See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/273677991

Ectoparasites and endoparasites of fish form networks with different

structures

Article  in  Parasitology · March 2015

DOI: 10.1017/S0031182015000128

CITATIONS READS

14 1,845

6 authors, including:

Sybelle Bellay Edson Fontes Oliveira

Universidade Estadual de Maringá Federal Technological University of Parana, Londrina, Brazil
27 PUBLICATIONS   243 CITATIONS    35 PUBLICATIONS   422 CITATIONS   

SEE PROFILE SEE PROFILE

Mário Almeida-Neto Marco Aurelio Ribeiro Mello

Universidade Federal de Goiás University of São Paulo
60 PUBLICATIONS   2,433 CITATIONS    89 PUBLICATIONS   1,147 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

A synthesis of the assembly rules of complex ecological systems View project

Diversity of proteocephalids (Eucestoda: Onchoproteocephalidea) parasites of freshwater fishes in South America: an integrated taxonomic approach View project

All content following this page was uploaded by Marco Aurelio Ribeiro Mello on 09 September 2016.

The user has requested enhancement of the downloaded file.

 901

Ectoparasites and endoparasites of fish form networks with

different structures

S. BELLAY 1 *, E. F. DE OLIVEIRA 2 , M. ALMEIDA-NETO 3 , M. A. R. MELLO 4 ,

R. M. TAKEMOTO 1 and J. L. LUQUE 5
1
Departamento de Ciências Biológicas, Programa de Pós-Graduação em Ecologia de Ambientes Aquáticos Continentais,
Núcleo de Pesquisas em Limnologia, Ictiologia e Aquicultura, Universidade Estadual de Maringá, Av. Colombo, Bloco G90,
Sala 13, 87020-900, Maringá, Paraná, Brazil
2
Departamento de Engenharia Ambiental, Programa de Pós-Graduação em Engenharia Ambiental, Universidade
Tecnológica Federal do Paraná, Campus Londrina, Londrina, Brazil
3
Departamento de Ecologia, Programa de Pós-Graduação em Ecologia e Evolução, Instituto de Ciências Biológicas,
Universidade Federal de Goiás, Goiânia, Brazil
4
Departamento de Biologia Geral, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo
Horizonte, Brazil
5
Departamento de Parasitologia Animal, Programa de Pós-Graduação em Ciências Veterinárias, Universidade Federal
Rural do Rio de Janeiro, Seropédica, Brazil

(Received 28 November 2014; revised 8 January 2015; accepted 12 January 2015; first published online 16 March 2015)

SUMMARY

Hosts and parasites interact with each other in a variety of ways, and this diversity of interactions is reflected in the net-
works they form. To test for differences in interaction patterns of ecto- and endoparasites we analysed subnetworks formed
by each kind of parasites and their host fish species in fish–parasite networks for 22 localities. We assessed the proportion of
parasite species per host species, the relationship between parasite fauna composition and host taxonomy, connectance,
nestedness and modularity of each subnetwork (n = 44). Furthermore, we evaluated the similarity in host species compo-
sition among modules in ecto- and endoparasite subnetworks. We found several differences between subnetworks of fish
ecto- and endoparasites. The association with a higher number of host species observed among endoparasites resulted in
higher connectance and nestedness, and lower values of modularity in their subnetworks than in those of ectoparasites.
Taxonomically related host species tended to share ecto- or endoparasites with the same interaction intensity, but the
species composition of hosts tended to differ between modules formed by ecto- and endoparasites. Our results suggest
that different evolutionary and ecological processes are responsible for organizing the networks formed by ecto- and
endoparasites and fish.

Key words: host–parasite metazoan networks, antagonistic networks, connectance, nestedness, modularity.

INTRODUCTION et al. 2011; Krasnov et al. 2012; Lima et al. 2012).

The network approach in studies about host–para-
Interaction networks are usually characterized by
site interactions stand out among the traditional
non-random topological patterns and some degree
approaches involving only a few species, to contrib-
of phylogenetic signal in the interactions (Rezende
ute to the elucidation of the mechanisms governing
et al. 2007; Bellay et al. 2011; Krasnov et al. 2012).
these systems (Bellay et al. 2013).
Different structures have been recorded and the
The diversity of host–parasite interactions is
specific configuration of an ecological network
reflected in the network structure and the simi-
depends mainly on the type of interaction (e.g.
larity between parasite faunas tends to increase
mutualistic vs antagonistic) and the level of intimacy
with host relatedness (Bellay et al. 2011; Krasnov
among species (e.g. symbiotic vs non-symbiotic
et al. 2012; Lima et al. 2012). In addition, a con-
interactions) (Guimarães et al. 2007; Fontaine et al.
vergence of ecological traits among phylogeneti-
2011). Interactions involving parasites and hosts
cally distant host species may also increase the
are a classical example of antagonistic network with
similarity among their parasite faunas (Krasnov
high intimacy, and they are often characterized by
et al. 2012). Parasite faunas may comprise species
a phylogenetic signal in the interactions (Fontaine
with different life strategies, which are grouped
* Corresponding author. Departamento de Ciências mainly as ectoparasites (with direct contact with
Biológicas, Programa de Pós-Graduação em Ecologia de the external environment) or endoparasites
Ambientes Aquáticos Continentais, Núcleo de Pesquisas (without direct contact with the external environ-
em Limnologia, Ictiologia e Aquicultura, Universidade
Estadual de Maringá, Av. Colombo, 5790, Bloco G90, ment) (Bush et al. 2001). Studies on the variations
Sala 13, 87020-900, Maringá, Paraná, Brazil. E-mail: in the structure of ecto- and endoparasite inter-
sybellebellay@yahoo.com.br actions with hosts are still scarce.

Parasitology (2015), 142, 901–909. © Cambridge University Press 2015

doi:10.1017/S0031182015000128
S. Bellay and others 902

Parasite life strategy and host phylogeny seem to phenomena, but they are not independent from each
affect specialization in host–parasite interactions, other, because they may occur in the same individual
which in turn explains why fish–parasite networks host. Therefore, we expected high similarity in host
are characterized by low levels of connectance and composition between the modules found in ecto- and
nestedness, and high levels of modularity (Bellay endoparasite networks.
et al. 2011, 2013; Krasnov et al. 2012; Lima et al.
2012; Poulin et al. 2013). Connectance is calculated
as the proportion of interactions that are actually MATERIALS AND METHODS
realised in relation to the total number of interactions
Data
that could be realised in the network (Pimm, 1982).
Nestedness occurs if the interactions of species with Twenty-two fish–parasite networks were obtained
fewer connections in a bipartite network represent a from the literature. The number of host and parasite
subset of the interactions made by species with species in the networks range from 6 to 91 and from
more connections (Almeida-Neto and Ulrich, 20 to 420, respectively (Table 1). We built the
2011). On the other hand, if there are subgroups of networks as adjacency matrices with host species in
species interacting with each other more than with the rows, parasite species in the columns and
other species of the same network (modules), the binary values in the cells (presence of absence of
network has a modular structure (Mello et al. interaction between a i row and a j column). To
2011). Although those archetypical topologies have control for an effect of spatial variations, we analysed
often been considered in studies on interaction pairs of subnetworks formed by either endo- or ecto-
patterns in different mutualistic and antagonistic parasites that belonged to the same complete
networks (see Lewinsohn and Prado, 2006), this network from a given locality.
approach is not commonly applied to studies on We restricted our analysis to metazoan parasites.
host–parasite networks (Poisot et al. 2013). The studied ectoparasites belong to the following
Interactions between hosts and parasites may taxonomic groups: Acari, Branchiura, Copepoda,
differ among parasite groups, if their biological Hirudinea, Isopoda, Mollusca, Monogenea and
traits and infestation processes result in distinct Myxosporea. The studied endoparasites were rep-
interaction constraints (Poisot et al. 2013). resented by Acanthocephala, Aspidobothrea,
Therefore, by looking at similarities and differences Cestoda, Digenea, Nematoda, Pentastomida, and
in the structure of interactions among ecto- and some species of Monogenea and Myxosporea (see
endoparasites, we can gain some insight on the Supplementary Material). In the studied networks,
underlying mechanisms of host-parasite networks. the larval and adult stages of a parasite species can
Lima et al. (2012) observed that variations in the have different niches (host species) in the same
specificity of interactions may be responsible for network. Therefore, different stages were regarded
differences in the structure of ecto- and endopara- as different ‘functional species’ in the network, as
site-fish networks. It is known that a single individ- in Vázquez et al. (2005) and Bellay et al. (2013).
ual host can harbour both ecto- and endoparasites.
Some processes that may lead fish species to share
ectoparasites may also lead them to share endopara- Network characteristics
sites, thereby putting those host species in the same To test for an influence of host taxonomy on host–
module within a network. parasite interactions, we calculated a correlation
Our goal in the present study was to understand the between the matrix of taxonomic distances (a
structure of networks formed by ecto- and endopara- proxy for phylogenetic distance; Koehler et al.
sites of fish. To fulfil this goal, we asked the following 2012) between fish species and the dissimilarity
questions (i) Do fish–ectoparasite and fish–endopara- matrix of parasite fauna composition with a Mantel
site subnetworks from the same locality differ in test, using 1000 randomizations and the Pearson
terms of host taxonomy and topology (proportion of method in the package vegan (Oksanen et al. 2014)
parasite species per host species, connectance, nested- for R 3·1·1 (R Development Core Team, 2014). To
ness and modularity)? (ii) Do host species that share build the dissimilarity matrix used in this analysis,
the same ectoparasites also share the same endopara- we used the Jaccard index available in the function
sites? First, we expected differences in the biology of vegdist in the package vegan. We calculated the
interactions between fish and their endo- and ectopar- matrix of taxonomic distance (MTD) for each
asites to result in networks with different structures. network using the following equation:
Second, the composition of modules in fish–parasite
networks was expected to reflect the taxonomic dis- MTD ¼ Md  ðCw þ Ow þ Fw þ Gw þ SwÞ
tance between host species, as host niches tend to be
phylogenetically conserved (e.g. distribution in the where Md is the maximum distance found in the fish
water column, foraging strategy). Furthermore, ecto- community (maximum distance = 5, referring to the
and endoparasites infestations are different taxonomic category class) and Cw, Ow, Fw, Gw and
Ectoparasites and endoparasites of fish form networks with different structure 903

Table 1. Fish–parasite networks analysed in the present study. H is the number of host species and P is the
number of parasite species of each network

Networka Country H P References

1. Middle Paraná River Argentina 54 93 Chemes and Takemoto (2011)
2. Floodplain of Upper Paraná River Brazil 65 311 Takemoto et al. (2009), Lima et al. (2012)
3.Smallwood Reservoir Canada 6 25 Chinniah and Threlfall (1978)
4. Parsnip River Canada 17 53 Arai and Mudry (1983)
5. McGregor River Canada 14 51 Arai and Mudry (1983)
6. Lake of the Woods Canada 30 144 Dechtiar (1972)
7. Cold Lake Canada 10 40 Leong and Holmes (1981)
8. Aishihik Lake Canada 7 29 Arthur et al. (1976)
9. Coastal Waters of Rio de Janeiro Brazil 59 420 Bellay et al. (2011, 2013)
10. Little Colorado River USA 11 20 Choudhury et al. (2004)
11. Lake Michigan Canada–USA 43 117 Muzzall and Whelan (2011)
12. Lake Superior Canada–USA 36 174 Muzzall and Whelan (2011)
13. Guandu River Brazil 22 85 de Azevedo et al. (2010)
14. Lake Huron Canada–USA 79 300 Muzzall and Whelan (2011)
15. Lake Erie Canada–USA 91 308 Muzzall and Whelan (2011)
16. Lake Ontario Canada–USA 61 257 Muzzall and Whelan (2011)
17. Gulf of Riga Latvia 52 94 Kirjušina and Vismanis (2007)
18. Lake Raznas Latvia 48 80 Kirjušina and Vismanis (2007)
19. Tres Palos Lagoon Mexico 13 40 Violante-González et al. (2007)
20. Mekong River Delta Vietnam 52 123 Arthur and Te (2006)
21. Gulf of Tonkin Vietnam 80 215 Arthur and Te (2006)
22. Coyuca Lagoon Mexico 10 34 Violante-González and Aguirre-Macedo (2007)

a
See Supplementary Material (online version only).

Sw are matrices for each taxonomic category (class, to calculate the degree of modularity (M) of
order, family, genus and species, respectively) gener- each network (Guimerà and Amaral, 2005). Values
ated by the function weight.taxo available in package of M = 0 indicate the absence of subgroups in the
ape (Paradis et al. 2004) for R. Nomenclature fol- network, whereas values near the maximum (M = 1)
lowed the taxonomic descriptions provided by indicate networks strongly divided into subgroups.
FishBase (Froese and Pauly, 2013). Therefore, Modularity was calculated in the program
species of the same genus exhibit a value of taxo- NETCARTO (Guimerà and Amaral, 2005). As
nomic distance (td) equal to 1, different genera NETCARTO does not include the Ce model for the
have td = 2, different families have td = 3, different estimation of significance, we used a function for R
orders have td = 4 and different classes have td = 5 (developed by Professor Nadson RS da Silva) to esti-
(see Rezende et al. 2007). mate the significance of M. With this function, we
We evaluated three general descriptors of network generated 1000 randomizations of each network
structure: connectance (C), nestedness (NODF), based on the null model Ce. For each matrix, M was
and modularity (M). To control the intrinsic nega- calculated in NETCARTO using a Fortran code
tive relationship between connectance and species (developed by Flávia M. D. Marquitti and first used
richness (Thébault and Fontaine, 2008), we used by Mello et al. 2011) to automate the calculation and
the residual connectance instead of absolute connec- compilation of M-values. For each network, the sig-
tance values. The residual connectance is calculated nificance (P) was obtained from the number of
by the residuals of the simple linear regression random matrices with M-values equal or higher than
between the log10-transformed values of observed the observed M-value, divided by the number of ran-
and possible interactions in each network (e.g. domized matrices. The R scripts are available from the
Fonseca and John, 1996). This analysis was carried authors upon request.
out in Statistica 7·0 (Statsoft, 2005).
The degree of nestedness was calculated using the
Data analysis
NODF index (nestedness metric based on overlap
and decreasing fill; Almeida-Neto et al. 2008). The Differences in the proportion of parasites per host,
significance of the observed NODF-values was esti- host taxonomy (Mantel r coefficient), residual con-
mated with a Monte Carlo procedure (1000 ran- nectance, nestedness and modularity between ecto-
domizations) based on the row–column probability and endoparasite networks were tested with a
null model, Ce, in the program Aninhado Wilcoxon test for paired samples. We applied a
(Guimarães and Guimarães, 2006). chi-squared test to compare the frequency of signifi-
To test for a modular structure in the host–parasite cant nested and modular structure between ecto- and
networks, we used a simulated annealing algorithm endoparasite networks.
S. Bellay and others 904

If the subnetworks of ecto- and endoparasites which suggests a weak relationship (r = 0·23 ± 0·08)
from the same locality were significantly modular, (Table 3).
we evaluated the similarity in the formation of
modules with a Mantel test (with the same procedure
DISCUSSION
mentioned above), considering only the host species
that were present in both networks. For this In the present study, we found that ecto- and endo-
purpose, we identified the host species in each parasite subnetworks from the same local assemblage
module of the network using the program differed in their topologies, thus implying that
NETCARTO, and built matrices whose rows and differences in the biology of parasitic interactions
columns corresponded to the host species present may lead to different interaction patterns at the com-
in both networks. The value ‘1’ was given to pairs munity level. Those differences were observed in all
of host species that occurred in the same module, topological metrics analysed.
and the value ‘0’ was given to pairs of host species One key point to consider is that the interior of
that did not occur in the same module. host can offer a higher diversity of sites (organs
and tissues) for parasite attachment than the external
surface of host. This might explain, for instance, the
RESULTS
greater number of endoparasite than ectoparasite
The species richness of ecto- and endoparasites species found. Another factor that potentially
varied in subnetworks from 6 to 181 and from 11 to influenced the richness patterns is the various
239, respectively. The values of all network descrip- routes of infection that are available to fish endopar-
tors obtained for each subnetwork are presented in asites (i.e. active penetration through the skin or
Table 2. The endoparasite subnetworks showed a trophic transmission). Those routes of infection con-
higher proportion of parasite species per host tribute to species diversity, because they increase the
species (PPHecto: mean = 1·49; PPHendo: mean = probability of host–parasite encounters and may
2·48; Wilcoxon T = 9; Z = 3·81; P < 0·001; Fig. 1a). reduce competition among parasite species (Poulin,
Thirty-three (75%) out of 44 subnetworks presented 1998; Dobson et al. 2008; Lima et al. 2012).
a positive and significant relation of parasite fauna Host characteristics, such as density, body size,
composition with host taxonomy. There were no diet and biogeographic distribution, may directly
differences in the Mr-values between ecto- and endo- influence parasite diversity (Takemoto et al. 2005;
parasite subnetworks (Mrecto: mean = 0·41; Mrendo: Poulin and Leung, 2011; Timi et al. 2011). Host
mean = 0·46; Wilcoxon T = 46; Z = 0·40; P = 0·683; species that are phylogenetically close tend to
Fig. 1b). We found significant differences in connec- present more similar parasite faunas than unrelated
tance, nestedness and modularity between ecto- and host species (Bellay et al. 2011, 2013; Krasnov
endoparasite subnetworks. Residual connectance was et al. 2012; Lima et al. 2012). This tendency would
higher in endoparasite subnetworks (Crecto: mean = result from parasite species persistence after specia-
12·53; Crendo: mean = 14·58; Wilcoxon T = 0; Z = tion events of the ancestral host and of the ecological
4·10; P < 0·001; Fig. 1c). similarity of these hosts (Poulin, 1998), and how we
Endoparasite subnetworks were more nested than observed, independent of the habitat type used by
ectoparasite subnetworks (NODFecto: mean = 16·23; parasites (ecto- or endoparasites).
NODFendo: mean = 23·11; Wilcoxon T = 39; Z = The residual connectance values were higher for
2·84; P = 0·004; Fig. 1d). In addition, nestedness endoparasite than ectoparasite networks. In net-
was significant in 17 (39%) out of 44 networks, and works, connectance provides important information
the number of significantly nested subnetworks that may allow the understanding of other structural
was higher among endoparasites (χ2 = 4·69; gl = 1; parameters, for example, an increase in connectance
P = 0·03). can reduce the possibility of nested and modular
The ectoparasite subnetworks were more modular structures simultaneously in a network (Fortuna
than the endoparasite networks (Mecto: mean = 0·62; et al. 2010). Due to the high specificity of host–para-
Mendo: mean = 0·47; Wilcoxon T = 11; Z = 3·74; P < site networks, the connectance values are generally
0·001; Fig. 1E). Twenty-eight (64%) out of 44 net- low (Bellay et al. 2013). In addition, several studies
works showed significant modularity, and the fre- have shown that the range of host species of the
quency of the significantly modular structures did endoparasites of fish may be wider than the range
not differ between subnetwork types (χ2 = 0; gl = 1; of host species of the ectoparasites (particularly
P = 1). monogeneans; Strona et al. 2013). The presence of
In 12 (55%) of the studied localities, both subnet- endoparasites in larval stages, that tend to be more
works (ecto- and endoparasites) were significantly generalist than adults (Bellay et al. 2013), also con-
modular. In eight of the localities with both tributed to increase the connectance.
modular subnetworks (67%) host module compo- In the present study, the nested structure of some
sition was correlated between ecto- and endoparasite networks was more closely related to the life strategy
subnetworks. However, we observed low Mr-values, of endoparasites, which suggests differences in the
 Ectoparasites and endoparasites of fish form networks with different structure
Table 2. Parameters calculated for host–parasite networks in 22 localities considering ecto- and endoparasites in separate subnetworks

Subnetworka S H Pa I PHP Mr P C rC NODF p(CE) M p(CE) mo

1 ecto 52 20 32 44 1·60 0·10 0·109 6·88 −0·23 4·72 0·980 0·81 0·001 13
endo 103 42 61 102 1·45 0·34 <0·001 3·98 −0·21 5·02 0·690 0·79 <0·001 20
2 ecto 169 40 129 155 3·23 0·04 0·147 3·00 −0·21 2·57 1·000 0·84 <0·001 32
endo 237 55 182 317 3·31 0·18 <0·001 3·17 −0·07 5·38 0·010 0·68 <0·001 24
3 ecto 13 6 7 11 1·17 0·29 0·226 26·19 −0·14 20·83 0·730 0·56 0·076 –
endo 24 6 18 42 3·00 0·93 0·022 38·89 0·19 33·18 0·960 0·29 0·449 –
4 ecto 22 12 10 23 0·83 0·25 0·029 19·17 −0·09 24·02 0·480 0·58 0·050 5
endo 60 17 43 135 2·53 0·60 <0·001 18·47 0·22 29·82 0·010 0·42 0·013 5
5 ecto 20 9 11 16 1·22 0·38 0·020 16·16 −0·20 13·74 0·740 0·74 0·005 5
endo 54 14 40 98 2·86 0·49 0·001 17·50 0·14 26·96 0·110 0·41 0·327 –
6 ecto 72 30 42 76 1·40 0·23 0·001 6·03 −0·16 6·94 0·720 0·73 0·011 13
endo 132 30 102 308 3·40 0·39 <0·001 10·07 0·21 17·25 <0·001 0·45 0·004 7
7 ecto 16 8 8 19 1·00 0·94 <0·001 29·69 −0·01 33·04 0·630 0·37 0·444 –
endo 42 10 32 72 3·20 0·72 <0·001 22·50 0·15 26·23 0·660 0·44 0·050 5
8 ecto 17 7 10 16 1·43 0·76 0·050 22·86 −0·11 15·91 0·890 0·46 0·356 –
endo 26 7 19 62 2·71 0·95 0·030 46·62 0·31 50·09 0·760 0·22 0·660 –
9 ecto 236 55 181 276 3·29 0·22 <0·001 2·77 −0·13 2·71 1·000 0·79 <0·001 20
endo 298 59 239 433 4·05 0·36 <0·001 3·07 −0·02 3·75 0·930 0·70 <0·001 15
10 ecto 17 8 9 14 1·13 −0·12 0·665 19·44 −0·17 15·63 0·810 0·60 0·075 –
endo 22 11 11 36 1·00 0·17 0·116 29·75 0·10 56·82 0·010 0·30 0·783 –
11 ecto 42 20 22 40 1·10 0·07 0·163 9·09 −0·17 10·11 0·660 0·69 0·050 9
endo 138 43 95 197 2·21 0·32 <0·001 4·82 −0·05 11·32 <0·001 0·63 0·002 16
12 ecto 72 24 48 71 2·00 0·40 <0·001 6·16 −0·17 6·63 0·770 0·77 0·003 10
endo 161 35 126 307 3·60 0·39 <0·001 6·96 0·12 12·91 <0·001 0·54 <0·001 9
13 ecto 60 22 38 59 1·73 0·50 <0·001 7·06 −0·17 6·46 0·930 0·77 0·002 9
endo 66 19 47 82 2·47 0·11 0·092 9·18 −0·05 11·23 0·650 0·61 0·038 11
14 ecto 164 60 104 208 1·73 0·27 <0·001 3·33 −0·13 5·44 0·070 0·72 <0·001 18
endo 269 73 196 765 2·68 0·30 <0·001 5·35 0·21 14·38 <0·001 0·45 <0·001 6

905
 S. Bellay and others
Table 2. (Cont.)
Subnetworka S H Pa I PHP Mr P C rC NODF p(CE) M p(CE) mo

15 ecto 154 57 97 207 1·70 0·18 <0·001 3·74 −0·10 8·67 <0·001 0·67 0·002 17
endo 300 89 211 910 2·37 0·39 <0·001 4·85 0·23 15·83 <0·001 0·43 <0·001 10
16 ecto 159 51 108 188 2·12 0·52 <0·001 3·41 −0·15 6·09 0·020 0·73 <0·001 18
endo 207 58 149 404 2·57 0·23 <0·001 4·67 0·07 12·32 <0·001 0·53 0·001 10
17 ecto 57 31 26 82 0·84 0·23 0·022 10·17 −0·02 34·23 <0·001 0·46 0·770 –
endo 120 52 68 371 1·31 0·39 <0·001 10·49 0·25 29·67 <0·001 0·41 <0·001 4
18 ecto 66 31 35 125 1·13 <−0·01 0·472 11·52 0·08 41·78 <0·001 0·41 0·735 –
endo 93 48 45 311 0·94 0·30 <0·001 14·40 0·30 48·76 <0·001 0·33 0·208 –
19 ecto 24 13 11 34 0·85 0·06 0·394 23·78 0·03 51·56 0·010 0·30 0·957 –
endo 42 13 29 98 2·23 0·63 <0·001 25·99 0·24 36·39 0·170 0·30 0·767 –
20 ecto 112 48 64 125 1·33 0·45 <0·001 4·07 −0·17 6·5 0·170 0·77 <0·001 14
endo 91 32 59 137 1·84 0·31 <0·001 7·26 −0·01 12·09 0·060 0·61 0·764 –
21 ecto 95 38 57 98 1·50 0·43 <0·001 4·52 −0·19 4·6 0·900 0·81 <0·001 20
endo 234 76 158 425 2·08 0·40 <0·001 3·54 0·01 11·33 <0·001 0·56 0·001 15
22 ecto 16 10 6 22 0·60 0·36 0·146 36·67 0·06 35 0·870 0·24 0·911 –
endo 38 10 28 82 2·80 0·52 0·035 29·29 0·24 37·85 0·370 0·30 0·672 –

Abbreviations: S, species richness; H, host species; Pa, parasite species; I, host–parasite interactions; PHP, proportion of parasite species per host species; Mr, Mantel r statistic obtained
between the host taxonomic distance matrix and the host–parasite dissimilarity matrix; C, connectance; rC, residual connectance; NODF, nestedness; M, modularity; mo, module
number; ecto, ectoparasite–host network; endo, endoparasite–host network.
a
The identity of networks by numbers corresponds to that in Table 1.

906
Ectoparasites and endoparasites of fish form networks with different structure 907

Fig. 1. Parameters calculated for host–parasite networks in 22 localities considering ecto- and endoparasites in separate
subnetworks. (a) Proportion of parasite species per host species; (b) Mantel r statistic (the influence of host taxonomy on
host–parasite interactions); (c) residual connectance; (d) nestedness and (e) modularity.

Table 3. Mantel r statistic (Mr) calculated for the correlation between the dissimilarity in host species
composition of the modules found in ecto- and endoparasites networks

Networka Mr P
1. Middle Paraná River −0·04 1·000
2. Floodplain of Upper Paraná River 0·02 0·518
4. Parsnip River 0·35 0·017
6. Lake of the Woods 0·08 0·108
9. Coastal Waters of Rio de Janeiro 0·25 <0·001
11. Lake Michigan 0·24 0·005
12. Lake Superior 0·21 0·007
13. Guandu River 0·14 0·087
14. Lake Huron 0·16 <0·001
15. Lake Erie 0·09 0·006
16. Lake Ontario 0·25 <0·001
21. Gulf of Tonkin 0·31 <0·001

a
See Supplementary Material (online version only)

organization of host–parasite networks as a function 2013). For example, the ecology and factors related
of host type (e.g. taxonomic group; aquatic or terres- to the parasite life cycle may contribute to the nest-
trial). Studies on terrestrial hosts showed that the edness pattern, particularly among endoparasites
endoparasites have a greater degree of host specifi- (Lima et al. 2012). The reason for this effect is that
city than the ectoparasites in the networks (see the larval stages of these parasites tend to be more
Brito et al. 2014). A high degree of specificity in generalist than the adult parasites (Bellay et al.
the use of host species may result in relatively low 2013), as mentioned above. Furthermore, the adult
levels of nestedness. This has been used as a basis stages may have been obtained by trophic trans-
to infer that mutualistic and antagonistic networks mission, and the host species may have nested
have similar organizations, particularly for a model diets, thus allowing the formation of nested parasitic
system in which ectoparasites use terrestrial hosts fauna.
(Graham et al. 2009). Several hypotheses have We observed no differences between subnetworks
been presented in previous studies to explain the in the frequency of a significantly modular structure,
nestedness structure in networks (see Suweis et al. but they differed from one another in their degree of
S. Bellay and others 908

modularity. The presence of specialized interactions Gerais Research Foundation (FAPEMIG, APQ-
is an important factor when interpreting the modular 01043-13), CNPq (472372/2013-0), Research
structure of ecological networks (Mello et al. 2011). Program on Atlantic Forest Biodiversity (PPBio-
Parasitism in general is expected to be highly special- MA/CNPq) and Ecotone Inc. (‘Do Science and
ized (Thompson, 1994), which may explain the lack Get Support Program’). MAN received research fel-
of difference in the frequency of modular structures. lowships (306843/2012-9 and 306870/2012-6,
But variations in specificity made endoparasite respectively) from CNPq. The funders had no role
subnetworks be more nested than modular, while in study design, data collection and analysis,
the opposite was observed for ectoparasites. The decision to publish or preparation of the manuscript.
studied ectoparasite subnetworks presented an
average high modularity values, probably because
REFERENCES
ectoparasites, particularly monogeneans, are more
specialized than endoparasites (Strona et al. 2013). Almeida-Neto, M. and Ulrich, W. (2011). A straightforward compu-
tational approach for measuring nestedness using quantitative matrices.
Although the co-occurrence of ecto- and endopar- Environmental Modelling and Software 26, 173–178.
asites in the parasitic fauna of hosts is common, we Almeida-Neto, M., Guimarães, P., Guimarães, P. R., Loyola, R. D.
observed a weak relationship (reflected in low Mr and Ulrich, W. (2008). A consistent metric for nestedness analysis in eco-
logical systems: reconciling concept and measurement. Oikos 117, 1227–
coefficients) between the host species composition 1239.
of modules found in ecto- and endoparasite subnet- Arai, H. P. and Mudry, D. R. (1983). Protozoan and metazoan parasites of
works. Thus, for example, hosts that shared the same fishes from the headwaters of the Parsnip and McGregor Rivers, British
Columbia: a study of possible parasite transfaunations. Canadian Journal
module, when only ectoparasites were evaluated, of Fisheries and Aquatic Sciences 40, 1676–1684.
normally tended to occur in distinct modules when Arthur, J. R. and Te, B. Q. (2006). Checklist of the Parasites of Fishes of
we evaluated only endoparasites. Consequently, we Viet Nam. FAO Fisheries Technical paper 369/2. Food and Agriculture
Organization of the United Nations, Rome, Italy. Available from: http://
may infer that different factors influence module www.fao.org/docrep/009/a0878e/a0878e00.htm
organization in ecto- and endoparasite networks. Arthur, J. R., Margolis, L. and Arai, H. P. (1976). Parasites of fishes of
However, the high specificity presented by the ecto- Aishihik and Stevens Lakes, Yukon Territory, and potential consequences
of their interlake transfer through a proposed water diversion for hydroe-
parasites suggests that host phylogeny is a key factor lectrical purposes. Journal of the Fisheries Research Board of Canada 33,
in module organization (see Krasnov et al. 2012), 2489–2499.
whereas host diet could have a stronger influence de Azevedo, R. K., Abdallah, V. D. and Luque, J. L. (2010).
Acanthocephala, Annelida, Arthropoda, Myxozoa, Nematoda and
for endoparasites than phylogeny, due to the Platyhelminthes parasites of fishes from the Guandu river, Rio de
trophic transmission of endoparasites (see Garrido- Janeiro, Brazil. Check List 6, 659–667.
Olvera et al. 2012). Future studies that include infor- Bellay, S., Lima, D. P., Jr., Takemoto, R. M. and Luque, J. L. (2011). A
host–endoparasite network of Neotropical marine fish: are there organiz-
mation of the phylogenetic, biological and ecological ational patterns? Parasitology 138, 1945–1952.
characteristics of host species may clarify which Bellay, S., de Oliveira, E. F., Almeida-Neto, M., Lima Junior, D. P.,
factors are most important to build up the modular Takemoto, R. M. and Luque, J. L. (2013). Developmental stage of
parasites influences the structure of fish–parasite networks. PloS ONE 8,
structure of host–parasite networks. e75710.
Brito, S. V., Corso, G., Almeida, A. M., Ferreira, F. S., Almeida, W.
O., Anjos, L. A., Mesquita, D. O. and Vasconcellos, A. (2014).
SUPPLEMENTARY MATERIAL Phylogeny and micro-habitats utilized by lizards determine the compo-
sition of their endoparasites in the semiarid Caatinga of Northeast Brazil.
To view supplementary material for this article, please Parasitology Research 113, 3963–3972.
visit http://dx.doi.org/10.1017/S0031182015000128 Bush, A. O., Fernández, J. C., Esch, G. W. and Seed, J. R. (eds.) (2001).
Parasitism: The Diversity and Ecology of Animal Parasites. Cambridge
University Press, Cambridge.
Chemes, S. B. and Takemoto, R. M. (2011). Diversity of parasites from
ACKNOWLEDGEMENTS Middle Paraná system freshwater fishes, Argentina. International Journal of
Biodiversity and Conservation 3, 249–266.
The authors thank Nadson R. S. da Silva for the elabor-
Chinniah, V. C. and Threlfall, W. (1978). Metazoan parasites of fish
ation of the function that allowed us to run the null from the Smallwood Reservoir, Labrador, Canada. Journal of Fish
model Ce in R, Flávia M. D. Marquitti for kindly provid- Biology 13, 203–213.
ing us with the Fortran code for the automation of the Choudhury, A., Hoffnagle, T. L. and Cole, R. A. (2004). Parasites of
modularity analysis, and Vanessa M. Algarte for her sug- native and nonnative fishes of the Little Colorado River, Grand Canyon,
gestions to an early version of the manuscript. Arizona. Journal of Parasitology 90, 1042–1053.
Dechtiar, A. O. (1972). Parasites of fish from Lake of the Woods, Ontario.
Journal of Fisheries Research Board of Canada 29, 275–283.
FINANCIAL SUPPORT Dobson, A., Lafferty, K. D., Kuris, A. M., Hechinger, R. F. and
Jetz, W. (2008). Homage to Linnaeus: how many parasites? How many
Manuscript funded by PEA/CAPES/PROEX, hosts? Proceedings of the National Academy of Sciences of the United
DIRPPG/UTFPR – Campus Londrina and the States of America 105(Suppl), 11482–11489.
Brazilian Research Council (CNPq). MAR Mello Fonseca, C. R. and John, J. L. (1996). Connectance: a role for community
allometry. Oikos 77, 353–358.
was sponsored by Ulm University, Humboldt Fontaine, C., Guimarães, P. R., Kéfi, S., Loeuille, N., Memmott, J.,
Foundation (AvH, 1134644), São Paulo Research van der Putten, W. H., van Veen, F. J. F. and Thébault, E. (2011).
Foundation (FAPESP, 06/00265-0, 05/00587-5; The ecological and evolutionary implications of merging different types
of networks. Ecology Letters 14, 1170–81.
07/50633-9), Federal University of Minas Gerais Fortuna, M. A., Stouffer, D. B., Olesen, J. M., Jordano, P.,
(UFMG, PRPq 01/2013, 14/2013, 02/2014), Minas Mouillot, D., Krasnov, B. R., Poulin, R. and Bascompte, J. (2010).
 Ectoparasites and endoparasites of fish form networks with different structure 909

Nestedness versus modularity in ecological networks: two sides of the same Paradis, E., Claude, J. and Strimmer, K. (2004). APE: analyses of phy-
coin? Journal of Animal Ecology 79, 811–817. logenetics and evolution in R language. Bioinformatics 20, 289–290.
Froese, R. and Pauly, D. (2013). FishBase. World Wide Web electronic Pimm, S. L. (1982). Food Webs. Chapman & Hall, London.
publication. http://www.fishbase.org Poisot, T., Stanko, M., Miklisová, D. and Morand, S. (2013).
Garrido-Olvera, L., Arita, H. T. and Pérez-Ponce De León, G. (2012). Facultative and obligate parasite communities exhibit different network
The influence of host ecology and biogeography on the helminth species properties. Parasitology 140, 1340–1345.
richness of freshwater fishes in Mexico. Parasitology 139, 1652–1665. Poulin, R. (1998). Evolutionary Ecology of Parasites. First. Chapman &
Graham, S. P., Hassan, H. K., Burkett-Cadena, N. D., Guyer, C. and Hall, London.
Unnasch, T. R. (2009). Nestedness of ectoparasite-vertebrate host net- Poulin, R. and Leung, T. L. F. (2011). Body size, trophic level, and the
works. PloS ONE 4, e7873. use of fish as transmission routes by parasites. Oecologia 166, 731–738.
Guimarães, P. R., Jr. and Guimarães, P. R. (2006). Improving the ana- Poulin, R., Krasnov, B. R., Pilosof, S. and Thieltges, D. W. (2013).
lyses of nestedness for large sets of matrices. Environmental Modelling and Phylogeny determines the role of helminth parasites in intertidal food
Software 21, 1512–1513. webs. The Journal of Animal Ecology 82, 1265–1275.
Guimarães, P. R., Rico-Gray, V., Oliveira, P. S., Izzo, T. J., dos R Development Core Team (2014). R: A Language and Environment for
Reis, S. F. and Thompson, J. N. (2007). Interaction intimacy affects Statistical Computing. R Foundation for Statistical Computing, Vienna.
structure and coevolutionary dynamics in mutualistic networks. Current http://www.r-project.org/
Biology: CB 17, 1797–1803. Rezende, E. L., Lavabre, J. E., Guimarães, P. R., Jordano, P. and
Guimerà, R. and Amaral, L. A. N. (2005). Functional cartography of Bascompte, J. (2007). Non-random coextinctions in phylogenetically
complex metabolic networks. Nature 433, 895–900. structured mutualistic networks. Nature 448, 925–928.
Kirjušina, M. and Vismanis, K. (2007). Checklist of the Parasites of Fishes Statsoft, Inc. (2005). Statistica (data analysis software system), version 7.1.
of Latvia. FAO Fisheries Technical paper 369/3. Food and Agriculture http://www.statsoft.com
Organization of the United Nations, Rome, Italy. Available from: http:// Strona, G., Galli, P. and Fattorini, S. (2013). Fish parasites resolve the
www.fao.org/docrep/010/a1078e/a1078e00.htm paradox of missing coextinctions. Nature Communications 4, 1718.
Koehler, A. V., Brown, B., Poulin, R., Thieltges, D. W., Suweis, S., Simini, F., Banavar, J. R. and Maritan, A. (2013).
Fredensborg, B. L. (2012). Disentangling phylogenetic constraints from Emergence of structural and dynamical properties of ecological mutualistic
selective forces in the evolution of trematode transmission stages. networks. Nature 500, 449–452.
Evolutionary Ecology 26, 1497–1512. Takemoto, R. M., Pavanelli, G. C., Lizama, M. A. P., Luque, J. L. and
Krasnov, B. R., Fortuna, M. A., Mouillot, D., Khokhlova, I. S., Poulin, R. (2005). Host population density as the major determinant of
Shenbrot, G. I. and Poulin, R. (2012). Phylogenetic signal in module endoparasite species richness in floodplain fishes of the upper Paraná
composition and species connectivity in compartmentalized host–parasite River, Brazil. Journal of Helminthology 79, 75–84.
networks. The American Naturalist 179, 501–511. Takemoto, R. M., Pavanelli, G. C., Lizama, M. A. P., Lacerda, A. C.
Leong, T. S. and Holmes, J. C. (1981). Communities of metazoan para- F., Yamada, F. H., Moreira, L. H. A., Ceschini, T. L. and Bellay, S.
sites in open water fishes of Cold Lake, Alberta. Journal of Fish Biology (2009). Diversity of parasites of fish from the upper Paraná River flood-
18, 693–713. plain, Brazil. Brazilian Journal Biology 69, 691–705.
Lewinsohn, T. M. and Prado, P. I. (2006). Structure in plant–animal Thébault, E. and Fontaine, C. (2008). Does asymmetric specialization
interaction assemblages. Oikos 113, 174–184. differ between mutualistic and trophic networks? Oikos 117, 555–563.
Lima, D. P., Jr., Giacomini, H. C., Takemoto, R. M., Agostinho, A. A. Thompson, J. N. (1994). The Coevolutionary Process. University of
and Bini, L. M. (2012). Patterns of interactions of a large fish–parasite Chicago Press, Chicago.
network in a tropical floodplain. Journal of Animal Ecology 81, 905–913. Timi, J. T., Rossin, M. A., Alarcos, A. J., Braicovich, P. E.,
Mello, M. A. R., Marquitti, F. M. D., Guimarães, P. R., Jr., Kalko, E. Cantatore, D. M. P. and Lanfranchi, A. L. (2011). Fish trophic level
K. V., Jordano, P. and de Aguiar, M. A. M. (2011). The missing part of and the similarity of non-specific larval parasite assemblages. International
seed dispersal networks: structure and robustness of bat-fruit interactions. Journal for Parasitology 41, 309–316.
PLoS ONE 6, e17395. Vázquez, D. P., Poulin, R., Krasnov, B. R. and Shenbrot, G. I. (2005).
Muzzall, P. M. and Whelan, G. (2011). Parasites of Fish from the Great Species abundance and the distribution of specialization in host–parasite
Lakes: A Synopsis and Review of the Literature, 1871–2010. Great Lakes interaction networks. Journal of Animal Ecology 74, 946–955.
Fishery Commission Miscellaneous Publication 2011–01. Ann Arbor, Violante-González, J. and Aguirre-Macedo, M. L. (2007). Metazoan
MI. Available from: http://www.glfc.org/pubs/SpecialPubs/2011-01.pdf. parasites of fishes from Coyuca Lagoon, Guerrero, Mexico. Zootaxa
Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., 1531, 39–48.
O’Hara, R. B., Simpson, G. L., Solymos, P., Stevens, M. H. H. and Violante-González, J., Aguirre-Macedo, M. L. and Mendoza-
Wagner, H. (2014). Vegan: Community Ecology Package. http://cran.r- Franco, E. F. (2007). A checklist of metazoan parasites of fish from Tres
project.org/web/packages/vegan/index.html Palos Lagoon, Guerrero, Mexico. Parasitology Research 102, 151–161.

View publication stats