Ecological Entomology (2020), 45, 155–166 DOI: 10.1111/een.

12785

Co-occurrence patterns and morphological similarity

of semiaquatic insects (Hemiptera: Gerromorpha)
in streams of Eastern Amazonia
A L A N A P . M . G U T E R R E S, 1,2 E R L A N E J . C U N H A, 2,3
B R U N O S . G O D O Y, 4 R O G É R I O R . S I L V A 1,5
and L E A N D R O J U E N 1,2,3 1 Programa de Pós-Graduação em Zoologia, Instituto de Ciências Biológicas,
Universidade Federal do Pará / Museu Paraense Emílio Goeldi, Belém, Brazil, 2 Laboratório de Ecologia e Conservação, Instituto de
Ciências Biológicas, Universidade Federal do Pará, Belém, Brazil, 3 Programa de Pós-Graduação em Ecologia, Instituto de Ciências
Biológicas, Universidade Federal do Pará, Belém, Brazil, 4 Instituto Amazônico de Agriculturas, Universidade Federal do Pará,
Belém, Brazil and 5 Museu Paraense Emílio Goeldi, Belém, Brazil

Abstract. 1. To understand how the semiaquatic bug communities are shaped, it was
first verified whether there was a pattern of co-occurrence between species. It was
subsequently tested whether the pattern found was related to environmental variables.
Lastly, it was verified whether morphological divergence between species was present
in the morphological and functional traits selected.
2. A pattern of species co-occurrence was found in the studied assemblage, related
to both environmental variables and interspecific biological interactions. Pairs of
species with negative, positive, and random co-occurrence were found. Traits related to
predation competition, sexual competition, and interaction with the environment showed
significant morphological divergence.
3. Therefore, more than one process defines species co-occurrence patterns in
semiaquatic bug communities. It is suggested that environmental influence is related
to species microhabitat preference. On the other hand, the morphological divergence
found is related to competition and sharing of food resources.
Key words. Niche, competition, morphological traits, functional traits, water striders,
small water striders.

Introduction use of resources in the environment, and are therefore proxies

for the niches occupied by the species (Weiher & Keddy, 1995;
The study of species distribution and patterns of co-occurrence is Pavoine et al., 2009; Mouillot et al., 2013).
one of the main themes in community ecology (Chesson, 2000; The two concepts that have received the most attention
Gotelli & McCabe, 2002; Downes & Reich, 2008; Ulrich & are: (i) environmental filters; and (ii) limiting similarity.
Gotelli, 2013; Ulrich et al., 2017a,b). To date, ecologists have These processes may have opposing effects in the shaping of
used species traits to deduce the shaping of assemblages (Per- assemblages but are not mutually exclusive and may occur
ronne et al., 2017; Rodrigues-Filho et al., 2017). A functional simultaneously (Weiher & Keddy, 1995). The structure of
trait can be defined as any characteristic of a species that can a community is determined by the amount of influence that
be measured at the individual level, be it morphological, phys- each process exerts (HilleRisLambers et al., 2012). Accord-
iological, or phenological (Mouillot et al., 2013; Arantes et al., ing to the environmental filters theory, species are selected
2019). Functional traits can also be used when attempting to through several biotic and abiotic filters (Lebrija-Trejos et al.,
quantify the niche overlap between species, as they reflect the 2010), and those species with inadequate trait combinations
cannot survive and reproduce in a given environment (Kluge
Correspondence: Alana Patricia M. Guterres, Avenida Perimetral, & Kessler, 2011; Herben & Goldberg, 2014; Velázquez et al.,
Cep 66075-110, ICB, Laboratório de Ecologia e Conservação. E-mail: 2015). From this perspective, the adaptations of the species
[email protected] to the characteristics of the environment should lead to a

© 2019 The Royal Entomological Society 155

156 Alana P. M. Guterres et al.

convergence of traits in the assemblages (Webb et al., 2002; pattern of co-occurrence and morphological divergence between
Velázquez et al., 2015). co-occurring species.
By contrast, limiting similarity would result in co-occurring
species with divergent traits (Hutchinson, 1959), as this reduces
the probability of concurring species that overlap in niche Materials and methods
occupation (Weiher et al., 1998; Webb et al., 2002; Abrams &
Study area
Rueffler, 2009). According to MacArthur and Levins (1967),
considering a limiting resources scenario, coexistence is possi- The study was carried out in streams of the Anapu River
ble when there is niche differentiation or specialisation among basin, in the Caxiuanã National Forest (Flona de Caxiuanã) and
species (Kneitel & Chase, 2004; Bode et al., 2011). In these surrounds, in the municipalities of Portel and Melgaço, Pará,
conditions, there would be no niche overlap (Pianka, 1974) and, Brazil (Fig. 1). The Flona de Caxiuanã is a Federal Protected
therefore, competition is minimised, avoiding the local exclu- Area totalling 3 330 000 ha of extensive preserved areas, 85%
sion of species with lower competitive potential (MacArthur & of which is covered by dense Terra Firme ombrophylous forest
Levins, 1967; Abrams & Rueffler, 2009). Thus, limiting similar- (Almeida et al., 1993). It has a humid tropical ‘Am’ climate,
ity generates morphological divergence that results in a commu- according to Köppen classification (Peel et al., 2007), with sea-
nity with less similar species (van Leeuwen & Etienne, 2013). sonal precipitation consisting of a rainy season between January
Several studies of aquatic environments have used aquatic and June and a dry season from July to December. Average
insects to assess species’ responses to environmental variation annual rainfall is 534 mm. The mean annual temperature is
(Cunha et al., 2015; Oliveira-junior et al., 2015; Valente-Neto 25.9 ∘ C, and the mean annual RH is c. 82% (Montag et al.,
et al., 2015; Cunha & Juen, 2017; Carter et al., 2017; de Castro 2019).
et al., 2017), as these are closely related to the abiotic conditions The Caxiuanã river network presents unique characteristics
of the streams (Merrit & Cummins, 1996; Resh, 2008). As with due to a mixture of lentic and lotic environments, called ‘lago
studies based on functional and morphological traits, they have de ria’ in Amazonia. This phenomenon was generated by the
demonstrated efficacy in evaluating assemblage mechanisms ‘drowning’ of the Anapú river valleys in the Holocene period
(Ulrich & Gotelli, 2012; Cadotte et al., 2015). With such studies, (Behling & Costa, 2000; Montag et al., 2019). As a result,
it is possible to identify or test hypotheses on what determines the river has lost its marginal floodplains and shows little
the distribution of aquatic insects or which forces govern the variation in water level over the dry and rainy seasons. These
structure of aquatic assemblages (Saito et al., 2016; de Castro changes also cause damming within the tributaries, such as the
et al., 2017; Valente-Neto et al., 2017). Caxiuanã, Pracupi, and Cariatuba rivers, through the loss of
Among the aquatic insects, Heteroptera (infraorder Gerro- lotic conditions, resulting in higher deposition of sediments that
morpha) are highly suitable as a model group for trait-based create natural dams, generating almost lacustrine environments
approaches (Múrria et al., 2018). They differ from other aquatic (Montag et al., 2019). Due to these processes, Caxiuanã streams
insects in that they present unique morphological modifications, have shallow (c. 30 cm), wide (c. 10 m wide) (Monteiro-júnior
such as hydrophobic coatings and leg modifications (Andersen, et al., 2016) and highly acidic waters (pH 5.5). The streambeds
1976; Armisén et al., 2015), which allow them to efficiently are formed by a large amount of organic matter, abundantly
explore the environment on the water’s surface (Andersen, covered by foliage and ligneous material. They have a main
1976). They are active predatory insects that chase prey on channel associated with an extensive flood plain, which can
the surface of the water but are also opportunistic, feeding reach up to 10 m wide (Montag et al., 2019).
on mosquito larvae and aquatic insects (Andersen & Weir, Streams of the Flona de Caxiuanã area present greater habi-
2004). They are able to detect vibration on the surface of the tat integrity due to greater preservation of forest and riparian
water caused by prey (Murphey, 1971), through perception vegetation, whilst surrounding streams are influenced by family
structures present in the tarsi (Murphey, 1971; Wilcox, 1995). farming practices from the populations living along the river-
They are commonly found living in large aggregations of banks (de Faria et al., 2017; Monteiro-júnior et al., 2016). This
has resulted in modification of the stream bed to allow the pas-
different species and families (Dias-silva et al., 2013; Ditrich
sage of canoes and modification of the riparian vegetation for
& Papáček, 2016), from lentic to lotic habitats (Dias-Silva
the planting of cassava and other subsistence cultivars.
et al., 2013).
In order to understand how the Gerromorpha community is
structured, we first tested whether there was a pattern of species Sample design and biological sampling
co-occurrence in the community and whether the environment
directly influences pattern formation. Concomitant to this, Collections were carried out in 32 streams during October
we verified if there was morphological divergence between and December 2012 and October 2013 (Fig. 1). In each stream,
species, using morphological and functional traits. Given that the collections were performed within a fixed length of 150 m,
these insects are predators (Andersen, 1982), with recognised divided into 10 sections of 15 m each, where each transect was
competition for environmental resources (Spence & Andersen, demarcated and labelled with the letters ‘A’ (downstream) to ‘K’
1994) in which some food resources may be limiting (Spence (upstream), in total consisting of 11 transects and 10 longitudinal
& Andersen, 1994; Ditrich & Papáček, 2016), our hypothesis sections of 15 m (A–B, B–C, C–D, … , J–K) (Fig. 2.1; adapted
is that the Gerromorpha assemblage should present a negative from Kaufmann et al., 1999). For biological sampling, the

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

 Co-occurrence patterns of the Gerromorpha 157

Fig. 1. Location of the 32 streams sampled during the dry season in 2012 and 2013 within Caxiuanã National Forest and surroundings. [Colour figure
can be viewed at wileyonlinelibrary.com].

longitudinal sections were divided into two subsections of 7.5 m distance of 75 m. For each stream, we determined the average of
each, in which the semiaquatic bugs were collected (Fig. 2). each variable. Characterisation of the physical habitat followed
Within each subsection, active collection was done on the the stream assessment protocol (see Peck et al., 2006 for more
surface of the water using an 18-cm-diameter net with a 1-mm details), where the following variables related to the physical
mesh (Cunha & Juen, 2017). For the purpose of comparison structure of each stream were measured: transect length (channel
in our data analysis, each stream in the study, composed of all width); mean total canopy cover (measured with a densitome-
subsamples (subsections) taken from the stream, was used as ter); mean channel cover from canopy vegetation (vegetation
a sample. Thus, the abundance of species per subsection was hanging over the banks, touching the water); percentage of rapid
summed and each stream represented one line in our dataset. The water movement within the stream (using a buoyant container to
collected Heteroptera were screened in the field and identified measure the time taken to travel 1 m); and percentage of all types
in the laboratory by means of dichotomous keys to species of standing water along the channel (perception of collector).
level (Kenaga, 1941; Kenaga, 1942; Drake & Hussey, 1951; The calculations of these variables are described in Kaufmann
Nieser & Melo, 1997; Moreira et al., 2011; Moreira & Barbosa, et al. (1999). Selection of the variables was according to those
2014; Aristizábal-garcía et al., 2015; Magalhães et al., 2016; defined by Cunha et al. (2015) and Juen et al. (2016) as impor-
Floriano et al., 2016). Voucher material was deposited in the tant for the group (Table 1).
Universidade Federal do Pará Museum of Zoology Scientific
Collection and Museu Paraense Emílio Goeldi.

Morphological measurements
Characterisation of the physicochemical properties of the
water and the physical structure of the habitat Six morphological traits (Table 2) were measured for all
species of Heteroptera collected. For the purposes of standard-
Five physicochemical variables of the water in each stream isation, and to avoid differences associated with sexual dimor-
were measured: dissolved oxygen, conductivity, pH, turbidity, phism, only males were measured. The measured morphologi-
and temperature. These variables were measured in each stream cal traits were selected based on their different functions in the
three times, in segments A, F and K, which were separated by a organism’s interaction with the environment (Table 2).

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

158 Alana P. M. Guterres et al.

Fig. 2. (a) Diagram of the sampling sections for physical characterisation of the habitat and collection of aquatic organisms in the stream/igarapé
sections (modified by Peck et al., 2006). The letters with a prime (K′ , F′ , A′ ) indicate the sections where physicochemical variables were measured.
(b) Diagram of the sampling sections, highlighting the division of the subsections. [Colour figure can be viewed at wileyonlinelibrary.com].

Data analysis which makes this model more suitable for detecting patterns of
interspecific interactions (Gotelli, 2000; Gotelli & Ulrich, 2010;
Species co-occurrence patterns. To verify if there was a pat- Ulrich et al., 2017b).
tern of species coexistence in the semiaquatic bug commu- We also employed the probabilistic model for the frequency
nity, we used the C-score index (Stone & Roberts, 1990) with of co-occurrence between pairs of species (Veech, 2013; Grif-
10 000 simulations. This is considered one of the best indices to fith et al., 2016). With this model, it is possible to identify which
determine species co-occurrence patterns (Gotelli, 2000). The species co-occur at a significantly high frequency (positive asso-
C-score consists of the average of all possible pairs of the ciations), at a significantly low frequency (negative associa-
checkerboard, calculated for species occurring at least once in tions), and at a frequency not significantly different from, or very
the matrix (Veech, 2014). In a competitively structured commu- close to, that expected at random (random association). The cal-
nity, the C-score should be significantly higher than expected by culation of the probability of occurrence of two species at the
chance, i.e. the observed index must be significantly higher than same sampling point is given by the formula:
the mean of the simulated index. A pattern of species segrega-
tion where the community is organised in the form {1,0}, {0,1}, ( ) ( )
N1 N−N1
×
may be the result of competition or habitat variability. On the j N2 −j
Pj =
other hand, a clustering pattern, where the community is organ- N
N2
ised in the form {1,1}, {1,1}, indicates the environmental filter
effect (Ulrich & Gotelli, 2012).
where N 1 is the number of sites where species 1 occurs, N 2 is
In order to generate the C-score, we used species occur-
the number of sites where species 2 (occurs,
) and N is the number
rence data (the presence–absence matrix). We obtained N
several random co-occurrence matrices (derived from the of sites where both occur. The term j1 represents the number
presence–absence matrix) to generate the C-score. We called of ways that j sites that have species 1 could be arranged among
the last matrix generated the co-occurrence matrix. the N 1)sites in the ‘population’ of all sample points. The term
(
N−N1
We used the fixed columns and fixed lines mode N2 −j
represents the number of ways that N 2 − j sites that have
(fixed – fixed – sim9), where the occurrence frequencies species 2, but not species 1, could be arranged among the N −N 1
(columns) and the species occurrence frequencies (lines) are sites. Multiplying these two quantities together (the numerator)
maintained. In this way, the differences between the sites were gives the total number of ways that j sites that have species 1 and
maintained, but the occurrences of the species were random, 2 could be arranged. N/N 2 represents the total number of ways

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

 Co-occurrence patterns of the Gerromorpha 159

Table 1. Environmental variables used in the study which were recorded at the 32 sampling points in Caxiuanã National Forest, Pará, Brazil.

All types of
% rapid water pools (standing water Mean total Conductivity Dissolved oxygen
Site movement† along the channel) (%) canopy coverage (μS cm –1 ) Temperature (∘ C) pH (mg litre –1 )

01 0.00 0.00 93.85 12.60 26.6 5.27 5.02

02 0.00 0.00 89.57 10.60 24.4 7.77 8.32
03 0.00 0.00 94.11 11.87 24.3 6.62 7.17
04 0.00 0.00 92.11 8.23 27.3 5.10 6.30
05 0.00 0.00 94.91 10.87 27.0 5.09 30.03
06 0.00 1.33 94.11 18.80 25.5 5.13 3.83
07 0.00 0.00 96.39 16.77 26.2 5.19 2.70
08 0.00 0.00 97.45 10.40 25.7 6.60 6.03
09 0.00 0.00 90.77 9.27 25.6 7.57 4.67
10 4.00 0.00 97.86 9.00 25.0 5.85 6.40
11 0.00 0.00 96.25 14.57 25.0 4.78 3.77
12 0.00 0.00 96.52 11.83 25.8 5.50 4.33
13 0.00 0.00 89.03 13.60 25.9 5.66 14.60
14 0.00 0.67 96.25 13.10 25.6 7.53 11.13
15 71.33 1.33 96.12 12.80 25.2 6.57 11.37
16 0.00 1.33 88.90 17.83 25.6 7.02 12.93
17 0.00 0.67 96.92 18.40 25.6 6.92 10.63
18 0.00 0.00 86.49 14.30 24.4 5.64 7.43
19 0.00 0.00 99.33 13.67 24.8 6.25 7.00
20 0.00 0.00 98.26 10.27 25.3 5.90 4.92
21 0.00 0.00 95.98 8.93 25.8 5.74 4.97
22 0.00 0.00 96.65 8.20 26.1 6.03 7.37
23 0.00 0.00 89.43 17.23 24.9 7.19 4.20
24 0.00 0.00 97.86 10.87 24.9 7.38 6.00
25 0.00 0.00 98.79 9.53 25.2 7.81 4.20
26 0.00 0.00 98.93 9.26 25.6 7.57 4.67
27 0.00 0.00 97.86 13.70 26.3 7.82 4.20
28 0.00 0.00 98.39 18.23 25.5 7.35 5.07
29 22.67 0.00 95.32 10.33 28.5 5.41 3.55
30 20.00 0.00 94.38 14.67 26.7 6.08 2.53
31 3.33 6.00 96.39 10.00 11.7 6.04 3.51
32 23.33 0.00 96.52 13.00 12.6 6.03 4.00

†Water velocity (falls + cascades + rapids + riffles → falls + cascades + rapids + riffles – % of channel length).

Table 2. Morphological traits, abbreviations, and their respective functions in the environment of individuals of semiaquatic bug (Gerromorpha)
species sampled in the Caxiuanã National Forest and surroundings, Pará, Brazil.

Morphological measurements Functions Reference

Body size (from the head to the end of the Sexual dimorphism and relationship between Preziosi and Fairbairn (1996); Khila
abdomen) dominant and inferior competitor (2014); Ditrich and Papáček (2016)
Foreleg size (left and right) (from the Body support, assists in feeding, prey Andersen (1982); Andersen, (1995);
trochanter to the end of the tarsus) preference, predation and copulation Andersen and Cheng (2004)
Middle leg size (left and right) (from the Balance of the body, important for the Andersen (1976, 1982, 1995)
trochanter to the end of the tarsus) flotation of the body beneath the water,
competition for alimentary resources
Hind leg size (from the trochanter to the end Competition for sexual partners and Toubiana and Khila (2019)
of the tarsus) oviposition habitats
Rostrum size (from the base to the Competition for alimentary resources Andersen (1982)
termination of the rostrum)

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

160 Alana P. M. Guterres et al.

that N 2 number of sites could be obtained from of a total of N aggregated distribution and overlap in resource use (Surkova
sites. Thus, the equation provides the proportion of N 2 sites that et al., 2018).
also have species 1 where the two species co-occur at j points For each of the communities we also calculated the standard-
(Griffith et al., 2016). ised effect size (SES), as follows:
SES = observed index-mean (simulated index)∕

Relationship between species co-occurrence and the SD (simulated index)

environment. In order to determine whether environmental
Following this, the values of 1000 randomised communities
variables affect the co-occurrence pattern of the Gerromorpha were averaged, where negative SES values indicate segregation
community, we first undertook a principal components analysis and positive SES values indicate aggregation.
(PCA; Legendre & Legendre, 2012) to summarise the environ- The size overlap analysis was performed with ecosim pro-
mental variation. We used the selected axis to carry out further fessional software (v.7.0) (Gotelli & Entsminger, 2004). The
analyses outlined in the following (Zwick & Velicer, 1986). remaining analyses were performed with r (R Development
We performed a linear regression (Zar, 2010) of the selected Core Team, 2013). To calculate the C-score we used the ecosimr
axis with the abundance of each species. The regression slope package (Gotelli & Ellison, 2013) and the ‘cooc_null_model’
coefficient values (b) and the respective standard errors of the function. To calculate the PCA we used the vegan (Oksanen
coefficients (Std) were extracted
√ and used to calculate the values et al., 2013), mass (Ripley et al., 2013), permute (Simpson,
of t, where t = bx – by/ Std bx + Std by. In this case, x and y 2012) and lattice (Sarkar, 2008) packages and the ‘princomp’
represent the pairs of species considered. Essentially, we calcu- function. Lastly, we used the vegan package and mantel func-
lated the difference of the regression slope/standard error of the tion for the Mantel analysis (Oksanen et al., 2013). All details
difference of the regression slope. This calculation was used to of the analyses can be found in Document S1.
generate the triangular matrix of species pairs, which indicates
the difference in response of these pairs to the environment
(for more details see Document S1). Finally, we performed the Results
Mantel test (Mantel, 1967) to test the correlation between the
co-occurrence matrix and the triangular matrix of species pairs Heteroptera community description
(Oksanen, 1983; Legendre & Legendre, 2012). The test results
A total of 973 adult individuals were collected from 27
in a value of r that indicates the Pearson correlation between
species, with 12 belonging to the family Gerridae, 13 to
the matrices.
the family Veliidae, one from the family Mesoveliidae and
one from the family Hebridae. The most abundant species
was Telmatometra fusca Kenaga, 1941 (Gerridae) with 255
Size overlap
individuals. For the family Veliidae, the most abundant species
To verify whether there was morphological divergence was Euvelia discala J. Polhemus & D. Polhemus, 1984 with
between pairs of species within the community, we used the 146 individuals. The least abundant species from the family
ecosim ‘size overlap’ module (Gotelli & Entsminger, 2004), to Gerridae were Limnogonus aduncus Drake & Harris 1982, with
determine whether the morphological overlap is smaller than only one individual, and the least abundant species from the
what would be expected in communities with these randomly family Veliidae was Rhagovelia humboldti D. Polhemus 1997,
distributed variables. The size overlap module allows for testing also with one individual. We collected two individuals belonging
to Hebridae and Mesoveliidae.
of unusual patterns in the body sizes of coexisting species and
comparing these patterns with what would be expected in an
assemblage that was not structured by competition. Species co-occurrence patterns
Thus, a matrix of the species’ mean morphological measures
was made. We used the program default model ‘variance in seg- We found a non-random pattern of heteropteran species
ment length’ and the uniform distribution parameter to calculate co-occurrence using the C-score co-occurrence index
the observed variance for each of the morphological measure- (C-score = 9.73; P < 0.001). The probability of co-occurrence
ments, and compared it with those of 1000 randomly assem- model identified 351 possible combinations of pairs of species.
bled communities generated by null models. The ‘data-defined’ Of these pairs, 242 (68.95%) were excluded from the analysis
option was used, where the program detects the maximum and as they presented a co-occurrence of < 1. Of the 109 pairs
minimum values of the data. For the data transformation we used remaining, nine were identified with significant co-occurrence,
four of which were positive co-occurrences and five of which
the ‘log(size ratios)’ option to test patterns on adjacent species.
were negative co-occurrences (Table 3).
This option indicates that the null model will test the propor-
tions, not the absolute values. Thus, in a community structured
by competition, the distribution of morphological measures is Relationship between species co-occurrence and the
indicated by the observed variation in length, being smaller than environment
expected by chance (observed < expected). In this case, we have
a segregated distribution of species morphology and probably The Mantel test between the random matrices generated by
resource sharing. The opposite (observed > expected) indicates the coexistence index and the triangular matrix of species versus

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

 Co-occurrence patterns of the Gerromorpha 161

Table 3. Results of the co-occurrence probability model of pairs of semiaquatic bug (Gerromorpha) species, sampled in the Caxiuanã National Forest
and surroundings, Pará, Brazil.

Sp1 Sp2 Cooc. obs Cooc.prob Cooc.exp P lt P gt

Cylindrostethus palmaris Ovatametra bella 1 0.130 4.2 0.009 0.999

C. palmaris Stridulivelia alia 4 0.034 1.1 0.999 0.003
C. palmaris Stridulivelia tersa 4 0.034 1.1 0.999 0.003
E. lata O. bella 0 0.093 3 0.005 1
Euvelia discala O. bella 6 0.297 9.5 0.014 0.997
O. bella Rheumatobates minutus flavidus 16 0.353 11.3 0.999 <0.001
O. bella Stridulivelia strigosa 0 0.074 2.4 0.020 1
O. bella S. tersa 0 0.093 3 0.005 1
O. bella T. fusca 17 0.427 13.7 0.998 0.010

‘Sp1’ and ‘Sp2’ represent the species of the pair in question; Cooc. Obs, Cooc.prob, and Cooc.exp represent the number of sites where species co-occur,
the probability of co-occurrence of the pair of species, and the number of expected sites where both co-occur, respectively; P lt and P gt indicate whether
there was a significantly negative or positive co-occurrence between the species pair, respectively.

Table 4. Values of Heteroptera community size overlap analysis.

Size overlap test values

Morphological measurements Observed index Mean of simulated index Variance of simulated index O<E O>E P(O < E) SES

Body size 0.005 0.006 < 0.001 553 447 0.447 −0.33
Left foreleg size 0.002 0.005 < 0.001 996 6 0.006 −1.31
Right foreleg size 0.002 0.007 < 0.001 987 13 0.013 −1.26
Left middle leg size 0.003 0.012 < 0.001 999 1 0.001 −1.44
Right middle leg size 0.003 0.012 < 0.001 985 15 0.015 −1.35
Left hind leg size 0.005 0.006 < 0.001 726 274 0.274 −0.64
Right hind leg size 0.004 0.009 < 0.001 934 66 0.066 −1.02
Rostrum size 0.001 0.002 < 0.001 700 300 0.300 −0.57

O < E (observed < expected), number of communities in which the observed index of size overlap was less than expected by chance; O > E, number
of communities in which the size overlap index was greater than expected by chance. Values of P in bold signify that for this trait there is significant
divergence between the species of the studied assemblage. SES, standardised effect size (negative values indicate trait divergence pattern).

species revealed that when considering the general pattern of Four species pairs exhibited aggregate co-occurrence (positive),
the community, co-occurrence was related to environmental and five pairs exhibited segregated co-occurrence (Table 3).
variables (r = 0.232, P = 0.01). Thus, we partially accept our hypothesis that species have a
negative co-occurrence pattern, as we found species pairs with
negative, positive, and random co-occurrence. Traits related to
Size overlap
predation competition, sexual competition, and interaction with
We measured a total of 407 individuals belonging to 21 the environment (forelegs and middle legs; see Table 4) showed
species. Limnogonus aduncus Drake & Harris 1992, Euvelia significant morphological divergence.
advena Drake 1957, Rhagovelia evidis Bacon 1948, Ovatametra Previous studies have related morphological and functional
minima Kenaga, 1942, and Neogerris sp. were not measured, traits to environmental characteristics (Dray et al., 2014). It
as only females were captured. We measured only the males is common to use ordinations to try to establish a relation-
of each species, due to the sexual dimorphism present in the ship between the environment and species traits (Legendre
species. The overlap test was performed with all morphological & Legendre, 2012). Morphological divergence tests between
measures, as each one represents a different relationship of the pairs of species are able to determine wide patterns in the
individual with the environment and other individuals. We found community but are poor at detecting variations at the local
morphological divergence between pairs of species for the left scale (Arregoitia et al., 2018). In addition, the index we used to
foreleg and middle leg traits only (Table 4). detect co-occurrence patterns can generate the same result for
a competition-structured community as for those communities
Discussion in environments with high heterogeneity. To minimise these
problems and have a more complete view of the organisation
We found a species co-occurrence pattern in the studied of this community, we used a series of statistical analyses
assemblage (see C-score result) that related to environmen- to test whether the patterns of co-occurrence were related
tal variables (see Mantel results) and interspecific biological to environmental variation. This methodology allowed us to
interactions (see model co-occurrence and size overlap results). answer our hypotheses with greater precision.

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

162 Alana P. M. Guterres et al.

In general, Gerromorpha species respond to environmental found for some traits (Cadotte et al., 2015; Ulrich et al., 2016).
variation. Previous works have shown that Gerromorpha is a Here we have seen that the forelegs and the middle legs, which
group closely related to a stream’s physicochemical characteris- are traits related to feeding (Andersen, 1982), sexual behaviour
tics (Dias-Silva et al., 2013; Cunha et al., 2015; Juen et al., 2016 (Andersen, 1982; Andersen & Cheng, 2004), and locomotive
; Cunha & Juen, 2017). Undisturbed environments, which tend efficiency (Andersen, 1976; Andersen, 1995), present morpho-
to exhibit greater environmental heterogeneity, generally present logical divergence between pairs of co-occurring species.
a greater diversity of semiaquatic bugs (Cunha et al., 2015). This In streams with a weak current, such as those of Caxiuanã,
type of response to the environment is demonstrated by Gerro- competition for food can generate selective pressure (Harada
morpha species, which possess morphological, ecological, and & Spence, 2000), which is expressed through the morpholog-
physiological characteristics directly related to specific compo- ical and capability differences in traits of some pairs of species.
nents of the streams (Hufnagel et al., 1999; Skern et al., 2010), Gerromorpha-derived offspring have more elongated middle
e.g. physicochemical and water velocity variables (Juen et al., legs and oar-like locomotion (Andersen, 1982). These individu-
2016; Cunha & Juen, 2017). als can move faster and more efficiently (Andersen, 1982; Cru-
Some of the species sampled in this study also showed distri- mière et al., 2016, 2019) and, thus, can capture prey at faster
butional patterns related to the environment seen in other studies rates with lower energy expenditure (Crumière et al., 2016).
(Dias-Silva et al., 2013; Cunha et al., 2015; Cunha & Juen, This locomotive efficiency allows for the occupation of new
2017). For example, individuals of the genus Microvelia were niches (Crumière et al., 2016) and may render species with these
previously associated with streams with a slower current at the characteristics more efficient predators and superior competitors
surface (Taylor & McPherson, 2003; Dias-silva et al., 2013). (Santos et al., 2017; Toubiana & Khila, 2019).
The relationship between these species and the particular envi- The co-occurrence of morphologically and competitively sim-
ronmental variables indicates that individuals of a species select ilar species is also possible, provided there is some difference in
habitats, or microhabitats, that have the most adequate charac- the use of resources (MacArthur & Levins, 1967). For example,
teristics for their development (St. Pierre & Kovalenko, 2014). in this study we have seen that species of the same subfamily
The species Euvelia discala, for example, has a preference for which are morphologically similar (Ovatametra bella/T. fusca)
leaf banks (Vilardi, 2015), where this species can be found along showed aggregated co-occurrence. Thus, throughout the evolu-
with other species of Euvelia or Microvelinae (Andersen, 1982). tion of these species, interspecific competition must have caused
The selection of microhabitats that contain ideal conditions some form of differentiation in resource use, enabling them to
and resources for the development of the species can gen- coexist (MacArthur & Levins, 1967).
erate both segregation and aggregation patterns among pairs Many pairs of species exhibited random co-occurrence pat-
of co-occurring species (D’Amen et al., 2018). In our study, terns or were not significantly different from those expected to
species with segregation patterns have distinct environmental be random. This apparent randomness may be related to the high
requirements (Andersen, 1982; Vilardi, 2015; Floriano et al., dispersibility of Gerromorpha species (Andersen, 1982); species
2016); therefore, they are unlikely to coexist in the same envi- with good dispersibility stay at a given location for a short time
ronment, generating the segregation patterns found. (Ulrich, 2004; Gotelli & McGill, 2006).
On the other hand, some species belonging to the Gerridae Other factors, such as temporal differences in the life cycle,
and Vellidae families, which have high niche overlap as well may also generate the pattern of co-occurrence. Individuals
as habitat preference (Andersen, 1982), presented an aggregate of Gerris argentatus, for example, can coexist on the same
co-occurrence pattern. The species pairs Cylindrostethus pal- substrate when in different life cycle stages (Vepsäläinen &
maris/Stridulivelia tersa and C. palmaris/Stridulivelia strigosa Järvinen, 1974). Younger instars coexist with older instars,
(Table 4) are associated with the same types of microhabitats, increasing efficiency in the use of the environment and available
such as shaded sites, exposed roots, wood debris, and macro- resources. Vepsäläinen and Järvinen (1974) also observed that
phyte banks (Andersen, 1982; Floriano et al., 2016). Cylin- young instars may select habitats along the riverbank to protect
drostethus and Stridulivelia are morphologically different and themselves from ‘rough’ water, as they do not yet possess all
phylogenetically distant genera (Damgaard, 2008); therefore, the layers of fully developed hydrophobic hairs seen in adults
despite having similar ecological requirements with high niche (Andersen, 1976) and are therefore more likely to drown or soak
overlap, co-occurrence is possible because of morphological when the water is agitated (Toubiana & Khila, 2019).
differences that allow differential use of the same resources Our study focused on the influence of morphological and
(Hutchinson, 1959). functional characteristics and environmental variation in species
At the local scale, the morphological divergence between co-occurrence patterns. However, other mechanisms, such as
species within the assemblage is commonly interpreted as evi- temporal differences in the life cycle (Vepsäläinen & Järvinen,
dence of interspecific competition (Bregman et al., 2015; Ulrich 1974), differences in behaviour, morphological variation, and
et al., 2016). However, we know that other processes can deter- intraspecific competition, are also important factors that should
mine these patterns in the community (Heino, 2013), such as be taken into account in future studies. We observed that species
predation (Giam & Olden, 2016) and facilitation (Gallien et al., co-occurrence patterns, at the scale studied, were related to
2018). According to Arregoitia et al. (2018), interspecific com- environmental variation and biological interactions. Therefore,
petition itself can generate divergence patterns and morphologi- it is not just one process that defines species co-occurrence
cal convergence. However, biological interactions, such as com- patterns in the Gerromorpha community. We suggest that the
petition, may be responsible for the morphological divergence influence of the environment is related to species’ preference of

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

 Co-occurrence patterns of the Gerromorpha 163

microhabitats. On the other hand, the morphological divergence Andersen, N.M. (1982) The semiaquatic bugs (Hemiptera: Gerromor-
found is related to competition and sharing of food resources. pha). Phylogeny, adaptations, biogeography and classification. Brill
Academic, Leinden.
Andersen, N.M. (1995) Cladistic inference and evolutionary scenarios:
locomotory structure, function, and performance in water striders.
Acknowledgements
Cladistics, 11, 279–295.
We extend our thanks to the Biodiversity Research Program Andersen, N.M. & Cheng, L. (2004) The marine insect Halobates
(Heteroptera: Gerridae). Oceanography and Marine Biology, 42,
of Eastern Amazonia (PPBIO) and the Fundação Amazônia
119–180.
de amparo a Estudos e Pesquisas (Research Foundation of
Andersen, N. M. & Weir, T. A. (eds) (2004) Australian water bugs: their
the State of Pará FAPESPA) through the FAPESPA project:
biology and identification (Hemiptera-Heteroptera Gerromorpha &
ICAAF 03/2011, for support. We are also grateful to the Nepomorpha). Apollo Books, Melbourne.
administrators and staff of the Estação Cientifica Ferreira Penna Arantes, C.C., Fitzgerald, D.B., Hoeinghaus, D.J. & Winemiller, K.O.
(Ferreira Penna Research Station), and the team that participated (2019) Impacts of hydroelectric dams on fishes and fisheries in
in field data collection, in particular, Dr Ana Paula Justino tropical rivers through the lens of functional traits. Current Opinion
de Faria and Dr Claúdio da Silva Monteiro Júnior. We also in Environmental Sustainability, 37, 28–40.
thank Dr Felipe Moreira for confirming the identification of Aristizábal-garcía, H., Floriano, C.F.B., Moreira, F.F.F. & Bispo, P.C.
some specimens. We also thank the Conselho Nacional de (2015) New species and notes on Microveliinae from northern
Desenvolvimento da Pesquisa – CNPq (National Council for South America (Hemiptera: Heteroptera: Veliidae). Zootaxa, 3980,
Scientific and Technological Development) for the grant of a 591–596.
master’s scholarship to APMG (no. 131284/2015-0), a PhD Armisén, D., Refki, P.N., Crumiere, A.J.J., Viala, S., Toubiana, W.
scholarship to EJC (no. 165908/2014-9) and a productivity grant & Khila, A. (2015) Predator strike shapes antipredator phenotype
to LJ (no. 307597/2016-4). We declare we have no competing through new genetic interactions in water striders. Nature Commu-
nications, 6, 8153.
interests.
Arregoitia, L.D.V., Hernández-Canchola, G., Santini, L., Schweizer, M.,
Cabrera-Garrido, M.Y. & León-Paniagua, L.S. (2018) Co-occurrence
and character convergence in two Neotropical bats. Journal of
Author contributions Mammalogy, 99, 1055–1064.
Behling, H. & Costa, L.D. (2000) Holocene environmental changes from
Project design: APMG, EJRC, LJ. Data collection: EJRC. Iden- the Rio Curuá record in the Caxiuana region, eastern Amazon Basin.
tification of species: APMG, EJRC. Measurement of speci- Quaternary Research, 53, 369–377.
mens: APMG. Data analysis: APMG, BSG, LJ. Writing of the Bode, M., Bode, L. & Armsworth, P.R. (2011) Different dispersal
manuscript: APMG, EJRC, BSG, RRS, LJ. abilities allow reef fish to coexist. Proceedings of the National
Academy of Sciences, 27, 16317–16321.
Bregman, T.P., Lees, A.C., Seddon, N., MacGregor, H.E., Darski, B.,
Supporting Information Aleixo, A. et al. (2015) Species interactions regulate the collapse
of biodiversity and ecosystem function in tropical forest fragments.
Additional supporting information may be found online in the Ecology, 96, 2692–2704.
Supporting Information section at the end of the article. Cadotte, M.W., Arnillas, C.A., Livingstone, S.W. & Yasui, S.L.E. (2015)
Predicting communities from functional traits. Trends in Ecology &
Document S1. Analysis script Guterres et al. 2019.
Evolution, 30, 510–511.
Table S1. All species collected from the 32 collection points in Carter, J.L., Resh, V.H. & Hannaford, M.J. (2017) Macroinvertebrates as
Caxiuanã National Forest, Pará, Brazil. biotic indicators of environmental quality. Methods in Stream Ecology
(Third Edition), 2, 293–318.
Table S2. Average traits of 21 species collected in Caxiuanã de Castro, D.M., Dolédec, S. & Callisto, M. (2017) Landscape variables
National Forest, Pará, Brazil. influence taxonomic and trait composition of insect assemblages in
Neotropical savanna streams. Freshwater Biology, 62, 1472–1486.
Table S3. Latitude and longitude. Chesson, P. (2000) Mechanisms of maintenance of species diversity.
Annual Review of Ecology and Systematics, 31, 343–366.
Crumière, A.J.J., Santos, M.E., Sémon, M., Armisén, D., Moreira, F.F.F.
References & Khila, A. (2016) Diversity in morphology and locomotory behavior
Abrams, P.A. & Rueffler, C. (2009) Coexistence and limiting similarity is associated with niche expansion in the semi-aquatic bugs. Current
of consumer species competing for a linear array of resources. Biology, 26, 1–7.
Ecology, 90, 812–822. Cunha, E.J. & Juen, L. (2017) Impacts of oil palm plantations on changes
Almeida, S.S., Lisboa, P.L.B. & Silva, A.S.L. (1993) Diversidade in environmental heterogeneity and Heteroptera (Gerromorpha and
florística de uma comunidade arbórea na Estação Cientifica Ferreira Nepomorpha) diversity. Journal of Insect Conservation, 21, 111–119.
Penna, em Caxiuanã (Pará). Boletim do Museu Paraense Emilio Cunha, E.J., Montag, L.F.A. & Juen, L. (2015) Oil palm crops effects
Goeldi, 9, 93–128. on environmental integrity of Amazonian streams and Heteroptera
Andersen, N.M. (eds) (1976) A comparative study of locomotion on (Hemiptera) species diversity. Ecological Indicators, 52, 422–429.
the water surface in semiaquatic bugs (Insecta, Hemiptera, Gerromor- D’Amen, M., Mod, H.K., Gotelli, N.J. & Guisan, A. (2018) Disentan-
pha). Videnskabelige Meddelelser fra dansk Naturhistoriske Foren- gling biotic interactions, environmental filters, and dispersal limita-
ing, 139, 337–396. tion as drivers of species co-occurrence. Ecography, 41, 1233–1244.

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

164 Alana P. M. Guterres et al.

Damgaard, J. (2008) Phylogeny of the semiaquatic bugs HilleRisLambers, J., Aldler, P.B., Harpole, W.S., Levine, L.M. &
(Hemiptera-Heteroptera, Gerromorpha). Insect Systematics & Mayfield, M.M. (2012) Rethinking community assembly through the
Evolution, 39, 431–460. lens of coexistence theory. The Annual Review of Ecology, Evolution,
Dias-Silva, K., Soares, H.S.R.C., Giehl, N.F.S. & Juen, L. (2013) Dis- and Systematics, 43, 110411–160411.
tribuição de Heteroptera Aquáticos (Insecta) em Diferentes Tipos de Hufnagel, L., Bakonyi, G. & Vásárhelyi, T. (1999) New approach
Substratos de Córregos do Cerrado Matogrossense. EntomoBrasilis, for habitat characterization based on species lists of aquatic and
6, 132–140. semiaquatic bugs. Environmental Monitoring and Assessment, 58,
Ditrich, T. & Papáček, M. (2016) Differences in prey capture in semi- 305–316.
aquatic bugs (Heteroptera: Gerromorpha). Entomological Science, 19, Hutchinson, G.E. (1959) Homage to Santa Rosalia or why are there so
34–41. many kinds of animals? American Naturalist, 93, 145–159.
Downes, B.J. & Reich, P. (2008) What is the spatial structure of stream Juen, L., Cunha, E.J., Carvalho, F.G., Ferreira, M.C., Begot, T.O.,
insect populations? Dispersal behaviour at different life-history Andrade, A.L. et al. (2016) Effects of oil palm plantations on the
stages. Aquatic insects: challenges to populations: Proceedings of the habitat structure and biota of streams in eastern Amazon. River
Royal (ed. by J. Lancaster and R. A. Briers), pp. 184, 184–203, 203. Research and Applications, 32, 2081–2094.
Lancaster and R. A. Briers, Londres. Kaufmann, P.R., Levine, P., Robison, G.E., Seeliger, C. & Peck,
Drake, C.J. & Hussey, R.F. (1951) Concerning some American Microv- D.V. (1999) Quantifying physical habitat in wadeable streams. U.S
elia (Hemiptera: Veliidae). Florida Entomologist, 34, 137–145. Environmental Protection Agency, Whashington, D.C.
Dray, S., Choler, P., Dolédec, S., Peres-Neto, P.R., Thuiller, W., Pavoine, Kenaga, E.E. (1941) The genus Telmatometra Bergroth
S. et al. (2014) Combining the fourth-corner and the RLQ methods (Hemiptera-Gerridae). University of Kansas, 17, 169–183.
for assessing trait responses to environmental variation. Ecology, 95, Kenaga, E.E. (1942) A new genus in the Holobatinae
14–21. (Gerridae-Hemiptera). Journal of the Kansas Entomological Society,
de Faria, A.P.J., Ligeiro, R., Callisto, M. & Juen, L. (2017) Response of 15, 136–141.
aquatic insect assemblages to the activities of traditional populations Khila, A. (2014) Comparative functional analyses of ultrabithorax reveal
in eastern Amazonia. Hydrobiologia, 802, 39–51. multiple steps and paths to diversification of legs in the adaptive
Floriano, C.F.B., Paladini, A. & Cavichioli, R.R.(2016) Systematics of radiation of semi-aquatic insects. Evolution, 68, 2159–2170.
the South American species of Cylindrostethus Mayr, 1865 (Hemip Kluge, J. & Kessler, M. (2011) Phylogenetic diversity, trait diversity and
tera : Heteropt era : Gerridae), with a new specie s from Amazoni niches: species assembly of ferns along a tropical elevational gradient.
anBrazil and Peru. Invertebrate Systematics, 30, 431–462. Journal of Biogeography, 38, 394–405.
Gallien, L., Zurell, D. & Zimmermann, N.E. (2018) Frequency and Kneitel, J.M. & Chase, J.M. (2004) Disturbance, predator and resource
intensity of facilitation reveal opposing patterns along a stress gra- interactions alter container community composition. Ecology, 85,
dient. Ecology and Evolution, 8, 2171–2181. 2088–2093.
Giam, X. & Olden, J.D. (2016) Environment and predation govern Lebrija-Trejos, E., Pérez-Garcia, E.A., Meave, J.A., Bongers, F. &
fish community assembly in temperate streams. Global Ecology and Poorter, L. (2010) Functional traits and environmental filtering drive
Biogeography, 25, 1194–1205. community assembly in a species-rich tropical system. Ecology, 91,
Gotelli, N.J. (2000) Null model analysis of species co-occurrence 386–398.
patterns. Ecology, 81, 2606–2621. van Leeuwen, E.V. & Etienne, R.S. (2013) Caught in the middle:
Gotelli, N. J., & Ellison, A. M. (2013) EcoSimR 1.00. [WWW doc- asymmetric competition causes high variance in intermediate trait
ument]. URL http://www.uvm.edu/~ngotelli/EcoSim/EcoSim.html abundances. Theoretical Population Biology, 85, 26–37.
[acessed on 12 January 2018]. Legendre, P. & Legendre, L. (eds) (2012) Numerical Ecology. Elsevier,
Gotelli, N.J. & Entsminger, G.L. (2004) EcoSim: Null models software New York.
for ecology. version 7. Acquired Intelligence Inc. & Kesey-Bear. MacArthur, R. & Levins, R. (1967) The limiting similarity, convergence
Jericho, VT 05465 [WWW document]. URL http://garyentsminger and divergence of coexisting species. American Naturalist, 101,
.com/ecosim/index.htm [acesssed on 10 January 2018]. 377–385.
Gotelli, N.J. & McCabe, D.J. (2002) Species co-occurrence: a Magalhães, O.M., Moreira, F.F.F. & Galvão, C. (2016) A new species of
meta-analysis of J. M. Diamond’s assembly rules model. Ecology, Rhagovelia Mayr, 1865 (Hemiptera: Heteroptera: Veliidae) from Pará
83, 2091–2096. state, with an updated key to Brazilian species of the robusta group.
Gotelli, N.J. & McGill, B.J. (2006) Null versus neutral models: What’s Zootaxa, 4171, 586–594.
the difference? Ecography, 29, 793–800. Mantel, N. (1967) The detection of disease clustering and a generalized
Gotelli, N.J. & Ulrich, W. (2010) The empirical Bayes approach as regression approach. Cancer Ressearch, 27, 209–220.
a tool to identify non-random species associations. Oecologia, 162, Merrit, R.W. & Cummins, K.W. (1996) An Introduction to the Aquatic
463–477. Insects of North America. Kendall Hunt Publishing Company, Lowa,
Griffith, D.M., Veech, J.A. & Marsh, C.J. (2016) Cooccur: probabilistic Dubuque.
species co-occurrence analysis in R. Journal of Statistical Software, Montag, L.F.A., Leão, H., Benone, N.L., Monteiro-Júnior, C.S., Faria,
69, 1–17. A.P.J., Nicacio, G. et al. (2019) Contrasting associations between
Harada, T. & Spence, J.R. (2000) Nymphal density and life histories of habitat conditions and stream aquatic biodiversity in a forest reserve
two water striders (Hemiptera: Gerridae). The Canadian Entomolo- and its surrounding area in the eastern Amazon. Hydrobiologia, 826,
gist, 132, 353–363. 263–277.
Heino, J. (2013) The importance of metacommunity ecology for envi- Monteiro-Júnior, C.S., Esposito, M.C. & Juen, L. (2016) Are the adult
ronmental assessment research in the freshwater realm. Biological odonate species found in a protected area different from those present
Reviews, 88, 166–178. in the surrounding zone? A case study from eastern Amazonia.
Herben, T. & Goldberg, D. (2014) Community assembly by limiting Journal of Insect Conservation, 20, 643–652.
similarity vs. competitive hierarchies: testing the consequences of Moreira, F.F.F. & Barbosa, J.F. (2014) A new Rhagovelia (Hemiptera:
dispersion o individual traits. Journal of Ecology, 102, 156–166. Heteroptera: Veliidae) from the Brazilian Amazon, with a key to

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

 Co-occurrence patterns of the Gerromorpha 165

species of the robusta group known from the country. Zootaxa, 3790, URL http://CRAN.R-project.org/package=MASS [accessed on 1 de
595–600. February 2018].
Moreira, F.F.F., Alecrim, V.P., Ribeiro, J.R.I. & Nessimian, J.L. (2011) Rodrigues-Filho, C.A.S., Gurgel Lourenço, R.C., Lima, S.M.Q.,
Identification key to the Gerridae (Insecta: Heteroptera: Gerromor- Oliveira, E.F. & Sánchez-Botero, J.I. (2017) What governs the
pha) from the Amazon River floodplain, Brazil, with new records for functional diversity patterns of fishes in the headwater streams of
the Brazilian Amazon. Zoologia, 28, 269–279. the humid forest enclaves: environmental conditions, taxonomic
Mouillot, D., Graham, N.A.J., Villéger, S., Mason, N.W.H. & Bellwood, diversity or biotic interactions? Environmental Biology of Fishes,
D.R. (2013) A functional approach reveals community responses to 100, 1023–1032.
disturbances. Trends in Ecology & Evolution, 28, 167–177. Saito, V.S., Cianciaruso, M.V., Siqueira, T., Fonseca-Gessner, A.A. &
Murphey, R.K. (1971) Sensory aspects of the control of orientation to Pavoine, S. (2016) Phylogenies and traits provide distinct insights
prey by the water strider, Gerris remigis. Journal of Comparative about the historical and contemporary assembly of aquatic insect
Physiology A: Neurothology, Sensory, Neural, and Behavioral Physi- communities. Ecology and Evolution, 6, 2925–2937.
ology, 72, 168–185. Santos, M.E., Le Bouquin, A., Crumière, A.J.J. & Khila, A. (2017)
Múrria, C., Dolédec, S., Papadopoulou, A., Vogler, A.P. & Bonada, Taxon-restricted genes at the origin of a novel trait allowing access
N. (2018) Ecological constraints from incumbent clades drive trait to a new environment. Science, 358, 1–4.
evolution across the tree-of-life of freshwater macroinvertebrates. Sarkar, D. (2008) Lattice: Multivariate Data Visualization with R.
Ecography, 41, 1049–1063. Springer, New York, NY.
Nieser, N. & Melo, A.L. (1997) Os heterópteros aquáticos de Minas Simpson, G. L. (2012). Restricted permutations; using the permute pack-
Gerais: guia introdutório com chave de identificação para as espécies age. [WWW document]. http://r.adu.org.za/web/packages/permute/
de Nepomorpha e Gerromorpha. Universidade Federal de Minas vignettes/permutations.pdf [acessed on 13 de January 2018].
Gerais. Minas Gerais, Belo Horizonte. Skern, M., Zweimüller, I. & Schiemer, F. (2010) Aquatic Het-
Oksanen, J. (1983) Ordination of boreal heath-like vegetation with prin- eroptera as indicators for terrestrialisation of floodplain habitats.
cipal component analysis, correspondence analysis and multidimen- Limnologica–Ecology and Management of Inland Waters, 40,
sional scaling. Vegetatio, 52, 181–189. 241–250.
Oksanen, J., Blanchet, FG, Kindt, R., Legendre, R., Minchin, PR, Spence, J.R. & Andersen, N.M. (1994) Biology of water striders:
O’Hara, PB et al. (2013) vegan: Community Ecology Package. Pacote interactions between systematics and ecology. Annual Review of
R, version 2.0–10. [WWW document]. URL http://CRAN.R-project Entomology, 39, 101–128.
.org/package=vegan [accessed on 1 February 2018]. St. Pierre, J.I. & Kovalenko, K.E. (2014) Effect of habitat complexity
Oliveira-Junior, J.M.B., Shimano, Y., Gardner, T.A., Hughes, R.M., attributes on species richness. Ecosphere, 5, 1–10.
De Marco, P. & Juen, L. (2015) Neotropical dragonflies (Insecta: Stone, L. & Roberts, A. (1990) The checkerboard score and species
Odonata) as indicators of ecological condition of small streams in the distributions. Oecologia, 85, 74–79.
eastern Amazon. Austral Ecology, 40, 733–744. Surkova, E.N., Vinarskaya-Korallo, N.P., Vinarski, M.V. & Mescht,
Pavoine, S., Vallet, J., Anne-Béatrice, D., Gachet, S. & Daniel, H. (2009) L.V.D. (2018) Body size distribution in flea communities har-
On the challenge of treating various types of variables: application boured by Siberian small mammals as affected by host species, host
for improving the measurement of functional diversity. Oikos, 118, sex and scale: scale matters the most. Evolutionary Ecology, 32,
391–402. 643–662.
Peck, D.V., Herlihy, A.T., Hill, B.H., Hughes, R.M., Kaufmann, P.R., Taylor, S.T. & McPherson, J.E. (2003) Voltinism and laboratory rearing
Klemm, D.J. et al. (2006) Environmental Monitoring and Assessment of Microvelia hinei (Heteroptera: Gerromorpha: Veliidae). The Great
Program-Surface Waters: Western Pilot Study Field Operations Man- Lakes Entomologist, 36, 1–9.
ual for Wadeable Streams. Environmental Protection Agency, Wash- Toubiana, W. & Khila, A. (2019) Fluctuating selection strength and
ington, DC. intense male competition underlie variation and exaggeration of a
Peel, M.C., Finlayson, B.L. & McMahon, T.A. (2007) Updated world water strider’s male weapon. BioRxiv, 286, 1–10.
map of the Köppen–Geiger climate classification. Hydrology and Ulrich, W. (2004) Species co-occurrences and neutral models: reassess-
Earth System Sciences, 11, 1633–1644. ing JM Diamond’s assembly rules. Oikos, 107, 603–609.
Perronne, R., Munoz, F., Borgy, B., Reboud, X. & Gaba, S. (2017) How Ulrich, W. & Gotelli, N.J. (2012) A null model algorithm for
to design trait-based analyses of community assembly mechanisms: presence-absence matrices based on proportional resampling.
insights and guidelines from a literature review. Perspectives in Plant Ecological Modelling, 244, 20–27.
Ecology, Evolution and Systematics, 25, 29–44. Ulrich, W. & Gotelli, N.J. (2013) Pattern detection in null model
Pianka, E.R. (1974) Niche overlap and diffuse competition. Proceedings analysis. Oikos, 122, 2–18.
of the National Academy of Sciences, 71, 2141–2145. Ulrich, W., Jabot, F. & Gotelli, N.J. (2016) Competitive interactions
Preziosi, R.F. & Fairbairn, D.J. (1996) Sexual size dimorphism and change the pattern of species co-occurrences under neutral dispersal.
selection in the wild in the waterstrider Aquarius remigis: body Oikos, 126, 91–100.
size, components of body size and male mating success. Journal of Ulrich, W., Kryszewski, W., Seweniak, P., Puchalka, R., Strona, G.
Evolutionary Biology, 9, 317–336. & Gotelli, N.J. (2017a) A comprehensive framework for the study
R Development Core Team. 2013. R: A language and environment of species co-occurrences, nestedness and turnover. Oikos, 126,
for statistical computing. R Foundation for Statistical Computing, 1607–1616.
Vienna, Austria. [WWW document]. URL http://www.R-project.org/ Ulrich, W., Kubota, Y., Kusumoto, B., Baselga, A., Tuomisto, H. &
[acessed on 12 January 2017]. Gotelli, N.J. (2017b) Species richness correlates of raw and standard-
Resh, V.H. (2008) Which group is best? Attributes of different biological ized co-occurrence metrics. Ecological soundgings, 27, 395–399.
assemblages used in freshwater biomonitoring programs. Environ- Valente-Neto, F., Roque, F.O., Rodrigues, M.L., Juen, L. & Swan,
mental Monitoring and Assessment, 138, 131–138. C.W. (2015) Toward a practical use of Neotropical odonates as
Ripley, B., Venables, B., Bates, DM, Hornik, K., Gebhardt, A., Firth, bioindicators: testing congruence across taxonomic resolution and life
D et al. (2013). Package ’MASS’ CRAN Repos. [WWW document]. stages. Ecological Indicators, 2, 952–959.

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166

166 Alana P. M. Guterres et al.

Valente-Neto, F., Durães, L., Siqueira, T. & Roque, F.O. (2017) Webb, C.O., Ackerly, D.D., Mcpeek, M.A. & Donoghue, M.J. (2002)
Metacommunity detectives: confronting models based on niche and Phylogenies and community ecology. Annual Review of Ecology and
stochastic assembly scenarios with empirical data from a tropical Systematics, 33, 475–505.
stream network. Freshwater Biology., 63, 86–99. https://doi.org/10 Weiher, E. & Keddy, P.A. (1995) Assembly rules, null models, and trait
.1111/fwb.13050. dispersion: new questions from old patterns. Oikos, 74, 159–164.
Veech, J.A. (2013) A probabilistic model for analysing species Weiher, E., Clarke, D.P. & Keddy, P.A. (1998) Community assembly
co-occurrence: probabilistic model. Global Ecology and Biogeogra- rules, morphological dispersion, and the coexistence of plant species.
phy, 22, 252–260. Oikos, 81, 309–322.
Veech, J.A. (2014) The pairwise approach to analysing species Wilcox, R.S. (1995) Ripple communication in aquatic and semiaquatic
co-occurrence. Journal of Biogeography, 41, 1029–1035. insects. Ecoscience, 2, 109–115.
Velázquez, E., Paine, C.T., May, F. & Wiegand, T. (2015) Linking trait Zar, J. H. (ed.) (2010) Biostatistical Analysis. Prentice Hall, Englewood
similarity to interspecific spatial associations in a moist tropical forest. Cliffs, NJ.
Ecological Society of America, 26, 1068–1079. Zwick, W.R. & Velicer, W.F. (1986) Comparison of five rules for
Vepsäläinen, K. & Järvinen, O. (1974) Habitat utilization of Ger- determining the number of components to retain. Psychological
ris argentatus (Het. Gerridae). Scandnavian Entomology., 5, Bulletin, 99, 432–442.
3–4.
Vilardi, G.C. (2015) Heteroptera aquáticos e semiaquáticos (Ger- Accepted 12 July 2019
romorpha e Nepomorpha) do sudoeste da Amazônia. UFSC, São First published online 30 July 2019
Carlos- SP. Associate Editor: Christopher Hassall

© 2019 The Royal Entomological Society, Ecological Entomology, 45, 155–166