Nestedness and β‐diversity in ectoparasite assemblages of small mammalian hosts: effects of parasite affinity, host biology and scale

Abstract

We asked whether (a) variation in species composition of parasite assemblages on the same host species follows a non-random pattern and (b) if so, manifestation of this non-randomness across space and time differs among parasites, hosts and scales. We assessed nestedness and its contribution to β-diversity of fleas and gamasid mite assemblages exploiting small mammals across three scales: (a) within the same region across different locations; (b) within the same location across different times and (c) across distinct geographic regions. We estimated (a) the degree of nestedness (NCOL) and (b) the proportional contribution of nestedness to the total amount of β-diversity across locations, times and regions (βNESP). In the majority of host species, parasite assemblages were nested significantly across all three scales. In mites, but not fleas, NCOL correlated with the contribution of nestedness to the total amount of β-diversity. In fleas, NCOL did not differ among assemblages at the two local scales, but was significantly lower at regional scale. In mites, NCOL was the highest in assemblages at local spatial scale. βNESP was significantly higher (a) in flea than in mite assemblages at both local scales and (b) in mite than in flea assemblages at regional scale. In fleas, βNESP was higher at both local scales, whereas in mites it was higher at both local temporal and regional scales. Sheltering habits and geographic range of a host species did not affect either NCOL or βNESP in flea assemblages, but both metrics significantly decreased with an increase of geographic range of a host species in mite assemblages. We conclude that flea and mite assemblages across host populations at smaller and larger spatial scales and at temporal scale were characterized by nestedness which, in turn, contributed to an important degree to the total amount of β-diversity of these assemblages.

Full text

Page 1

630
Nestedness and b -diversity in ectoparasite assemblages of small
mammalian hosts: effects of parasite affi nity, host biology and scale
Boris R. Krasnov , Michal Stanko , Irina S. Khokhlova , Georgy I. Shenbrot , Serge Morand ,
Natalia P. Korallo-Vinarskaya and Maxim. V. Vinarski
B. R. Krasnov (krasnov@bgu.ac.il) and G. I. Shenbrot, Mitrani Dept. of Desert Ecology, Inst. for Dryland Environmental Research,
Jacob Blaustein Inst. for Desert Research, Ben-Gurion Univ. of the Negev, Sede-Boqer Campus, IL – 84990 Midreshet Ben-Gurion, Israel.
– M. Stanko, Inst. of Zoology and Parasitological Inst., Slovak Academy of Sciences, Loffl erova 10, SK – 04001 Kosice, Slovakia.
– I. S. Khokhlova , Wyler Dept. of Dryland Agriculture, French Associates Inst. For Agriculture and Biotechnology of Drylands, Jacob Blaustein
Inst. for Desert Research, Ben-Gurion Univ. of the Negev, Sede-Boqer Campus, IL – 84990 Midreshet Ben-Gurion, Israel. – S. Morand, Inst. des
Sciences de l ’ Evolution, CNRS-UM2, CC65, Universit é de Montpellier 2, FR – 34095 Montpellier, France. – N. P. Korallo-Vinarskaya ,
Laboratory of Arthropod-Borne Viral Infections, Omsk Research Inst. of Natural Foci Infections, Mira str. 7, RU – 644080 Omsk, Russia.
– M. V. Vinarski , Dept. Zoology and Physiology, Faculty of Chemistry and Biology, Omsk State Pedagogical Univ., Tukhachevskogo emb. 14,
RU – 644099 Omsk, Russia.
We asked whether (a) variation in species composition of parasite assemblages on the same host species follows a non-
random pattern and (b) if so, manifestation of this non-randomness across space and time diff ers among parasites, hosts
and scales. We assessed nestedness and its contribution to β -diversity of fl eas and gamasid mite assemblages exploiting
small mammals across three scales: (a) within the same region across diff erent locations; (b) within the same location across
diff erent times and (c) across distinct geographic regions. We estimated (a) the degree of nestedness (N COL ) and (b) the
proportional contribution of nestedness to the total amount of β -diversity across locations, times and regions ( β NESP ). In
the majority of host species, parasite assemblages were nested signifi cantly across all three scales. In mites, but not fl eas,
N COL correlated with the contribution of nestedness to the total amount of β -diversity. In fl eas, N COL did not diff er among
assemblages at the two local scales, but was signifi cantly lower at regional scale. In mites, N COL was the highest in assem-
blages at local spatial scale. β NESP was signifi cantly higher (a) in fl ea than in mite assemblages at both local scales and (b)
in mite than in fl ea assemblages at regional scale. In fl eas, β NESP was higher at both local scales, whereas in mites it was
higher at both local temporal and regional scales. Sheltering habits and geographic range of a host species did not aff ect
either N COL or β NESP in fl ea assemblages, but both metrics signifi cantly decreased with an increase of geographic range of a
host species in mite assemblages. We conclude that fl ea and mite assemblages across host populations at smaller and larger
spatial scales and at temporal scale were characterized by nestedness which, in turn, contributed to an important degree to
the total amount of β -diversity of these assemblages.
A set of co-occurring species may either be assembled accord-
ing to certain rules or represent an unstructured, random
collection of species. Vigorous search for non-randomness
in biological communities over several decades has revealed
a variety of structural patterns (Diamond 1975, Hanski
1982, Patterson and Atmar 1986). As a result, it is com-
monly accepted that, in general, biological communities are
structured. Most of the studies have dealt with free-living
organisms, whereas parasites attracted less attention. When
parasite communities were investigated, their structure did
not appear to fi t into a general rule (Gotelli and Rohde 2002,
Gouy de Bellocq et al. 2003, Krasnov et al. 2006a, 2010a).
Instead, parasite communities fell along the entire gradient
from being randomly assembled (Gotelli and Rohde 2002)
to being structured (Krasnov et al. 2006a). When the struc-
ture in these communities was revealed, its strength appeared
to vary (a) among diff erent parasites exploiting the same host
(Gonz á lez and Oliva 2009) and (b) among the same parasites
exploiting diff erent hosts (Poulin and Valtonen 2001, Krasnov
et al. 2005a, 2010a). Furthermore, communities of the same
parasites exploiting the same hosts in the same location were
structured at some times, but were randomly assembled at
other times (Krasnov et al. 2006a). Although inconsistencies
in detection of community structure of parasites could be
associated with the methodological issues (Timi and Poulin
2008), we reasoned that the above-mentioned contradictory
results might also be related to parasite affi nity, host biology
and scale of consideration.
Species composition of a community of any organisms,
including parasites, varies in space and time. Consequently,
a structure of a community at any given location and time
represents only a snapshot of a more general pattern. Th ere-
fore, the questions arise: (a) whether variation in species
composition of a community is random or it follows some
Oikos 120: 630–639, 2010
doi 10.1111/j.1600-0706.2010.19072.x
© 2011 Th e Authors. Oikos © 2011 Nordic Society Oikos
Subject Editor: Werner Ulrich. Accepted 6 September 2010

Page 2

631
non-random pattern such as, for example, the nested subsets
of species and (b) does the manifestation of non-randomness
in variation of species composition of a community across
space and time diff er among parasites, hosts and scales
(Gonz á lez and Oliva 2009)?
Here, we addressed these questions using data on varia-
tion in species composition of two taxa of haematophagous
arthropods, fl eas and gamasid mites, exploiting small Palae-
arctic mammals across three scales, namely (a) local spatial
scale, that is, across diff erent locations within the same
region; (b) local temporal scale, that is, within the same
location across diff erent times and (c) regional scale, that
is, across distinct geographic regions. We assessed nested-
ness of species composition of fl ea and mite assemblages of
the same host species and its relative contribution to total
amount of change in a community composition, that is,
β -diversity (Baselga 2010). We used NODF, a new metric of
nestedness proposed recently by Almeida-Neto et al. (2008).
Th is metric possesses better statistical properties than ear-
lier metrics. Although the concept of β -diversity is related
originally to variation of community composition among sites
(Whittaker 1960), it can also be applied to variation in com-
munity composition within a site among times (DeVries
et al. 1997). Two additive antithetic processes, namely spa-
tial species turnover and nestedness, contribute to the total
amount of β -diversity (Baselga et al. 2007). Spatial turnover
is the replacement of some species by others (Harrison et al.
1992), while nestedness is a pattern in which species com-
prising poorer assemblages constitute non-random subsets of
the species occurring in successively richer assemblages
(Patterson and Atmar 1986, Wright et al. 1998). Any case of
non-identical assemblages may be described using only these
two patterns or their combinations (see Baselga 2010 for
details). Th e reason behind the selection of nestedness as an
indicator of parasite community structure was that nested-
ness proved to be a common pattern in communities of insu-
lar or fragmented habitats (Patterson and Atmar 1986), and
parasites can be viewed as inhabitants of biological islands
represented by their hosts. In fact, the search for nested pat-
terns has been applied often to parasite communities (Poulin
1996, Poulin and Valtonen 2001, Timi and Poulin 2003,
Krasnov et al. 2005a).
Fleas are obligate haematophagous ectoparasites. In most
fl ea species, all stages of the life cycle are spent off the host,
except for the adults that feed intermittently on the host.
In contrast, gamasid mites vary in their ecology and feeding
modes. Here, we focused on mites collected from host body
surfaces which include facultative and obligatory haemato-
and/or lymphophagous as well as phoretic (that is, using a
host as a vehicle rather than a food source) mites.
Recently, we evaluated structure of ectoparasite commu-
nities in rodent hosts from South Africa, South America and
west Siberia (Krasnov et al. 2010a). We found that the com-
munity structure in terms of species co-occurrences was bet-
ter expressed (a) in fl eas than in mites and (b) in hosts with
permanent shelters than in hosts with ephemeral shelters (see
discussion of mechanisms in Krasnov et al. 2010a). Given
negative relationships between nestedness and species co-
occurrence (see details in Ulrich et al. 2009), we predicted
that nestedness will be less pronounced and its contribution
to variation of species composition of parasite assemblages
will be lower in (a) fl eas than in mites and (b) in hosts with
well developed permanent burrows (e.g. Microtus voles) than
in hosts with ephemeral shallow burrows and/or above-
ground nests (e.g. Myodes voles, Apodemus mice and shrews).
Another earlier study that implied measurement of nested-
ness via matrix temperature suggested that host species that
occupy larger geographic ranges were more likely to have
a nested distribution of parasites (specifi cally, fl eas) among
their populations (Krasnov et al. 2005a). Consequently,
we predicted that (c) the degree of nestedness measured via
NODF and contribution of nestedness to β -diversity will
increase with an increase of host ’ s geographic range.
Comparison of nestedness pattern among scales has rarely
been done. Sfenthourakis et al. (2004) compared the nest-
edness values for communities of terrestrial isopods among
three spatial scales and found that these values were similar.
Nevertheless, we predicted that (d) nestedness of parasite
assemblages will be expressed stronger at regional scale. Th is
is because parasite assemblages of the same host across diff er-
ent locations within the same region or within the same loca-
tion across times are governed by epidemiological processes
such as birth and death dynamics acting at the level of each
parasite species (Morand et al. 2002), while parasite assem-
blages of the same host species across diff erent regions might
be aff ected by biogeographical processes such as, for instance,
ecological fi tting (Brooks et al. 2006). Consequently, the
regional scale seems to be more relevant to the original idea
of the nested subsets pattern (Patterson and Atmar 1986).
Material and methods
Data sets
We used three sets of data on fl eas and gamasid mites parasitic
on small mammals (rodents and soricomorphs). Data for the
‘ local spatial ’ set were collected during 81 trapping sessions
between 1983 and 2001 in seven locations across Slovakia,
while data for the ‘ local temporal ’ set were collected during
26 trapping sessions between 1988 and 1990 near the town
of Trebi š ov (48 ° 41 ′ E, 21 ° 45′ N) in the east Slovakian low-
lands. For both sets, we used data collected during the warm
season (April – October). We used data on fl eas and mites col-
lected from six most common host species that included fi ve
rodents ( Apodemus agrarius , Apodemus fl avicollis , Apodemus
uralensis , Myodes glareolus and Microtus arvalis ) and a shrew
( Sorex araneus ). In total, the ‘ local spatial ’ set comprised
data on 9505 individual fl eas belonging to 28 species and
51 212 individual mites belonging to 89 species collected from
8566 individual hosts. Th e ‘ local temporal ’ set comprised
data on 3933 individual fl eas belonging to six species and
8949 individual mites belonging to 52 species collected from
3049 individual hosts. In the ‘ local spatial ’ set, data collected
in the same location were pooled across time periods (see
Krasnov et al. 2006b for explanations).
Trapping sessions (on average, 700 traps per session, rang-
ing from 100 to 2000 traps) lasted one to three nights. Mam-
mals were captured using traps that were deposited following
the same protocol at each trapping session. Each trapped
animal was identifi ed, sexed, weighed and euthanized with
sulfur ether. Th us, diff erent host individuals were captured

Page 3

632
during diff erent trapping sessions, allowing the avoidance of
pseudoreplications. Th en, an animal was placed in an indi-
vidual pre-marked plastic or cloth bag and transferred to a
laboratory where it was examined for ectoparasites. Th e ani-
mal ’ s fur was combed thoroughly, using a toothbrush, over a
plastic pan and fl eas and mites were carefully collected.
Data for the ‘ regional ’ set were obtained from published
surveys (60 surveys in 52 regions for fl eas and 25 surveys
in 25 regions for mites) and unpublished data (four surveys
in one region for mites) that reported the number of fl eas
or mites of each species found on a given number of indi-
viduals of each host species in a particular location (see map
of the regions, lists of parasite and host species and refer-
ences in Krasnov et al. 2010b). We included in the dataset
hosts (a) for which at least 10 individuals were examined
per region, (b) that occurred in at least four regions and (c)
for which data on both fl eas and mites were available. Th is
resulted in 14 rodent and three soricomorph species (Table
2) exploited by 129 fl ea and 51 mite species (Supplementary
material Appendix 2). Among these hosts, six species (four
Microtus species, Cricetus cricetus and Arvicola amphibius )
construct deep, complex and permanent underground shel-
ters, whereas the remaining 11 hosts build ephemeral shal-
low burrows or above-ground nests.
Lists of fl ea and mite species for the ‘ local spatial/local
temporal ’ datasets and the ‘ regional ’ dataset are presented in
Supplementary material Appendix 1 and 2 respectively. Raw
data (matrices of presence/absence) are provided in Supple-
mentary material Appendix 3.
Data organization and analyses
Th e data were organized as presence/absence matrices in which
rows represented fl ea or mite species, whereas columns repre-
sented locations (for local spatial scale), time periods (for local
temporal scale) or regions (for regional scale). Presence/absence
matrices were constructed for each host species separately.
For each matrix, we estimated (a) nestedness and (b) the con-
tribution of nestedness to the total amount of β -diversity across
locations, times and regions. A number of nestedness metrics has
been developed such as the matrix temperature T (the amount
of disorder in the system relative to perfect nestedness; Attmar
and Patterson 1993), the C metric (the standardized number
of times a species presence correctly predicts its occurrence at
equally rich or richer sites; Wright and Reeves 1992), and the
discrepancy measure d (the number of absences or presences
that must be erased to produce a perfectly nested matrix; Brualdi
and Sanderson 1999). Each of these metrics has its own merits
and shortcomings (reviewed by Almeida-Neto et al. 2008). Th e
most important problem with these metrics is that they often
erroneously detect a nested pattern, and are thus prone to type
I statistical errors (Almeida-Neto et al. 2008). Almeida-Neto
et al. (2008) proposed a new metric, NODF (nestedness met-
ric based on overlap and decreasing fi ll), that has been proved
to possess better statistical properties. As implied by its
name, this metric is based on standardized diff erences in row
and column fi lls and paired matching of occurrences. One
of the most important and appealing features of NODF
is that it allows the calculation of nestedness indepen-
dently among rows or columns, thus providing a possibil-
ity to evaluate nestedness only among sites or only among
species. In addition, absolute values and Z-transformed scores
of NODF were found to be insensitive to matrix shape and
size, although sensitive to matrix fi ll (except for Z-scores under
the fi xed-fi xed null model) (Almeida-Neto et al. 2008). In
our dataset, absolute values of NODF did not correlate with
matrix fi ll except for fl ea assemblages at regional scale (after
angular transformations; Spearman ’ s correlation coeffi cients
for ‘ local spatial ’ , ‘ local temporal ’ and ‘ regional ’ datasets were
– 0.19 (p ? 0.05), – 0.38 (p ? 0.05) and 0.67 (p ? 0.05),
respectively, for fl ea assemblages and 0.57, – 0.16, 0.30
(p ? 0.05 for all), respectively, for mite assemblages). Visual
investigation of the scatterplot of the absolute NODF value
and matrix fi ll for fl ea assemblages at regional scale indicated
that signifi cance of the correlation between these two vari-
ables was due to a single host species ( Apodemus fl avicollis ).
After removal of this species, correlation between absolute
NODF and matrix fi ll for this dataset became non-signifi cant
(Spearman ’ s correlation coeffi cient was 0.39, p ? 0.05). We
used NODF and evaluated nestedness among columns (that
is, nestedness in species composition among locations, times
or regions; N COL ). N COL were calculated using program
ANINHADO (Guimaraes and Guimaraes 2006) (note that
in the original presentation of ANINHADO and in contrast
to our data organization, columns represent species, while
rows represent locations). For each matrix, we also calcu-
lated a standardized eff ect size (SES) as a Z-transformed score
and compared the observed index to the distribution of indices
calculated for 1000 randomly assembled null matrices (Gure-
vitch et al. 1992), thereby measuring the tail probability that
an observed index is larger than expected by chance. Simu-
lated matrices were assembled by Monte Carlo procedures.
For the sake of biological realism, presences were randomly
assigned within the columns (that is, within a location, time
or region for each host species), but not within the rows (that
is, among locations, times or regions) of simulated matrices.
Th is is because the presence of a fl ea or mite species on a given
host within a locality/region or at a time may be caused by
horizontal transfer from co-occurring hosts (Krasnov et al.
2004a), while the presence of a fl ea or mite species on a given
host among localities/times/regions is determined by parasite ’ s
environmental preferences (Krasnov et al. 2006b, Vinarski
et al. 2007).
To estimate contribution of nestedness to spatial patterns
of variation in species composition of fl ea and mite assem-
blages, we used a technique developed by Baselga (2010).
Th is approach allows the separation of the contribution of
spatial turnover and nestedness to β -diversity pattern. Given
that β -diversity is a measure of similarity between sites
(Koleff et al. 2003), the total amount of β -diversity may
be estimated using multiple-site metric ( β SOR ) based on the
S ø rensen dissimilarity measure (Baselga et al. 2007, Baselga
2010). Th e latter encompasses both spatial turnover and dif-
ferences in species richness. Th is total amount of β -diversity
may further be partitioned into two components. Measure of
multi-site spatial turnover free from the infl uence of richness
( β SIM ) is based on the Simpson dissimilarity index (Baselga
et al. 2007, Baselga 2010) and involves construction of mul-
tiple-site equivalents of the matching components of indices
(that is, species shared and not shared by assemblages; see
Baselga et al. 2007 for details). If assemblages are composed
from equal number of species, then their dissimilarity is due

Page 4

633
and standardized contrasts in host geographic range size were
computed as forced through the origin.
Prior to the analyses, N COL and β NESP were angular trans-
formed, while geographic range size was log-transformed.
Th ese transformations produced distributions that did not
deviate signifi cantly from normality (Kolmogorov – Smirnov
tests, p ? 0.20). Untransformed data are presented in fi g-
ures (except for illustrations of the results of the independent
contrasts method).
Results
Th e degree of nestedness and its contribution to variation in
species composition of parasite assemblages are presented in
Table 1. Local fl ea and mite assemblages were signifi cantly
nested across spatial scale in fi ve and four of six host spe-
cies, respectively (except for mites in A. uralensis and both
parasite taxa in S. araneus ). At temporal scale, assemblages
of both fl eas and mites were signifi cantly nested in all
hosts except for S. araneus . Regional fl ea or mite assemblages
were signifi cantly nested in 15 of 17 host species. Further-
more, in fi ve host species ( A. uralensis , C. cricetus , M. agrestis ,
N. fodiens and S. betulina ) signifi cant nestedness was found
in assemblages of only one of the two parasite taxa, while the
structure of assemblages of both taxa did not signifi cantly
deviate from randomness in M. gregalis and S. caecutiens .
In fl ea assemblages, the degree of nestedness did not cor-
relate with either β NES or β NESP independently of whether
host species with non-signifi cant nestedness of these assem-
blages were included in the analysis (Table 2). In mite assem-
blages, however, these variables correlated signifi cantly and
positively (Table 2). After removal of host species in which
nested structures of mite assemblages were not signifi cant,
positive correlations between N COL and β NES or β NESP
remained signifi cant for regional, but not local spatial and
temporal mite assemblages (Table 2).
Results of repeated measure ANOVAs of the eff ects of
scale and parasite taxon on N COL and β NESP are presented
in Table 3. In general, the degree of nestedness was signifi -
cantly aff ected by scale (univariate tests for signifi cance for
planned comparison; F ? 4.96 for fl eas and F ? 20.08
for mites; p ? 0.05 for both). In fl eas, N COL did not dif-
fer between the assemblages at the two local scales (Tukey ’ s
HSD test for unequal samples, p ? 0.72), while N COL at
regional scale was signifi cantly lower than those at both local
scales (Tukey ’ s HSD tests for unequal samples, p ? 0.002 –
0.04; Fig. 1). In mites, N COL was signifi cantly higher in the
assemblages at local spatial scale than in those at the two
other scales (Tukey ’ s HSD tests for unequal samples, p ?
0.002 – 0.008; Fig. 1), but did not diff er between assemblages
at local temporal and regional scales (Tukey ’ s HSD test for
unequal samples, p ? 0.98). Interaction between these two
factors was signifi cant, suggesting that the eff ect of parasite
affi nity was scale-dependent. It was signifi cantly higher in
mite than fl ea assemblages at local spatial scale (univari-
ate test for signifi cance for planned comparison F ? 5.45,
p ? 0.04), but did not diff er between the two taxa at local
temporal and regional scales (univariate tests for signifi cance
for planned comparison F ? 3.42 and F ? 0.91, respec-
tively, p ? 0.09 for both) (Fig. 1). When the entire ‘ regional ’
to ‘ pure ’ turnover because nestedness is not possible. Con-
sequently, the total amount of β -diversity will be equal to
the amount of turnover ( β SOR ? β SIM ). As a result, diff er-
ence between β SOR and β SIM can be considered as nestedness-
resulted dissimilarity or β NES ? β SOR – β SIM (Baselga 2010).
It should be stressed that β NES it is not a measure of nested-
ness but rather a measure of the dissimilarity among assem-
blages due to nestedness (Baselga 2007). For each matrix, we
calculated β SOR , β SIM and β NES using function ‘ β -multi.R ’
for R software environment (R Development Core Team
2009) compiled by Baselga (2010). Host species diff ered in
the number of localities, time periods or regions in which
they occurred (4 – 8, 7 – 25 and 4 – 21, respectively). To make
measures of dissimilarity computed for these hosts compa-
rable, we followed Baselga (2010) and calculated β SOR , β SIM
and β NES for hosts that occurred in more than four localities,
seven time periods or four regions using resampling proce-
dures. For each of these hosts, we took 100 random samples
of fi ve localities, seven time periods or four regions and aver-
aged metrics across these samples. Th en, we calculated the
proportional contribution of nestedness to spatial variation
of parasite species composition across localities, times and
regions as a quotient of β NES and β SOR (thereafter β NESP ).
To test whether either β NES or β NESP is related to the degree
of nestedness of these assemblages, we carried out correlation
analyses. Th is was done separately for fl eas and mites and for
the ‘ local spatial ’ , ‘ local temporal ’ and ‘ regional ’ datasets.
To test for the eff ect of parasite affi nity and scale on the
degree of nestedness (N COL ) and its contribution to the total
amount of β -diversity ( β NESP ), we compared these variables
between parasite taxa and among scales using repeated mea-
sures ANOVAs (with parasite taxon as a between-groups fac-
tor and scale as within-subject factor). Th ese analyses were
conducted using six host species for which we have data col-
lected at all three scales. In addition, we compared N COL and
β NESP between fl eas and mites within the entire ‘ regional ’ set
using ANOVAs.
Th e eff ect of host sheltering habits on N COL and β NESP
was tested within parasite taxon and between hosts with per-
manent and ephemeral shelters using t-tests for independent
samples, while the eff ect of the size of the geographic range
of a host on the degree of nestedness and β NESP was analyzed
within parasite taxon and among hosts with diff erent geo-
graphic ranges using the method of independent contrasts
(Felsenstein 1985). Th is was done for the ‘ regional ’ dataset
only due to the lack of host species with permanent shelters
in both local datasets (except for M. arvalis ).
For the implementation of the method of independent
contrasts, we used the global phylogenetic supertree for
mammals of Bininda-Emonds et al. (2007) as the main
source of phylogenetic information (topology and branch
length). We transformed topology of some of the branches
to resolve the relationships within Microtus and Myodes gen-
era as well as among subfamilies within Muridae according
to Jaarola et al. (2004), Busan et al. (2008) and Steppan
et al. (2004), respectively. We computed independent con-
trasts using the PDAP:PDTREE program (Midford et al. 2008)
implemented in the mesquite modular system for evolutionary
analysis (Maddison and Maddison 2009). Th en, correlations
between standardized contrasts in the degree of nestedness or its
proportional contribution to the total amount of β -diversity

Page 5

634
Table 1. Summary of the results of nestedness and β -diversity analyses of fl ea and mite assemblages on six small mammalian host species at
local spatial (LS), local temporal (LT) and regional (R) scales (see text for explanations). N COL – nestedness among columns, SES – standardized
effect size, LCL and UCL – lower and upper 95% confi dence limits of the null distribution, β NESP – proportional contribution of nestedness
to spatial or temporal variation of parasite species composition within a host species.
Fleas Mites
ScaleHost N COL SESLCLUCL β NESP N COL SES LCLUCL β NESP
LS Apodemus agrarius
Apodemus fl avicollis
Apodemus uralensis
Microtus arvalis
Myodes glareolus
Sorex araneus
Apodemus agrarius
Apodemus fl avicollis
Apodemus uralensis
Microtus arvalis
Myodes glareolus
Sorex araneus
Apodemus agrarius
Apodemus fl avicollis
Apodemus uralensis
Arvicola amphibius
Cricetus cricetus
Eutamias sibiricus
Microtus agrestis
Microtus arvalis
Microtus gregalis
Microtus oeconomus
Myodes glareolus
Myodes rufocanus
Myodes rutilus
Neomys fodiens
Sicista betulina
Sorex araneus
Sorex caecutiens
14.68 * *
21.17 * * *
30.00 * *
30.95 * * *
41.37 * * *
25.00 ns
33.89 * * *
35.50 * * *
32.17 * * *
41.81 * * *
36.26 * * *
35.71 ns
6.94 * * *
33.04 * * *
4.90 *
4.81 *
11.11 ns
10.00 * * *
9.40 * * *
5.58 * * *
7.34 ns
12.89 * *
7.96 * * *
5.50 ns
10.36 * * *
20.29 * * *
15.09 ns
4.87 * *
15.87 ns
2.40
3.99
2.40
11.59
5.43
0.00
7.63
5.41
5.23
4.56
4.79
0.99
3.02
2.68
1.83
2.05
?2.23
2.66
4.43
8.23
0.74
2.20
7.94
0.23
7.55
5.90
?0.06
2.21
1.08
10.57
13.33
23.27
23.50
11.02
24.69
26.76
28.05
27.09
28.56
31.02
28.50
5.14
24.81
3.73
2.97
24.24
4.55
5.41
2.84
6.42
10.45
4.09
5.27
6.17
10.89
15.09
3.61
9.71
10.77
13.57
23.61
23.90
11.23
25.33
26.87
28.22
27.21
28.76
31.31
29.34
5.21
25.18
3.81
3.08
24.99
4.79
5.52
2.88
6.56
10.59
4.15
5.37
6.24
11.09
15.58
3.68
10.38
0.56
0.45
0.83
0.45
0.49
0.33
0.81
0.87
0.94
0.81
0.80
0.36
0.09
0.13
0.08
0.09
0.08
0.12
0.16
0.14
0.11
0.09
0.19
0.11
0.12
0.13
0.09
0.18
0.08
58.79 * * *
39.54 * * *
38.68 ns
57.22 * * *
59.45 * * *
18.73 ns
32.66 * * *
26.73 * * *
25.31 * * *
34.40 * * *
21.08 * * *
0.00 ns
10.97 * * *
16.67 *
11.79 ns
12.24 * * *
47.50 * * *
8.04 * * *
16.00 ns
9.22 * * *
9.56 ns
7.99 * * *
14.06 * * *
16.24 * * *
14.58 * * *
27.00 ns
12.50 *
5.24 *
15.56 ns
6.03
8.28
1.08
6.68
7.76
?0.93
24.30
20.21
14.00
10.78
8.56
?1.24
3.48
1.95
0.72
4.65
4.15
2.93
1.07
3.17
1.38
9.07
4.94
8.91
7.85
?0.41
2.29
1.80
0.58
44.04
30.31
35.57
44.82
37.29
23.52
14.74
10.92
12.70
14.74
17.40
14.04
6.53
7.65
10.48
7.37
25.76
3.82
9.90
6.00
8.07
3.15
8.02
7.41
8.12
30.10
5.38
3.14
11.85
44.34
30.45
35.91
45.05
37.65
24.20
14.83
11.02
12.81
14.83
17.60
15.53
6.69
8.20
10.68
7.50
26.40
4.00
10.57
6.13
8.20
3.22
8.17
7.54
8.22
31.22
5.75
3.28
12.56
0.16
0.18
0.14
0.16
0.14
0.07
0.40
0.46
0.35
0.46
0.22
0.00
0.21
0.53
0.21
0.26
0.66
0.35
0.24
0.30
0.28
0.31
0.24
0.31
0.26
0.16
0.24
0.26
0.18
LT
R
* * * – p ? 0.001, * * – p ? 0.01, * – p ? 0.05, ns – non-signifi cant.
Table 2. Pearson product-moment correlations between the degree of nestedness and absolute ( β NES ) and proportional ( β NESP ) contributions
of nestedness to the total amount of β -diversity in fl ea and mite assemblages at local spatial, local temporal and regional scales for all
studied hosts (A) and after removal of host species in which nested structures of parasite assemblages were not signifi cant (B).
Local spatial scaleLocal temporal scaleRegional scale
Parasites Host subset β NES β NESP β NES β NESP β NES β NESP
FleasA
B
A
B
0.10 ns
0.02 ns
0.84 *
0.26 ns
0.07 ns
0.04 ns
0.81 *
– 0.03 ns
0.27 ns
– 0.61 ns
0.98 *
0.76 ns
– 0.19 ns
0.64 ns
0.98 *
0.79 ns
– 0.08 ns
0.05 ns
0.49 *
0.76 *
– 0.05 ns
0.02 ns
0.51 *
0.78 *
Mites
* – p ? 0.05, ns – non-signifi cant.
dataset was analyzed separately from the ‘ local spatial ’ and
‘ local temporal ’ datasets, N COL did not diff er between fl ea
and mite assemblages (F ? 2.56, p ? 0.11).
Th e proportional contribution of nestedness to the total
amount of β -diversity depended on both parasite taxon and
scale as well as on the interaction between the two factors
(Table 3). It was signifi cantly higher in fl ea than in mite
assemblages at both local scales (univariate tests for signifi -
cance for planned comparison F ? 22.39 and F ? 12.77,
p ? 0.01 for both; Fig. 2) and in mite than in fl ea assem-
blages at regional scale (univariate test for signifi cance for
planned comparison F ? 9.24, p ? 0.01; Fig. 2). β NESP of
both fl ea and mite assemblages diff ered signifi cantly among
scales (univariate tests for signifi cance for planned compari-
son F ? 22.05 and F ? 6.29, respectively; p ? 0.03 for both).
In fl ea assemblages, nestedness contributed similarly to the
total amount of β -diversity at both local scales (Tukey ’ s HSD
test for unequal samples, p ? 0.08), but contributed signifi -
cantly less at regional scale (Tukey ’ s HSD tests for unequal
samples, p ? 0.0006 – 0.006) (Fig. 2). In mite assemblages,
β NESP was signifi cantly lower at local spatial scale than at both
local temporal and regional scales (Tukey ’ s HSD tests for
unequal samples, p ? 0.05 for both comparisons), whereas
it did not diff er between the two latter scales (Tukey ’ s HSD
test for unequal samples, p ? 0.46) (Fig. 2). Removal of
host species in which nested structures of assemblages were

Page 6

635
Table 3. Summary of the repeated measures ANOVAs of the effects of parasite taxon (PT; fl eas or mites) and scale (local spatial, local tempo-
ral and regional) on the degree of nestedness (N COL ) of parasite assemblages and the proportional contribution of nestedness to the total
amount of β -diversity ( β NESP ) of these assemblages.
Independent variableEffect Sum of squaresDFFp
N COL PT
Error
Scale
Scale ? PT
Error
PT
Error
Scale
Scale ? PT
Error
0.01
0.45
0.44
0.22
0.28
0.49
0.29
0.63
1.13
0.79
10.220.65
10
2
2
15.65
7.72
? 0.001
0.003
20
β NESP
1 16.350.002
10
2
2
7.95
14.33
0.003
? 0.001
20
0.0
10.0
20.0
30.0
40.0
50.0
60.0
Local spatialLocal temporal
Scale
Regional
Degree of nestedness
Figure 1. Th e degree of nestedness (means ? SE) in fl ea (black
columns) and mite (white columns) assemblages on small mam-
malian hosts at local spatial, local temporal and regional scales (see
text for explanations).
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Proportional contribution of nestedness
to beta-diversity
Local spatial Local temporal
Scale
Regional
Figure 2. Th e proportional contribution of nestedness (means ? SE)
to the total amount of β -diversity in fl ea (black columns) and mite
(white columns) assemblages on small mammalian hosts at local spa-
tial, local temporal and regional scales (see text for explanations).
not signifi cant did not generally change these results, except
that β NESP of fl ea assemblages diff ered signifi cantly between
the two local scales, being higher at temporal scale (Tukey ’ s
HSD tests for unequal samples, p ? 0.001 for all). When
the ‘ regional ’ dataset was analyzed separately, β NESP was sig-
nifi cantly lower in fl ea than in mite assemblages (0.11 ?
0.01 vs 0.29 ? 0.03; F ? 40.46, p ? 0.001). Th is remained
true after removal of hosts in which nested structure of either
fl ea or mite assemblage was not signifi cant (0.12 ? 0.01 for
fl eas vs 0.33 ? 0.04 for mites; F ? 32.53, p ? 0.001).
Sheltering habits of a host species aff ected neither N COL
nor β NESP of fl ea and mite assemblages (F ? 0.93 and
F ? 0.20 for fl eas and F ? 0.24 and F ? 1.33 for mites,
respectively; p ? 0.26 for all). No correlation between
N COL or β NESP with the size of the geographic range of
a host was found (r ? – 0.12 and r ? 0.26, respectively,
p ? 0.50 for both) in fl ea assemblages. Th is remained true
after host species with non-signifi cant nestedness of fl ea
assemblages were omitted from the analyses (r ? – 0.26 and
r ? 0.24, respectively, p ? 0.40 for both). In contrast, both
metrics signifi cantly decreased with an increase in the size of
host geographic range in mite assemblages (r ? – 0.57 and
r ? – 0.55, respectively, p ? 0.02 for both; Fig. 3). Removal
of host species with non-signifi cant nestedness of mite assem-
blages resulted in even stronger correlation (r ? – 0.83 and
r ? – 0.80, respectively, p ? 0.002 for both).
Discussion
Results of this study suggest the following. Th e degree of
nestedness was strongly aff ected by scale, being the lowest at
the regional scale, but was only aff ected by parasite taxon at
local spatial scale (stronger in mites). In addition, the degree
of nestedness was similar between host species with diff erent
sheltering habits, did not depend on the size of host geo-
graphic range in fl eas, but negatively correlated with the size
of host geographic range in mites. Th e proportional contri-
bution of nestedness to β -diversity diff ered between fl eas
and mites as well as among scales, was not aff ected by hosts ’
sheltering habits and decreased with an increase of the host
geographic range size. Taken together, patterns of the eff ect
of parasite taxon, host biology and scale on the degree of
nestedness and its contribution to β -diversity suggest that

Page 7

636
Contrasts in geographic range size
-0.06
-0.18 -0.16 -0.14 -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06
Contrasts in geographic range size
-0.18 -0.16 -0.14 -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
Contrasts in NCOL
-0.08
-0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
Contrasts in βNESP
(a)
(b)

Figure 3. Relationships between phylogenetically independent con-
trasts in the size of geographic range of a host species and phyloge-
netically independent contrasts in (a) the degree of nestedness and
(b) the proportional contribution of nestedness to the total amount
of β -diversity of mite assemblages.
the relationships between nestedness and β -diversity are
complicated.
Nestedness of parasite assemblages
Earlier studies reported that nestedness in parasite
assemblages (including, for example, fl eas) is rather uncom-
mon (Poulin and Valtonen 2001, Krasnov et al. 2005a). In
contrast, signifi cant nestedness was found in the majority of
fl ea and mite assemblages we tested. Th ere can be at least
two reasons for this inconsistency. Firstly, the majority of
earlier studies considered parasite assemblages across host
individuals (i.e. infracommunities), whereas this study
considered parasite assemblages across host populations
(i.e. component communities). Indeed, studies of nested-
ness of parasite assemblages across host populations often
demonstrated signifi cant nestedness. For example, Gouy de
Bellocq et al. (2003) reported signifi cant nestedness in hel-
minth communities across populations of the woodmouse
Apodemus sylvaticus . Component communities of metazoan
parasites of adult Argentinian anchovy Engraulis anchoita ,
were found to be nested, although this was true for some
but not other seasons (Timi and Poulin 2003). In contrast,
Krasnov et al. (2005a) reported that component commu-
nities of fl eas were signifi cantly nested in only six of 25
mammalian host species. Moreover, when parasite assem-
blages of the same host were tested for nestedness at both
hierarchical levels (that is, component communities and
infracommunities), the former appeared to be generally
nested, while the latter did not (Simkova et al. 2001 for
monogenean assemblages on the roach Rutilus rutilus ).
Secondly, in contrast to other metrics, NODF allows
not only to evaluate nestedness of an entire matrix, but also
to disentangle nestedness among columns (that is, among
locations, times or regions in our matrices) from nested-
ness among rows (that is, among species in our matrices).
Th e former corresponds to the nestedness in species com-
position among sites, whereas the latter corresponds to spe-
cies occupancy. Furthermore, the degree of nestedness for
species composition was shown to be higher than that for
species occupancy in the same matrices (Almeida-Neto
et al. 2008). In this study, we evaluated nestedness in col-
umns only, whereas earlier studies on parasite assemblages
evaluated nestedness of the entire matrices.
Parasite taxon-dependence
Diff erences in structure among parasite taxa have been
reported earlier. For example, Krasnov et al. (2010a) found
that at local spatial scale, community structure was expressed
stronger in fl eas than in mites in terms of positive species
co-occurrences. Th is diff erence was explained by diff erence
in feeding ecology between these taxa with fl eas being obli-
gate haematophages and mites possessing a variety of feeding
modes. A lower dependence of mites on the host as a food
source could be the reason underlying lower frequency of
detection of non-randomness in their assemblages. Ulrich
et al. (2009) demonstrated relatedness between nestedness
and species co-occurrence with Z-scores of the co-occurrence
and nestedness metrics (NODF) being negatively correlated.
Although this correlation was strong under the equiprobable
model (no constraints on marginal totals), it disappeared
under the fi xed-fi xed model (fi xed numbers of occurrences
within rows and columns; see Ulrich et al. 2009 for details).
Th e null model used in our study was less liberal than the
null equiprobable model, but less conservative than the null
fi xed-fi xed model. A higher nestedness in mite assemblages
and a lower frequency of their positive co-occurrences at the
local spatial scale represent thus an example of the relation-
ships between the two patterns of structure found by Ulrich
et al. (2009).
Th ere is evidence that more vagile taxa tend to demon-
strate higher degrees of nestedness (Simaiakis and Martinez-
Morales 2010; but see Wright et al. 1998). Although active
dispersal abilities of gamasid mites are undoubtedly inferior
to those of fl eas, their passive dispersal (e.g. phoresy on
a variety of hosts including accidental ones which is not
characteristics for fl eas) may facilitate their transfer over
long distances (Vinarski et al. 2007). Higher passive disper-
sal of mites could result in their relatively higher nestedness
at local spatial scale. Th is mechanism most likely does not
act on local temporal and regional scales, so diff erences in
the degree of nestedness between mite and fl ea assemblages
vanish. In addition, both fl ea and mite assemblages at
regional scale may have a high number of idiosyncratic species
(Attmar and Patterson 1993) merely because the regional

Page 8

637
matrices contain more species. Indeed, the degree of nested-
ness at regional scale was lower than that at local scales in
both taxa (Fig. 1).
Nevertheless, nestedness contribution to β -diversity was
higher in fl ea than in mite assemblages at both local scales,
while the opposite was true at regional scale. It is important
to note that the contribution of nestedness to β -diversity does
not necessarily correlate with the degree of nestedness (Baselga
2010). High contribution of nestedness to β -diversity suggests
that variation in species composition from richer to poorer
assemblages involves the progressive disappearance of some
species without their replacement by other species, while the
lower contribution of nestedness to β -diversity (that is, higher
contribution of species turnover) involves many such replace-
ments (Baselga 2010). In other words, higher β NESP likely
indicates predominance of extinctions, whereas lower (but not
too low) β NESP indicates an equalization between extinctions
and colonizations. Flea assemblages harboured by host popu-
lations were shown to be saturated (Krasnov et al. 2006c). Th is
might preclude colonization of these communities by new
species and, thus, lead to equalization between extinctions and
colonizations. It is unknown and deserves special investiga-
tion whether component communities of mites are saturated
or not, but our results hint that the latter might be the case.
A higher contribution of nestedness to β -diversity in
mite than fl ea assemblages at regional scale suggests that
mechanisms acting at this scale diff er from those acting at
local scale. Th ese mechanisms will be discussed in the next
subsection. Furthermore, this result can also be envisaged as
higher among-region species turnover in fl ea than in mite
assemblages. Indeed, fl ea assemblages demonstrate distance
decay of similarity (that is, similarity in species composition
decreasing with geographic distance; Krasnov et al. 2005b,
2010c), whereas no distance decay of similarity has been
found in mite assemblages (Vinarski et al. 2007). In other
words, regional fl ea assemblages seem to be less similar
than regional mite assemblages, so that fl ea species replace
each other at a greater rate. Interestingly, distance decay of
similarity has been reported mainly for moderately and low
host-specifi c, but not for highly host-specifi c fl eas (Krasnov
et al. 2010c). Although fl eas vary greatly in the degree of
their host specifi city, majority of species used in this study
are moderately and low host-specifi c, while only a few fl eas
are highly host-specifi c (Krasnov et al. 2004b).
Host biology-dependence
Th e only host-related eff ect on the degree of nestedness and
its contribution to β -diversity found in our study was related
to the size of host geographic range and was characteris-
tic for mite, but not fl ea assemblages. In our earlier study,
where nestedness was estimated via matrix temperature, we
found that the tendency for fl ea assemblages to approach
nestedness increased with increasing host geographic range
size (Krasnov et al. 2005a). Th e results of this study do
not support this conclusion. Given better properties of
the NODF metric for nestedness estimation, comparison
of the results of the present study with those of Krasnov
et al. (2005a) supports the idea of Ulrich et al. (2009) that
a critical reassessment of previous studies that used earlier
metrics of nestedness is needed. Indeed, Timi and Poulin
(2008) using data on helminth communities of fi sh hosts
demonstrated that the probability of detection of a nested
pattern in a parasite community depended on the metric
and null model used.
Host species occupying larger geographic ranges had less
nested distribution of mite species among their populations.
Th e mechanism behind this pattern may be related to the rela-
tionships between the size of geographic range and other traits
of a host species such as environmental niche breadth (Brown
1984) and abundance (Gaston 1996). Broadly distributed spe-
cies are expected to be (a) more abundant and (b) demonstrate
a broader range of environmental tolerances at local scale than
species with restricted geographic ranges. Th is, in turn, may
result the increased frequency of contacts with a variety of
other host species and thus substantial horizontal transfer of
parasites. Under such conditions, a nested pattern of parasite
assemblages is highly unlikely. Th e reason why this mechanism
does not act on fl ea assemblages may be related to the fact that
all fl eas are obligate haematophages, while the majority of mites
are not, so that mites are, in general, more host opportunistic
than even fl eas (Radovsky 1985, Krasnov 2008).
Scale-dependence
Scale-dependence of the degree of nestedness as well as its
contribution to β -diversity indicates that parasite communi-
ties at diff erent scales are governed by diff erent mechanisms.
Nestedness across insular or fragmented habitats of free-living
organisms is commonly thought to result from diff erential
colonization/extinction dynamics among species (Patterson
and Atmar 1986). Th e same can be true for parasites (Poulin
and Valtonen 2001). However, proximate mechanisms aff ect-
ing colonizations and extinctions of parasites seem to be diff er-
ent at diff erent scales. At local scale(s), colonization/extinction
dynamics are likely ruled by epidemiological processes (Morand
et al. 2002). In particular, they involve diff erential birth and
death dynamics (Krasnov 2008) and diff erential responses to
fl uctuations of host density (Krasnov et al. 2002, Stanko et al.
2006 for fl eas) among parasite species. At regional scale, colo-
nization/extinction dynamics are infl uenced by biogeographic
and environmental processes. For example, environmental
factors that limit distribution of some but not other parasites
may produce a nested pattern (Gonz á lez and Oliva 2009). Th e
role of biogeographic processes is indicated by spatial patterns
such as the latitudinal gradient of species richness (reviewed
by Rohde 2010 for parasites). Th e latitudinal gradient of
diversity may aff ect nestedness via the eff ect of latitude on spe-
cies co-occurrence pattern (Ulrich and Gotelli 2007, Ulrich
et al. 2009).
Nevertheless, some factors can contribute to nestedness at
both local and regional scales. For instance, diff erential host
specifi city among parasite species may aff ect nested structure
because it determines whether a parasite is able to colonize a
host (Morand et al. 2002). In fl eas, host specifi city has been
shown to be scale-invariant (Krasnov et al. 2008) and, thus, its
eff ect on nestedness may be scale-independent.
Nestedness and b -diversity
Baselga (2010) noted that the β NES metric “ is not a mea-
sure of nestedness in absolute terms but a measure of the

Page 9

638
Bininda-Emonds, O. R. P. et al. 2007. Th e delayed rise of present-
day mammals. – Nature 446: 507 – 512.
Brooks, D. R. et al. 2006. Ecological fi tting as a determinant of
the community structure of platyhelminth parasites of anurans.
– Ecology 87: S76 – S85.
Brown, J. H. 1984. On the relationship between abundance and
distribution of species. – Am. Nat. 124: 255 – 279.
Brualdi, R. A. and Sanderson, J. G. 1999. Nested species subsets,
gaps, and discrepancy. – Oecologia 119: 256 – 264.
Busan, E. V. et al. 2008. Mitochondrial phylogeny of Arvicolinae
using comprehensive taxonomic sampling yields new insights.
– Biol. J. Linn. Soc. 94: 825 – 835.
DeVries, P. J. et al. 1997. Species diversity in vertical, horizontal,
and temporal dimensions of a fruit-feeding butterfl y commu-
nity in an Ecuadorian rainforest. – Biol. J. Linn. Soc. 62:
343 – 364.
Diamond, J. M. 1975. Assembly of species communities. – In:
Cody, M. L. and Diamond, J. M. (eds), Ecology and evolution
of communities. Harvard Univ. Press, pp. 342 – 444.
Felsenstein, J. 1985. Phylogenies and the comparative method.
– Am. Nat. 125: 1 – 15.
Gaston, K. J. 1996. Th e multiple forms of the interspecifi c abun-
dance ? distribution relationship. – Oikos 76: 211 – 220.
Gonz á lez, M. T. and Oliva, M. E. 2009. Is the nestedness of meta-
zoan parasite assemblages of marine fi shes from the southeastern
Pacifi c coast a pattern associated with the geographical distribu-
tional range of the host? – Parasitology 136: 401 – 409.
Gotelli, N. J. and Rohde, K. 2002. Co-occurrence of ectoparasites
of marine fi shes: a null model analysis. – Ecol. Lett. 5:
86 – 94.
Gouy de Bellocq, J. et al. 2003. A comparison of the structure of
helminth communities in the woodmouse, Apodemus sylvati-
cus , on islands of the western Mediterranean and continental
Europe. – Parasitol. Res. 90: 64 – 70.
Guimaraes, P. R. and Guimaraes, P. 2006. Improving the analyses
of nestedness for large sets of matrices. – Environ. Model.
Software 21: 1512 – 1513.
Gurevitch, J. et al. 1992. A meta-analysis of fi eld experiments on
competition. – Am. Nat. 140: 539 – 572.
Hanski, I. 1982. Communities of bumblebees: testing the core-
satellite hypothesis. – Ann. Zool. Fenn. 19: 65 – 73.
Harrison, S. et al. 1992. Beta-diversity on geographic gradients in
Britain. – J. Anim. Ecol. 61: 151 – 158.
Jaarola, M. et al. 2004. Molecular phylogeny of the speciose vole
genus Microtus (Arvicolinae, Rodentia) inferred from mito-
chondrial DNA sequences. – Mol. Phylogenet. Evol. 33:
647 – 663.
Koleff , P. et al. 2003. Measuring beta diversity for presence–absence
data. – J. Anim. Ecol. 72: 367 – 382.
Krasnov, B. R. 2008. Functional and evolutionary ecology of fl eas:
a model for ecological parasitology. – Cambridge Univ. Press.
Krasnov, B. R. et al. 2002. Th e eff ect of host density on ectopara-
site distribution: an example with a desert rodent parasitized
by fl eas. – Ecology 83: 164 – 175.
Krasnov, B. R. et al. 2004a. Flea species richness and parameters
of host body, host geography and host “ milieu ” . – J. Anim.
Ecol. 73: 1121 – 1128.
Krasnov, B. R. et al. 2004b. Ectoparasitic “ jacks-of-all-trades ” : rela-
tionship between abundance and host specifi city in fl eas
(Siphonaptera) parasitic on small mammals. – Am. Nat. 164:
506 – 515.
Krasnov, B. R. et al. 2005a. Nested pattern in fl ea assemblages across
the host ’ s geographic range. – Ecography 28: 475 – 484.
Krasnov, B. R. et al. 2005b. Spatial variation in species diversity
and composition of fl ea assemblages in small mammalian
hosts: geographic distance or faunal similarity? – J. Biogeogr.
32: 633 – 644.
dissimilarity of communities due to the eff ect of nestedness
patterns ” . Consequently, one cannot a priori expect a higher
contribution of nestedness to the total amount of β -diversity
in communities characterized by a higher degree of nested-
ness as compared with communities characterized by a lower
degree of nestedness. Indeed, the correlation between the
degree of nestedness and its contribution to β -diversity in
mites and the lack of this correlation in fl eas suggested that
the relationships between these two measures are compli-
cated. By defi nition, NODF of Almeda-Neto et al. (2008)
attains maximum values when the value of matrix fi ll is
intermediate and decreases with an increase and a decrease
of fi ll, while β NES continuously increases with an increased
dissimilarity between nested assemblages. In other words,
a plot of NODF against β NES is unimodal (see Fig. 4 in
Baselga 2010; note that this has been shown for matrices
with no unexpected presences or absences). Th erefore, (a) a
positive correlation between nestedness metrics and β NES can
be expected when the β NES values are smaller than 0.5; (b)
a negative correlation can be expected when the β NES values
are greater than 0.5; and (c) no correlation can be expected
when the value of 0.5 takes some median position in the
range of the β NES values. Absolute values of the β NES in our
datasets ranged from 0.07 to 0.50 in fl eas and from 0.00 to
0.39 in mites. Consequently, the diff erence between parasite
taxa in the relationships between the degree of nestedness
and its contribution to β -diversity cannot be explained by
mathematical properties of the metrics. Some other, still
unknown, factor(s) may be the reason underlying this dif-
ference. Relationships between β NES and nestedness warrant
further investigation.
In conclusion, fl ea and mite assemblages across host
populations within host species at smaller and larger spa-
tial scales, as well as at temporal scale, was characterized
by nestedness which, in turn, contributed to an impor-
tant degree to the total amount of β -diversity of these
assemblages.
Acknowledgements – Th e sampling protocol used in
Slovakia complied with the laws of the Slovak Republic.
Th is study was partly supported by the Slovak Grant
Committee VEGA (grants no. 2/0043/09 and 2/0042/10
to MS) and by the Israel Science Foundation (grant no.
27/08 to BRK and ISK). We thank Allan Degen for help-
ful comments on the earlier version of the manuscript.
Th is is publication no. 688 of the Mitrani Dept of Desert
Ecology.
References
Almeida-Neto, M. et al. 2008. A consistent metric for nestedness
analysis in ecological systems: reconciling concept and meas-
urement. – Oikos 117: 1227 – 1239.
Atmar, W. and Patterson, B. D. 1993. Th e measure of order and
disorder in the distribution of species in fragmented habitat.
– Oecologia 96: 373 – 382.
Baselga, A. 2010. Partitioning the turnover and nestedness compo-
nents of beta-diversity. – Global Ecol. Biogeogr. 19: 134 – 143.
Baselga, A. et al. 2007. A multiple-site similarity measure inde-
pendent of richness. – Biol. Lett. 3: 642 – 645.

Page 10

639
Rohde, K. 2010. Marine parasite diversity and environmental gra-
dients. – In: Morand, S. and Krasnov, B. R. (eds), Th e bioge-
ography of host-parasite interactions. Oxford Univ. Press, pp.
73 – 88.
Sfenthourakis, S. et al. 2004. From sampling stations to archipela-
gos: investigating aspects of the assemblage of insular biota.
– Global Ecol. Biogeogr. 13: 23 – 35.
Simaiakis, S. M. and Martinez-Morales, M. A. 2010. Nestedness
in centipede (Chilopoda) assemblages on continental islands
(Aegean, Greece). – Acta Oecol. 36: 282 – 290.
Simkova, A. et al. 2001. Order and disorder in ectoparasite com-
munities: the case of congeneric gill monogeneans ( Dactylogy-
rus spp.). – Int. J. Parasitol. 31: 1205-1210.
Stanko, M. et al. 2006. Relationship between host abundance and
parasite distribution: inferring regulating mechanisms from
census data. – J. Anim. Ecol. 75: 575-583.
Steppan, S. J. et al. 2004. Phylogeny and divergence-date estimates
of rapid radiations in muroid rodents based on multiple
nuclear genes. – Syst. Biol. 53: 533 – 553.
Timi, J. T. and Poulin, R. 2003. Parasite community structure
within and across host populations of a marine pelagic
fi sh: how repeatable is it? – Int. J. Parasitol. 33:
1353 – 1362.
Timi, J. T. and Poulin, R. 2008. Diff erent methods, diff erent
results: temporal trends in the study of nested subset patterns
in parasite communities. – Parasitology 135: 131 – 138.
Ulrich, W. and Gotelli, N. J. 2007. Disentangling community pat-
terns of nestedness and species co-occurrence. – Oikos 116:
2053 – 2061.
Ulrich, W. et al. 2009. A consumer ’ s guide to nestedness analysis.
– Oikos 118: 3 – 17.
Vinarski, M. V. et al. 2007. Decay of similarity of gamasid mite
assemblages parasitic on Palaearctic small mammals: geo-
graphic distance, host species composition or environment?
– J. Biogeogr. 34: 1691 – 1700.
Whittaker, R. H. 1960. Vegetation of the Siskiyou Mountains,
Oregon and California. – Ecol. Monogr. 30: 280 – 338.
Wright, D. A. and Reeves, J. H. 1992. On the meaning and meas-
urement of nestedness of species assemblages. – Oecologia 92:
416 – 428.
Wright, D. H. et al. 1998. A comparative analysis of nested subset
patterns of species composition. – Oecologia 113: 1 – 20.
Krasnov, B. R. et al. 2006a. Are ectoparasite communities struc-
tured? Species co-occurrence, temporal variation and null
models. – J. Anim. Ecol. 75: 1330 – 1339.
Krasnov, B.R. et al. 2006b. Habitat variation in species composi-
tion of fl ea assemblages on small mammals in central Europe.
– Ecol. Res. 21: 460-469.
Krasnov, B. R. et al. 2006c. Relationships between local and regional
species richness in fl ea communities of small mammalian hosts:
saturation and spatial scale. – Parasitol. Res. 98: 403 – 413.
Krasnov, B. R. et al. 2008. Scale-invariance of niche breadth in
haematophagous ectoparasites. – Ecography 31: 630 – 635.
Krasnov, B. R. et al. 2010a. Co-occurrence of ectoparasites on
rodent hosts; null model analyses of data from three conti-
nents. – Oikos 119: 120 – 128.
Krasnov, B. R. et al. 2010b. Similarity in ectoparasite faunas of
Palaearctic rodents as a function of host phylogenetic, geo-
graphic, or environmental distances: which matters the most?
– Int. J. Parasitol. 40: 807 – 817.
Krasnov, B. R. et al. 2010c. Deconstructing spatial patterns in
species composition of ectoparasite communities: the relative
contribution of host composition, environmental variables and
geography. – Global Ecol. Biogeogr. 19: 515 – 526.
Maddison, W. P. and Maddison, D. R. 2009. Mesquite: a modular
system for evolutionary analysis. Ver. 2.72. – ? http://
mesquiteproject.org ?
Midford, P. E. et al. 2008. PDAP:PDTREE package for Mesquite,
ver. 1.14. – ? http://mesquiteproject.org/pdap_mesquite/
index.html ?
Morand, S. et al. 2002. Order in ectoparasite communities of
marine fi sh is explained by epidemiological processes. – Para-
sitology 124: S57 – S63.
Patterson, B. D. and Atmar, W. 1986. Nested subsets and the struc-
ture of insular mammalian faunas and archipelagos. – Biol. J.
Linn. Soc. 28: 65 – 82.
Poulin, R. 1996. Richness, nestedness and randomness in parasite
infracommunity structure. – Oecologia 105: 545 – 551.
Poulin, R. and Valtonen, E. T. 2001. Nested assemblages resulting
from host-size variation: the case of endoparasite communities
in fi sh hosts. – Int. J. Parasitol. 31: 194 – 1204.
Radovsky, F. J. 1985. Evolution of mammalian mesostigmatid
mites. – In: Kim, K. C. (ed.), Coevolution of parasitic arthro-
pods and mammals. Wiley, pp. 441 – 504.
Supplementary material (available online as Appendix
O19072 at ? www.oikosoffi ce.lu.se/appendix ? ). Appendix
1 – 3.