Grazing effects on the species-area relationship: Variation along a climatic gradient in NE Spain

Abstract

Questions: Does grazing have the same effect on plant species richness at different spatial scales? Does the effect of spatial scale vary under different climatic conditions and vegetation types? Does the slope of the species‐area curve change with grazing intensity similarly under different climatic conditions and vegetation types?
Location: Pastures along a climatic gradient in northeastern Spain.
Methods: In zones under different regimes of sheep grazing (high‐, low‐pressure, abandonment), plant species richness was measured in different plot sizes (from 0.01 to 100 m2) and the slope of the species‐area curves was calculated. The study was replicated in five different locations along a climatic gradient from lowland semi‐arid rangelands to upland moist grasslands.
Results: Species richness tended to increase with grazing intensity at all spatial scales in the moist upland locations. On the contrary, in the most arid locations, richness tended to decrease, or remain unchanged, with grazing due to increased bare soil. Grazing differentially affected the slope ( z ) of the species‐area curve (power function S = c A z ) in different climatic conditions: z tended to increase with grazing in arid areas and decrease in moist‐upland ones. ß‐diversity followed similar pattern as z .
Conclusions: Results confirm that the impact of grazing on plant species richness are spatial‐scale dependent. However, the effects on the species‐area relationship vary under different climatic conditions. This offers a novel insight on the patterns behind the different effects of grazing on diversity in moist vs. arid conditions reported in the literature. It is argued that the effect of spatial scale varies because of the different interaction between grazing and the intrinsic spatial structure of the vegetation. Variations in species‐area curves with grazing along moisture gradients suggest also a different balance of spatial components of diversity (i.e. a‐ and ß‐diversity).

Full text

- Grazing effects on the species-area relationship - 25
Journal of Vegetation Science 18: 25-34, 2007
© IAVS; Opulus Press Uppsala.
Grazing effects on the species-area relationship:
Variation along a climatic gradient in NE Spain
de Bello, Francesco1*; Lepš, Jan2 & Sebastià, Maria-Teresa1,3
1Laboratory of Plant Ecology and Forest Botany, Forestry and Technology Centre of Catalonia, E-25280 Solsona, Spain;
2Department of Botany, Faculty of Biological Sciences, University of South Bohemia, and Institute of Entomology, Czech
Academy of Sciences, CZ-370 05 České Budějovice, Czech Republic; E-mail suspa@bf.jcu.cz;
3Agronomical Engineering School, University of Lleida, E-25198 Lleida, Spain; E-mail teresa.sebastia@ctfc.es;
*Corresponding author; Fax +34 973481392; E-mail fradebello@ctfc.es
Abstract
Questions: Does grazing have the same effect on plant species
richness at different spatial scales? Does the effect of spatial
scale vary under different climatic conditions and vegetation
types? Does the slope of the species-area curve change with
grazing intensity similarly under different climatic conditions
and vegetation types?
Location: Pastures along a climatic gradient in northeastern
Spain.
Methods: In zones under different regimes of sheep grazing
(high-, low-pressure, abandonment), plant species richness
was measured in different plot sizes (from 0.01 to 100 m2)
and the slope of the species-area curves was calculated. The
study was replicated in five different locations along a climatic
gradient from lowland semi-arid rangelands to upland moist
grasslands.
Results: Species richness tended to increase with grazing
intensity at all spatial scales in the moist upland locations.
On the contrary, in the most arid locations, richness tended to
decrease, or remain unchanged, with grazing due to increased
bare soil. Grazing differentially affected the slope (z) of the
species-area curve (power function S = c Az) in different
climatic conditions: z tended to increase with grazing in arid
areas and decrease in moist-upland ones. β-diversity followed
similar pattern as z.
Conclusions: Results confirm that the impact of grazing on
plant species richness are spatial-scale dependent. However,
the effects on the species-area relationship vary under different
climatic conditions. This offers a novel insight on the patterns
behind the different effects of grazing on diversity in moist vs.
arid conditions reported in the literature. It is argued that the
effect of spatial scale varies because of the different interac-
tion between grazing and the intrinsic spatial structure of the
vegetation. Variations in species-area curves with grazing along
moisture gradients suggest also a different balance of spatial
components of diversity (i.e. α- and β-diversity).
Keywords: Competition; Heterogeneity; Land use; Manage-
ment; Mediterranean; Power law species–area curve; Sheep-
grazing.
Nomenclature: Bolòs et al. (1993).
Introduction
The role of herbivores in controlling plant spe-
cies richness is a critical issue in the conservation and
management of grazed systems (Olff & Ritchie 1998;
Landsberg et al. 2002; Guo 2004). Syntheses and models
suggest that herbivore effects on plant diversity vary
across environmental gradients of soil fertility and pre-
cipitation (Milchunas et al. 1988; Huston 1994; Proulx
& Mazumder 1998; Cingolani et al. 2005). A moderate
grazing pressure is thought to enhance plant diversity by
the direct consumption of competitively dominant plant
species, thus indirectly affecting plant competition and
promoting species coexistence (Grime 1973; Al-Mufti
et al. 1977). However, in water- and nutrient-limited
environments, increased grazing is expected to increase
plant mortality and ultimately decrease species richness
(Huston 1994; Proulx & Mazumder 1998).
The change in species richness after grazing cessation
may be consequently different under different climatic
conditions. However the conclusions of previous stud-
ies differ to a certain extent (Perevolotsky & Seligman
1998; Osem et al. 2002; Rook et al. 2004) causing some
uncertainty in terms of conservation purposes and for
the generalization of synthetic theories of biodiversity.
Discrepancies could be partially caused by the fact that
richness depends on the spatial scales considered (Lepš
& Štursa 1989; Canals & Sebastià 2000; Magurran
2004).
Given the dependence of species richness on spatial
scale, the number of species in a community is better
described by the species-area relationship rather than by
a single number (Rejmánek & Rosén 1992; Rosenzweig
1995; Gotelli & Colwell 2001; He & Legendre 2002).
The species–area relationship, describing the increase
in the number of species (S) with increasing area (A),
is one of the most robust patterns in ecology (Huston
1994; Ostling et al. 2003; Magurran 2004). Species-area
curves are based on the evidence showing that the number

26 de Bello, F. et al.
of species increases with increasing spatial extent of a
community and that the rate of increase differs among
communities (Lepš & Štursa 1989; Rosenzweig 1995;
Wilson & Chiarucci 2000; Desiltes & Houle 2005).
The curve is generally formulated as a power function,
S = c Az, although the semi-logarithmic form S = a + b
log A has also been used by botanists (Lepš 2005). The
rate at which the species number increases with area (as
expressed by z) has been associated to a great variety of
theoretical mechanisms suggesting, at least, differences
in the processes regulating diversity (Huston 1994). It is
generally accepted that z relates to different aspects of
spatial heterogeneity (Lepš & Štursa 1989; Huston 1994):
resource distribution (Desiltes & Houle 2005), spatial
distribution of the species (He & Legendre 2002; Ostling
et al. 2003), and spatial display of the species, as, e.g.,
size and vertical complexity (Huston 1994; Ovaskainen
& Hanski 2003; Lepš 2005)
Species-area curves might thus be useful when ana-
lyzing the control of disturbance (e.g. grazing) on species
richness at different ranges of scales, as they allow the
study of the partition of spatial components of diversity
(Huston 1999; Loreau 2000). Olff & Ritchie (1998)
hypothesized that species-area curves in grazed areas
could be less steep compared to curves from ungrazed
areas and that these curves should finally intersect.
Hence, species richness may increase at small scales
while decreasing at wider-regional scales. Intersec-
tions of species curves have been reported (Lepš &
Štursa 1989; Lande et al. 2000) and they have important
consequences for extrapolation of biodiversity patterns
(Magurran 2004). More experimental data are however
needed to confirm if this different effect of grazing on
different spatial scale will hold for different vegetation
types and determine the implications for the mutual
dependency of local and regional diversity (Rosenzweig
1995; Huston 1999).
In this study, we assessed, under different climatic
conditions and vegetation types, whether (1) the effect
of grazing on species richness was similar at different
spatial scales and (2) the slope of the species-area curve
changed with grazing intensity. Changes in plant species
richness in grazed (high and low pressure) and abandoned
areas were analysed along a climatic and altitudinal
gradient in northeastern Spain, from semi-arid lowland
to moist-upland areas. Different plot sizes were sampled
ranging from 0.01 to 100 m2 and the species-area curve
was calculated. To our knowledge, this study is the first
attempt to investigate the spatial-dependent effects of
pastoralism on plant richness with a comparable design
for different climatic conditions.
Methods
Experimental design
Five locations were selected along an altitudinal
and climatic gradient in northeastern Spain (Fig. 1)
including the transition from the Mediterranean to the
Boreo-alpine biogeographical region (Bolòs et al. 1993).
Climatic parameters covary along the gradient in such
a way that selected locations ranged from semi-arid to
humid conditions (Moisture index in Fig. 1; Anon. 1992)
and each was placed in a different vegetation belt (Vigo
& Ninot 1987). The number of species shared between
adjacent locations along the gradient was comparable,
as it was the similarity in species composition (Fig. 1)
calculated with the Jaccard similarity index:
J = a/(a+b+c) (1)
where a is the number of species shared between the
two neighbour locations, b is the number found only
in the first location and c the number found only in the
second location.
There is a century-long history of livestock (mainly
sheep and goat) grazing in the study area. As in most
areas of the northern rim of the Mediterranean Basin, rural
abandonment is causing successional changes (de Bello
et al. 2005). Modernization of livestock production has
resulted in a decline of the use of extensive rangelands
and grasslands in the last few decades (Rook et al. 2004).
Sheep herds in the region have approximately 500-700
animals and the maximum stocking rates are normally
encountered next to corrals and water points. In some
limited cases traditional transhumance is still practised:
shepherds may behave as semi-nomads and flocks of
animals are moved up or down the altitudinal gradient
according to the state of the vegetation (Etienne 1996).
Local shepherds, farmers and technicians were inter-
viewed in each location to identify a gradient of sheep
grazing intensity. We limited our study to sheep-grazed
systems. The selected areas were (1) abandoned for more
than 10 years, (2) with low or (3) high grazing pressure.
It was not possible to quantify grazing pressure more pre-
cisely and thus broad categories were used. Distance of
areas from water points and corrals as well as evidences of
fresh excrements were taken into account in determining
the grazing pressure (Jauffret & Lavorel 2003; Landsberg
et al. 2002). Recently burnt land and rock outcrops were
avoided. Following Proulx & Mazumber (1998) we ap-
plied categorical data on grazing pressure rather than
a quantitative scale of grazing pressure. Nevertheless,
to allow comparison with other grazing systems and
vegetation types, the maximum stocking rates generally
encountered in each location was estimated by comparing
results from a regional survey (Taüll & Casals unpubl.

- Grazing effects on the species-area relationship - 27
data) with data from different vegetation types and graz-
ing systems from Etienne (1996). This estimation gives
an approximate range of grazing pressures considered
in each location, ranging from abandoned grazing to a
maximum stocking rate (Fig. 1).
Four independent plots were established per grazing
intensity (2 replicates on south and 2 on north facing
slopes). Only sloping areas were included as, usually,
flat areas are used for cropping. The experiment was
established using a factorial design: 3 sheep grazing
intensities × 5 localities × 2 aspects × 2 replicates = 60
plots. Slope inclination varied between 19° and 35°.
Plots were placed in the approximate centre of a rela-
tively homogeneous zone and were surveyed at the time
of the expected peak of vegetation development in the
corresponding location.
Vegetation was sampled using 10 m × 10 m plots,
divided into 100 1-m2 subplots. We recorded all vascular
plant species whose vertical projection was included in
each subplot. Presence/absence of each species was re-
corded in every subplot. In addition, one of the subplots,
systematically positioned in a corner with respect to the
centre of the plot, was further divided into 100 10 cm × 10
cm sub-subplots, and the species present were recorded in
each of them. Further details on the experimental design
and vegetation changes in the study region are presented
in de Bello et al. (2005).
The percentage of bare soil was estimated in each
plot within the 3 driest locations (in the moist-upland
locations, the bare soil was rare). This was done by the
point-quadrat method (Daget & Poissonet 1971) with
5 linear transects of 10 m repeated at regular distances
and distributed along the main direction of the slope. A
total of 100 points separated by 30 cm were recorded in
the five lines.
For every grazing treatment in each location, the
Whittaker index of beta diversity was calculated. This
was done by dividing the total number of species in one
treatment (for instance, the four abandoned plots in the
most arid location) with the average number of species
in the 10 m × 10 m plots of that treatment.
Fig. 1. Climate and vegetation of the study locations (PET = potential evapotranspiration). Changes in the relative proportions of
different growth forms with grazing intensity in different locations along the climatic gradient (Subshrubs = chamaephytes, small
shrubs shorter than 40 cm). The maximum stocking rates (AU = Animal Units) for each location is shown (see Methods). In the
species names P. stands for Pinus and Q. for Quercus.

28 de Bello, F. et al.
Data analysis
For each 10 m × 10 m plot, the average number of
species was calculated for quadrats of increasing size:
1 × 1, 2 × 2, 3 × 3..... up to 10 m × 10 m. The average
was calculated over all possible different locations for
quadrats of a given size in the 10 m × 10 m plot. This
was done to reduce the effect of the starting position in
the sampling. Clearly, the number of possible positions
decreases with increased quadrat size (100 non-overlap-
ping plots of 1 m × 1 m; 81, partially overlapping plots
of 2 m × 2 m, …; one 10 m × 10 m plot). We consider
each average value to be the best possible estimate of spe-
cies richness for a given quadrat size, but are aware that
no other statistical characteristics (e.g. any measure of
variability) would be correctly estimated. Consequently,
only the averages were used in further calculations. The
same process was repeated for the 1 m × 1 m subplot
divided into 10 m × 10 cm quadrats (ʻ1×1 smallʼ), with
the number of species calculated in quadrats of increas-
ing size from 10 m × 10 cm to 1-m2.
The species-area curve parameters were calculated
based on the number of species within each quadrat size.
This was expressed as the power function:
S = c Az (2)
where S is the number of species, A is the sampled area
and c and z are the parameters estimated by linear re-
gression after log transformation of both variables. The
power function starts at the origin (no species present in
plot size zero) and implies a linear dependence of the log
transformed variables; c is number of species in a plot of
a given unit size; z measures the rate of increase: when
doubling the plot size, the number of species increases
2z times (z usually ranges from 0.15 to 0.3; Lepš 2005).
Species-area curve parameters were calculated on three
ranges of scale: (a) for quadrats ranging from 0.01 to 1
m2 (ʻ1 × 1 m smallʼ plot); (b) for quadrats ranging from
1 to 100-m2; and (c) for quadrats ranging from 0.01 to
100 m2 (combining 10 cm × 10 cm sub-subplots with
the main plot).
A repeated measures ANOVA was performed (after
log transformation of the number of species), to test
the effect of grazing regime (abandonment, low- and
high-pressure), location along the climatic gradient (five
locations; Fig. 1) and aspect (north-, south-facing slope)
on species richness in different plot sizes. The repeated
measures ANOVA was applied because the nested plot
design meant that different plot sizes would be correlated.
Plot size was used as the repeated measure (=ʻwithin
subjectʼ) factor. Whereas the effect of size itself is trivial
(the number of species increases with plot size), the inter-
actions with size are of interest. Because the number of
species was log-transformed, the interaction tests for the
deviation from multiplicative effects. We would expect
grazing to have the same effect at all spatial scales if it
increases species richness equally by the same proportion
(e.g. one third) at all spatial scales. The relative increase
would be different for different spatial scales, if the Size
× Grazing interaction were significant.
Three-way ANOVAs were used to test the effect
of grazing, location and aspect (all introduced as fixed
factors) on the calculated z parameters and on bare soil
percentage (this variable only in the three driest loca-
tions). Standard errors of the z values presented in the
graphs refer only to the variability among independent
plots (they are not derived from the regression analyses
within a plot).
Results
The effect of grazing on the average number of spe-
cies changed with plot size and along the climatic gradi-
ent: all of the first order interactions with Size and also
the second order interaction Size × Location × Grazing
were highly significant (Table 1). In moist-upland areas
(i.e. the two locations at higher altitude), species richness
tended to increase with grazing at all scales considered
while, in more arid areas, it tended increase only at the
largest plot sizes (Fig. 2). In the smallest plot sizes (i.e.
0.01 to 1 m2) in the most arid location, species richness
tended to decrease with grazing (Fig. 2). The percentage
of bare soil increased more strongly with grazing with
increased arid conditions (Fig. 3).
The z parameters of the species-area curve were
Table 1. Results of the repeated measures ANOVA. Changes
in the number of species with increasing plot size (quadrats
from 0.01 m2 to 100 m2) and its relation with the environmental
factors considered (locations along the climatic gradient, aspect
orientation, grazing intensity).
n of species
df F p
Between-subjects factors
Location 4 35.3 < 0.001
Aspect 1 6.9 0.013
Grazing intensity 2 9.4 < 0.001
Aspect × Location 4 5.3 0.002
Aspect × Grazing 2 0.3 0.709
Location × Grazing 8 2.4 0.040
Location × Grazing × Aspect 8 1.0 0.460
Within-subjects factors
Size 18 1142.5 < 0.001
Size × Location 72 11.5 < 0.001
Size × Aspect 18 3.1 < 0.001
Size × Grazing 36 5.1 < 0.001
Size × Aspect × Location 72 1.3 0.055
Size × Aspect × Grazing 36 0.8 0.810
Size × Location × Grazing 144 1.9 < 0.001
Size × Loc. × Grazing × Aspect 144 1.3 0.019

- Grazing effects on the species-area relationship - 29
differently related to grazing intensity under different
climatic conditions (Table 2). The results were similar
for z calculated between different ranges of scales (0.01
to 1-m2; 1 to 100-m2 and the combined 0.01 to 100-m2)
even if the adjusted R2 increased at greater scales (Table
2). Overall, z tended to increase with grazing in arid
locations and decrease in moist-temperate ones (Fig. 4).
β-diversity covaried with z , showing similar patterns of
variations (Figs. 4 and 5). The z parameters calculated
between at the 0.01 to 1-m2 and the 1 to 100-m2 scale
were correlated (R = 0.54) even if the z-values calculated
at the smaller sizes were significantly higher (paired
sample T-test; p < 0.001). Overall, species-area curves
in arid areas tended to have steeper power functions (Fig.
4). The effect of aspect orientation on z also changed
in different locations (Table 2) but no interaction with
grazing was noted.
Table 2. Results of the 3-way ANOVA for the slope (z) of the species-area curve (S = cAz). The slope was calculated for three ranges
of scales: quadrats doubling from 0.01 to 1 m2 (ʻ1×1 smallʼ plot ), 1 to 100 m2 (main plot) and from 0.01 to 100-m2 (combining the
ʻ1×1 smallʼ plot with the main plot). The adjusted R2 (adj. R2) for each ANOVA is shown.
z (0.01 m2 - 1 m2) z (1 m2 - 100 m2) z (0.01 m2 -100 m2)
d.f. F p F p F p
Location 4 8.2 <0.001 10.2 <0.001 24.0 <0.001
Aspect 1 0.7 0.409 2.5 0.127 1.7 0.195
Grazing intensity 2 2.2 0.123 1.2 0.316 1.9 0.162
Aspect × Location 4 1.3 0.292 3.2 0.025 11.4 <0.001
Aspect × Grazing 2 0.3 0.729 0.1 0.966 1.1 0.341
Location × Grazing 8 2.6 0.029 3.8 0.003 5.1 <0.001
Location × Grazing × Aspect 8 0.7 0.682 1.3 0.260 0.4 0.929
Error 30
Fig. 2. Number of species under different grazing regimes (legend) and different locations along the climatic gradient (see also Fig.
1). Separate graphs for different plot sizes (10 cm × 10 cm; 50 cm × 50 cm; ʻ1 × 1 smallʼ upper graphs; 1 m × 1 m; 5 m × 5 m; 10
m × 10 m lower) are shown. The error bars denote mean ± 1 SE. Note that the graphs have different scales. The ʻsmall 1 m × 1 mʼ
is the single 1-m2 quadrat divided into 100 10 cm × 10 cm sub-subplots. See Table 1 for the ANOVA model.

30 de Bello, F. et al.
Fig. 4. Slope (z) of the species-area curves for different grazing
intensity in different locations along the climatic gradient. The
error bars denote mean ± 1 SE (see Table 2 for the ANOVA
model). The slope was calculated using different ranges of
scales (a; b; c; see Methods) Note that the three graphs have
different scales.
Fig. 3. Changes in the percentage of bare soil with grazing
regime in different locations. Bare soil was estimated by the
point-quadrat method. The error bars denote mean ± 1 SE. The
p-value refers to the results of location × grazing interaction
in the ANOVA model that showed a significant effect of (R2
= 0.76 for the whole model). The study was restricted to the
three most arid locations.
Discussion
This study shows that grazing has different effects
on the species-area relationship under different climatic
conditions. This offers a novel insight into the patterns
behind the effect of grazing on plant species diversity
in moist vs. arid conditions, which has been proposed
by several authors as a critical issue in the conservation
and management of landscapes (Milchunas et al. 1988;
Huston 1994; Proulx & Mazumder 1998; Cingolani et
al. 2005; Lepš 2005).
In the most arid locations included in our study, we
found that grazing reduced species richness at small
scales while promoting it at larger scales (Fig. 2). The
vegetation in arid regions is normally clumped in patches
that can be separated by bare soil (Cipriotti & Aguiar
2005). In these conditions, grazing might cause an
increase in plant mortality (Milchunas et al. 1988) and,
thus, increase the proportion of bare soil (Landsberg et al.
2002). These observations are consistent with our results,
which indicate that the percentage of bare soil increased
more strongly with grazing in arid conditions. Thus,
in arid environments, the chance of encountering only
bare soil increases in small-scale plots and the number
of species would tend to decrease, or remain the same,
with grazing. This was also found by Osem et al. (2002)
in 20 cm × 20 cm plots.
The increase of bare soil caused by grazing in arid
areas might also increase the degree of patchiness and,

- Grazing effects on the species-area relationship - 31
thus, the spatial heterogeneity for species establishment
(Huston 1994; Alados et al. 2004) and resource distribu-
tion (Adler et al. 2001; Desiltes & Houle 2005). This will
ultimately increase the slope (z) of the species area curves.
This implies that the negative effect of grazing on species
richness at small scales in more arid conditions might
be reversed at larger scales (Fig. 6). The suppression of
potential dominants (preventing competitive exclusion)
probably functions at all spatial scales, but this effect is
more than compensated for at small spatial scales by the
increased amount of bare soil.
In moist-upland locations, species richness tended to
increase at all scales considered with grazing (Fig. 2),
while the slope of the species area curve (z) decreased
(Fig. 4). The decrease of z with grazing suggests a shift
towards a more homogeneous spatial plant distribu-
tion. The vegetation in these moist regions is normally
composed by a majority of herbaceous species, with
similar size and reproductive strategies (de Bello et al.
2005), that form a relatively compact vegetation layer
with scarce bare soil. In these environments grazing
might increase plant species richness by creating gaps
necessary for establishment (Rook et al. 2004; Pakeman
& Small 2005), decrease dominance of more competi-
tive species (Grime 1973; Huston 1994; Olff & Ritchie
1998; Lepš 2005) and decrease the spatial heterogeneity
of the vegetation. In fact, the presence of shrubs would
increase this spatial heterogeneity, due to their vertical
complexity and size (Huston 1994; Lepš 2005), but this
life form is relatively infrequent in the species pool of
these moist locations (Fig. 1, de Bello et al. 2005) and
Fig. 6. Grazing effect on plant species richness at different
spatial scales. (a) Olff & Ritchieʼs model (1998) with arbitrary
axes scales; (b) Schematic species-area relationship for grazed
(dashed line) and abandoned/ungrazed (solid line) communities
in the most arid location; (c) Schematic species-area relation-
ship for the most moist and cold location.
it almost disappeared under grazed conditions.
These differential effects of grazing on the species-
area relationship under different climatic conditions (i.e.
semi-arid vs. moist-temperate; Fig. 6) might depend on
the interactions between grazing and the pre-existing
spatial patterns of the vegetation. Alados et al. (2004)
found that the degree of spatial heterogeneity of the
vegetation was responsible for different trajectories in the
changes in species richness along grazing gradients under
different climatic conditions. In this sense, Adler et al.
(2001) noted also that selective grazing (e.g. sheep graz-
ing) operating on a patchy vegetation should ultimately
result in an enhanced spatial heterogeneity and contrast
between vegetation types, while on more homogeneous
vegetation the opposite pattern could be expected.
Our data confirm the observation by Adler et al.
Fig. 5. Variation in β-diversity (calculated with the Whittaker
index, see Methods) with grazing intensity on different loca-
tions.

32 de Bello, F. et al.
(2001), which also suggests that, under different climatic
conditions, the various effects of grazing on the spatial
heterogeneity of the vegetation might result in a different
partition of spatial components of diversity (i.e. α and
β). β-diversity might thus increase with grazing where
pre-existing vegetation is more heterogeneous (i.e. in
more arid conditions) and vice versa in more homogenous
vegetation (i.e. upland-moist conditions; Figs. 4 and
5). It should also be noted that in the conditions where
grazing had only positive effects on alpha diversity (i.e.
moist upland-locations in plots up to 100 m2), β-diversity
decreased. This confirms that the spatial components of
diversity (i.e. α and β) are mutually dependent (Rosen-
zweig 1995; Huston 1999) and possibly complementary
(Loreau 2000).
The hypotheses that the slope of the species-area
curve might be affected by the grazing regime (Olff &
Ritchie 1998) and that disturbance may cause an inter-
section of species-area curves of grazed and ungrazed
zones in a given location (Lande et al. 2000) were also
confirmed by our data. However, our results did not
match the prediction from the Olff & Ritchie model (Fig.
6) that grazing enhances richness at small scales (due
to reduced competition) and depletes richness at larger
scales (due to a selection of grazing-tolerant species
within the species pool). The lower z found in grazed
areas in moist-upland locations might partially support
Olff & Ritchie predictions. However the crossing of the
curves might occur far from the range of scales consid-
ered (Fig. 6) and extrapolations are unjustified outside
the range at which species area curve were originally
assessed (Lepš & Štursa 1989; Lepš 2005), because, for
example, at the landscape scale community replacement
with topographical changes are likely to occur (Sebastià
2004).
Indeed the effect of grazing on the species-area rela-
tionship might be also determined by the degree at which
the local community is linked to the species and trait
pools of the surrounding landscape (Olff & Ritchie 1998;
Pärtel 2002; Frank 2005; Reilly et al. 2006), opening
up the field for further studies. Nevertheless, deviations
from the Olff & Ritchie model can be expected for arid
areas, as the selection effect of grazing on the species
pool could be less likely to occur in these conditions. As
a matter of fact, in arid environments species often show
a suite of traits that confer common sets of adaptations
to both grazing and water limited environments (Osem
et al. 2004; de Bello et al. 2005). Facilitation processes
are also common there (Pugnaire et al. 2004). A study in
arid environments in Australia (Landsberg et al. 2002),
for example, did not show any negative effect of grazing
on species richness at regional scales (rather an increase
of diversity within 0.5-km2 paddocks).
Overall, our study shows that, similarly to predic-
tions (Milchunas et al. 1988; Huston 1994; Proulx &
Mazumder 1998; Lepš 2005), the effects of grazing on
diversity patterns changed along climatic gradients. In
our study it is further suggested that the different effects
of grazing along moisture gradients are detectable in
different spatial-area relationships. This might be basi-
cally related to a different interaction of grazing with the
pre-existing spatial heterogeneity of the vegetation, in
terms of species distribution and, possibly (as envisaged
by Adler et al. 2001), in resource distribution (Desilets
& Houle 2005). This interaction also produces different
partition of spatial components of diversity. More work is
certainly needed to analyse the evolutionary implications
of the link between the effect of grazing at the community
scale with the regional diversity and the regional pool of
species/traits (Olff & Ritchie 1998; Pärtel 2002; Díaz et
al. 2004; de Bello et al. 2005).
Indeed, the results shown in this study deserve
further comparisons with other systems. Employing an
altitudinal gradient leaves uncertainty of what exactly
drives the major changes across climatic gradients, even
assuming the fact that in our case aridity was the most
likely (Anon. 1992). At the same time, topography (Osem
et al. 2002; Sebastià 2004), grazing selectivity (Adler et
al. 2001), historical evolution of disturbance (Milchunas
et al. 1988) and time and intensity of grazing regimes
(Pakeman & Small 2005) might give further insight on
the variations of the species-area relationship and their
relevance for generalizing ecological patterns. Overall,
the main implication of these results is that to attain
general theories of biodiversity we should encourage
comparative studies in terms of spatial scaling law pat-
terns. The differential spatial effect of grazing along the
productive-moisture gradient in this study is an exam-
ple.

- Grazing effects on the species-area relationship - 33
Acknowledgements. We thank T. Torrigiani, C. Dal Zennaro
and A. Pardini for collaborating in the analysis of the point-
quadrat method and M. Taüll and P. Casals for sharing data
for stocking rates. J.M. Ninot and J.A. Conesa collaborated
in species identification. K. Edwards reviewed the English
language. Three anonymous reviewers provided important
insights thereby improving the manuscript significantly.
The research was partly funded by grants to FdB from the
University of Nuoro, the Government of Catalonia (DURSI;
FI-2002-2004 programme) and was partially developed within
the CARBOCAT and CARBOPAS projects. The Fundació Ter-
ritori i Paisatge facilitated field sampling in the Alinyà Valley.
Participation of JL in this study was partially supported by the
MSMT 600-766-5801 grant.
The generosity and cooperation of local shepherds made
this study possible.
References
Anon. (UNEP) 1992. World atlas of desertification. UNEP,
Edward Arnold, London, UK.
Adler, P.B., Raff, D.A. & Lauenroth, W.K. 2001. The effect of
grazing on the spatial heterogeneity of vegetation. Oeco-
logia 128: 465-479.
Al-Mufti, M.M., Sydes, C.L., Furness, S.B., Grime, J.P. &
Band, S.R. 1977. A quantitative analysis of shoot phenol-
ogy and dominance in herbaceous vegetation. J. Ecol. 65:
769-791.
Alados, C.L., ElAich, A., Papanastasis, V.P., Ozbek, H., Nav-
arro, T., Freitas, H., Vrahnakis, M., Larrosi, D. & Cabezudo,
B. 2004. Changes in plant spatial patterns and diversity
along successional gradients on Mediterranean grazing
ecosystems. Ecol. Model. 180: 523-535.
Bolòs, O., Vigo, J., Masalles, R.M. & Ninot, J.M. 1993. Flora
manual dels Paisos Catalans. 2nd ed. Pòrtic, Barcelona,
ES.
Canals, R.M. & Sebastià, M.-.T. 2000. Analyzing mechanisms
regulating diversity in rangelands through comparative
studies: a case in the south-western Pyrennees. Biodiv.
Conserv. 9: 965-984.
Cingolani, A.M., Noy-Meir, I. & Díaz, S. 2005. Grazing ef-
fects on rangeland diversity: a synthesis of contemporary
models. Ecol. Appl. 15: 757:-773.
Cipriotti, P.A. & Aguiar, M.R. 2005. Effects of grazing on patch
structure in a semi-arid two-phase vegetation mosaic. J.
Veg. Sci. 16: 55-66.
Daget, P. & Poissonet, J. 1971. Une méthode dʼanalyse phyto-
logique des praires. Critères dʼapplication. Ann. Agron.
22: 5-41.
Davis, M.A., Curran, C., Tietmeyer, A. & Miller, A. 2005. Dy-
namic tree aggregation patterns in a species-poor temperate
woodland disturbed by fire. J. Veg. Sci. 16: 167-174.
de Bello, F., Lepš J. & Sebastià, M.-T. 2005. Predictive value
of plant traits to grazing along a climatic gradient in the
Mediterranean. J. Appl. Ecol. 42: 824-833.
Desilets, P. & Houle, G. 2005. Effects of resource availability and
heterogeneity on the slope of the species-area curve along a
floodplain-upland gradient. J. Veg. Sci. 5: 487-496.
Díaz, S., Hodgson, J.G., Thompson, K., Cabido, M., Cor-
nelissen, J.H.C., Jalili, A., Montserrat-Marti, G., Grime,
J.P., Zarrinkamar, F., Asri, Y., Band, S.R., Basconcelo,
S., Castro-Diez, P., Funes, G., Hamzehee, B., Khoshnevi,
M., Perez-Harguindeguy, N., Perez-Rontome, M.C.,
Shirvany, F.A., Vendramini, F., Yazdani, S., Abbas-Azimi,
R., Bogaard, A., Boustani, S., Charles, M., Dehghan, M.,
de Torres-Espuny, L., Falczuk, V., Guerrero-Campo, J.,
Hynd, A., Jones, G., Kowsary, E., Kazemi-Saeed, F.,
Maestro-Martinez, M., Romo-Diez, A., Shaw, S., Siavash,
B., Villar-Salvador, P. & Zak, M.R. 2004. The plant traits
that drive ecosystems: Evidences from three continents. J.
Veg. Sci. 15: 295-304.
Etienne, M. (ed.) 1996. Western European silvopastoral sys-
tems. INRA Editions, Versailles, FR.
Frank, D.A. 2005. The interactive effects of grazing ungulates
and aboveground production on grassland diversity. Oeco-
logia 143: 629-634.
Grime, J.P. 1973. Competitive exclusion in herbaceous vegeta-
tion. Nature 242: 344-347.
He, F. & Legendre, P. 2002. Species diversity patterns derived
from species-area models. Ecology 83: 1185-1198.
Huston, M.A. 1994. Biological diversity: the coexistence of
species on changing landscapes. Cambridge University
Press, Cambridge, UK.
Huston, M.A. 1999. Local processes and regional patterns:
appropriate scales for understanding variation in diversity
of plants and animals. Oikos 86: 393-401.
Gotelli, N.J. & Colwell, R.K. 2001. Quantifying biodiversity:
procedures and pitfalls in the measurement and comparison
of species richness. Ecol. Lett. 4: 379-391.
Guo, Q. 2004. Slow recovery in desert perennial vegetation
following prolonged human disturbance. J. Veg. Sci. 15:
757-762.
Lande, R., DeVries, P.J. & Walla, T.R. 2000. When species
accumulation curves intersect: implications for ranking
diversity using small samples. Oikos 89: 601-605.
Landsberg, J., James, C.D., Maconochie, J., Nicholls, A.O.,
Stol, J. & Tynan, R. 2002. Scale-related effects of grazing
on native plant communities in an arid rangeland region
of South Australia. J. Appl. Ecol. 39: 427-444.
Lepš, J. 2005. Diversity and ecosystem function. In: van der
Maarel, E. (ed.) Vegetation ecology, pp. 199-237. Blackwell
Publishing, Oxford, UK.
Lepš, J. & Štursa, J. 1989. Species-area relationship, life his-
tory strategies and succession – a field test of relationships.
Vegetatio 83: 249-257.
Loreau, M. 2000. Are communities saturated? On the relation-
ship between a, b and g diversity. Ecol. Lett. 3: 73-76.
Magurran, A.E. 2004. Measuring biological diversity. Black-
well Publishing, Oxford, UK.
Milchunas, D.G., Sala, O.E. & Lauenroth, W.K. 1988. A gener-
alized model of the effects of grazing by large herbivores on
grassland community structure. Am. Nat. 132: 87-106.
Olff, H. & Ritchie, M.E. 1998. Effects of herbivores on grass-
land plant diversity. Trends Ecol. Evol. 13: 261-265.
Osem, Y., Perevolotsky, A. & Kigel, J. 2002. Grazing effect on
diversity of annual plant communities in a semi-arid range-
land: interactions with small-scale spatial and temporal

34 de Bello, F. et al.
variation in primary productivity. J. Ecol. 90: 936-946.
Ostling, A., Harte, J., Green, J.L. & Kinzig, A.P. 2003. A com-
munity-level fractal property produces power-law species
area relationships. Oikos 103: 218-224.
Ovaskainen, O. & Haski, I. 2003. The species-area relationship
derived from species-specific incidence functions. Ecol.
Lett. 6: 903-909.
Pakeman, R.J. & Field, C. B. 2005. The role of seed bank, seed
rain, and the timing of disturbance in gap regeneration. J.
Veg. Sci. 16: 121-130.
Pärtel, M. 2002. Local plant diversity patterns and evolutionary
history at the regional scale. Ecology 83: 2361-2366.
Perevolotsky, A. & Seligman, N.G. 1998. Role of grazing
in Mediterranean rangeland ecosystems – Inversion of a
paradigm. BioScience 48: 1007-1017.
Proulx, M. & Mazumder, A. 1998. Reversal of grazing impact
on plant species richness in nutrient-poor vs. nutrient-rich
ecosystems. Ecology 79: 2581-2592.
Reilly, M.J., Wimberly, M.C. & Newell, C.L. 2006. Wildfire
effects on plant species richness at multiple spatial scales
in forest communities of the southern Appalachians. J.
Ecol. 94: 118-130.
Rejmánek, M. & Rosén, E. 1992. Influence of colonizing shrubs
on species-area relationship in alvar plant-communities. J.
Veg. Sci. 3: 625-630.
Rook, A.J., Dumont, B., Isselstein, J., Osoro, K., WallisDeVries,
M.F., Parente, G. & Mills, J. 2004. Matching type of
livestock to desired biodiversity outcomes in pastures – a
review. Biol. Conserv. 119: 137-150.
Rosenzweig, M.L. 1995. Species diversity in space and time.
Cambridge University Press, Cambridge, UK.
Sebastià, M.-T. 2004. Role of topography and soils in grassland
structuring at the landscape and community scales. Basic
Appl. Ecol. 5: 331-346.
Vigo, J.P. & Ninot, J.M. 1987. Pirineos. In: Peinado M. &
Rivas-Martinez, S. (eds.) La vegetación de España, pp.
349-384. Servicio de Publicaciones de la Universidad de
Alcalá de Henares, Madrid, ES.
Wilson, J.B. & Chiarucci, A. 2000. Do plant communities ex-
ist? Evidence from scaling-up local species-area relations
to the regional level. J. Veg. Sci. 11: 773-775.
Received 21 September 2005;
Accepted 12 July 2006;
Co-ordinating Editor: M. Pärtel.