large populations) and for woody plants
326
(because of large generation times)). However, the general linear modelling analysis did not
327
find a significant relationship between area and species richness.
328
In our literature search, we found 104 SNS for which the area was recorded or could easily be
329
inferred; these occur in all inhabited continents. To these we added the 22 sacred groves in
330
Epirus that we mapped, including the 8 whose biodiversity we studied in detail. The
331
histogram for this ensemble (Fig 4a) shows that the size of SNS varies greatly, ranging from
332
a few square metres to over 100,000 km
2
, with the groves that we studied falling in the
333
smaller part of the range. By contrast, National Parks are always at least 10 km
2
(Fig. 4b).
334
335
4. Discussion
336
Globally, this is the first study to evaluate the conservation capacity of SNS by use of a large
337
and taxonomically broad set of species. Regarding Hypothesis (I), our study shows that while
338
sacred groves contained more species overall, the difference between them and control sites
339
was not statistically significant unless the north-facing conifer sites were omitted from the
340
analysis. Similar statistical issues have arisen in a previous study comparing protected and
341
unprotected areas for several taxonomic groups (Gray et al., 2016), despite the expected
342
differences between such areas. These results suggest that the advantage of protected over
343
unprotected areas becomes blurred when more than one taxonomic group is examined
344
(Khumbongmayum et al., 2005; Gao et al., 2013). To avoid the bias of masking differences
345
when pooling together data from different taxonomic groups, in the present study,
346
biodiversity was assessed for each group separately. While species richness was higher for
347
most groups in sacred groves, only for fungi was this difference significant. This lends
348
support to Hypothesis (II), except that for lichens, the other taxon that should benefit from the
349
presence of older trees, the differences were not significant. For plants, this lack of strong
350
distinction contrasts with an earlier study (Frascaroli et al., 2016) reporting significantly more
351
species in sacred groves than in reference sites. In contrast to the nuanced difference in
352

12
species richness between sacred groves and control sites, there was a clear biodiversity
353
benefit when beta diversity was considered (Hypothesis III). Its higher value for sacred
354
groves suggests that there is a greater distinction (in the sets of species) between sacred
355
groves than between control sites. This might be explained by the groves different histories of
356
usage, which have a significant effect on sacred grove’s vegetation structure and therefore on
357
the ecological community structure, thus increasing the dissimilarities between groves.
358
Different patterns of land abandonment could also play a role. By contrast, the non-sacred
359
control areas arose largely through natural regeneration in the last 100 years and thus have a
360
more uniform structure.
361
Given the lack of evidence of a strong difference in species richness or composition between
362
sacred groves and control sites, other factors were explored to explain the results found. The
363
most obvious candidate was vegetation type, as the eight pairs of sites were stratified between
364
topographic locations, with three different vegetation types being distinguished, dominated
365
by coniferous, evergreen broadleaved or deciduous broadleaved trees. In all of the analyses,
366
and for many of the species groups examined separately, a clear distinction was found in
367
species richness and composition between the six site pairs dominated by broadleaved trees
368
(with either similar overall richness between the site types or higher richness in the sacred
369
groves) in contrast to the two site pairs with conifer-dominated vegetation (where control
370
sites had higher richness). Other than the nature of coniferous forests per se, a number of
371
features might also contribute to the distinct biodiversity pattern in these two site pairs.
372
Firstly, these two groves and their control sites are in closer proximity to the nearest village
373
than is the case for the other sites. This could have led to more intense anthropogenic
374
influence or, alternatively, it might have increased the effectiveness of the protection
375
associated with religious prohibitions (Frosch et al., 2016). Secondly, they are located on
376
very steep slopes, so these groves would require strict protection to fulfil the role of erosion
377
or landslide control.
Looking closely at each sacred grove, it becomes apparent that its current
378
status has been individually shaped by its history. For example, despite a long history of
379
protection, one of the conifer groves is the forest of Konitsa (4S) was heavily logged for
380
timber and fuel wood in the 1940s, during the Second World War and the following Greek
381
Civil War. Subsequently, in 1953, the municipality decided to manage the forest by removing
382
mature trees in an effort to raise funds for enforcing its protection, particularly of its most
383
degraded parts. Our review of the management history of the eight sacred groves also reveals
384
site-specific variation in the enforcement of restrictions on tree cutting or livestock grazing,
385

13
which are likely to have influenced considerably the habitat properties and, hence,
386
conservation capacity.
387
Land abandonment is another driving force in the evolution of the landscapes of this area. In
388
the postwar period, as agriculture in Western Europe entered a productivity-orientated phase,
389
agricultural change in the study area coincided with decline of agricultural activity or simply
390
of its abandonment. Crop fields disappeared and grasslands gradually developed into
391
shrublands and forests due to a decrease in animal grazing and subsequent natural succession.
392
An exception to that is Konitsa, where the surrounding fertile lowlands remain agricultural to
393
this day (Zomeni et al., 2008). This homogenization of the landscape may explain the
394
differences between sacred and control sites being only marginal. Photos from 1945 and 2007
395
(Fig. H.1) reveal a changing forest landscape with the forest areas around the groves most
396
often expanding. Thus, a possible hypothesis is that the sacred groves acted as nuclei of
397
expansion and dispersal of biodiversity into newly regenerated forest areas.
398
Because sacred groves along the mountainsides of Epirus were established for their benefits
399
in terms of cultural and religious beliefs, hill-slope protection, recreation or even scenery
400
(visual amenity), rather than for biodiversity conservation per se, they can be described as
401
suffering from a kind of “rocks and ice syndrome” (Terborgh, 1999). Biodiversity
402
conservation was not the priority in delimiting these areas; this has emerged as a secondary
403
benefit. For that reason, the sites chosen for sacred status were not selected according to
404
conservation criteria. This is especially the case with respect to their size. Size is a major
405
factor limiting conservation capacity (Halpern, 2003; Ramesh et al., 2016), both with respect
406
to the number of species that can be supported in the long-term and in the length of time an
407
extinction debt can be sustained following isolation (Fig. 4). However, people establishing
408
sacred groves might settle for much smaller areas than are necessary in conservation terms, as
409
can be seen at a global scale in Fig 4.
410
No size dependence was observed for the diversity of sacred groves. This was initially
411
surprising, given the expected dependence of species richness and relaxation time on area.
412
However, as the actual sampling area (given any taxonomic group) is the same in each site
413
we expect this to increase only weakly with site area (Phillips et al., 2017). Furthermore, we
414
should not think of these groves as islands of forest in a landscape of cultivation. The groves
415
have always existed in a matrix of habitable or partially-habitable landscape, so for this
416
reason also, it is not so surprising that measurements of diversity fail to show the limiting
417

14
effect of size expected from Eq. 1. Finally, consistent with historical and photographic
418
evidence, the area of groves is not constant. Most have expanded since 1945 while some were
419
not isolated even in 1945. Also, the variability of areas is not so great (Fig 4a), so that area
420
dependence is not easily detectable if statistical power is low. Thus, while Eqs (1-3), based on
421
isolated fixed-area island models, can illuminate our understanding of conservation capacity
422
and relaxation time, they must be used in conjunction with historical and landscape
423
information when their basic assumptions are not met.
424
These results show a conservation benefit of SNS, which is variable amongst taxa and is
425
affected by the type of grove and by management history. Other SNS in Epirus or elsewhere
426
are likely to behave similarly, particularly if they are of similar size. Thus, in the wider
427
context, if SNS are to play a role in modern conservation, these factors must be carefully
428
assessed. Extension of the analyses reported here should prioritize a landscape-scale
429
assessment of the relative fragmentation of the different sacred groves and control sites, and
430
the extent to which this explains the variation in their species composition and diversity
431
(Echeverría et al., 2007; Daye and Healey, 2015). A fuller knowledge of the historical context
432
can help in this, especially regarding changes in management regime. The issue of vegetation
433
type should be also addressed so as to clarify if it really plays an important role in
434
conservation efficiency.
435
The sacred groves studied here are small in size and have been affected by changing degrees
436
of protection and management throughout their history. Many of them could not function as a
437
reserves or conservation areas by themselves. However, following another modern paradigm,
438
that of the European Natura 2000 system (Official Journal of the European Union, 2011), a
439
network of protected areas existing in an agricultural matrix (following the “countryside
440
SAR” principle) (Pereira et al., 2014) offers an alternative approach. If SNS were
441
incorporated into wider parks or networks, the small conservation advantage that we
442
observed here could become more important. Moreover, a conservation network based
443
around such areas might gain local recognition more readily than a park or network
444
developed on a purely scientific basis. As a large proportion of SNS are small, this approach
445
is likely to be important globally.
446
447
448
449

15
Acknowledgements
450
We would like to acknowledge the inspiration of the late Oliver Rackham. We also thank
451
Stuart Pimm and Thanasis Kallimanis for discussions and helpful comments on the
452
manuscript. This research was co-financed by the European Union (European Social Fund –
453
ESF) and Greek national funds through the Operational Program Education and Lifelong
454
Learning" of the National Strategic Reference Framework (NSRF) - Research Funding
455
Program: THALIS. Investing in knowledge society through the European Social Fund.
456
References
457
Aerts, R., Van Overtveld, K., Haile, M., Hermy, M., Deckers, J. and Muys, B., 2006. Species
458
composition and diversity of small Afromontane forest fragments in northern Ethiopia. Plant
459
Ecology, 187, 127-142.
460
Austad, S.N., Fischer, K.E. 1991. Mammalian aging, metabolism, and ecology: evidence
461
from the bats and marsupials. J. Gerontol. 46, 47–53.
462
Berkes, F., Colding, J., Folke, C., 2000. Rediscovery of traditional ecological knowledge as
463
adaptive management. Ecol. Appl. 10, 1251-1262.
464
Bhagwat, S.A., Rutte, C., 2006. Sacred groves: Potential for biodiversity management. Front
465
Ecol. Environ. 4, 519-524.
466
Birtel, J., Walser, J.-C., Pichon, S., Bürgmann, H., Matthews, B., 2015. Estimating bacterial
467
diversity for ecological studies: methods, metrics and assumptions. PLoS ONE 10(4):
468
e0125356. https://doi.org/10.1371/journal.pone.0125356.
469
Blondel, J., Aronson, J., 1999. Biology and wildlife of the Mediterranean region, Oxford
470
University Press, Oxford, UK.
471
Bongers, T., 1994. De nematoden van Nederland, Pirola: Schoorl, the Netherlands.
472
Boraiah, K.T., Vasudeva, R., Bhagwat, S.A., Kushalappa, C.G., 2003. Do informally
473
managed sacred groves have higher richness and regeneration of medicinal plants than state-
474
managed reserve forests? Curr. Sci. 84, 804–808.
475
Bossart, J., Opuni-Frimpong, E., Kuudaar, S., Nkrumah, E., 2006. Complementarity of
476
fruitfeeding butterfly species in relict sacred forests and forest reserves of Ghana. Biodivers.
477
Conserv. 15, 333–359.
478
Byers, B.A., Cunliffe, R., Hudak, A.T., 2001. Linking the conservation of culture and nature:
479
a case study of sacred forests in Zimbabwe. J. Hum. Ecol. 29, 187-218.
480
Caldasa, A., Robbins, R.K., 2003. Modified Pollard transects for assessing tropical butterfly
481
abundance and diversity. Biol. Conserv. 110, 211-219.
482
Clarke, K. R., P. J. Somerfield, Chapman, M.G., 2006. On resemblance measures for
483
ecological studies, including taxonomic dissimilarities and a zero-adjusted Bray-Curtis
484
coefficient for denuded assemblages. J. Exp. Mar. Biol. Ecol. 330:55–80.
485

16
Clauzade, G., Roux, C., 1985. Likenoj de Okcidenta Europo. Bull. Soc. Bot. Centre-Ouest 7,
486
1-893.
487
Colding, J., Folke, C., 2001. Sacred taboos: ‘invisible’ systems of local resource management
488
and biological conservation. Ecol. Appl. 11, 584–600.
489
Corrigan, C., Hay-Edie, T., 2013. A Toolkit to Support Conservation by Indigenous Peoples
490
and Local Communities: Building Capacity and Sharing Knowledge for Indigenous Peoples’
491
and Community Conserved Territories and Areas (ICCAS), UNEP-WCMC, Cambridge, UK.
492
Council Directive, 1992. Council Directive 92/43/EEC of 21 May 1992 on the Conservation
493
of Natural Habitats and of Wild Fauna and Flora. (Available at:). http://eur-lex.
494
europa.eu/legal-content/EN/ALL/?uri=CELEX:01992L0043-20070101.
495
Daye, D.D., Healey, J., 2015. Impacts of land-use change on sacred forests at the landscape
496
scale. Global Ecol. Conserv. 3, 349-358.
497
Decher, J., 1997. Conservation, small mammals, and the future of sacred groves in West
498
Africa. Biodivers. Conserv. 6, 1007-1026.
499
Derhé, M.A., Murphy, H., Monteith, G., Menéndez, R., 2016. Measuring the success of
500
reforestation for restoring biodiversity and ecosystem functioning. J. Appl. Ecol. 53(6), 1714-
501
1724.
502
Diamond, JM., 1972. Biogeographic kinetics: estimation of relaxation times for avifaunas of
503
southwest Pacific islands. PNAS 69, 3199–3203.
504
Didham, R.K., Hammond, P.M., Lawton, J.H., Eggleton, P., Stork, N.E., 1998. Beetle species
505
responses to tropical forest fragmentation. Ecol. Mon. 68, 295-323.
506
Dudley, N., Higgins-Zogib, L., Mansourian, S., 2009. The links between protected areas,
507
faiths, and sacred natural sites. Conserv. Biol. 23, 568-577.
508
Echeverría, C., Newton, A.C., Lara, A., 2007. Impacts of forest fragmentation on species
509
composition and forest structure in the temperate landscape of southern Chile. Glob. Ecol.
510
Biogeogr. 16, 426-439.
511
Frascaroli, F., 2013. Sacred sites for the conservation of biodiversity. PhD Thesis. University
512
of Zurich, Zurich.
513
Frascaroli, F., Bhagwat, S., Guarino, R., Chiarucci, A., Schmid, B., 2016. Shrines in Central
514
Italy conserve plant diversity and large trees. Ambio 45, 468-479.
515
Frosch, B., Holger, J., Mhamdi, A., Achhal El Kadmiri, A., Rudner,
Μ
., Deil, U., 2016.
516
Sacred sites in north-western Morocco – naturalness of their vegetation and conservation
517
value for vulnerable plant species. Feddes. Repert. 127, 83–103.
518
Gadgil, M., Vartak, V.D., 1976. Sacred groves of Western Ghats of India. Econ. Bot. 30,
519
152–160.
520
Gao, H., Quyang, Z., Chen, S., van Koppen, C.S.A., 2013. Role of culturally protected forests
521
in biodiversity conservation in Southeast China. Biodiv. Conserv. 22, 531-544.
522

17
Gray, C.L., Hill, S.L.L., Newbold, T., Hudson, L.N., Börger, L., Contu, S., Hoskins, A.J.,
523
Ferier, S., Purvis, A., Schalermann, J.P.W., 2016. Local biodiversity is higher inside than
524
outside terrestrial protected areas. Nat. Commun. 7, 12306.
525
Githitho, A.N., 2003. The sacred Mijikenda Kaya Forests of Coastal Kenya and biodiversity
526
conservation. The importance of sacred natural sites for biodiversity conservation.
527
Proceedings of the International Workshop held in Kumming and Xishuangbanna Biosphere
528
Reserve, People’s Republic of China. UNESCO, Paris, France.
529
Halley, J.M., Monokrousos, N., Mazaris, A.D., Newmark, W.D., Vokou, D., 2016. Dynamics
530
of extinction debt across five taxonomic groups. Nat. Commun. 7:12283.
531
Halley, J.M., Sgardeli, V., Monokrousos, N., 2013. Species-area relationships and extinction
532
forecasts. Ann. N.Y. Acad. Sci. 1286, 50-61.
533
Halpern, B.S., 2003. The impact of marine reserves: do reserves work and does reserve size
534
matter. Ecol. Appl. 13, 117-137.
535
Haridasan, K., Rao, R.R., 1985. Forest flora of Meghalaya, Vol. 1. Bishen Singh and
536
Mahendrapal Singh, Dehra Dun, Scientific Publishers.
537
Hughes, J.D., Chandran, M.D.S., 1998. Chapter 5: Sacred groves around the earth: An
538
overview, in: Ramakrishnan, P.S., Saxena, K.G., Chandrashekara, U.M. (Eds.), Conserving
539
the Sacred for Biodiversity Management. Oxford & IBH, New Delhi and UNESCO, Paris.
540
Jim, C.Y., 2003. Conservation of soils in culturally protected woodlands in rural Hong Kong.
541
For. Ecol. Manage. 175, 339–353.
542
Jones, K.E., Altringham, J.D., Deaton, R., 1996. Distribution and population densities of
543
seven species of bat in northern England. J. Zool. 240, 788–798.
544
Junsongduang, A., Balslev, H., Inta, A., Jampeetong, A., Wangpakapattanawong, P., 2013.
545
Medicinal plants from swidden fallows and sacred forest of the Karen and the Lawa in
546
Thailand. J. Ethnobiol. Ethnomed. 9, 1–10.
547
Karthikeyan, C., Dhamotharan, R., 2015. Density, species richness and diversity of Trees in a
548
Sacred Grove-Tropical Dry Deciduous Forest, Dharmapuri District, Tamil Nadu, South India.
549
Scholars Acad. J. Biosci. 3, 536-540.
550
Keppel, G., Mokany, K., Wardell-Johnson, G.W., Phillips, B.L., Welbergen, J.A., Reside,
551
A.E., 2015. The capacity of refugia for conservation planning under climate change. Front.
552
Ecol. Environ. 13:106–112.
553
Khumbongmayum, A.D., Khan, M.L., Tripathi, R.S., 2005. Survival and growth of seedlings
554
of a few tree species in the four sacred groves of Manipur, Northeast India. Curr. Sci. 88,
555
1781-1788.
556
Khumbongmayum, A.D., Khan, M.L., Tripathi, R.S., 2006. Biodiversity conservation in
557
sacred groves of Manipur, northeast India: population structure and regeneration status of
558
woody species. Biodivers. Conserv. 15, 2439–2456.
559

18
Klein, C.J., Wilson, K.A., Watts, M., Stein, J., Carwardine, J., Mackey, B., Possingham, H.P.,
560
2009. Spatial conservation prioritization inclusive of wilderness quality: a case study of
561
Australia’s biodiversity. Biol. Conserv. 142, 1282–1290.
562
Klepeis, P., Orlowska, I.A., Kent, E.F., Cardelús, C.L., Scull, P., Wassie Eshete, A., Woods,
563
C., 2016. Ethiopian Church Forests: A Hybrid Model of Protection. Hum. Ecol. 44, 715-730.
564
Koh, L.P., Ghazoul, J., 2010. A matrix-calibrated species-area model for predicting
565
biodiversity losses due to land-use change. Conserv. Biol. 24, 994–1001.
566
Kosambi, D.D., 1962. Myth and Reality. Popular Press, Bombay, India.
567
Kyriakidou-Nestoros, A., 1989 Laografika meletimata (Folklore Studies). Hellenic Literary
568
and Historical Archive, Athens. [in Greek]
569
Larsen, T.B., Adusei-Poku, K., Sáfián, S. 2009. The Butterflies of Boabeng-Fiema Monkey
570
Sanctuary - Biodiversity and Extinction in a Forest Fragment in Ghana. Afr. Entomol. 17,
571
131-146.
572
Lazarina, M., Sgardeli, V., Kallimanis, A.S., Sgardelis, S.P. 2013. An effort-based index of
573
beta diversity. Methods Ecol. Evol. 4, 217-225.
574
Lee, D.L., 2002. The biology of nematodes. CRC Press.
575
Magurran, A.E., 2004. Measuring biological diversity. Blackwell, Oxford.
576
Mallarach, J.M., Papayannis, T., 2006. Protected areas and spirituality. Proceedings of the 1
st
577
Workshop of the Delos Intiative. IUCN and Nontserrat, Spain. Publicacions de l’Abadia de
578
Montserrat.
579
Mallarach, J.M., Papayannis, T., 2010. Sacred natural sites in technologically developed
580
countries: reflections from the experience of the Delos Initiative, in: Verschuuren, B., Wild,
581
R., McNeeley, J., Oviedo, G. (Eds.), Sacred Natural Sites: Conserving Nature and Culture.
582
Earthscan, London.
583
Malhotra, K.C., Gokhale, Y., Chatterjee, S., Srivastava, S., 2007. Sacred Groves in India.
584
Aryan Books International, New Delhi, India.
585
Mihailaris, P.D., 2004. Aforismos: I prosarmogi mias poinis stis anagkaiotites tis
586
Tourkokratias (Excommunication: Adaptation of a punishment to the needs of the Ottoman
587
Occupation), second ed. National Research Foundation- Center of Modern Greek Research,
588
Athens. [in Greek]
589
Mishra, B.P., Tripathi, O.P., Tripathi, R.S., Pandey, H.N., 2004. Effects of anthropogenic
590
disturbance on plant diversity and community structure of a sacred grove in Meghalaya,
591
northeast India. Biodivers. Conserv. 13,421-436.
592
Nganso, B.T., Kyerematen, R., Obeng-Ofori, D., 2012) Diversity and abundance of butterfly
593
in the Abiriw and Odumante sacred groves in the Eastern Region of Ghana. Res. Zool. 2, 38-
594
46.
595
Nicol, E., Stevens, J.R., Jobling, S., 2017. Riverine fish diversity varies according to
596
geographical isolation and land use modification. Ecol. Evol. 7, 7872-7883.
597

19
Nimis, P.L., 1987. I macrolicheni d'Italia: chiavi analitiche per la determinazione. Atti
598
Musem Friul Storia Nat, grafiche fulvio spa, Udine.
599
Nitsiakos, V., 2016. Peklari, Social Economy in a Greek Village. Lit Verlag, Zurich.
600
Nixon, L., 2006. Making a landscape Sacred. Outying Churches and Icon Stands in Sphakia,
601
Southwestern Crete. Oxbow Books, Oxford.
602
O’Neill-Campbell, M., 2004. Traditional forest protection and woodlots in the coastal
603
savannah of Ghana. Environ. Conserv. 31, 225–232.
604
Official Journal of the European Union, 2009. Directive 2009/147/EC of the European
605
Parliament and of the Council of 30 November 2009 on the Conservation of Wild Birds.
606
(Available at:). http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX: 32009L0147.
607
Official Journal of the European Union, 2011. Commission Implementing Decision of 11
608
July 2011 Concerning a Site Information Format for Natura 2000 Sites. (Available at:).
609
http://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX%3A32011D0484.
610
Oksanen, J., Guillaume Blanchet, F., Kindt, R., Legendre, P., Minchin, P.R., O'Hara, R.B.,
611
Simpson, G.L., Solymos, P., Stevens, M.H.H., Szoecs, E., Helene Wagner, H., 2016. vegan:
612
Community Ecology Package. R package version 2.3-3. https://CRAN.R-
613
project.org/package=vegan.
614
Orange, A., James, P.W., White, F.J., 2001.
Microchemical methods for the identification
615
of lichens. British Lichen Society, London.
616
Ormsby, A., 2013. Analysis of local attitudes toward the sacred groves of Meghalaya and
617
Karnataka, India. Conserv. Soc. 11, 187–197.
618
Papanastasis, V.P., Yiakoulaki, M.D., Decandia, M., Dini-Papanastasi, O., 2008. Integrating
619
woody species into livestock feeding in the Mediterranean areas of Europe. Anim. Feed. Sci.
620
Technol. 140, 1- 17.
621
Pereira H.M, Ziv, G., Miranda, M., (2014) Countryside species-area relationship as a valid
622
alternative to matrix-calibrated species-area model. Conserv. Biol. 28, 874-876.
623
Phillips, H.R.P., Halley, J.M., Urbina
‐
Cardona, N.J., Purvis, A., 2017. The effect of fragment
624
area on site
‐
level biodiversity. Ecography. doi:10.1111/ecog.02956.
625
Pimm, S.L., Russell, G.J., Gittleman, J.L., Brooks, T.M., 1995. The future of biodiversity.
626
Science 269, 347–350.
627
Politis, N., 1904 Meletai peri tou viou kai tis glossis tou ellinikou laou: Paradoseis (Studies
628
on life and language of Greeks: Traditions). PD Sakellariou editions, Athens. [in Greek]
629
Purvis, W.O., Coppins, B.J., Hawksworth, D.L., James, P.W., Moor, D.M., 1992. The Lichen
630
Flora of Great Britain and Ireland.
The British Lichen Society, London.
631
R Core Team, 2015. R: A language and environment for statistical computing. R Foundation
632
for Statistical Computing, Vienna, Austria. https://www.R-project.org/Ramanujam, M.P.,
633
Praveen, K., Cyril, K., 2003. Woody species diversity of four sacred groves in the
634
Pondicherry region of South India. Biodiver. Conserv. 12, 289–299.
635

20
Ramesh, T., Kalle, R., Downs, C.T., 2016. Predictors of mammal species richness in
636
KwaZulu-Natal South Africa. Ecol. Indicat. 60, 385-393.
637
Reed, L.W., Carol, J.P.C., 2004. Sacred forest, hunting, and conservation in west Kalimantan,
638
Indonesia. Hum. Ecol. 32, 313–338.
639
S’Jacob, J.J., van Bezooijen, J., 1984. A manual for practical work in nematology.
640
Department of Nematology, Wageningen Agricultural University: Wageningen, the
641
Netherlands.
642
Stewart, C., 1993. Demons and the devil. Moral imagination in Modern Greek Culture.
643
Princeton University Press, Princeton, New Jersey.
644
Soury, A., van Koppen, K., Tchibozo, M.S., Cotonou, B., 2007. Sacred forests: a sustainable
645
conservation strategy? The case of sacred forests in the Quémé valley, Benin. Wageningen
646
University, Wageningen.
647
Stara, K., Tsiakiris, R., Nitsiakos, V., Halley, J.M., 2016. Religion and the management of
648
the commons - The sacred forests of Epirus, in: Agnoletti, M., Emanueli, F. (Eds.),
649
Biocultural diversity in Europe. Springer International Publishing, Switzerland, pp. 283-302.
650
Stara, K., Tsiakiris, R., Wong, J., 2012. Sacred trees and groves in Zagori, northern Pindos
651
National Park, Greece, in: Pungetti, G., Oviedo, G., Hooke, D. (Eds.), Sacred species and
652
sites. Advances in biocultural conservation. Cambridge University Press, Cambridge, pp.
653
392–396.
654
Stara, K., Tsiakiris, R,, Wong, J.L.G., 2015. The trees of the Sacred Natural Sites of Zagori,
655
NW Greece. Landsc. Res. 40, 884-904.
656
Swamy, P.S., Kumar, M., Sundarapandian, S.M., 2003. Spirituality and ecology of sacred
657
groves in Tamil Nadu. Unasylva 54, 53–58.
658
Terborgh, J., 1999. Requiem for Nature, Island Press, Washington DC.
659
Tsiakiris, R., Betsis, A., Stara, K., Nitsiakos, V., Roubas, V., Halley, J.M., 2013. Defining
660
sacred forests in Konitsa and Zagori: A method of randomly selecting research and control
661
sites for biodiversity study and further statistical analysis. Report of 1
st
and 4
th
Working
662
Groups. Project THALIS-SAGE “Conservation through religion. The sacred groves of
663
Epirus”, pp 9.
664
Venables, W.N., Ripley, B.D., 2002. Modern Applied Statistics with S, fourth ed. Springer,
665
New York.
666
Verschuuren, B., Wild, R., McNeely, J.A., Oviedo, G., 2010. Sacred Natural Sites:
667
Conserving Nature and Culture. Earthscan, London.
668
Virtanen, P., 2002. The role of customary institutions in the conservation of biodiversity:
669
Sacred forests in Mozambique. Environ. Values 11, 227–241.
670
Wadley, R.L., Colfer, C.J.P., 2004. Sacred forest, hunting and conservation in West
671
Kalimantan, Indonesia. Hum. Ecol. 32, 313-338.
672

21
Wassie, A., Sterck, F.J., Bongers, F., 2010. Species and structural diversity of church forests
673
in a fragmented Ethiopian Highland landscape. J. Veg. Sci. 21, 938–948.
674
Newmark, W.D., Jenkins, C.N., Pimm, S.L., McNeally, P.B., Halley, J.M., 2017. Targeted
675
habitat restoration can reduce extinction rates in fragmented forests. PNAS 114, 9635-9640.
676
Wirth, V., 1995. Die Flechten Baden-Württembergs. 2nd ed. Ulmer, Stuttgart.
677
World Wide Fund for Nature, Equilibrium and Alliance of Religions and Conservation, 2005.
678
Beyond Belief: Linking Faiths and Protected Areas to Support Biodiversity Conservation.
679
WWF, London.
680
Wortley, L., Hero, J.-M., Howes, M., 2013. Evaluating ecological restoration success: a
681
review of the literature. Rest. Ecol. 21(5), 537-543.
682
Yuan, J., Liu, J., 2009. Fengshui forest management by the Buyi ethnic minority in China.
683
For. Ecol. Manage. 257, 2002–2009.
684
Zomeni, M., Tzanopoulos, J., Pantis, J.D., 2008. Historical analysis of landscape change
685
using remote sensing techniques: An explanatory tool for agricultural transformation in
686
Greek rural areas. Landsc. Urban Plan. 86, 38–46.
687
688

22
689
690
Figure 1
Identified sacred groves (circles) in the broad area of Zagori and Konitsa. For the
691
current study, biodiversity was measured in eight of these sacred groves (green circles) and in
692
eight corresponding control sites (squares). Shown in the inset is the location of the Epirus
693
study area in Greece. Red lines denote major roads.
694
695

23
696
697
698
699
700
701
702
703
704
705
706
Figure 2
Representations of biodiversity in the sacred and control sites for various taxonomic
707
groups:
(a)
Total species richness (genus richness for nematodes) in each group of species
708
across all eight sacred groves and their respective control sites (with mean and standard error
709
bars). Taxonomic groups are: NM, nematodes; IN, insects; PB, passerine birds; BT, bats; FN,
710
fungi; LC, lichens; HP, herbaceous plants; WP, woody plants. (
b
)
Scatterplot of species
711
richness recorded in sacred groves and their respective control sites. The fitted line
712
(y=0.727x+30.56, R² = 0.912) was calculated after the two pairs of sites dominated by
713
conifers (4 and 7) were excluded. Open diamonds are deciduous broadleaved sites, closed
714
diamonds evergreen broadleaved sites and closed triangles coniferous sites.
715
716
(a) (b)

24
717
718
Figure 3
Species turnover measured as the beta diversity between the local scale (sites) and
719
global scale (union of sites):
(a
) Whittaker index and (
b
)
N* index for the sets of eight sacred
720
groves (black) and respective control sites (gray), by taxonomic group (NM, nematodes; IN,
721
insects; PB, passerine birds; BT, bats; FN, fungi; LC, lichens; HP, herbaceous plants; WP,
722
woody plants) with error bars corresponding to the standard deviation of the species
723
accumulation curve used to estimate the N* index. In the case of nematodes, genus turnover
724
is shown.
725
726
(a)
(b)

25
0
10
20
30
40
0
10
20
30
40
Size
1
10
100
1000
10000
100000
1000000
Half-life (years)
FrequencyFrequency
727
Figure 4 (a)
Histogram of area for 126 SNS: 22 mapped in Epirus and 104 found in our
728
literature search. Superimposed on this is the expected half-life of species loss following
729
habitat isolation using Eq. 3 for all taxonomic groups except fungi and lichens for areas
730
ranging from 0.01 hectare to 100,000 km
2
. The taxonomic group name appears below the line
731
except for nematodes and woody plants for which it is above the line. The sizes of the eight
732
sacred groves of Epirus in this study are shown as black dots just above the horizontal axis.
733
(b)
Histogram of area for the national parks in Greece (light blue), Great Britain (red) and the
734
USA (dark blue). The main divisions (powers of 10) in the horizontal axis are the same for
735
both panels.
736
737
(a)
(b)

26
Table 1.
Location, area and vegetation type of the eight sacred groves (S) and their respective
738
control sites (C). For vegetation type, D = deciduous broadleaf, E = evergreen broadleaf, C =
739
coniferous forests.
740
741
Associated
village
Vegetation
type
Sacred groves Control sites
Code Area
(ha) Code Area
(ha)
Aidonohori D 1S 19.8 1C 16.24
Elafotopos
E
2S
29.11
2C
69.09
Kato Pedina E 3S 10.33 3C 55.23
Konitsa C 4S 115.7 4C 538.9
Mazi D 5S 10.37 5C 54.24
Mesovouni D 6S 17.02 6C 22.01
Molista C 7S 43.29 7C 41.29
Vitsa D 8S 4.87 8C 41.38
742

27
Appendix A – Selection of sacred groves, control sites and sampling points
750
Sacred sites
751
We identified sacred grove sites across the landscape based on archival and ethnographic
752
fieldwork. We further identified and mapped the borderline of these groves using ortho-
753
rectified aerial photographs from the year 1945, which is the oldest complete set of aerial
754
photographs of the area. From these identified sacred groves, eight were selected on the basis
755
of a number of criteria. Firstly, we excluded those less than 3.5 hectares in size, as estimated
756
for 1945, so as to secure at least one permanent bird observation point of 100 m radius (see
757
Appendix B) in each grove. Secondly, we excluded all sites for which there was evidence of
758
substantive felling of trees during the last 60 years, according to the Forestry Department
759
management plans and records or earlier ethnographic or field research. To the remaining
760
sites, we applied the criteria of a minimum threshold of 70% current tree cover and lack of
761
degradation, based on recent ethnographic and field data. From the initial shortlist of sites, a
762
stratified set of sacred groves was selected so as to cover a range of cultural diversity
763
(cultural units, ritual praxes and management regimes) according to ethnographic data (See
764
Appendix E). Where possible (all criteria being satisfied), groves closer to roads were chosen
765
so as to reduce field work and allow more time for sampling.
The final set of sacred groves
766
that were selected was limited to eight because of time constraints.
767
768
Fig A1. View from inside three sacred groves of different types: (left) Elafotopos, a broadleaved evergreen
769
forest (2S), (middle) Molista, a coniferous forest (7S), and (right) Aidonochori, a deciduous forest (1S). (Photos
770
K. Stara 2015)
771
772
Control sites
773
Since our main hypotheses concern biodiversity, we define a control site for each sacred
774
grove so as to assess the biodiversity difference relative to a non-sacred, reference forest.
775
This approach has been used widely in similar studies of biodiversity comparisons (Wortley
776
et al. 2013, Derhé et al. 2016). Here, the selection priority is to find a non-sacred forest for
777
which the environmental factors are as close as possible to the sacred grove. Thus, for each of
778
the eight sacred groves, we identified the best matched control site (without sacred status but
779
with similar site environment and vegetation characteristics) according to a series of criteria:
780
(a) the site had to be close to the respective sacred site (less than 4 km), (b) its area should be
781

28
as large or larger than the respective sacred site, (c) tree cover in it should be no less than
782
70%, (d) it should be of the same vegetation type (dominated by coniferous, evergreen
783
broadleaved or deciduous broadleaved trees) as the sacred grove, (e) it should have the same
784
geological bedrock, and also (f) similar slope and aspect. Selection was based on the analysis
785
of ortho-rectified aerial photos from 2007, existing forest vegetation maps, digitized
786
geological maps of 1:5000 scale, and the Google Earth digital elevation model, supplemented
787
by observations during field visits. Their boundaries were defined using all of the above
788
criteria. Control sites were usually part of larger contiguous woodland areas and except for
789
one, they were larger in area than the sacred sites.
790
791
Fig A2. View from the outside of two sacred groves of different types: (left) Molista, a coniferous forest (7S)
792
lies behind slope above village, and (right) the evergreen broadleaved grove of Kato Pedina (3S) rises upwards
793
to the right along the slope above the village. See also Appendix H. (Photos K. Stara 2015)
794
Sampling points
795
Inside each sacred site and in each corresponding control site, a set of points was chosen by
796
random placement. These points were subject to the additional constraints that they should be
797
located at least 100 m from the woodland edge and separated from any other by at least 300
798
m. These criteria define a maximum number of independent sample plots that can fit in each
799
site. A heuristic algorithm [Generate Random Points, provided by the online software
800
‘Geospatial Modelling Environment’ (www.spatialecology.com)] was employed to provide
801
the sequence of potential sampling points for each taxon. A common sequence of random
802
points was generated for each site and provided to all the teams working on different
803
taxonomic groups. However, the teams were not constrained to use the same points. For each
804
taxonomic group, the same sampling effort was used in all sites and the total number of
805
species that were found in the site was recorded.
806
807
Appendix B – Sampling protocols
808
Nematodes
809
Four sampling points were chosen at each site (sacred and control) and at each a plot of 100
810
m
2
was established. In each plot, a composite soil sample of five soil cores, 3 cm in diameter
811

29
and 12 cm in depth, was collected, so that four composite samples were taken from each site.
812
In all cases, the litter layer was removed before sampling. Nematodes were extracted from
813
200 cm
3
of each composite soil sample. For extraction, the modified Cobb’s sieving and
814
decanting method (S’Jacob and van Bezooijen, 1984) was employed. After counting total
815
abundance of nematodes, samples were fixed with 4% formaldehyde solution. From each
816
sample, 150 nematodes were selected and identified to the genus level using an identification
817
key (Bongers, 1994). In cases where the number of specimens of a sample was less than 150,
818
we identified them all.
819
Insects
820
One sampling point was chosen at each site. Insect sampling was conducted using a modified
821
Pollard sampling scheme (Caldasa and Robbins, 2003), following transects in four directions
822
(N, S, E, W) of 200 m, with a width of 10 m on each side of the center line, lasting exactly 45
823
minutes. Sites were visited twice (early summer 2013 and late summer 2014) for five days
824
each time, in order to include species that appear in different periods during the year, while
825
the order at which sites were sampled differed each time, so as to avoid a bias induced by the
826
specific time of the day. Flying adult insects were collected in nets, whereas soil dwelling and
827
wood-boring adult insects were retrieved with the help of a knife and a tweezer. Specimens
828
were then put into plastic bags and were given a label that described the site, the time and the
829
number of individuals observed for each species. Identification was conducted at the
830
Laboratory of Forest Entomology (Forest Research Institute - HAO Demeter, Greece) using
831
the appropriate morphological keys for each insect order.
832
Passerine birds
833
One sampling point was chosen at each site. Point counts of a fixed radius of 100 m were
834
carried out, recording all bird species observed or identified from their calls and breeding
835
songs for a fixed time period of 10 minutes. One point-count was conducted per site, at the
836
same fixed point, in early morning (from 30 min before dawn and for a duration of 3 h) on
837
two dates, in early and late spring (with the interval between replicates being less than 30
838
days). Breeding songs were considered to indicate a pair of birds, whereas all other
839
observations indicated one individual. The sum of individuals that were recorded on the two
840
sampling dates, in each site, were taken as the measure of abundance in the analysis.
841
Bats
842
One sampling point was chosen at each site. Starting from there, another four sampling points
843
were selected on a line with an approximate distance of 100 m between them. Echolocation
844
calls of bats were recorded at each point for 15 minutes as well as between points (while
845
walking from one point to the next), using the ultrasound receiver Batcorder (ecoObs).
846
Recordings started half an hour after sunset and lasted approximately one and a half hours in
847
each site. Sampling was conducted from mid to late summer and was repeated twice in each
848
site, in 2013 and 2014. Calls were analyzed and species were identified by use of the
849
ultrasound analysis software bcAnalyze v.2 (ecoObs).
850

30
Fungi
851
Sampling was conducted at eight sampling points within each sacred and control site. At each
852
point, a plot of 200 m² was clearly marked along its edges and carefully examined for fungal
853
carpophores. The area was visited twice during the year: in autumn, when most
854
Basidiomycetes fruit, and again in spring in order to observe the fruiting Ascomycetes.
855
Sampling was thus carried out four times in each of the 16 sites: autumn 2013, spring and
856
autumn 2014 and autumn 2015. The exact timing of the visits relied on the information given
857
by local collaborators about the occurrence of fruiting. Carpophores on all substrates (soil,
858
leaf litter, dead wood) were sampled. Their identification was based on their macroscopic
859
features in the field. Specimens of each species were counted and recorded. Specimens whose
860
identification was in doubt were kept in portable coolers and taken to the Laboratory of
861
Forest Pathology & Mycology (Forest Research Institute - HAO Demeter, Greece) for further
862
laboratory examination and verification.
863
Lichens
864
In each site, one sampling point was chosen as the centroid of a 250 m
2
sample plot. Lichen
865
sampling was carried out on tree trunks up to 2 m above ground, on five individuals of each
866
tree species present in the plot. The sampling followed a random time- and species recovery-
867
constrained strategy: on the set of sampled trees, all crustose, foliose or fruticose species
868
observed were collected until no additional species could be detected. All collecting sites
869
were visited once. The identification of the lichen material was carried out using stereo-
870
(Zeiss Stemi) and light-microscopes (Zeiss Axioscope). Standard chemical spot tests, based
871
on potassium hydroxide, bleach, iodine and para-phenylenediamine, and thin layer
872
chromatography (Orange et al., 2001) were applied, and results were compared with those
873
from literature (Clauzade and Roux, 1985; Nimis, 1987; Purvis et al., 1992; Wirth, 1995).
874
Specimens are stored at the GZU Herbarium of the Institute of Plant Science, Karl-Franzens
875
University of Graz (Austria).
876
Herbaceous and Woody Plants
877
In each site, two sampling points were selected. At each, a plot of 250 m
2
was set up. Within
878
these plots, every vascular plant, whether a seed plant (Spermatophyta) or a fern
879
(Pteridophyta), was identified to species level and recorded. Species were further divided into
880
herbaceous and woody plants.
881
Appendix C - Species richness and its analysis
882
The location of the eight selected sacred groves of Konitsa and Zagori, in Epirus,
883
northwestern Greece, and of their matching control sites are presented in Table C.1. Given
884
are for each site (sacred grove or control) the number of species that were recorded for each
885
of the eight taxonomic groups examined per site and overall.
886

31
A generalized linear regression model was built to test the effect of site area (area containing
887
the sampling locations within each site), site type (sacred or control) and vegetation type
888
(dominated by coniferous, evergreen broadleaved or deciduous broadleaved trees) on the total
889
species richness (S) and on the species richness within each taxonomic group (for nematodes
890
this was genus richness). The model used is S ~ area + type + vegetation type + area:type,
891
with a Poisson response and a logarithmic link function. The results are summarized in Table
892
C.2. The significance of each predictor variable is judged on a 5% significance level.
893

32
Table C.1. Number of species* recorded in the eight sacred groves (S) and their respective control sites (C) by taxonomic group. Total
894
corresponds to the total species richness across all sites of each type for each taxonomic group (columns), and across all species groups for each
895
site (rows). The grand total is the number of species in each group found across all 16 sites. For vegetation type, D = deciduous broadleaf, E =
896
evergreen broadleaf, C = coniferous forests.
897
898
Type Site
Number of species
Nematodes*
Insects
Passerine
birds Bats Fungi Lichens
Herbaceous
plants
Woody
plants Total
sacred groves
Aidonohori (1S) 39 9 14 2 33 48 70 11 226
Elafotopos (2S) 48 10 7 3 14 19 46 9 156
Kato Pedina (3S)
32
11
6
2
21
12
47
7
138
Konitsa (4S)
35
7
9
4
13
20
30
8
126
Mazi (5S) 37 9 10 5 8 33 46 11 159
Mesovouni (6S) 39 11 9 1 20 21 49 15 165
Molista (7S) 37 11 14 2 22 24 58 16 184
Vitsa (8S) 35 7 15 3 27 50 61 11 209
Total
64 45 29 10 116 113 213 39 629
control sites
Aidonohori (1C) 31 8 9 2 11 42 74 17 194
Elafotopos (2C) 46 10 6 5 12 15 28 8 130
Kato Pedina (3C) 42 9 7 6 12 23 28 9 136
Konitsa (4C)
36
11
8
4
17
29
48
10
163
Mazi (5C) 25 5 9 6 10 23 57 11 146
Mesovouni (6C) 31 8 5 2 20 27 50 16 159
Molista (7C) 38 11 9 2 12 42 63 19 196
Vitsa (8C) 34 11 10 4 12 39 59 15 184
Total
58 49 20 14 78 109 189 43 560
Grand total
72 69 29 14 159 152 270 51 816
*Number of genera for nematodes.
899

33
Table C.2. Summary statistics and ANOVA results of the generalized linear regression model
900
predicting species richness (total and per taxonomic group) from the site area (extent of
901
sampling area), the site type (S for sacred; control is baseline) and the vegetation type (E,
902
evergreen broadleaved forest; D, deciduous broadleaved forest; coniferous forest is baseline).
903
The model coefficient estimates (Estimate), standard error of the estimate (Std. error),
904
associated p-value (Pr(>|z|)) and ANOVA p-values (Pr(>Chi)) are given.
905
Taxonomic group
Summary statistics ANOVA
Estimate Std.
Error Pr(>|z|) Pr(>Chi)
All species
(Intercept) 2.358
0.444
1.09E-07
Area -0.042
0.037
0.252
Area 0.582
type S -0.843
0.475
0.076
Type 0.788
vegetation D 0.262
0.230
0.253
vegetation
0.001
vegetation E -0.500
0.331
0.131
area:type
0.018
area:type S 0.123
0.052
0.017
Nematodes
(Intercept) 3.571
0.150
0
Area 0.00004
0.0004
0.931
Area 0.994
type S 0.088
0.108
0.418
Type 0.412
vegetation D -0.090
0.150
0.549
vegetation 0.097
vegetation E 0.126
0.150
0.399
area:type 0.771
area:type S -0.001
0.002
0.772
Insects
(Intercept) 2.398
0.276
0
Area -1.17E-05
0.001
0.988
Area 0.644
type S 0.186
0.218
0.393
Type 0.751
vegetation D -0.319
0.281
0.256
vegetation 0.694
vegetation E -0.140
0.285
0.623
area:type 0.255
area:type S -0.005
0.005
0.269
Passerine birds
(Intercept) 2.398
0.276
0
Area -1.17E-05
0.001
0.988
Area 0.644
type S 0.186
0.218
0.393
Type 0.751
vegetation D -0.319
0.281
0.256
vegetation 0.694
vegetation E -0.140
0.285
0.623
area:type 0.255
area:type S -0.005
0.005
0.269
Bats
(Intercept) 2.398
0.276
0
Area -1.17E-05
0.001
0.988
Area 0.492
type S 0.186
0.218
0.393
Type 0.277
vegetation D -0.319
0.281
0.256
vegetation 0.584
vegetation E -0.140
0.285
0.623
area:type 0.278
area:type S -0.005
0.005
0.269
Fungi
(Intercept) 2.589
0.232
0
Area 0.0004
0.001
0.506
Area 0.671
type S 0.585
0.163
0.0003
Type
0.001
vegetation D -0.021
0.227
0.927
vegetation 0.357

34
vegetation E -0.186
0.236
0.431
area:type 0.130
area:type S -0.005
0.003
0.139
Lichens
(Intercept) 3.635
0.153
0
Area -0.001
0.0004
0.270
Area 0.443
type S 0.104
0.121
0.389
Type 0.390
Vegetation D -0.065
0.154
0.672
vegetation
1.01E-07
Vegetation E -0.754
0.179
2.61E-05
area:typeS -0.007
0.003
0.015
area:type
0.0113
Herbaceous plants
(Intercept) 4.173
0.117
0
Area -0.001
0.0005
0.093
Area 0.166
type S 0.147
0.092
0.109
Type 0.676
vegetation D -0.126
0.117
0.285
vegetation
7.6E-06
vegetation E -0.542
0.130
3.03E-05
area:type
0.001
area:type S -0.006
0.002
0.002
Woody plants
(Intercept) 3.051
0.215
0
Area -0.001
0.001
0.055
Area 0.377
type S -0.119
0.190
0.531
Type 0.125
vegetation D -0.336
0.221
0.129
Vegetation
0.012
vegetation E -0.778
0.255
0.002
area:type 0.211
area:type S -0.005
0.004
0.225
906
907
Appendix D - Ordination analysis
908
To visualize the difference in composition between sites, multidimensional scaling analysis
909
based on Bray-Curtis dissimilarity was conducted for each taxon, separately, and for all taxa
910
combined. The analysis was implemented in R 3.2.3 (R Core Team, 2015) using function
911
isoMDS of the MASS package (Venables and Ripley, 2002) and function vegdist of the
912
VEGAN package (Oksanen et al., 2016).
913
For herbaceous and woody plant species, ordination showed a surprising lack of
914
differentiation in floristic composition between the three vegetation types corresponding to
915
different topographic positions (Fig. D.1). This is possibly due to the fact that we have only
916
presence counts for these taxonomic groups. For the other species, patterns of species
917
composition amongst sites varied notably by taxonomic group. For lichens, insects and bats,
918
there was no clear pattern, with much overlap amongst the pairs and the vegetation types. For
919
passerine birds, there was a clear distinction amongst the three vegetation types, but the two
920
sites within each pair were not closely clustered. Notably, for passerine birds there is a
921
separation between sacred groves and control sites. For the remaining two taxonomic groups,
922
the conifer-dominated sites were distinct from the broadleaf tree-dominated ones, but
923
whereas for the fungi the two sites within each pair were quite well clustered, for the
924
nematodes they tended to be split. For all species combined, there is a clear distinction in the
925
species composition of the three vegetation types and for the majority of the eight pairs (Fig.
926
D.2). However, the ordination analyses did not reveal any consistent differences in
927

35
composition between the two types of sites (sacred groves and control) for the individual
928
groups of species or for all species combined.
929
930

36
931
Figure D.1. Ordination of sacred groves (S) and respective control sites (C) using
932
multidimensional scaling with the Bray-Curtis dissimilarity index as a measure of the
933
distance between sites for (a) nematodes, (b) insects, (c) passerine birds, (d) bats, (e) fungi,
934
(f) lichens, (g) herbaceous plants, (h) woody plants. Ellipses define 90% intervals of the
935
distribution of scores within the three vegetation types dominated by different tree types (C,
936
coniferous; E, evergreen broadleaved; D, deciduous broadleaved). The analysis was
937
implemented in R using function isoMDS in the MASS package.
938
939

37
940
941
Figure D.2. Ordination of sacred groves (S) and respective control sites (C) using
942
multidimensional scaling with the Bray-Curtis dissimilarity index as a measure of the
943
distance between sites. Data from all taxa were reduced to presence-only before carrying out
944
the analysis. Ellipses define 90% intervals of the distribution of scores within the three
945
vegetation types dominated by different tree types (coniferous, evergreen broadleaved and
946
deciduous broadleaved). The analysis was implemented in R using function isoMDS in the
947
MASS package.
948
949
950

38
Appendix E – SCI species
951
Table E.1. Species of Community Interest (SCI) identified in each sacred grove and
952
respective control site of this study. Of the 8 taxa investigated, SCI species were identified
953
only for bats, insects, passerine birds (Passer.), herbaceous plants (P-herb) and woody plants
954
(P-Wood), as nematodes were identified at the genus level and in some cases also lichens and
955
fungi.
956
957
958
959
Appendix F – Size of sacred natural sites and national parks size worldwide
960
Table F.1. Sacred natural sites (SNS) included in the comparative analysis. SNS mapped by
961
us in the study area are in italics (for these, the names are in two parts: [village name]-[sacred
962
forest name]). Those whose biodiversity we surveyed also are in bold.
963
Name Area (ha) Country Continent Reference
Tsodilo Hills 9,000.0
Botswana Africa WWF 2005
Zaïpobly 12.3
Côte d’Ivoire Africa WWF 2005
Gufae 33.5
Ethiopia Africa Daye & Healey 2015
Tele 12.6
Ethiopia Africa Daye & Healey 2015
Osha-Ocha 5.3
Ethiopia Africa Daye & Healey 2015
Akasie 4.9
Ethiopia Africa Daye & Healey 2015
Ula 1.8
Ethiopia Africa Daye & Healey 2015
Type
Location Number of SCI
Code Bats Insects Passerine
birds
Herbaceous
plants
Woody
plants Total
sacred groves
Aidonohori 1S 0 0 1 0 0 1
Elafotopos 2S 1 0 0 0 1 2
Kato Pedina
3
S
1
0
1
0
1
3
Konitsa 4S 0 0 1 0 1 2
Mazi 5S 0 0 3 0 1 4
Mesovouni
6S
0
0
2
0
1
3
Molista 7S 0 0 3 0 0 3
Vitsa 8S 0 1 1 0 1 3
Total
1
1
8
0
1
11
control sites
Aidonohori 1C 0 0 1 0 1 2
Elafotopos 2C 1 0 0 0 1 2
Kato Pedina 3C 1 0 1 0 1 3
Konitsa 4C 0 0 1 0 0 1
Mazi 5C 0 0 0 0 1 1
Mesovouni 6C 0 0 1 0 1 2
Molista 7C 0 0 2 0 0 2
Vitsa 8C 0 2 1 0 1 4
Total
2
2
4
0
1
9
Grand Total 2 2 8 0 1 13

39
Qimme 0.7
Ethiopia Africa Daye & Healey 2015
Bortianor 164,892.0
Ghana Africa O'Neal Campbell 2005
Oshiye 772.0
Ghana Africa O'Neal Campbell 2005
Asantemanso Sacred grove 295.0
Ghana Africa Bossart et al. 2006
Boabeng-Fiema 190.0
Ghana Africa Larsen et al. 2009
Gyakye Sacred grove 11.5
Ghana Africa Bossart et al. 2006
Bonwire Sacred grove 8.0
Ghana Africa Bossart et al. 2006
Kajease forest 6.0
Ghana Africa Bossart et al. 2006
Kokrobite 0.1
Ghana Africa O’Neill Campbell 2005
Abiriw 0.04
Ghana Africa Nganso et al. 2012
Odumante 0.03
Ghana Africa Nganso et al. 2012
Mount Kenya 142,020.0
Kenya Africa Dudley et al. 2009
Mijikenda Kaya forests 6,000.0
Kenya Africa Githitho 2003
Nyika National Park 313,400.0
Malawi Africa Dudley et al. 2009
Sacred groves of Oshogbo 55.0
Nigeria Africa Dudley et al. 2009; WWF 2005
Limpopo’s Modjadji Reserve
439.0
South Africa Africa Dudley et al. 2009; WWF 2005
Misali Island marine
conservation area
2,158.0
Tanzania Africa Dudley et al. 2009
Mude Lhong 330.0
Thailand Asia Junsongduang et al. 2013
Jigme Dorji Wildlife
Sanctuary
790,495.0
Bhutan Asia Dudley et al. 2009
Angkor 40,000.0
Cambodia Asia WWF 2005
Xishuangbanna 247,439.0
China Asia Dudley et al 2009; WWF 2005
Meghalaya 100,000.0
India Asia Mishra et al. 2004
Periyar Tiger reserve 77,700.0
India Asia Dudley et al. 2009
Mawsmai Syiem 122.0
India Asia Ormsby 2013
Law Lyngdoh 77.0
India Asia Ormsby 2013
Ayappa 41.7
India Asia Ormsby 2013
Ayyapa devarakadu 16.6
India Asia Ormsby 2013
Betekurubara devarakadu 15.9
India Asia Ormsby 2013
Khloo Langdoh 15.7
India Asia Ormsby 2013
Khloo Blai Phlong 10.0
India Asia Ormsby 2013
Ayyapa Kadanoor 10.0
India Asia Ormsby 2013
Poonya Bhagavathi 7.0
India Asia Ormsby 2013
Law Lyngdoh 4.4
India Asia Ormsby 2013
Battemaki 3.6
India Asia Ormsby 2013
Periya Mudaliar 3.2
India Asia Ramanujan et al. 2003
Karekud 3.0
India Asia Ormsby 2013
Koorvale 3.0
India Asia Ormsby 2013
Bhagavathi temple 2.0
India Asia Ormsby 2013
Kadenkad 1.6
India Asia Ormsby 2013
Kundachappa 1.4
India Asia Ormsby 2013
Kilialamman 1.0
India Asia Ramanujan et al. 2003
Keezhbuvanagiri 1.0
India Asia Ramanujan et al. 2003
Mahadevara 1.0
India Asia Ormsby 2013
Kikut Aiyappa 1.0
India Asia Ormsby 2013
Pammangalathamme 0.8
India Asia Ormsby 2013

40
Aiyappa (Mythadi) 0.8
India Asia Ormsby 2013
Ayappa Temple 0.6
India Asia Ormsby 2013
Chamundi 0.6
India Asia Ormsby 2013
Kalath Bhagavathi 0.5
India Asia Ormsby 2013
Periya Kattupalayam Chavadi
0.4
India Asia Ramanujan et al. 2003
Bhagavathi temple Kadanoor
1.0
India Asia Ormsby 2013
Alagar hills 4,500.0
India-Tamil Nadu Asia Swamy et al. 2003
Kandanur 33.0
India-Tamil Nadu Asia Swamy et al. 2003
Solai-Anadaver kovil 12.0
India-Tamil Nadu Asia Swamy et al. 2003
Ayaanar kovil 10.0
India-Tamil Nadu Asia Swamy et al. 2003
Danau Sentarum National
Park
80,000.0
Indonesia Asia Wadley and Colfer 2004
Mount Hakusan 14,826.0
Japan Asia Dudley et al. 2009
The sacred forest of Kashima
1,500.0
Japan Asia WWF 2005
Kii Mountain range 265.0
Japan Asia Mallarach & Papayannis 2006
Kinabalu National Park 75,370.0
Malaysia Asia Dudley et al. 2009
Khovsgol Lake 838,070.0
Mongolia Asia WWF 2005
Sagarmatha National Park 114,800.0
Nepal Asia Dudley et al. 2009; WWF 2005
Peak wilderness park 22,380.0
Sri Lanka Asia Dudley et al. 2009
Mihintale 1,000.0
Sri Lanka Asia WWF 2005
Mae tae hai 325.0
Thailand Asia Junsongduang et al. 2013
Kata Tjuta National Park 132,566.0
Australia Australasia Dudley et al. 2009
Deen Maar 453.0
Australia Australasia WWF 2005
Tongarino National Park 76,504.0
New Zealand Australasia Dudley et al. 2009
Hunstein Range Wildlife
Management Areas
220,000.0
Papua New Guinea
Australasia WWF 2005
Č
ertova st
ě
na 105.0
Czech republic Europe WWF 2005
Gammelstadsviken 435.0
Estonia Europe Mallarach et al. 2010
Hiiemägi 25.0
Estonia Europe Mallarach et al. 2010
Northern Karelia 350,000.0
Finland Europe Dudley et al. 2009
Pyätunturi National Park 4,340.0
Finland Europe WWF 2005
Mt Athos 33,563.0
Greece Europe WWF 2005
Meteora 375.0
Greece Europe WWF 2005
Greveniti – Eftapapado
117.2
Greece Europe Tsiakiris et al. 2013
Konitsa – Kouri (4S)
115.7
Greece Europe Tsiakiris et al. 2013
Manasi- Livadi
53.7
Greece Europe Tsiakiris et al. 2013
Kalouta – Livadi
51.7
Greece Europe Tsiakiris et al. 2013
Molista – Trafos (7S)
43.3
Greece Europe Tsiakiris et al. 2013
Tristeno – Livadi
39.1
Greece Europe Tsiakiris et al. 2013
Kalovrisi - Ag. Nikolaos
38.8
Greece Europe Tsiakiris et al. 2013
Elafotopos-Kri Panagias
(2S)
29.1
Greece Europe Tsiakiris et al. 2013
Aristi – Pournaria
25.1
Greece Europe Tsiakiris et al. 2013
Palioseli - Mereáo
24.4
Greece Europe Tsiakiris et al. 2013
Kapesovo – Gradista
23.6
Greece Europe Tsiakiris et al. 2013
Leptokaria - Ekklisiastiko
23.3
Greece Europe Tsiakiris et al. 2013
Aidonochori-
Aidonolalousa (1S)
19.8
Greece Europe Tsiakiris et al. 2013
Mesovouni-Ag Charálampos
17.0
Greece Europe Tsiakiris et al. 2013

41
(6S)
Iliochori - Proph. Elias
16.6
Greece Europe Tsiakiris et al. 2013
Kavasila – Panagia
13.0
Greece Europe Tsiakiris et al. 2013
Vrysochori – Livadi
11.4
Greece Europe Tsiakiris et al. 2013
Mazi – Panagia (5S)
10.4
Greece Europe Tsiakiris et al. 2013
Kato Pedina – Anilia (3S)
10.3
Greece Europe Tsiakiris et al. 2013
Aetopetra - Ag. Paraskevi
8.6
Greece Europe Tsiakiris et al. 2013
Vovousa - Ag. Paraskevi
6.8
Greece Europe Tsiakiris et al. 2013
Vitsa – Livadakia (8S)
4.9
Greece Europe Tsiakiris et al. 2013
Mt Carmel 26,600.0
Israel Europe Dudley et al. 2009
Benedictine monastery Monte
Oliveto Maggiore
500.0
Italy Europe Frascarolli 2013
Quercus ilex
forest 100.0
Italy Europe Frascarolli 2013
Yuganskiy Kanthy 648,700.0
Russia Europe Dudley et al. 2009; WWF 2005
Laponian area 940,000.0
Sweden Europe Dudley et al. 2009
Coconino National Forest 747,061.0
USA N. America Dudley et al. 2009; WWF 2005
Wupatki National Monument
14,267.0
USA N. America Dudley et al. 2009; WWF 2005
Lanin National Park 379,000.0
Argentina S. America Dudley et al. 2009; WWF 2005
Kaa-lya del Gran Chaco 1,954,875.0
Bolivia S. America WWF 2005
Isiboro-sécure 1,200,000.0
Bolivia S. America WWF 2005
Sajama National Park 100,230.0
Bolivia S. America WWF 2005
Tumucumaque 2,700,000
Brasil S. America WWF 2005
Laguna De la cocha 39,000.0
Colombia S. America Dudley et al. 2009; WWF 2005
Arenal 12,010.0
Costa Rica S. America WWF 2005
Cayapas Mataje 51,300.0
Ecuador S.America WWF 2005
Tikal 55,005.0
Guatemala S. America WWF 2005
Lagunas de Montebello 60,022.0
Mexico S. America WWF 2005
Kuna Park 60,000.0
Panama S. America WWF 2005
Lake Titikaka 460,000.0
Peru S. America WWF 2005
Machu Pichu 32,592.0
Peru S. America Dudley et al. 2009
964
Table F.2. National Parks (NP) in Greece, UK and the USA used in the analysis and their size
965
(in km
2
).
966
Name km
2
Country
Lakes Volvi & Koroneia 2,120
Greece
Northern Pindos National Park 1,970
Greece
Rodopi Mountain Range National Park 1,731
Greece
National Park of East Macedonia - Thrace 930
Greece
Lake Kerkini National Park 831
Greece
National Park of Tzoumerka, Peristeri and
Arachthos Gorge
820
Greece
Chelmos-Vouraikos National Park 544
Greece
Dadia – Lefkimi – Soufli Forest National Park 428
Greece
Axios-Loudias-Aliakmon National Park 338
Greece

42
Prespa National Park 327
Greece
Olympus National Park 238
Greece
Evros Delta 200
Greece
Parnitha National Park 180
Greece
Mt Oiti National Park 70
Greece
Parnassos National Park 36
Greece
Ainos National Park 29
Greece
National Park of Schinias – Marathon 14
Greece
Cairngorms 4,528
UK
Lake District 2,362
UK
Yorkshire Dales 2,179
UK
Snowdonia 2,176
UK
Loch Lomond and the Trossachs 1,865
UK
South Downs 1,624
UK
Peak District 1,437
UK
North York Moors 1,434
UK
Brecon Beacons 1,344
UK
Northumberland 1,048
UK
Dartmoor 953
UK
Exmoor 694
UK
Pembrokeshire Coast 621
UK
New Forest 570
UK
Broads 303
UK
Wrangell - St. Elias 53,370
USA
Gates of the Arctic 34,398
USA
Denali 24,398
USA
Katmai 16,552
USA
Lake Clark 16,370
USA
Death Valley 13,759
USA
Glacier Bay 13,275
USA
Yellowstone 8,991
USA
Kobuk Valley 7,082
USA
Everglades 6,105
USA
Grand Canyon 4,927
USA
Glacier 4,102
USA
Olympic 3,731
USA
Sequoia & Kings Canyon 3,495
USA
Big Bend 3,242
USA
Joshua Tree 3,213
USA
Yosemite 3,027
USA
North Cascades 2,768
USA
Kenai Fjords 2,456
USA
Isle Royale 2,314
USA
Great Smoky Mountains 2,110
USA
Canyonlands 1,366
USA
Grand Teton 1,255
USA

43
Rocky Mountain 1,076
USA
Channel Islands 1,009
USA
Badlands 989
USA
Capitol Reef 979
USA
Mount Ranier 954
USA
Voyageurs 882
USA
Hawaii Volcanoes 880
USA
Shenandoah 794
USA
Crater Lake 741
USA
Biscayne 700
USA
Zion 593
USA
Redwood 439
USA
Great Sand Dunes 433
USA
Lassen Volcanic 430
USA
Petrified Forest 379
USA
Saguaro 370
USA
Guadalupe Mountains 350
USA
Great Basin 312
USA
Arches 309
USA
Theodore Roosevelt 285
USA
Dry Tortugas 262
USA
Mammoth Cave 214
USA
Mesa Verde 211
USA
Acadia 193
USA
Carlsbad Caverns 189
USA
Bryce Canyon 145
USA
Cuyahoga Valley 134
USA
Black Canyon of the Gunnison 123
USA
Haleakala 122
USA
Wind Cave 115
USA
Pinnacles 108
USA
Congaree 90
USA
Virgin Islands 52
USA
American Samoa 43
USA
Hot Springs 22
USA
967
968
Appendix G – Ethnographic Research
969
Methods
970
Ethnographic study of the sacred groves of Epirus aiming to describe people's valuation and
971
perception of different tree species and to identify the sacred natural sites and their
972
emblematic trees (Stara et al., 2015) started in 2005, involving initially 23 villages in Zagori.
973
Work resumed in 2012 and covered the rest of Zagori and the adjacent area of Konitsa.
974

44
Research for this study of the archives of municipalities, the Forestry Service and the Church,
975
and of local libraries targeted at finding references to the sacred groves and their history, in
976
general, and of those selected for the study, in particular. Ethnographic research involved
977
interviews with local people. They were asked about their community’s sacred groves, the
978
reasons for their maintenance, also about their history and the ritual activities, the
979
supernatural guardians, acceptable and non-acceptable uses, and stories or taboos about
980
trespassing in the groves (Stara et al., 2016).
981
Management regimes in the sacred groves of Epirus
982
The groves appear either in the form of protective forests above or close to villages or as
983
groups of veteran trees that accompany outlying churches ("xoklissia") or icon stands
984
(“eikonismata”, shrines comprising boxes containing icons and an oil lamp that remains lit
985
most evenings; Stewart, 1993; Nixon, 2006) retaining a protection value through association
986
with various Orthodox saints (Politis, 1904; Kyriakidou-Nestoros, 1989). Management
987
regimes in the sacred groves of Epirus vary from strict protection to controlled management.
988
These regimes are site-dependent and related to the specific reasons for which these groves
989
were established and maintained, to the type of religious dedication, the perceived personality
990
of the protector saint or saints, historical circumstances and community needs. When a
991
church with a sacred grove was founded on the epiphany of the divine, then the protection
992
was strict. For example, for the grove in Vovoussa in East Zagori, dedicated to the saint Agia
993
Paraskevi, local people argue that Agia Paraskevi herself chose the exact point, where the
994
church should be built, through various manifestations, such as repeatedly moving her icon
995
there. The local cult remains very much alive today linked to that grove and all harvests (e.g.
996
from hunting, collecting honey from wild bee hives, plants, mushrooms, dead wood etc.) are
997
still strictly prohibited (Stara et al., 2016). Strict regimes also tended to prevail for protective
998
forests on very steep slopes (e.g. at Molista, site 7S; Table S1, Fig S3). In contrast, the regime
999
in some groves is much more relaxed (e.g. at Mazi, 5S; Table S1, Fig. S3); for instance,
1000
grazing is allowed without restrictions during certain time periods. Harvesting of branches
1001
(“shredding”) of evergreen tree species during harsh winters (for fuelwood or animal fodder)
1002
was allowed occasionally by church and community councils, whereas shredding of
1003
deciduous tree species during early spring was always considered a trespass. In extreme
1004
cases, controlled management might permit timber harvesting for necessary public works.
1005
Some tolerance of breaking these rules was extended to members of lower social strata.
1006
Finally, collective trespassing could be allowed in abnormal situations. For example, in times
1007
of war or during festivals that are characterized by the ceremonial reversal of social order, the
1008
collection of dead wood and flammable branches of shrubs for use might be allowed (e.g. at
1009
Christmas or for carnival bonfires). Several hamlets in the area were consolidated during the
1010
16
th
to 17th century forming the present villages. Where settlements are abandoned, their
1011
associated sacred groves are often gradually neglected and only mature trees in the vicinity of
1012
the church itself are protected (Stara et al., 2016).
1013
1014

45
Excommunication
1015
Excommunication is the exclusion of a person from the Church and the deprival of its
1016
mysteries. In the Orthodox Church, it is the heaviest punishment that can be imposed on a
1017
Christian. From the later Byzantine period, and particularly under the Ottoman rule,
1018
excommunication was commonly employed for offenses of economic or social character, as
1019
are cases of theft, rape, livestock stealing, defamation, trespassing etc. It was also used as a
1020
threat in order to protect trees and other natural resources from trespassing and interference
1021
(Mihailaris, 2004; Stara et al., 2012).
1022
1023
Appendix H – Aerial Photos of sacred groves in 1945 and 2007
1024
The exact borders of the sacred groves studied were identified and mapped using ortho-
1025
rectified aerial photographs from the year 1945, the oldest complete set of aerial photographs
1026
of the area (source: Hellenic Military Geographical Service, digital aerial photo 1945 -
1027
orthorectified) and compared with the most recent set of 2007 (Hellenic Cartographic and
1028
Cadastral Organisation, digital orthorectified image 2007). The scale for all photographs is
1029
1:7,500 except for the site 4S (Konitsa) for which it is 1:20,000.
1030
These photos reveal a changing forest landscape, with the forest areas around the groves
1031
often expanding. Sites 1S, 4S, 5S and 7S were not isolated from the surrounding forest areas
1032
even in 1945.
1033
Sites 1S and 5S show little net change in cover but exhibit a pattern of patchy increase or
1034
decrease in tree cover within the sacred grove. The area surrounding site 5S changed in
1035
vegetation structure, from a dense scrubland to a young forest as grazing by goats decreased.
1036
In sites 2S, 4S and 7S, forest cover remained high within the sacred grove but with
1037
substantial changes in the surrounding matrix. Whereas the sacred groves in 1945 were
1038
largely isolated (surrounded mainly by rangelands, scrublands or wood-pastures with
1039
minimum tree cover), by 2007, much of this surrounding matrix was covered by trees. This is
1040
predominantly because of the cessation of grazing that allowed the regrowth of forests. In the
1041
case of conifer forests (sites 4S, 7S), trees in the sacred groves could have been an important
1042
seed source, while for the other types, existing shrubs (e.g. around site 2S) can take tree form
1043
once grazing stops.
1044
Sites 3S and 6S show a similar trend of a large increase in tree cover between the two dates,
1045
both inside the sacred grove and in the surrounding matrix.
1046
Around site 8S, there is substantial increase in tree cover in the surrounding matrix, with just
1047
patchy changes in tree cover inside the sacred grove, as grazing (goats, sheep and cows) is
1048
still active forming an open extensive wood pasture characterized by scattered trees and
1049
scrubs.
1050

46
1051

47
1052
Figure H.1. Changes in and around sacred groves between 1945 (left panels) and 2007 (right
1053
panels). Sacred groves are marked by the green line. Lettering inside is from the official state
1054
agency that issued the 2007 maps. The X’s in groves 1S, 2S, 3S and 7S correspond to the
1055
vantage points from which the photos in figures A1 and A2 were taken.
1056