Environmental Science and Pollution Research

https://doi.org/10.1007/s11356-020-11137-9

REVIEW ARTICLE

The use of Odonata species for environmental assessment:

a meta-analysis for the Neotropical region
María Gómez-Tolosa 1 & Gustavo Rivera-Velázquez 2 & Tamara M. Rioja-Paradela 3 & Luis F. Mendoza-Cuenca 4 &
César Tejeda-Cruz 2 & Sergio López 3

Received: 8 May 2020 / Accepted: 4 October 2020

# Springer-Verlag GmbH Germany, part of Springer Nature 2020

Abstract
The order Odonata has been regularly used as an indicator of the ecosystem’s condition. The objective of this review was to analyze the
importance of Odonata for environmental assessments (assessment types, statistical approach, life stages, and sampling method, or
particular metric), summarizing the current state, the trends, and identifying related research issues in the Neotropical region. Therefore,
we selected 62 articles from 2007 to 2018 based on published research to monitor Odonata assessments in the Neotropical region. We
compiled a database and ran statistical analyses for the observed frequencies. We found that ecosystem health was the most frequent
assessment type and quality the most used objective. In the case of statistical tests and metrics, multivariate analyses and species
richness were most used in these papers. However, because there is a great diversity of habitats in this region, there is no unique
monitoring protocol to assess the quality of ecosystem health and it is needed to create a proposal for a standard evaluation protocol.
Consequently, guidelines for monitoring are presented, and we suggest three stages to establish a specific protocol for each site, which
records the set of species most sensitive to the exchange rate evaluated, as well as the use of rarefaction methods, the index of diversity
based on the area under the curve, and multivariate analysis, among other recommendations.

Keywords Biomonitoring . Biotic index . Damselflies . Diversity . Dragonflies . Habitat index . Indicators

Introduction
Responsible Editor: Thomas Hein
Electronic supplementary material The online version of this article Up to one million plant and animal species face extinction
(https://doi.org/10.1007/s11356-020-11137-9) contains supplementary because of human activities according to the
material, which is available to authorized users.
Intergovernmental Science-Policy Platform on Biodiversity
and Ecosystem Services (IPBES) (Tollefson 2019). In addi-
* María Gómez-Tolosa
[email protected] tion, an increase in loss or degradation of aquatic and terres-
trial ecosystems is a reality in our current world (Corlett
* Sergio López
[email protected] 2016). For these reasons, accurate assessment of habitats
should be a priority for the management and conservation of
1
Programa de Doctorado en Ciencias en Biodiversidad y public policy. There are many approaches to assess the level
Conservación de Ecosistemas Tropicales, Instituto de Ciencias of degradation, and certain species have been selected as in-
Biológicas, Universidad de Ciencias y Artes de Chiapas, Libramiento dicator groups. Therefore, in the aquatic ecosystem, inverte-
Norte-Poniente 1150, 29018 Tuxtla Gutiérrez, Chiapas, México
2
brates have been widely used to monitor aquatic habitats; for
Instituto de Ciencias Biológicas, Universidad de Ciencias y Artes de example, the Biological Monitoring Working Party (BMWP)
Chiapas, Libramiento Norte-Poniente 1150, 29018 Tuxtla
Gutiérrez, Chiapas, México is one of the most commonly used biotic indices (Hawkes
3 1997). It is based on the tolerance of organisms to pollution
Cuerpo Académico Sustentabilidad y Ecología Aplicada,
Universidad de Ciencias y Artes de Chiapas, Libramiento classified on a scale of 1 to 10. In addition to the indices,
Norte-Poniente 1150, 29018 Tuxtla Gutiérrez, Chiapas, México Poulton et al. (2003) proposed the use of water quality param-
4
Facultad de Biología, Universidad Michoacana de San Nicolás de eters and their correlations with the species richness of
Hidalgo, Avenida Francisco J. Múgica S/N, Ephemeroptera, Plecoptera, Trichoptera, and Odonata
58030 Morelia, Michoacán, México (EPTO).
 Environ Sci Pollut Res

Odonata has also been used as a biological indicator of Methods

microhabitat quality (Corbet 1993) or ecological restoration
success in Mexico (Gómez-Anaya and Novelo-Gutiérrez Data collection
2015) and Indonesia (Cleary et al. 2004). Odonata also has
been used to evaluate different climate change scenarios in We searched the ISI Web of Knowledge (Web of Science
such contrasting regions as the Andean region (Montemayor Core Collection), Google Scholar, and open-source journals
et al. 2016), eastern Australia (Bush et al. 2012), and Europe (Directory of Open Access Journals) for papers on Odonata in
(Termaat et al. 2019) and to test ecological hypotheses and the Neotropical region. Given that the review by Oertli (2008)
theories, like the intermediate disturbance (Stewart and spans studies published until 2007, we only screened for stud-
Samways 2008), ecological niche (Mendes et al. 2015), and ies that were published as of 2007. We used the combinations
island biogeography (Oertli et al. 2002). Additionally, of the following keywords: Odonata OR dragonflies OR dam-
Odonata has been used to model land use intensification selflies OR Anisoptera OR Zygoptera AND indicator AND
(Rocha-Ortega et al. 2019), species dispersal patterns habitat index AND biotic index AND Neotropical countries.
(Angélibert and Giani 2003), assess connectivity along corri- We found 76 publications from seven different countries that
dors (Khazan 2014), or for the use of different habitat types met the search criteria.
(Pither and Taylor 1998). In other eco-regions, studies were
conducted to test indices for monitoring and evaluating fresh- Criteria for data selection
water with odonates, such as the dragonfly biotic index (DBI)
proposed by Simaika and Samways (2009) for South Africa, Specific exclusion criteria were used for publication selection
and the habitat association index (Chovanec et al. 2015) for in this review. We selected published papers that presented the
Eastern Europe. following categories of assessment types, as stated by Oertli
The monetary cost-benefit (the information quality) consti- (2008): Biodiversity, Ecosystem Health, Management
tutes one of the greatest advantages of using Odonata for hab- Practices, and Climatic Change from Neotropical countries
itat assessments, and particularly, adult life stage can be mea- in the period between 2007 and 2018; we also considered
sured at a low cost (Oertli 2008). However, in the Neotropical the Odonata life stages (larvae, adults, adults/larvae) and other
region, anthropogenic disturbances have caused drastic taxa. We selected 62 publications that met these criteria.
changes to their aquatic biota (Nessimian et al. 2008). Then, We created a database (Appendix 1) with the selected pub-
it is important to study Odonata species because they are dis- lications. This database has the following fields: ID (a reference
tributed in a great variety of habitats (Paulson 2004, 2006) due number for each paper), reference (first author’s last name),
to the complex orography (Peel et al. 2007) and in terms of year (2007–2018), habitat (lentic, lentic–lotic, lotic), assess-
climate complexity in the region. Therefore, this region is ment type (biodiversity, ecosystem health, management prac-
considered one of the most diverse and complex in the world tices, and climatic change), geographical location (the country
(Macías-Ordoñez et al. 2014). where the study took place), larvae (presence = 1), adult (pres-
In this review, we analyzed 62 published studies about the ence = 1), other taxa (presence = 1), life stage/other taxa (larvae,
use of Odonata for environmental assessment. We compiled adult, adult/larvae and other taxa), assessment objective (agri-
all published Odonata monitoring studies from 2007 to 2018 culture, conservation value and genetic diversity, forest conver-
including the seven Neotropical countries that have pub- sion, fragmentation, hydromorphology, magnitude-composi-
lished studies of this issue (Brazil, Colombia, Costa Rica, tion, plantation, precipitation, quality, restoration, structure
Ecuador, Mexico, Panama, and Peru). The primary objective and function and urbanization), and sampling method (cap-
of this review was to analyze the importance of Odonata ture-mark-recapture, net/transect, metal sieves, no data, fixed
assessment, summarizing the current state of ecosystem as- area scanning, tissue sampling, and transect). Also, we grouped
sessment, and identifying related research issues in the the publications in four main study subjects: richness (it in-
Neotropical region. We also intend to identify research cludes rarefaction curves and richness estimation at family,
topics, using Odonata, related to the next questions: (1) is genera and species level), diversity (that includes alpha, beta,
there any trend on assessment type, the year or life stages of and gamma indices), conservation value (we considered only
the analyzed period and in which countries? (2) What is the one genetic diversity metric: intraspecific genetic divergence),
assessment type objective, statistical test, sampling method, and particular metrics (including functional diversity, allome-
and metric more frequently applied? (3) Is there any relation try, population traits, landscape analysis, microhabitat quality,
between objectives, assessment type, sampling method, life and pollution levels). Finally, we counted the application of
stages, or particular metric? Also, for this work, we consid- habitat and biotic indices in two separate columns.
ered different assessment studies and we discuss if it is pos- For the analysis of statistical tests used (Appendix 2), we
sible to establish monitoring guidelines using Odonata for classified them in multivariate, parametric, non-parametric,
the Neotropical region. modeling, and simulation analysis. Multivariate analyses
Environ Sci Pollut Res

include correspondence analysis (CA), principal components Neotropical region. The year 2015 recorded the highest num-
(PC), clustering (C, except similarity indexes), multiple linear ber of publications about ecosystem health assessment, and
regression analysis (MLRA), multivariate analysis of variance 12.9% (n = 8) of the publications were concentrated in this
(MAV), and deterrent analysis (DA). Parametric tests are two period. However, 2018 was the year with more publications
samples comparison (TSCP, Student’s t test mainly), multiple for all assessment types (17.7%, n = 11; Fig. 1a; X2 = 22.4, df
samples comparison (MSCP, analysis of variance), and simple = 11, p = 0.021). Papers that used only adult stage organisms
linear regression (LR). Non-parametric tests include two sam- account for 41.9 % of the total (n = 26, X2 = 26.5, df = 4, p <
ples comparison (TSCNP) like Mann–Whitney test, multiple 0.001); meanwhile, larvae stage studies comprehend the
sample comparison (MSCNP) as Kolmogorov–Smirnoff and 22.5% (n = 14), larvae and other taxa the 19.3% (n = 12),
similar tests, non-parametric correlation (NPC), and sampling larvae and adult 14.5% (n = 9), and only one study included
distribution comparison (SDC) likelihood tests. Also, we con- the adult stage and other taxa (1.6%; Fig. 1b). We found only
sidered a group for analyses like modeling spatial distributions three papers that involved more than one country. Brazil and
or numeric simulations (M-S). Mexico showed at least three different assessment types (bio-
diversity, ecosystem health, and management practices);
Data analyses Costa Rica showed the only climatic change study
(Gutiérrez-Fonseca et al. 2018) (Fig. 1c). Based on our results,
We counted the number of papers per year and the assessment Brazil ranked first on the use of Odonata for environmental
type. Also, we counted the papers according to the life stage assessment in the Neotropical region, between 2007 and 2018.
and if included other taxa. Additionally, we made a map of the This may be explained by the fact that Brazilian research
Neotropical region to show the number of published papers institutions have received numerous incentives and well-
per country between 2007 and 2018. funded graduate and research programs, in addition to the
For the published paper’s total, we tested two null hypoth- recent growth in the number of research organizations, and
eses: (i) that observed frequencies are equally to expect ones very successful international exchange schemes for Brazilian
and (ii) that examined traits are independent. We ran the chi- researchers and students (Miguel et al. 2017a, b). However,
squared test (chisq.test) function in R when expected values recent changes in science policies in Brazil can stop this trend
were greater than 5. When the expected value was equal to or (de Area Leão Pereira et al. 2019).
less than 5, we used the likelihood ratio test (G test) function Because of the high deforestation and land degradation
of the DescTools package in R (R Core Team 2018). rates in the Neotropical region, environmental assessments
We evaluated differences between observed and expected are urgently needed (Laurance et al. 2009; Seymour and
numbers of published papers according to the type of assess- Harris 2019), particularly in countries where environmental
ment (biodiversity, ecosystem health, management practices, assessments have not been done. In contrast, Oertli (2008)
and climatic change) and life stage (adult and larvae). We also reported only six papers from the tropical region, with only
analyzed the differences between the frequency of use for two papers belonging to the Neotropics, we report 62 papers
metrics (richness, diversity, conservation value, and particu- in the current review.
lars) and the statistical test used (modeling and simulation,
multivariate, parametric, and non-parametric). For indepen- Trends in assessment type and objective
dent analyses, we compared the frequency of use for sampling
and life stage versus index type (habitat or biotic), assessment We found that ecosystem health represented 61.2% (n = 38) of
type versus objective, assessment type, life stage, and sam- the total publications and is more frequent than expected by
pling versus metrics. We also analyzed the goodness of fit to chance (X2 = 50.9, df = 3, p < 0.001). This may be because
relate the objectives with the assessment type and environ- most of studies made rapid assessments considering habitat
mental indices with sampling methods. A contingency table changes in (a) structure and function (Fulan and Henry 2007;
was elaborated in order to relate each pair of analyzed assess- Altamiranda et al. 2010; Altamiranda and Ortega 2012; Fulan
ment types, metrics, or sampling. and dos Anjos 2015; Soares et al. 2015) and (b)
hydromorphology (De Marco et al. 2015; Vilela et al. 2016),
and the major proportion was on (c) quality (Corbi et al. 2008;
Results and discussion De Paiva-Silva et al. 2010; Corbi et al. 2011; Dominguez-
Granda et al. 2011; Pinto et al. 2012; Forero-Céspedes et al.
The importance of Odonata assessment in the 2013; Renner et al. 2013; Teles et al. 2013; Brasil et al. 2014,
Neotropical region 2017; Juen et al. 2014; Martínez-Sanz et al. 2014; De
Oliveira-Junior et al. 2015; Dutra and De Marco 2015;
Between 2007 and 2018, we found a total of 62 papers using Gómez-Tolosa et al. 2015; Mendes et al. 2015, 2017, 2018;
odonates and other taxa for environmental assessment in the Monteiro-Júnior et al. 2015, 2016; Damanik et al. 2016;
 Environ Sci Pollut Res

Published papers a) c)

b) Year

Life-stage

Fig. 1 Odonate peer-reviewed publications found by year and category. a Neotropical region. Pie chart size represents the number of published
The number of published papers by assessment types along 11 years papers. Black color corresponds to Biodiversity, dark grey color to
(from 2007 to 2018). b Number of published papers by Odonata life Climatic Change, grey color to Ecosystem Health, and white color to
stages and other taxa by assessment type. c Map of countries in the Management Practices

Valente-Neto et al. 2016; García-García et al. 2016; Gómez- quality (Renner et al. 2013), one for conservation value, and
Anaya et al. 2017; Hanh Tien et al. 2017; Miguel et al. 2017b; one for genetic diversity (Koroiva et al. 2017) related to biodi-
Brito et al. 2018; Silva et al. 2018; Silva-Franca et al. 2018; versity assessment.
Torres-Olvera et al. 2018; Valente-Neto et al. 2018). Management practices assessment has a similar frequency to
Interestingly, or results coincide with those reported by random and both climatic change, and biodiversity had lower
Oertli (2008), highlighting the relevance of the use of frequencies. Unfortunately, within management practices, they
Odonates for the study of ecosystem health. have only been used in one study to assess agriculture (Rizo-
In contrast, studies about Management practices had 25.8% (n Patrón et al. 2013), urbanization (Monteiro-Júnior et al. 2014),
= 16) focused mainly on forest conversion (Fincke and Hedström and ecological restoration (Gómez-Anaya and Novelo-Gutiérrez
2008; Hofhansl and Schneeweihs 2008; Bücker et al. 2010; 2015) projects. This is a research avenue that must be addressed
Carvalho et al. 2013; Calvão et al. 2016; Renner et al. 2016; soon. The use of biodiversity assessments and climatic change
Rodrigues et al. 2016; Rodrigues et al. 2018), and only two has been poorly exploited because there are few studies related to
studies each on habitat fragmentation (Feindt et al. 2013; detecting odonates diversity hotspots, as well as long-term re-
Khazan 2014) and plantation (Carvalho et al. 2018; Dalzochio cords along with a complete hydrological cycle to analyze cli-
et al. 2018). While, agriculture (Rizo-Patrón et al. 2013), urban- matic effects.
ization (Monteiro-Júnior et al. 2014), and restoration (Gómez- When analyzing the objectives by assessment type, the result
Anaya and Novelo-Gutiérrez 2015) only have one study each, of this review showed 13 principal groups (agriculture, conser-
whereas biodiversity and climatic change represented 11.2% (n = vation value and genetic diversity, forest conversion, fragmenta-
7) and 1.6% (n = 1), respectively (Fig. 2a). In the case of biodi- tion, hydromorphology, magnitude-composition, plantation, pre-
versity, the magnitude-composition assessment objective had a cipitation, quality, restoration, structure and function, urbaniza-
higher frequency (Juen et al. 2007; Novelo-Gutiérrez and tion). The exclusive objectives by assessment type are agricul-
Gómez-Anaya 2009; Dalzochio et al. 2011; Pires et al. 2013; ture, forest conversion, plantation, restoration and urbanization
De Souza et al. 2017). We only found one study for habitat for management practices; conservation value and genetic
Environ Sci Pollut Res

Fig. 2 Observed frequency of a) 40

Odonate peer-reviewed publica-
tions found by category. a
Assessment types, b statistical
tests, and c sampling methods. 30
Dotted lines correspond to ex-
pected frequencies for each

Frequency
category
20

10

0
Biodiversity Climatic change Ecosystem health Management practices
Assessment

b) 40

30
Frequency

20

10

0
Modelling and Multivariate Parametric Non Parametric
Simulation
Statistical tests

c) 20

15
Frequency

10

0
C-M-R No data Area scanning Tissue sample Transect

Sampling method
 Environ Sci Pollut Res

diversity and magnitude-composition for biodiversity; structure and incomplete observations, and because Jackknife resam-
and function for ecosystem health; and precipitation for climate pling shows better results through the Nelson–Aalen estimator
change. (Ramírez-Montoya et al. 2016). Nevertheless, confidence in-
Hydromorphology and quality have shared objectives for tervals obtained through bootstrap resampling are less dis-
at least two different assessment types. Quality is the most persed, and quadratic error correction is faster using bootstrap
frequently objective of the reviewed papers and includes both than Jackknife (Ramírez-Montoya et al. 2016). Because there
aquatic and terrestrial ecosystems (Table 1). are different sampling techniques (capture with nets, areas or
transects), we suggest that observed and estimated species
richness be reported against the number of individuals cap-
Statistical approach
tured to each day of sampling. This will allow comparing the
results of the rarefaction, regardless of the type of sampling
Because most studies were focused on ecosystem health assess-
used.
ment, specifically quality assessment, we found that multivariate
analysis is the main statistical test applied than other tests (X2 =
12.4, df = 3, p = 0.006; Fig. 2b). Therefore, multivariate analysis Life stages and sampling methods
was used to study more complex data sets than those that can
handle univariate analysis methods (Forero-Céspedes et al. 2013; Papers that used only adult stage organisms account for
Carvalho et al. 2018). Pearson’s correlation was the most used 41.9% of the total (n = 26, X2 = 26.5, df = 4, p < 0.001);
parametric statistical test. This test could be used to measure the meanwhile, larvae stage studies comprehend the 22.5% (n =
relationship between environmental variables and species traits 14), larvae and other taxa the 19.3% (n = 12), larvae and
(Altamiranda and Ortega 2012; Martínez-Sanz et al. 2014; De adult 14.5% (n = 9), and only one study included the adult
Oliveira-Junior et al. 2015; Monteiro-Júnior et al. 2015; stage and other taxa (1.6%; Fig. 1b). The number of papers
Damanik et al. 2016; Miguel et al. 2017a, b; Rodrigues et al. that used the transects sampling method for the adult life
2018; Silva et al. 2018; Valente-Neto et al. 2018; Brito et al. stage was higher than expected (X2 = 30.6, df = 4, p <
2018). However, Pinto et al. (2012) used Pearson’s correlation 0.001; Fig. 2c).
coefficient to analyze only allometric relationships. Principal There is an ongoing trend to use Odonata adult life stage for
component analysis (PCA) was the most applied multivariate rapid assessment monitoring in terrestrial environments/
analysis (Pinto et al. 2012; Martínez-Sanz et al. 2014; Khazan riparian corridors. Because the study of their taxonomy is well
2014; Vilela et al. 2016). We noted that correlation analyses are known (Miguel et al. 2017a) with the exception of new Argia
related to a few environmental variables, while multivariate anal- species (Hofhansl and Schneeweihs 2008; Monteiro-Júnior
yses are applied to determine species closely related to specific et al. 2014; Brasil et al. 2014; Gómez-Tolosa et al. 2015;
environmental conditions at a local scale. Monteiro-Júnior et al. 2015; García-García et al. 2016;
Jackknife was the most used resampling rarefaction tech- Calvão et al. 2016; Miguel et al. 2017b), logistically easier,
nique, maybe because it can be employed with small samples less expensive, and were, therefore, used to identify local

Table 1 Number of published papers by assessment type and objective

Assessment

Objectives Biodiversity Climatic change Ecosystem health Management practices

Agriculture 0 0 0 1
Conservation value and genetic diversity 1 0 0 0
Forest conversion 0 0 0 8
Fragmentation 0 0 0 2
Hidromorphology 0 0 2 1
Magnitude-composition 5 0 0 0
Plantation 0 0 0 2
Precipitation 0 1 0 0
Quality 1 0 31 0
Restoration 0 0 0 1
Structure and function 0 0 5 0
Urbanization 0 0 0 1
Total 7 1 38 16
Environ Sci Pollut Res

hotspots of diversity (González Soriano et al. 2011) or, on the et al. 2010; Gómez-Anaya and Novelo-Gutiérrez 2015;
opposite, to identify degraded sites (Oertli 2008; Simaika and Gómez-Anaya et al. 2017), and gain and loss of species in a
Samways 2011; Šigutová et al. 2019). Several studies con- community’s gradient. However, traditional diversity indices
cluded that both larvae and adult stages are affected by envi- focus on species richness or equity and cannot summarize
ronmental changes (Kutcher and Bried 2014; Dutra and De species diversity, so a composite statistical measure is needed
Marco 2015; Monteiro-Júnior et al. 2015; Valente-Neto et al. (Guisande et al. 2017). In this context, Di Battista et al. (2017)
2016), and there is also a relation between life stages. For proposed three functional measures (the first and second de-
example, the species Calopteryx splendens prefers optimal rivative, the radius of curvature and the length of the curva-
habitats over suboptimal to reduce egg mortality and aid egg ture), providing a scalar measure that leads to a classification
development (Siva-Jothy et al. 1995). The above suggests that of the community unit, denominated beta diversity profiles
life stages are interconnected, and the conditions experienced (Area). These functional measures allow communities to order
by one stage of the life cycle may affect the later stages, de- both by species richness and equity; therefore, the result of the
pending on the selection pressures on a given characteristic analysis is not biased by dominant species (Simpson), rare
through metamorphosis (Stoks and Córdoba-Aguilar 2012). species (Shannon–Wiener), gain and loss of species in a gra-
However, EPT and odonates larval are suitable for the assess- dient of communities (Cody’s, it is not independent of species
ment of the water quality of aquatic environments (Corbi et al. richness), or by dominant and rare species (Renyi entropy).
2011; Martini et al. 2013). Currently, there is a discussion We suggest applying the area under beta diversity profiles as a
about the use of Odonata larval versus adult life stage and diversity estimation that allows us to compare results with
versus other Taxa as Ephemeroptera, Plecoptera, and different sampling methods.
Trichoptera or other aquatic macroinvertebrates. As Simaika
and Samways (2011) have stated, it is more efficient (in terms Related research issues
of cost-benefit-accuracy) to use adult Odonata over other mac-
roinvertebrate taxa. According to the G test of independence; we did not detect
Transect sampling method was more frequently used for any link between assessment type and objectives (G test = 4.7,
adult life stage, whereas D-net trap was the most used sam- df = 9, p = 0.854; Table 1). Then, there is not a clear associ-
pling method for studies with larvae (G = 16.8, df = 6, p = ation between particular objectives and assessment types. In
0.009; Fig. 3b). Linear transects were used because this meth- the case of metric category, we also did not find any associa-
od can record the changes that take place throughout the study tion with assessment type (G = 8.4041, df = 9, p = 0.494;
site and that are related to a particular gradient along which the Table 2), nor life stage (G = 4.9418, df = 6, p = 0.551;
communities of riparian plants and adult odonates change Table 3) or sampling methods (G = 17.16, df = 21, p =
(Hofhansl and Schneeweihs 2008; Khazan 2014; Rodrigues 0.701; Table 4).
et al. 2018; Valente-Neto et al. 2018). Also, for larvae sam- Our results showed that biotic indices were more frequently
pling, Ghani et al. (2016) showed that the square net was the used than other habitat indices (G = 12.23, df = 2, p = 0.002)
most efficient of the three sampling methods used for when the sampling method is net and transect, whereas habitat
collecting a greater diversity of macroinvertebrate taxa and a indices are more frequently used when fixed areas or transects
greater number of specimens. are employed (Fig. 4a). In the case of life stage, the statistical
test showed that adult and other taxa are related to habitat
Used metrics indices (G = 14.11, df = 2, p < 0.001), whereas larvae and
other taxa were related to biotic indices (Fig. 4b).
The most frequently used metric category was richness, Environmental indices used were related to both sampling
followed by particular metrics, which include a variety of methods (transects and nets) and life stages (larvae and
traits, ranging from population traits (such as allometric indi- adults). For example, biotic indices are more frequently
ces, mortality rates, oviposition patterns, and sex ratio) to employed when the larval stage is captured with an entomo-
functional diversity (X2 = 43.12, df = 3, p < 0.001; Fig. 3a). logical net, because this is a more specific way to obtain the
Therefore, the results of the reviewed documents indicate species of interest for the purpose of the study. In contrast,
that the most prevalent metric is species richness, but this when habitat indices are used, the area scanning and transects
metric does not take into account the relative abundances for are the methods more frequently applied, because they focus
each species, as the species diversity indices do. The most- on environmental elements and not only on adult odonates.
reported diversity indices were Shannon–Wiener, followed by Unfortunately, only two papers used both types of indices:
Renyi, Simpson, and Cody because they are accurate to iden- biotic and habitat (Silva-Franca et al. 2018; Torres-Olvera
tify rare species (Corbi et al. 2011; Forero-Céspedes et al. et al. 2018); then, it is not possible to establish a comparison
2013; Teles et al. 2013; Martínez-Sanz et al. 2014; Bücker between them. We expect that in the near future, more re-
et al. 2010; Renner et al. 2013), dominant species (Bücker search would be done considering both types of indices
 Environ Sci Pollut Res

Fig. 3 Observed frequency of a) 50

Odonate peer-reviewed publica-
tions found by category. a Metrics
and b sampling devices. Dotted
40
lines correspond to expected fre-
quencies for each category

Frequency
30

20

10

0
Richness Diversity Particular Metrics Conservation Value

Metric
b) 15
Frequency

10

0
Conical Net D-net Kick-net Metal sieves No data Pond-net (FBA Surber-net
estándar)

Sampling device

because the advantages of one type could overcome the fail- Additionally, the most frequently biotic index was the
ures of the other, and also, one index complements the infor- biological monitoring working party (BMWP, six times),
mation obtained from the other. which used the larval stage in concurrence with other

Table 2 Number of published papers by metrics and assessment type

Metrics

Assessment type Richness Diversity Specific metrics Conservation value

Biodiversity 7 4 3 1
Climatic change 1 1 1 0
Ecosystem health 24 14 26 0
Management practices 13 3 12 1
Total 45 22 42 2
Environ Sci Pollut Res

Table 3 Number of published papers by metrics and life stage

Metric

Life stage Richness Diversity Specific metrics Conservation value

Adult/other taxa 23 8 19 1
Larvae/other taxa 17 12 19 0
Larvae/adult 5 2 5 1
Total 45 22 42 2

taxa. This biotic index allows a fast and accurate evalua- We found that habitat indices use the adult life stage, for
tion of lotic water quality, due to its flexibility. Also, the example, when assessing riparian conditions, such as HPI (De
BMWP index can be easily modified to meet local condi- Paiva-Silva et al. 2010), HII (Nessimian et al. 2008) which is
tions, as shown in Costa Rica (Rizo-Patrón et al. 2013), frequently applied (Pinto et al. 2012; Carvalho et al. 2013;
Ecuador (Dominguez-Granda et al. 2011; Damanik et al. Monteiro-Júnior et al. 2014; Brasil et al. 2014; Juen et al.
2016), Brazil (Corbi et al. 2011), and Colombia 2014; De Oliveira-Junior et al. 2015; Dutra and De Marco
(Altamiranda et al. 2010; Forero-Céspedes et al. 2013). 2015; Miguel et al. 2017a, b; Brasil et al. 2017), and the
The EPT index can be combined with Odonata and could VBHA (Barbour et al. 1999) as described by Gómez-Tolosa
be expressed as a percentage of sensitive orders (E = et al. (2015). Through a meta-analysis, the HII showed to be a
Ephemeroptera, P = Plecoptera, T = Trichoptera) with measure of the environmental integrity of the Amazon
respect to the total taxa found. This index is commonly rainforest streams, like in Brazilian and other countries
used as an indicator of water quality and has a direct (Brasil et al. 2020). This index has predicted the Odonata
application in rapid resource assessments (Forero- composition, for example, species composition decreases as
Céspedes et al. 2013; Teles et al. 2013; Martínez-Sanz riparian vegetation and native vegetation decline, or the integ-
et al. 2014; Valente-Neto et al. 2018; Brito et al. 2018). rity index presents a positive relationship with the abundance
We found that TITAN was the only biotic index that uses of potentially useful indicators (Dalzochio et al. 2011; Pinto
adult life stage and habitat conditions (Rodrigues et al. et al. 2012; Carvalho et al. 2013; Brasil et al. 2014; Renner
2016). This index has been applied in central-west et al. 2016; Monteiro-Júnior et al. 2015; Rodrigues et al. 2016)
Brazil to determine which species are affected by the loss (Table 5).
of native vegetation. The species affected were Argia There are several indices that have been modified for the
smithiana (Calvert, 1909), A. modesta (Selys, 1865), Neotropical region, such as the BMWP, EPT, and VBHA,
A. subapicalis (Calvert, 1909), Enallagma among others. Nevertheless, Neotropical ecosystems are
novaehispaniae (Calvert, 1902), Hetaerina mortua much more complex than temperate ecosystems, due to the
(Hagen in Selys, 1853), and Epipleoneura venezuelensis diversity of habitats and species richness (Macías-Ordoñez
(Rácenis, 1955), which potentially can be used as et al. 2014). For example, Gonçalves and de Menezes
bioindicators for this particular area. In this way, species (2011) found that functional feeding group analysis was more
conservation priorities can be drawn based on these spe- accurate, to evaluate river disturbance, than other biotic indi-
cies (Samways 2008). ces, in do Pinto River, Brazil. Moreover, Simaika and

Table 4 Number of published papers by metrics and sampling technique

Metric

Sampling Richness Diversity Specific metrics Conservation value

C-M-R 1 0 2 0
Net/transect 12 10 12 0
Metal sieves 7 3 6 0
No data 4 2 4 1
Scanning fixed area 6 2 3 0
Tissue sample 1 0 1 1
Transect 14 5 14 0
Total 45 22 42 2
 Environ Sci Pollut Res

Fig. 4 Observed frequency of a) 15

Odonate peer-reviewed publica-
tions found by category. a
Sampling methods, b life stage
12
and other taxa. Dark grey bars

Frequency
correspond to habitat indices,
whereas grey bars correspond to
biotic indices 9

0
Net/transect Area scanning Transect

Sampling method

b) 15

12
Frequency

0
Adult/other Taxa Larvae/Adult Larvae/other Taxa

Life-stage/Taxa

Samways (2011) compared the DBI with the average taxo- Recommendations for future environmental
nomic distinctness index (AvTD), as well as with a standard assessments
freshwater benthic macroinvertebrate-based freshwater health
index (South African Scoring System, using average score per The niche evolution theory expects habitat specialization and
taxon (ASPT)). They concluded that DBI has the potential to species vulnerability to habitat disturbance (Marvier et al.
replace other benthic macroinvertebrate-based freshwater 2004), based on the assumption of niche breadth (Devictor and
quality assessments. However, a DBI with native indicator Robert 2009); then species with narrow niches would be more
species has not yet been developed for the Neotropical region. affected than broad niche species. For this reason, habitat degra-
Habitat indices were frequently applied for local conditions in dation should affect specialists negatively (Devictor and Robert
the Neotropical region, commonly using adult stage individ- 2009; Šigutová et al. 2019).
uals (for example, HPI, HII). These indices are used due to Because Neotropical ecosystems present a high complex-
their low investment cost and accurate estimations for habitat ity of environments and biotic relationships, it is not possible
assessment. to set up a single set of species that function as an indicator
Environ Sci Pollut Res

Table 5 Comparison of habitat indices. Cells in grey color correspond to characteristics that the index does not have; the meaning of the italic letters
remarks exclusive characteristics that only are presented on one index

VHBAS HPI HII

Land use pattern beyond the riparian zone Land use pattern beyond the riparian zone
(Cecropia and Vismia spp.)
Riparian vegetative zone width Width of riparian forest Width of riparian forest
Completeness of riparian forest Completeness of riparian forest
Bank vegetative protection Vegetation of riparian zone within 10 m of channel Vegetation of riparian zone within 10 m of channel
Epifaunal substrate/available cover Retention devices Retention devices
Channel flow status Channel sediments Channel sediments
Channel Alteration Bank structure Bank structure
Bank stability Bank undercutting Bank undercutting
Embeddedness/pool substrate Stream bottom Stream bottom
Frequency of riffles/channel sinuosity Riffles and pools, or meanders Riffles and pools, or meanders
Aquatic vegetation Aquatic vegetation
Sediment deposition Detritus Detritus
Velocity/depth regimes pool variability

for different contexts. In this sense, and based on our review, A second step is to identify key species, possibly specialist,
we propose a set of steps to be followed when doing envi- which responds to alterations on the focused environmental
ronmental quality assessments, according with Jackson et al. quality selected. Based on the previous results, a set of species
(2000), Dale and Beyeler (2001), and Niemeijer and de that are more sensitive to environmental changes will be iden-
Groot (2008): tified, with the information inferred from the multivariate
The first step is to define the research question (i.e., what analyses. Then, this selected set of species could be proposed
environmental quality aim will be evaluated?). Once the objec- for local suitable monitoring protocols. At the same time, it is
tive has been set up, it is necessary to determine (i) the kind of necessary to evaluate methods, costs, logistics, and other is-
available information, (ii) the scale at which to work, (iii) where sues before implementation. But sometimes, this analysis
in the driving force–pressure–state–impact–response (DPSIR) leaves two strategies, or additional indicators will be
framework the focus lies, and (iv) whether an environment or demanded that would provide grounds to turn out or correct
anthropogenic perspective is required. To our knowledge, the other factors causing changes in the indicator, or it can go back
Odonata assemblages are relevant for the assessment of environ- to identify only those species that have the same original cause
mental quality and ecological function in Neotropical ecosys- of change. Then, it is possible to look at the employment of
tems. Then, it is possible to apply a rapid assessment method other life cycle stages, i.e., larvae, in combination with specif-
using Odonata adult life stage sampling, and a few specific en- ic physical and chemical parameters. For example, the use of a
vironmental variables (depending on the objective) that can be set of specialized species on a particular environmental feature
correlated. In the case of sampling multiple environmental vari- (as canopy cover and water quality), using both larvae and
ables, in combination with physical and chemical parameters, it adults, could improve the assessment accuracy. Another strat-
is recommended to relate them with species composition using egy is to look out for relationships between the environmental
multivariate methods of ordination (for example, CCA or another traits and species assemblages at different taxonomic levels, as
sort of multivariate method). Also, we suggest the use of rarefac- suborders, families, or genera. To explore these relations, we
tion methods (Bootstrap and Chao, preferably) to analyze the suggest the use of general linear models (GLMs) in combina-
relationship of accumulated species richness with the number tion with the best-fitted data distribution.
of individuals captured per sample, as an estimate of sampling And thirdly, once that key species have been identified
representativeness in the study area. Likewise, we recommend associated to general indicators, the last step is to identify
analyzing the relationship between abundance, species richness, the best concrete indicators to use for long-term monitoring
and diversity. Alpha diversity should be estimated using diversity protocols. At this point, it is crucial to analyze possible
profiles (the area under the curve) as an indicator of diversity sources of errors (human, methodological, or instrumental)
since this method allows for comparisons between samples ob- in order to include on the estimations with the proper confi-
tained using different methods. Alpha diversity should then be dence interval (data quality objectives and constrains). Sample
related to rapid assessment scores or other environmental size, monitoring duration and temporality, and other variables
variables. could affect the precision and confidence levels of reported
 Environ Sci Pollut Res

results. Then, we require to continually checking the results reviewers who improved the final manuscript. Also, we thank Francisca
Méndez-Harclerode, Allen Wedel, Jonathan Gering, and Robert
for optimize the monitoring program goals and aims. A con-
Milliman of Bethel College, Kansas, for the facilities granted for the
tinued discussion on the selected indicator may improve the completion of this paper. This paper is dedicated to the memory of
efficiency tracking on environmental alterations and can pro- Camerina Mendoza.
mote alternatives to monitoring programs to switch from spe-
cies assemblages to individual traits, as wing symmetry or Author contributions María Gómez-Tolosa: conceptualization,
method, software, formal analysis, investigation, and writing -
genetic variation. This involves frequent feedback actions to original draft. Gustavo Rivera-Velázquez: writing - review.
maintain useful the indicator and providing support for users, Tamara M. Rioja-Paradela: writing - review. Luis F. Mendoza-
resource managers, and decision-makers. Cuenca: conceptualization, method, investigation, writing-review
Finally, we expect that this feedback information will be & editing, and visualization. César Tejeda-Cruz: writing - review
& editing and visualization. Sergio López: conceptualization,
available in a database of sensitive and tolerant odonates spe-
method, software, formal analysis, resources, and writing - review
cies, in order to generate biotic and habitat indices for & editing.
Neotropical region, such as the DBI (Simaika and
Samways 2009) for South Africa, and the habitat association Funding María de Lourdes Gómez Tolosa is a doctoral student
index (Chovanec et al. 2015) for Eastern Europe. from Ciencias en Biodiversidad y Conservación de Ecosistemas
Tropicales,Universidad de Ciencias y Artes de Chiapas
(UNICACH) and received a fellowship from CONACYT
(Scholar No. 630424; CVU: 208813). Sergio López was granted
a fellowship to an academic stay to write this paper by the
Conclusions Mexican Program to Professorship Improving of the Public
Education Ministry (PRODEP-SEP).
This review presents evidence that Odonata species may
be used to evaluate and monitor not only freshwater Data availability Available in the supplementary material appendices 1
bodies but also environmental changes like deforesta- and 2.
tion, urbanization, land use, climate changes, and man-
agement practices in the Neotropical region. Odonata Compliance with ethical standards
group has value for environmental assessment purposes
Competing interests The authors declare that they have no competing
since they are sensitive to environmental changes. interests.
Ecosystem health was the most frequent type of en-
vironmental assessment, and quality was the most used
research objective for the period analyzed. We found a
relationship between biotic and habitat indices with References
sampling methods. Scanning areas and transects are
more frequently used, than expected by chance, to de- Altamiranda M, Ortega O (2012) Estructura poblacional de Polythore
termine habitat indices. While biotic indices are more gigantea (Odonata: Polythoridae) en sistemas lóticos con diferentes
estados de conservación en Antioquia, Colombia. Rev Biol Trop 60:
associated with net/transect sampling. 1205–1216. https://doi.org/10.15517/rbt.v60i3.1772
Adult life stage sampling was preferred over larvae Altamiranda M, Pérez L, Gutiérez L (2010). Composición y preferencia
and exuviae because of the lower cost of fieldwork and de microhábitat de larvas de odonata (insecta), en la ciénaga san juan
time employed, but to include larval stages may result de tocagua (Atlántico, Colombia) Caldasia, 32: 2357-3759 (https://
in a more detailed evaluation. However, we strongly revistas.unal.edu.co/index.php/cal/article/view/36231/37765, ISSN
electrónico 2357-3759)
recommend employing the three life stages to obtain a Angélibert S, Giani N (2003) Dispersal characteristics of three odonates
more accurate assessment. species in a patchy habitat. Ecography 26:13–20. https://doi.org/10.
Due to the great diversity of habitats and complexity 1034/j.1600-0587.2003.03372.x
in the Neotropical region, there is no single habitat as- Barbour MT, Faulkner C, Gerritsen J (1999) Rapid bioassessment proto-
cols for use in streams and wadeable rivers: periphyton, benthic
sessment pattern and it is complicated to set up a
macroinvertebrates and fish Second Edition. EPA 841-B-99-002
unique protocol; thus, we suggest three steps that should US Office of Water, Washington, D.C http://www.epa.gov/
focus the environmental quality assessments. The long- OWOW/monitoring/techmon
term goal is to establish an index (Odonata Neotropical Brasil LS, Batista JD, Giehl NF d S et al (2014) Environmental integrity
index) that accounts for the value of dragonflies and and damselfly species composition in Amazonian streams at the “arc
of deforestation” region, Mato Grosso, Brazil. Acta Limnol Bras 26:
damselflies to assess environmental integrity and 278–287. https://doi.org/10.1590/s2179-975x2014000300007
degradation. Brasil LS, Vieira TB, de Oliveira-Junior JMB, Dias-Silva K, Juen L
(2017) Elements of metacommunity structure in Amazonian
Acknowledgments We thank Eduardo Cuevas García who improves an Zygoptera among streams under different spatial scales and environ-
early version of this manuscript. We thank Andre Swartley and Katie mental conditions. Ecol Evol 7:3190–3200. https://doi.org/10.1002/
Branam-Swartley for linguistic corrections, and two anonymous ece3.2849
Environ Sci Pollut Res

Brasil LS, de Lima EL, Spigoloni ZA, Ribeiro-Brasil DRG, Juen L Damanik A, Lock M, Boets K et al (2016) Ecological water quality
(2020) The habitat integrity index and aquatic insect communities analysis of the Guayas river basin (Ecuador) based on macroinver-
in tropical streams: a meta-analysis. Ecol Indic 116:0–2. https://doi. tebrates indices. Limnologica 57:27–59
org/10.1016/j.ecolind.2020.106495, 106495 de Area Leão Pereira EJ, Silveira Ferreira PJ, de Santana Ribeiro LC et al
Brito JG, Martins RT, Oliveira VC, Hamada N, Nessimian JL, Hughes (2019) Policy in Brazil (2016–2019) threaten conservation of the
RM, Ferraz SFB, de Paula FR (2018) Biological indicators of diver- Amazon rainforest. Environ Sci Policy 100:8–12. https://doi.org/
sity in tropical streams: congruence in the similarity of invertebrate 10.1016/j.envsci.2019.06.001
assemblages. Ecol Indic 85:85–92. https://doi.org/10.1016/j. De Marco P, Batista JD, Cabette HSR (2015) Community assembly of
ecolind.2017.09.001 adult odonates in tropical streams: an ecophysiological hypothesis.
Bücker A, Sondermann M, Frede HG, Breuer L (2010) The influence of PLoS One 10:e0123023. https://doi.org/10.1371/journal.pone.
land-use on macroinvertebrate communities in montane tropical 0123023
streams—a case study from Ecuador. Fundam Appl Limnol 177: De Oliveira-Junior JMB, Shimano Y, Gardner TA et al (2015)
267–282. https://doi.org/10.1127/1863-9135/2010/0177-0267 Neotropical dragonflies (Insecta: Odonata) as indicators of ecologi-
Bush A, Nipperess D, Turak E, Hughes L (2012) Determining vulnera- cal condition of small streams in the eastern Amazon. Austral Ecol
bility of stream communities to climate change at the landscape 40:733–744. https://doi.org/10.1111/aec.12242
scale. Freshw Biol 57:1689–1701. https://doi.org/10.1111/j.1365- De Souza MM, Pires EP, Brunismann ÂG et al (2017) Dragonflies and
2427.2012.02835.x damselflies (Odonata) from the wetland of the Rio Pandeiros, north-
Calvão LB, Nogueira DS, de Assis Montag LF, Lopes MA, Juen L ern region of Minas Gerais State, Brazil, with a description of the
(2016) Are Odonata communities impacted by conventional or re- male of Archaeogomphus vanbrinki Machado (Anisoptera:
duced impact logging? For Ecol Manage 382:143–150. https://doi. Gomphidae). Int J Odonatol 20:13–26. https://doi.org/10.1080/
org/10.1016/j.foreco.2016.10.013 13887890.2017.1281848
Calvert PP (1892-1908) Odonata. In F.D. Godwin. (ed.). Biologia Devictor V, Robert A (2009) Measuring community responses to large-
Centrali-Americana, p 17-420 scale disturbance in conservation biogeography. Divers Distrib 15:
Calvert PP (1909) Contributions to a knowledge of the Odonata of the 122–130. https://doi.org/10.1111/j.1472-4642.2008.00510.x
Neotropical region, exclusive of Mexico and Central America. Ann Di Battista T, Fortuna F, Maturo F (2017) BioFTF: an R package for
of the Carn Mus 6:73–280. https://www.biodiversitylibrary.org/ biodiversity assessment with the functional data analysis ap-
page/39791419#page/97/mode/thumb proach. Ecol Indic 73:726–732. https://doi.org/10.1016/j.ecolind.
Carvalho FG, Pinto N, Júnior et al (2013) Effects of marginal vegetation 2016.10.032
removal on Odonata communities. Acta Limnol Bras 25:10–18.
Dominguez-Granda L, Lock K, Goethals PLM (2011) Application of
https://doi.org/10.1590/S2179-975X2013005000013
classification trees to determine biological and chemical indicators
Carvalho FG, de Oliveira RF, Barbosa L et al (2018) Oil palm plantation
for river assessment: case study in the Chaguana watershed
is not a suitable environment for most forest specialist species of
(Ecuador). J Hydroinformatics 13:489–499. https://doi.org/10.
Odonata in Amazonia. Anim Conserv 21:526–533. https://doi.org/
2166/hydro.2010.082
10.1111/acv.12427
Dutra S, De Marco P (2015) Bionomic differences in odonates and their
Chovanec A, Schindler M, Waringer J, Wimmer R (2015) The dragonfly
influence on the efficiency of indicator species of environmental
association index (Insecta: Odonata)-a tool for the type-specific as-
quality. Ecol Indic 49:132–142. https://doi.org/10.1016/j.ecolind.
sessment of lowland rivers. River Res Appl 31:627–638. https://doi.
2014.09.016
org/10.1002/rra.2760
Cleary D, Mooers A, Karl A et al (2004) Diversity and community com- Feindt W, Fincke O, Hadrys H (2013) Still a one species genus? Strong
position of butterflies and odonates in an ENSO-induced fire affect- genetic diversification in the world’s largest living odonate, the
ed habitat mosaic: a case study from East Kalimantan, Indonesia. Neotropical damselfly Megaloprepus caerulatus. Conserv Genet
Oikos 105:426–446 15:469–481. https://doi.org/10.1007/s10592-013-0554-z
Corbet PS (1993) Are Odonata useful as bioindicators? Libellula 12:91– Fincke OM, Hedström I (2008) Differences in forest use and colonization
102 (http://www.libellula.org/wp-content/uploads/2015/12/12_ by Neotropical tree-hole damselflies (Odonata: Pseudostigmatidae):
Corbet.pdf) implications for forest conversion. Stud Neotrop Fauna Environ 43:
Corbi JJ, Trivinho-Strixino S, Dos Santos A (2008) Environmental eval- 35–45. https://doi.org/10.1080/01650520701504597
uation of metals in sediments and dragonflies due to sugar cane Forero-Céspedes AM, Reinoso-Flórez G, Gutiérrez C (2013) Water qual-
cultivation in neotropical streams. Water Air Soil Pollut 195:325– ity assessment of the Opia River (tolima-Colombia), using macro-
333. https://doi.org/10.1007/s11270-008-9749-1 invertebrates and physicochemical parameters. Caldasia 35:371–
Corbi JJ, dos Santos FA, Zerlin R et al (2011) Assessment of chromium 387 (http://www.scielo.org.co/scielo.php?script=sci_arttext&pid=
contamination in the Monte Alegre stream: A case study. Brazilian S0366-52322013000200012, ISSN 0366-5232)
Arch Biol Technol 54:613–620. https://doi.org/10.1590/S1516- Fulan JÂ, dos Anjos MR (2015) Predation by Erythemis nymphs
89132011000300023 (Odonata) on Chironomidae (Diptera) and Elmidae (Coleoptera) in
Corlett RT (2016) Tropical rainforests and climate change. Reference different conditions of habitat complexity. Acta Limnol Bras 27:
Module in Earth Systems and Environmental Sciences 1-5 p. 454–458. https://doi.org/10.1590/s2179-975x2415
https://doi.org/10.1016/B978-0-12-409548-9.09501-4 Fulan JÂ, Henry R (2007) Distribuição temporal de imaturos de Odonata
Dale VH, Beyeler SC (2001) Challenges in development and use of (Insecta) associados a Eichhornia azurea (Kunth) na Lagoa do
radiopharmaceuticals. J Nucl Med 42:3–10 Camargo, Rio Paranapanema, São Paulo. Rev Bras Entomol 51:
Dalzochio MS, Costa JM, Uchôa MA (2011) Diversity of Odonata 224–227. https://doi.org/10.1590/S0085-56262007000200013
(Insecta) in lotic systems from Serra da Bodoquena, Mato Grosso García-García PL, Vázquez G, Novelo-Gutiérrez R, Favila ME (2016)
do Sul State, Brazil. Rev Bras Entomol 55:88–94. https://doi.org/10. Effects of land use on larval Odonata assemblages in cloud forest
1590/S0085-56262011000100014 streams in central Veracruz, Mexico. Hydrobiologia 785:19–33.
Dalzochio MS, Périco E, Renner S, Sahlén G (2018) Effect of tree plan- https://doi.org/10.1007/s10750-016-2900-x
tations on the functional composition of Odonata species in the Ghani W, Salmah Md Rawi C, Abd Hamid S, Abdo Al-Shami S (2016)
highlands of southern Brazil. Hydrobiologia 808:283–300. https:// Efficiency of different sampling tools for aquatic macroinvertebrate
doi.org/10.1007/s10750-017-3431-9 collections in Malaysian streams. Trop Life Sci Res 27:115–133
 Environ Sci Pollut Res

Gómez-Anaya JA, Novelo-Gutiérrez R (2015) A case of successful res- Laurance WF, Goosem M, Laurance SGW (2009) Impacts of roads and
toration of a tropical wetland evaluated through its Odonata (Insecta) linear clearings on tropical forests. Trends Ecol Evol 24:659–669.
larval assemblage. Rev Biol Trop 63:1043–1058. https://doi.org/10. https://doi.org/10.1016/j.tree.2009.06.009
15517/rbt.v63i4.15738 (https://revistas.ucr.ac.cr/index.php/rbt/ Macías-Ordoñez R, Machado G, Macedo RH (2014) Macroecology of
article/view/15738) sexual selection: large-scale influence of climate on sexually se-
Gómez-Anaya JA, Novelo-Gutiérrez R, Astudillo-Aldana MR (2017) lected traits. In: Machado G, Macedo RH (eds) Sexual Selection
Efecto de las descargas domésticas y de beneficio de café sobre la Perspectives and Models from the Neotropics. Elsevier, USA, pp
calidad del agua y la diversidad de larvas de Odonata (Insecta) en 1–10
un arroyo de bosque mesófilo de montaña en Veracruz, México. Martínez-Sanz C, Puente-García SM, Rebolledo ER, Jiménez-Prado P
Rev Mex Biodivers 88:372–380. https://doi.org/10.1016/j.rmb. (2014) Macroinvertebrate richness importance in coastal tropical
2017.03.004 streams of esmeraldas (Ecuador) and its use and implications in
Gómez-Tolosa M, Mendoza-Cuenca LF, Rioja-Paradela TM et al (2015) environmental management procedures. Int J Ecol 2014:1–11.
Odonata (Insecta) de tres cuencas en la costa de Chiapas: Lista de https://doi.org/10.1155/2014/253134
especies y registro nuevo. Rev Mex Biodivers 86:41–47. https://doi. Martini A, Resende D, Silva L, Duarte M (2013) Distribuição espacial e
org/10.7550/rmb.48665 temporal da fauna de invertebrados bentônicos na APA do
Gonçalves FB, de Menezes MS (2011) A comparative analysis of biotic município de Coqueiral, MG, com ênfase em Odonata. Rev Bras
indices that use macroinvertebrates to assess water quality in a coast- Zoociências 15:183–194
al river of Paraná state, southern Brazil. Biota Neotrop 11:27–36. Marvier M, Kareiva P, Neubert MG (2004) Habitat destruction, fragmen-
https://doi.org/10.1590/s1676-06032011000400002 tation, and disturbance promote invasion by habitat generalists in a
González Soriano E, Noguera F, Oñate Ocaña L (2011) A biodiversity multispecies metapopulation. Risk Anal 24:869–878
hotspot for odonates in mexico: the huasteca potosina, san luis Mendes TP, Cabette HSR, Juen L (2015) Setting boundaries: environ-
potosí. Odonatologica 40:179–190 mental and spatial effects on Odonata larvae distribution (Insecta).
An Acad Bras Cienc 87:239–248. https://doi.org/10.1590/0001-
Guisande C, Heine J, García-Roselló E, González-Dacosta J, Vilas LG,
3765201520130477
Perez-Schofield BJG (2017) DER: an algorithm for comparing spe-
Mendes TP, Oliveira-Junior JMB, Cabette HSR, Batista JD, Juen L
cies diversity between assemblages. Ecol Indic 81:41–46. https://
(2017) Congruence and the biomonitoring of aquatic ecosystems:
doi.org/10.1016/j.ecolind.2017.05.049
are odonate larvae or adults the most effective for the evaluation of
Gutiérrez-Fonseca PE, Ramírez A, Pringle CM (2018) Large-scale cli-
impacts. Neotrop Entomol 46:631–641. https://doi.org/10.1007/
matic phenomena drive fluctuations in macroinvertebrate assem- s13744-017-0503-5
blages in lowland tropical streams, Costa Rica: the importance of
Mendes TP, Luiza-Andrade A, Cabette HSR, Juen L (2018) How does
ENSO events in determining long-term (15y) patterns. PLoS One
environmental variation affect the distribution of dragonfly larvae
13:1–24. https://doi.org/10.1371/journal.pone.0191781
(Odonata) in the Amazon-Cerrado transition zone in Central Brazil?
Hanh Tien N, Boets P, Lock K et al (2017) Water quality related macro- Neotrop Entomol 47:37–45. https://doi.org/10.1007/s13744-017-
invertebrate community responses to environmental gradients in the 0506-2
Portoviejo River (Ecuador). Ann Limnol - Int J Limnol 53:203–219. Miguel TB, Calvão LB, Vital MVC, Juen L (2017a) A scientometric
https://doi.org/10.1051/limn/2017007 study of the order Odonata with special attention to Brazil. Int J
Hawkes H (1997) Origin and development of the biological monitoring Odonatol 20:27–42. https://doi.org/10.1080/13887890.2017.
working party score system. Water Research 32:964–968. https:// 1286267
doi.org/10.1016/S0043-1354(97)00275-3 Miguel TB, Oliveira-Junior JMB, Ligeiro R, Juen L (2017b) Odonata
Hofhansl FP, Schneeweihs S (2008) Banderillas : effects of deforestation (Insecta) as a tool for the biomonitoring of environmental quality.
on dragonflies (Insecta, Odonata) in the Pacific lowland of Costa Ecol Indic 81:555–566. https://doi.org/10.1016/j.ecolind.2017.06.
Rica Banderillas : Efectos de la deforestación sobre libélulas 010
(Insecta, Odonata) en la tierra baja Pacífica en Costa Rica. Nat Monteiro-Júnior CS, Juen L, Hamada N (2014) Effects of urbanization on
Cult Hist Golf Dulce Reg Costa Rica 768 pp 80:237–247 stream habitats and associated adult dragonfly and damselfly com-
Jackson LE, Kurtz JC, Fisher WS (2000) Evaluation guidelines for eco- munities in central Brazilian Amazonia. Landsc Urban Plan 127:28–
logical indicators. Research Triangle Park, NC 40. https://doi.org/10.1016/j.landurbplan.2014.03.006
Juen L, Cabette HSR, De Marco P (2007) Odonate assemblage structure Monteiro-Júnior CDS, Juen L, Hamada N (2015) Analysis of urban im-
in relation to basin and aquatic habitat structure in Pantanal wet- pacts on aquatic habitats in the central Amazon basin: adult odonates
lands. Hydrobiologia 579:125–134. https://doi.org/10.1007/ as bioindicators of environmental quality. Ecol Indic 48:303–311.
s10750-006-0395-6 https://doi.org/10.1016/j.ecolind.2014.08.021
Juen L, de Oliveira-Junior JMB, Shimano Y et al (2014) Composição e Monteiro-Júnior CDS, Esposito MC, Juen L (2016) Are the adult odonate
riqueza de odonata (Insecta) em riachos com diferentes níveis de species found in a protected area different from those present in the
conservação em um ecótone cerrado-floresta amazônica. Acta surrounding zone? A case study from eastern Amazonia. J Insect
Amaz 44:223–233. https://doi.org/10.1590/S0044- Conserv 20:643–652. https://doi.org/10.1007/s10841-016-9895-5
59672014000200008 Montemayor S, Melo MC, Elizabeth Scheibler E (2016) Forecasting the
Khazan ES (2014) Tests of biological corridor efficacy for conservation fate of high mountain ponds in the Andean region under future
of a Neotropical giant damselfly. Biol Conserv 177:117–125. climate change. Austral Ecol 41:983–992. https://doi.org/10.1111/
https://doi.org/10.1016/j.biocon.2014.06.006 aec.12392
Koroiva R, Pepinelli M, Rodrigues ME, Roque FO, Lorenz-Lemke AP, Nessimian JL, Venticinque EM, Zuanon J, de Marco P Jr, Gordo M,
Kvist S (2017) DNA barcoding of odonates from the Upper Plata Fidelis L, D’arc Batista J, Juen L (2008) Land use, habitat integ-
basin: database creation and genetic diversity estimation. PLoS One rity, and aquatic insect assemblages in Central Amazonian
12:1–15. https://doi.org/10.1371/journal.pone.0182283 streams. Hydrobiologia 614:117–131. https://doi.org/10.1007/
s10750-008-9441-x
Kutcher T, Bried J (2014) Adult Odonata conservatism as an indicator of
Niemeijer D, de Groot RS (2008) A conceptual framework for selecting
freshwater wetland condition. Ecol Indic 38:31–39. https://doi.org/
environmental indicator sets. Ecol Indic 8:14–25. https://doi.org/10.
10.1016/j.ecolind.2013.10.028
1016/j.ecolind.2006.11.012
Environ Sci Pollut Res

Novelo-Gutiérrez R, Gómez-Anaya JA (2009) A comparative study of Rodrigues ME, de Oliveira RF, Quintero JMO et al (2016) Nonlinear
Odonata (Insecta) assemblages along an altitudinal gradient in the responses in damselfly community along a gradient of habitat loss
sierra de Coalcomán Mountains, Michoacán, Mexico. Biodivers in a savanna landscape. Biol Conserv 194:113–120. https://doi.org/
Conserv 18:679–698. https://doi.org/10.1007/s10531-008-9533-y 10.1016/j.biocon.2015.12.001
Oertli B (2008) Dragonflies in the assessment and the monitoring of Rodrigues ME, Roque FDO, Guillermo-Ferreira R, Saito VS, Samways
aquatic habitats. In: Córdoba A (ed) Dragonflies and Damselflies, MJ (2018) Egg-laying traits reflect shifts in dragonfly assemblages
Model Organisms for Ecological and Evolutionary Research. in response to different amount of tropical forest cover. Insect
Oxford University Press, pp 79–95 Conserv Divers 12:231–240. https://doi.org/10.1111/icad.12319
Oertli B, Auderset Joye D, Castella E et al (2002) Does size matter? The Samways MJ (2008) Dragonflies as focal organisms in contemporary
relationship between pond area and biodiversity. Biol Conserv 104: conservation biology. In: Córdoba A (ed) Dragonflies and
59–70. https://doi.org/10.1016/S0006-3207(01)00154-9 Damselflies, Model Organisms for Ecological and Evolutionary
Paiva-Silva D, De Marco P, Resende DC (2010) Adult odonate abun- Research. Oxford University Press, pp 97–108
dance and community assemblage measures as indicators of stream Seymour F, Harris NL (2019) Reducing tropical deforestation. Science
ecological integrity: a case study. Ecol Indic 10:744–752. https:// 365(80):756LP–757LP. https://doi.org/10.1126/science.aax8546
doi.org/10.1016/j.ecolind.2009.12.004 Šigutová H, Šipoš J, Dolný A (2019) A novel approach involving the use
Paulson DR (2004) Critical species of Odonata in the Neotropics. Int J of Odonata as indicators of tropical forest degradation: when family
Odonatol 7:163–188 matters. Ecol Indic 104:229–236. https://doi.org/10.1016/j.ecolind.
Paulson DR (2006) The importance of forests to neotropical dragonflies. 2019.05.001
In Cordero R (ed) Forests and Dragonflies, 4er WDA International Silva L d FR, Machado FS, Resende DLMC, Neiss UG (2018) Immature
Symposium of Odonatology, Pontevedra (Spain), pp 79-101 Odonata community in streams: diversity, season variation and hab-
Peel MC, Finlayson BL, McMahon TA (2007) Updated world map of the itat preference in different levels of degradation. North West J Zool
Köppen-Geiger climate classification. Hydrol Earth Syst Sci 11: 14:232–236 (http://biozoojournals.ro/nwjz/content/v14n2/nwjz_
1633–1644. https://doi.org/10.5194/hess-11-1633-2007 e172101_Silva.pdf)
Pinto NS, Juen L, Cabette HSR, de Marco P (2012) Fluctuating asym- Silva-Franca SR, Hughes RM, Callisto M (2018) Student monitoring of
metry and wing size of Argia tinctipennis Selys (Zygoptera: the ecological quality of neotropical urban streams. Ambio 48:867–
Coenagrionidae) in relation to riparian forest preservation status. 878. https://doi.org/10.1007/s13280-018-1122-z
Neotrop Entomol 41:178–185. https://doi.org/10.1007/s13744-
Simaika JP, Samways MJ (2009) An easy-to-use index of ecological
012-0029-9
integrity for prioritizing freshwater sites and for assessing habitat
Pires MM, Kotzian CB, Spies MR, Neri DB (2013) Diversity of Odonata
quality. Biodivers Conserv 18:1171–1185. https://doi.org/10.1007/
(Insecta) larvae in streams and farm ponds of a montane region in
s10531-008-9484-3
southern Brazil. Biota Neotrop 13:259–267. https://doi.org/10.1590/
Simaika JP, Samways MJ (2011) Comparative assessment of indices of
s1676-06032013000300028
freshwater habitat conditions using different invertebrate taxon
Pither J, Taylor PD (1998) An experimental assessment of landscape
sets. Ecol Indic 11:370–378. https://doi.org/10.1016/j.ecolind.
connectivity. Oikos 83:166–174
2010.06.005
Poulton BC, Wildhaber ML, Charbonneau CS, Fairchild JF, Mueller BG,
Siva-Jothy MT, Gibbons DW, Pain D (1995) Female oviposition-site
Schmitt CJ (2003) A longitudinal assessment of the aquatic macro-
preference and egg hatching success in the damselfly Calopteryx
invertebrate community in the channelized lower Missouri River.
splendens xanthostoma. Behav Ecol Sociobiol 37:39–44. https://
Environ Monit Assess 85:23–53. https://doi.org/10.1023/A:
doi.org/10.1007/BF00173897
1023301016001
Rácenis J (1955) Los representantes Venezolanos de la familia Soares JAC, Batista-Silva VF, Boneto DD, Bailly D, Abelha MCF,
Protoneuridae (Odonata). Memoria de la Sociedad de Ciencias Oliveira IADV (2015) Assemblage of immature Odonata (Insecta,
Naturales La Salle 15:48–62 Anisoptera) in streams of the Mato Grosso do Sul State: spatial
R Core Team (2018) R: a language and environment for statistical com- implications. Iheringia Série Zool 105:325–332. https://doi.org/10.
puting. R Foundation for Statistical Computing, Vienna https:// 1590/1678-476620151053325332
www.R-project.org Stewart DAB, Samways MJ (2008) Conserving dragonfly (Odonata) as-
Ramírez-Montoya J, Osuna-Vergara I, Rojas-Mora J, Guerrero-Gómez S semblages relative to river dynamics in an African savanna game
(2016) Remuestreo Bootstrap y Jackknife en confiabilidad: Caso reserve. Conserv Biol 12:683–692. https://doi.org/10.1111/j.1523-
Exponencial y Weibull. Rev Fac Ing 25:55. https://doi.org/10. 1739.1998.96465.x
19053/01211129.4137 Stoks R, Córdoba-Aguilar A (2012) Evolutionary ecology of Odonata: a
Renner S, Perico E, Sahlén G (2013) Dragonflies (Odonata) in subtrop- complex life cycle perspective. Annu Rev Entomol 57:249–265.
ical Atlantic forest fragments in Rio Grande do Sul, Brazil : seasonal https://doi.org/10.1146/annurev-ento-120710-100557
diversity and composition. Sci Plena 9:1–8 http://www.diva-portal. Teles HF, Seabra Linares M, Rocha PA, Ribeiro AS (2013)
org/smash/get/diva2:713079/FULLTEXT01.pdf Macroinvertebrados Bentônicos como Bioindicadores no Parque
Renner S, Périco E, Sahlén G (2016) Effects of exotic tree plantations on Nacional da Serra de Itabaiana, Sergipe, Brasil. Rev Bras
the richness of dragonflies (Odonata) in Atlantic Forest, Rio Grande Zoociências 15:123–137 (https://periodicos.ufjf.br/index.php/
do Sul, Brazil. Int J Odonatol 19:207–219. https://doi.org/10.1080/ zoociencias/article/view/24509, e-ISSN: 2596-3325)
13887890.2016.1251343 Termaat T, van Strien AJ, van Grunsven RHA et al (2019) Distribution
Rizo-Patrón VF, Kumar A, McCoy Colton MB et al (2013) trends of European dragonflies under climate change. Divers Distrib
Macroinvertebrate communities as bioindicators of water quality 25:936–950. https://doi.org/10.1111/ddi.12913
in conventional and organic irrigated rice fields in Guanacaste, Tollefson J (2019) One million species face extinction. Nature 569:171.
Costa Rica. Ecol Indic 29:68–78. https://doi.org/10.1016/j.ecolind. https://doi.org/10.1038/d41586-019-01448-4
2012.12.013 Torres-Olvera MJ, Durán-Rodríguez OY, Torres-García U et al (2018)
Rocha-Ortega M, Rodríguez P, Córdoba-Aguilar A (2019) Can dragonfly Validation of an index of biological integrity based on aquatic mac-
and damselfly communities be used as bioindicators of land use roinvertebrates assemblages in two subtropical basins of central
intensification? Ecol Indic 107. https://doi.org/10.1016/j.ecolind. Mexico. Lat Am J Aquat Res 46:945–960. https://doi.org/10.3856/
2019.105553 vol46-issue5-fulltext-8
 Environ Sci Pollut Res

Valente-Neto F, De Oliveira RF, Rodrigues ME et al (2016) Toward a Vilela DS, Ferreira RG, Del-Claro K (2016) A comunidade de odonatos
practical use of Neotropical odonates as bioindicators: testing con- de uma vereda Brasileira: Padrões sazonais, diversidade de espécies
gruence across taxonomic resolution and life stages. Ecol Indic 61: e raridade em um ambiente de buritizal. Biosci J 32:486–495.
952–959. https://doi.org/10.1016/j.ecolind.2015.10.052 https://doi.org/10.14393/BJ-v32n2a2016-30491
Valente-Neto F, Rodrigues ME, Roque F d O (2018) Selecting indicators
based on biodiversity surrogacy and environmental response in a Publisher’s note Springer Nature remains neutral with regard to jurisdic-
riverine network: bringing operationality to biomonitoring. Ecol tional claims in published maps and institutional affiliations.
Indic 94:198–206. https://doi.org/10.1016/j.ecolind.2018.06.066