Ic

PDa

Mb

R

a

ARRAA

KMBBAN

1

mibidegwg

h1

Ecological Indicators 69 (2016) 606–616

Contents lists available at ScienceDirect

Ecological Indicators

jo ur nal ho me page: www.elsev ier .com/ locate / ecol ind

ncorporating natural variability in the bioassessment of streamondition in the Atlantic Forest biome, Brazil

riscilla S. Pereiraa, Natália F. Souzaa, Darcilio F. Baptistaa, Jaime L.M. Oliveirab,aniel F. Bussa,∗

Fundac ão Oswaldo Cruz, Instituto Oswaldo Cruz, Laboratório de Avaliac ão e Promoc ão da Saúde Ambiental (LAPSA/IOC/FIOCRUZ), Av. Brasil 4365,anguinhos, Pavilhão Lauro Travassos, Rio de Janeiro, RJ CEP 21040-360, BrazilFundac ão Oswaldo Cruz, Escola Nacional de Saúde Pública Sérgio Arouca (DSSA/ENSP/FIOCRUZ), Rua Leopoldo Bulhões 1480, Manguinhos, Rio de Janeiro,J CEP 21041-210, Brazil

r t i c l e i n f o

rticle history:eceived 24 August 2015eceived in revised form 13 May 2016ccepted 19 May 2016vailable online 28 May 2016

eywords:ultimetric index

iological monitoringioassessment protocolquatic insectseotropical streams

a b s t r a c t

Most bioassessment programs in Brazil face difficulties when scaling up from small spatial scales becauselarger scales usually encompass great environmental variability. Covariance of anthropogenic pressureswith natural environmental gradients can be a confounding factor in the evaluation of biologic responsesto anthropogenic pressures. The objective of this study was to develop a multimetric index (MMI) withmacroinvertebrates for two stream types and two ecoregions in the Atlantic Forest biome in Rio de Janeirostate, Brazil. We hypothesized that by using two approaches – (1) testing and adjusting metrics to land-scape parameters, and (2) selecting metrics using a cluster analysis to avoid metrics redundancy – thefinal MMI would perform better than the traditional approach (unadjusted metrics, one metric repre-senting each category). Four MMIs were thus developed: MMI-1 – adjusted MMI with metrics selectedafter cluster analysis); MMI-2 – adjusted MMI with one metric from each category; MMI-3 – unadjustedMMI with metrics selected after cluster analysis; MMI-4 – unadjusted MMI with one metric from eachcategory. We used three decision criteria to assess MMI’s performance: precision, responsiveness andsensitivity. In addition, we tested the MMI’s by using an independent set of sites to validate the results.Although all MMIs performed well in the three criteria, adjusting metrics to natural variation increasedMMI response and sensitivity to impairment. In addition, the selected MMI-2 was able to classify sites

of two stream types and two ecoregions. The use of cluster analysis, however, did not avoid high redun-dancy between metrics of different branches. The MMI-4 had the poorest performance among all testedMMIs and it was not able to distinguish adequately reference and impaired sites from both ecoregions.We present some considerations on the use of metrics and on the development of MMI’s in Brazil and elsewhere.

. Introduction

Multimetric indices (MMIs) with macroinvertebrates are theost widely used approach for biological assessment and monitor-

ng of aquatic ecosystems. This approach has been widely acceptedecause: (a) it is based on ecological concepts and processes, (b)

t has the potential to assess ecological functions, and (c) it caniscriminate human impacts (Helson and Williams, 2013; Meretat al., 2013). However, one difficulty many bioassessment pro-

rams in Brazil face is scaling up from small spatial scales – fromhere most indices are developed (e.g. a river basin or hydro-

raphic region) – to larger scales. Buss et al. (2015), analyzing 13

∗ Corresponding author.E-mail addresses: dbuss@ioc.fiocruz.br, dbuss.rio@gmail.com (D.F. Buss).

ttp://dx.doi.org/10.1016/j.ecolind.2016.05.031470-160X/© 2016 Elsevier Ltd. All rights reserved.

© 2016 Elsevier Ltd. All rights reserved.

large-scale macroinvertebrate protocols from around the worlddiscussed that limitations for scaling up may be associated withlack of logistics and funding, a reluctance to change establishedtechniques or gear, or the fact that locally developed methodssometimes yield more accurate results than regionally applica-ble ones. Despite these difficulties, many countries succeeded inbuilding widely applicable monitoring programs by using the sameand/or compatible sampling protocols and by selecting metrics thatare sensitive on large-scales (e.g. Stoddard et al., 2008; Moya et al.,2011; Jun et al., 2012).

Large-scale studies, however, may encompass great environ-mental variability. One approach to describe and account for

variability in ecological studies is to classify areas in ecore-gions. Ecoregions are usually defined as relatively homogeneousareas that have similar environmental conditions (Omernik, 1995).Ecoregions can be defined at different spatial scales, and aim to

al Indi

salSb(eahec(d(ncdictw

mo2nsarsdTFutafm

2

2

gtttvadapd1Armn2

2

r

P.S. Pereira et al. / Ecologic

erve as a territory for investigation, assessment, managementnd monitoring of ecosystems, including the development of bio-ogical criteria and water quality standards (Kong et al., 2013).ome authors have found relationships between the macroinverte-rate fauna and environmental conditions at the ecoregional levelBarbour et al., 1996; Reynoldson et al., 1997). Other authors, how-ver, have found macroinvertebrate fauna to be more stronglyssociated with micro/local scales, such as substrate and micro-abitat (Gerth and Herlihy, 2006; Costa and Melo, 2008; Ligeirot al., 2013). Acknowledging that, some countries incorporate localharacteristics (stream types) in different biomonitoring protocolsAQEM, 2002; Buss et al., 2015). For example, New Zealand hasifferent protocols for hard-bottomed and soft-bottomed streamsStark et al., 2001). The authors argue that this separation isecessary because of significant differences in the structure andomposition of aquatic communities among stream types, and thusifferent methods are required for sample collection and process-

ng to be cost-effective. The definition of stream types allows theorrect establishment of reference conditions that are comparableo the ecological status classifications within each group of riversith similar characteristics (Stark et al., 2001).

Covariance of anthropogenic pressures with natural environ-ental gradients can be a confounding factor in the evaluation

f biologic responses to anthropogenic pressures (Stoddard et al.,008; Hawkins et al., 2010; Moya et al., 2011). One simple tech-ique for normalizing metrics for natural gradients is to remove thetressor gradient from the data by focusing on reference-site datand to quantify the remaining correspondence between the met-ic value and the natural gradient (Stoddard et al., 2008). Recenttudies aimed to test this and other alternative approaches for theevelopment of MMIs (e.g. Chen et al., 2014; Macedo et al., 2016).he objective of this study was to develop a MMI for the Atlanticorest biome in Rio de Janeiro state, Brazil. We hypothesized that bysing two of these approaches – (1) testing and adjusting metricso landscape parameters, and (2) selecting metrics using a clusternalysis to avoid metrics redundancy – the final MMI would per-orm better than the traditional approach (unadjusted metrics, one

etric representing each category).

. Materials and methods

.1. Study area

The geomorphology of Rio de Janeiro state is composed of aroup of coastal plains separated by hills and two mountain chainshat run parallel to the ocean (Serra do Mar, ranging from alti-udes 0–2000 ma s.l and Serra da Mantiqueira, ranging from 800o 2500 m.a.s.l). In between the two mountain chains, lies thealley of the state’s main river, Paraiba do Sul (at an altitude ofround 800 ma s.l.). According to Alvares et al. (2013) 44% of Rioe Janeiro state’s mid-lower portions is classified as tropical with

summer rainy season, with the most mountainous regions andlateaus classified as humid subtropical with hot summer, withoutry season or with a dry winter. Temperatures oscillate between5 ◦C and 28 ◦C and annual rainfall is around 1000–1500 mm. Thetlantic Forest biome, which originally covered virtually the entireegion, now represents less than 12% of its original extent, and isostly spread in the higher parts of the mountains and in rem-

ants interspersed with agriculture and pasture (Ribeiro et al.,011).

.2. Sampled sites

We sampled 73 sites (once, during the dry season, in streamsanging from 1 st to 4th order, according to Strahler classification

cators 69 (2016) 606–616 607

using 1:50,000 scale maps) representing two stream types, twoecoregions and three classes of impairment. Ecoregions were basedon the classification of ’dominions’ of RadamBrasil (Brasil, 1983).Stream types were classified as “transitional/sedimentary areas”(stream type 1) and “rocky substrates” (stream type 2; see below fordetails). We chose sites based on ad hoc indication and/or by previ-ous knowledge of the area to represent sites classified as reference,intermediate or impaired. Sites classified as “reference” had “Opti-mal” or “Good” environmental condition according to the HabitatAssessment Protocol (HAP; see below for logic and measurement),absence of channelization and <25% upstream urban or industrialareas. Impaired stream reaches were classified as “Poor” conditionaccording to HAP and >40% of upstream area affected by urbanareas or intense farming or livestock grazing. Intermediate siteshad characteristics between these two classes.

We sampled forty-nine sites in the sedimentary deposits ecore-gion (SD). The SD ecoregion is located at the piedmont of Serrado Mar mountain range, with altitudes about 200 ma.s.l. andis a depositional zone formed by marine, lacustrine and flu-vial sedimentation processes (Brasil, 1983). Being a transitionalzone between erosive/depositional zones, sampled streams weredivided into two predominant substrate types: reaches with >80%sand and clay (“transitional/sedimentary areas”; stream type (1)and reaches with >70% particles of gravel size or greater (“rockysubstrates”; stream type (2). In this region, land use is dominated bypatches of small-scale agriculture and livestock grazing, and mini-mally impacted areas are scarce. Reference areas in this ecoregionwere classified as “least disturbed areas”, according to the refer-ence condition approach (RCA; Stoddard et al., 2006). Twenty-twosites of stream type 1 were sampled; of which six were reference,six intermediates and ten impaired. Twenty-seven sites of streamtype 2 were sampled, of which fifteen were reference, three inter-mediate and nine impaired (Fig. 1).

We sampled twenty-four sites in the mountainous scarps ecore-gion (MS). This ecoregion is located at higher altitude (from>200 ma.s.l. to around 1,800 ma.s.l) in a mountainous region withhigh slope and steep scarps. Streams in MS have a > 80% predom-inance of rocky substrates (stream type 2) – bedrocks in somereaches – with few patches of sand and formation of pools inter-twined with riffles or runs. All sites were sampled within or nearprotected areas (conservation units), which were classified as “min-imally disturbed areas” or “best attainable” (RCA; Stoddard et al.,2006). The latter occurred in areas outside conservation units, buthad low to moderate impact by rural activities, and full or partialriparian vegetation and forest fragments.

2.3. Environmental and biological data

We sampled macroinvertebrates by using a kick-net samplerwith mesh size of 500 �m. Twenty samples (around 20 m2) weretaken proportional to the substrates available in each site, follow-ing the multi-habitat method (Barbour et al., 1999). The percentageof available habitats was estimated by visual inspection. Substrateswith less than 5% of the site area were not sampled. Samples wereobtained from a site length of approximately 20 times the chan-nel width. Samples were composited and conserved in the fieldin 80% ethanol and taken to the laboratory for further inspec-tion. In the laboratory, samples were washed to remove coarseorganic matter, such as leaves and twigs. The remaining mate-rial was placed in a sub-sampler (64 × 36 cm), divided into 24quadrats, each measuring 10.5 × 8.5 cm (Patent application numberPCT/BR2011/000144). Sub-sampling is a procedure widely used in

formulating multimetric indices, to assure randomness of the pro-cedure, making it less subject to inherent variations from changesin team members (Oliveira et al., 2011). Eight quadrats were cho-sen at random, following the procedures described in Oliveira

608 P.S. Pereira et al. / Ecological Indicators 69 (2016) 606–616

Fig. 1. Sampling sites indicating the two ecoregions (sedimentary deposits, SD and mountain scarps, MS), two stream substrate typologies (rocky, R and transi-t e, Int

ei1anDl

pd(MprcAhgHaectai5e

ional/sedimentary, S) and three classes of impairment (reference, Ref; intermediat

t al. (2011). Two sites from stream type 1 (one reference and onempaired) were excluded from the analyses because fewer than50 individuals were found in the entire sample, following Carternd Resh (2001). Specimens were identified to the lowest taxo-omic level possible (mostly genus) – except for Hemiptera andiptera, which were identified to family level, and Annelida to class

evel.In each sampling occasion, physicochemical and environmental

arameters were recorded in the field. Dissolved oxygen (DO) wasetermined by using a YSI 550A analyzer, pH with a MPA 210pLabConte) and conductivity and total dissolved solids by using

CA 150p (LabConte). Water samples were preserved in sterilelastic bags (whirl-pak) according to APHA (2000). In the labo-atory, ammonia was determined by using a HACH (DR 2500),hloride and total alkalinity by the titrimetric method followingPHA (2000). Sampling sites were classified using the visual-basedabitat assessment protocol (HAP; Barbour et al., 1999) for low-radient (ecoregion SD) and high-gradient (MS ecoregion). TheAP analyzes ten environmental parameters, such as substratevailability for colonization by benthic fauna, water velocity andmbeddedness (pool variability for low-gradient streams), channelondition (sinuosity for low-gradient streams), sediment deposi-ion, margin stability and riparian vegetation. For each parameter,

score between 0 and 20 is assigned. Sites are classified accord-ng to the mean score obtained in the HAP, as follows: 0–5 “Poor”,.1–9.9 “Marginal”, 10–14.9 “Suboptimal” and 15–20 an “Optimal”nvironmental condition (Barbour et al., 1999).

and impaired, Imp).

2.4. MMI development and validation

2.4.1. Metric selectionWe examined 42 benthic macroinvertebrate assemblage met-

rics commonly used in biomonitoring protocols (e.g. Barbour et al.,1999; Baptista et al., 2007; Helson and Williams, 2013). These met-rics comprise five categories, based on their logic (according toBarbour et al. (1999) and Stoddard et al. (2008)):

1) Richness – express the number of taxa of the whole macroin-vertebrate assemblage or of its subgroups, in family or genuslevel.

2) Diversity/evenness – diversity indices include information onthe taxonomic richness and the relative abundances of taxo-nomic groups. This category includes diversity indices such asShannon-Wiener, Margalef and Dominance (1-Simpson index),as well as equitability/evenness indices.

3) Composition – measures of the relative abundance of selectedgroups, in taxonomic level of order or lower.

4) Tolerance – metrics that consider the sensitivity and toler-ance of organisms to impairment. This category includes bioticindices and information on other sensitive or tolerant indica-

tor groups. We included indices developed for Brazil such as theBiological Monitoring Working Party (BMWP-CETEC; Junqueiraand Campos, 1998) and the Indice Biotico Estendido (IBE-IOC;Mugnai et al., 2008).

al Indi

5

afwtie(sevso2

irafi(e(dBABatalbcrw(vrrtasetmlamacasueMcscret

P.S. Pereira et al. / Ecologic

) Trophic – encompass functional feeding groups and provideinformation on the balance of feeding strategies (food acqui-sition and morphology) in the benthic assemblage and theirtrophic roles (shredders, collectors, scrapers, filterers or preda-tors).

The first step for metric selection was to evaluate if metrics metll these four criteria: (a) present similar results for reference areasrom typologies 1 and 2 using Mann-Whitney U tests. Our intentionas to have a single index for both typologies because often the two

ypologies are mixed in a site and it is difficult to determine whichs predominant. (b) Discriminate impairment conditions (refer-nce, intermediate and impaired) according to Kruskal-Wallis test.c) Have a linear response to impairment – i.e. intermediate siteshould present intermediate values between those found for ref-rence and impaired classes. (d) Do not present a narrow range ofalues (e.g., richness metrics based on only a few taxa) or that haveimilar values at most sites (e.g., most sites have values = 0), or manyutliers (following Stoddard et al., 2008; Helson and Williams,013), analyzed after inspection of Box-and-Whisker plots.

With the approved metrics, we developed and tested four MMIsn order to evaluate the potentially confounding effects of natu-al variation into biological response to anthropogenic impacts,nd to reduce the effects of redundancy between metrics. For therst target, we used four GIS-extracted environmental variablesbased on Chen et al. (2014) and Macedo et al. (2016)). Two param-ters were related to climate, available from WorldClim databasehttp://www.worldclim.org/): seasonality of temperature (i.e. stan-ard deviation × 100), coded Bio04, and annual precipitation, codedio12. The two others were related to topography, available fromMBDATA (http://www.dpi.inpe.br/Ambdata/): altitude and slope.oth geographic information systems had spatial resolution of 30rc-seconds (∼1 km). After verifying and correcting for normal dis-ribution according to Shapiro-Wilk test, metrics were regressedgainst environmental variables by stepwise (forward) multipleinear regression procedures and the Akaike information criteria touild the simplest possible model that adequately explained theandidate metric (Moya et al., 2007; Chen et al., 2014). If natu-al environmental gradients were significantly associated (p < 0.05)ith the variation in metric values, the residuals of each model

residual = observed value − predicted value) were used as metricalues. MMI-1 and MMI-2 used the residuals of metrics (hereaftereferred as “adjusted” metrics and MMIs). MMI-3 and MMI-4 usedaw metric values (hereafter “unadjusted” metrics and MMIs). Forhe second target, we measured metric redundancies by means of

cluster analysis using Pearson’s r correlation coefficients as theimilarity measure and Ward linkage as the clustering method (Caot al., 2007; Chen et al., 2014). Metrics in a cluster branch are poten-ially redundant (Stoddard et al., 2008). In order to decide which

etric from each cluster should be retained for the MMI we calcu-ated discrimination efficiency (DE) and t-tests between referencend impaired sites (Barbour et al., 1999; Stoddard et al., 2008). Foretrics expected to decrease at impaired sites, DE was calculated

s the percentage of impaired sites with a metric value <25th per-entile of the reference site values. For metrics expected to increaset impaired sites, DE was calculated as the percentage of impairedites with a metric value >75th percentile of the reference site val-es. The metric with the highest DE and significant t-test scores inach cluster was retained for the MMI. This approach was used forMI-1 and MMI-3. For MMI-2 and MMI-4, we calculated Pearson

orrelations between metrics of the same category (richness, diver-ity/evenness, composition, tolerance and trophic). Metrics were

onsidered redundant if r > 0.70 (Stoddard et al., 2008). We thenetained the metrics with the greater power to discriminate refer-nce and impaired conditions according to t-tests, and if similar,he one with lower cost and/or easier to calculate. As a summary:

cators 69 (2016) 606–616 609

MMI-1 – adjusted MMI with metrics selected after cluster analysis;MMI-2 – adjusted MMI with one metric from each category; MMI-3 – unadjusted MMI with metrics selected after cluster analysisMMI-4 – unadjusted MMI with one metric from each category.

2.4.2. MMI standardizationThe selected metrics were included in the four MMIs, using a

continuous system, following the recommendation of Blocksom(2003). For positive raw metrics (values that increase as the envi-ronment improves), Eq. (1) was used, and for negative raw metrics(values that decrease as the environment improves), Eq. (2) wasused, as follows:

Metric − 25%impaired75% reference − 25% impaired

× 10 (1)

Metric − 75%impaired25% reference − 75% impaired

× 10 (2)

where 25% impaired is the 25th percentile of impaired sites, 75%reference is the 75th percentile of reference sites, 75% impaired isthe 75th percentile of impaired sites, and 25% reference is the 25thpercentile of reference sites.

Negative values were assigned a value of zero and those above10 were assigned a value of 10. The sum of all metrics scores wasthen adjusted to fit a scale from 0 to 100, which was used as thefinal MMI score. The MMI score was used to define the class of bio-logical integrity, according to the arbitrary scale: severely impaired(0–19), impaired (20–39), regular (40–59), good (60–79), excellent(80–100).

2.4.3. MMI evaluationThree criteria were used to evaluate the performance of the

MMIs: precision, responsiveness and sensitivity to anthropogenicpressures (Chen et al., 2014; Macedo et al., 2016). For the precision,we calculated the standard deviation of each MMI for the refer-ence sites. Lower standard deviation meant higher precision. Theresponsiveness was tested with t-tests to verify if each MMI dis-criminated the pre-determined impairment conditions. To evaluatethe sensitivity, we ran stepwise-forward regression between eachMMI and anthropogenic-related variables (pH, dissolved oxygen,conductivity, total dissolved solids, chloride and ammonia).

2.4.4. MMI validationValidation tests consisted on evaluating MMI’s ability to cor-

rectly classify independent sites (not used to develop the index)and separate reference and impaired conditions. For that, we used24 reference sites from the MS ecoregion (stream type 2), and nineintermediately impaired sites from SD ecoregion (six of stream type1 and three of stream type 2). In addition, two Principal ComponentAnalysis (PCA) were performed – one for each set of sites (the 38sites used to develop and the 33 sites used to validate the index)– with environmental variables related to impairment (total dis-solved solids, conductivity, alkalinity, ammonia and HAP). Prior toanalysis, data were standardized by subtracting each value from itsmean and dividing it by its standard deviation to reduce the effectsof different scales used in the variables. The values for PCA axis 1were correlated (Pearson) with each corresponding MMI value.

2.4.5. Data analysisAll statistical analyses were conducted in R (version 3.2.3; R

Development Core Team, 2015, http://www.r-project.org/).

610 P.S. Pereira et al. / Ecological Indicators 69 (2016) 606–616

Table 1Stream order range, mean value and standard deviation of physicochemical parameters and Habitat Assessment Protocol (HAP) value and class, collected in streams used todevelop and to validate the MMI.

Environmental parameters Sites used to develop the MMI (SDecoregion, stream types 1 and 2)

Sites used to validate the MMI (SD and MSecoregions, stream types 1 and 2)

Reference (n = 20) Impaired (n = 18) Reference (n = 24) Intermediate (n = 9)

Stream order (range) 1st–4th 1st–4th 1st–4th 1st–4thMean width (m) 9.5 ± 5.8 9.4 ± 5.6 9.1 ± 3.4 12.1 ± 3.8Total dissolved solids (mg/L) 14.1 ± 6.5 31.5 ± 31.7 9.8 ± 2.8 18.3 ± 15.1pH 6.7 ± 0.8 7.01 ± 0.68 6.9 ± 0.4 7.24 ± 0.57Conductivity (�S/cm) 28.4 ± 13.4 66.4 ± 68.4 19.3 ± 5.0 36.8 ± 30.1Alkalinity (mg/L) 17.7 ± 8.6 46.5 ± 38.1 17.7 ± 9.6 22.5 ± 15.9Chloride (mg/L) 13.7 ± 8.3 14.5 ± 3.3 11.5 ± 4.0 8.7 ± 4.1Dissolved oxygen (mg/l) 7.4 ± 0.9 6.6 ± 1.8 7.5 ± 0.7 6.9 ± 1.1

0.113.8 ±Poo

3

3

uc

iitaec

3

WrfKehtrt

vmomtnsaf

sea

wTI

(s

NH4+ (mg/l) 0.02 ± 0.03

Habitat Assessment Protocol (HAP) 16.2 ± 2.3

Class (HAP) Excellent-Good

. Results

.1. Environmental and biological data

Significant differences between reference and impaired sitessed to develop the MMIs were found for total dissolved solids,onductivity, alkalinity, ammonia and the HAP index (Table 1).

Regarding the biological samples, 44,892 individuals (78 fam-lies and 141 genera) were identified. From this total, 9083ndividuals (65 families and 83 genera) were collected in transi-ional/sedimentary areas (stream type 1) and 19,088 (62 familiesnd 97 genera) in rocky substrate (stream type 2), both in the SDcoregion. A total of 16,721 (60 families and 101 genera) wereollected in stream type 2 in the MS ecoregion.

.2. MMI development and validation

All metrics, except ‘Number of Ephemeroptera families’ (Mann-hitney U test, p < 0.05), were approved for not discriminating

eference areas of stream types 1 and 2 (Table 2). Twelve metricsailed in discriminating the two classes of impairment (Table 2;ruskal-Wallis tests, p < 0.05) and fourteen other metrics werexcluded because they presented narrow ranges of values and/orad non-linear responses to the stress gradient (Table 2). Thus, fif-een metrics were approved for further screening, of which tenequired identification to family level, two to genus level and threehat required identification to order level (Table 2).

The approved metrics were regressed against environmentalariables (altitude, Bio04, Bio12, slope). All metrics followed a nor-al distribution and no transformation was necessary. Only six

f the fifteen metrics were significantly associated with environ-ental variables (four with temperature seasonality, Bio04, and

wo with annual precipitation, Bio12; Table 3). No metric was sig-ificantly associated with altitude or slope. The residuals of thoseix metrics (hereafter added a “ re” after the name of the metric)nd the unadjusted values for the other nine metrics were used inurther analysis.

For MMI-1, the selected metrics were based on a cluster analy-es (Fig. 2a) and the ones with the highest DE and t-test scores inach cluster were retained (Shanf, %EPTsBH, IBE-IOC re, BMWP rend%Ple). Only two of the five metrics had to be adjusted.

For MMI-2, the selected metrics for each category were the onesith the highest t-test scores to discriminate impairment classes.

wo of the four metrics were adjusted: Margf, EPTsBHf re, %Ple and

BE-IOC re.

For MMI-3, selected metrics were based on a cluster analysesFig. 2b) and we selected the ones with the highest DE and t-testcores in each cluster (Shanf, %Ple, BMWP, IBE-IOC and Domf). How-

± 0.15 0.03 ± 0.03 0.02 ± 0.02 1.5 17.9 ± 1.5 8.4 ± 2.4

r Excellent-Good Good-Regular

ever, using only the reference sites, we found a high correlationbetween two metrics that were in different clusters (Domf × Shanfr = −0.91). The original MMI-3 with five metrics had precision of12.11, responsiveness t-score of 11.04 (p < 0.0001) and sensitiv-ity of r2 = 0.86 (p < 0.0001). First, we decided to substitute thesecond-best performance metric in the cluster, provided it had nohigh correlation (r > 0.70) with the ones already selected. However,when substituting Shanf by%EPT or Equif, or when substitutingDomf by %Coleo, it reduced MMI precision, responsiveness and sen-sitivity. We then tested excluding one of these metrics. Both MMIshad better performance than the original index, so we retainedShanf because it yielded slightly better responsiveness (higher t-score, 13.12 versus 13.03 with Domf), although with slightly lowerprecision (standard deviation, 10.85 versus 9.95), and sensitivity (r20.91 versus 0.92). The final MMI-3 with four metrics (%Ple, BMWP,IBE-IOC and Shanf) was used in further analysis (Table 4).

For the MMI-4, when metrics were correlated (r > 0.70), weselected the ones with the lowest p-value according to Kruskal Wal-lis tests to separate the three impairment classes. In the richnesscategory, EPTsBHf was selected over Trif. In the diversity category,Margalef at family and genus levels were selected over Shannon-wiener index at family-level and Equitability index at family-level.In the composition category, %EPTsBH was selected over%EPT/%Chi,%Ple, and%EPT. In the tolerance category, BMWP-CETEC (familylevel) and IBE-IOC (genus level) were selected over Dominanceindex in families and genus level, respectively. For the latter, wedecided to retain both metrics because they have different logicand evaluate different aspects of macroinvertebrate assemblage.In addition, they were not highly correlated.

The four MMI performed similarly (Table 4). MMI-3 had betterprecision (lowest standard deviation in reference site scores) andthe lowest mean correlation between metrics in reference sites. TheMMI-2 had a slightly better response (highest t-score to discrim-inate impairment classes) and sensitivity (higher association withanthropogenic variables), but all MMIs performed well (p < 0.0001)for both criteria. The MMI-1, which adopted all the hypothesisapproaches, did not outperform the other MMIs in any criteria, norhad the lowest mean correlation between metrics considering onlythe reference sites (Table 4).

Considering the responsiveness criteria, MMI-2 (Table 5) had thebest performance, for both the development database and the val-idation databases (Table 4; Fig. 3). MMI-4 performed poorly whenconsidering the validation database, with a very narrow range ofvalues separating reference and impaired sites (Fig. 3). For MMI-2,metric ranges were similar and showed no significant differences

between reference sites of the two stream types, and in both casesit was able to discriminate reference and impaired sites (Fig. 4).MMI-2 had a 76% correct classification of validation sites (25 of

P.S. Pereira et al. / Ecological Indicators 69 (2016) 606–616 611

Table 2Candidate metrics to integrate the MMIs, predicted responses to impairment, evaluation if metric range was similar for both stream types (p > 0.05, Mann-Whitney test),if metric was able to discriminate between classes of impairment (p < 0.05, Kruskal-Wallis test), if metric had a linear response to impairment and final evaluation. EPTs-BHf = Ephemeroptera, Plecoptera and Trichoptera excluding Baetidae and Hydropsychidae; MOLD = Mollusca + Diptera; IBE-IOC = Índice Biótico Estendido Instituto OswaldoCruz; BMWP-CETEC = adaptation of Biological Monitoring Working Party score system, Dominance = 1-Simpson index.

Category Metrics (codesfor valid metrics)

Predictedresponse toimpairment

Metric range issimilar for bothstream types?

Metricdiscriminatebetween classesof impairment?

Metric present alinear responseto impairment?

Final evaluation

Diversity Number of genera Decrease Yes Yes No –Number of families Decrease Yes Yes No –Number of Plecoptera genera Decrease Yes Yes No –Number of Ephemeroptera genera Decrease Yes Yes No –Number of Ephemeroptera families Decrease No – – –Number of Trichoptera genera Decrease Yes Yes No –Number of Trichoptera families (Trif) Decrease Yes Yes Yes ValidNumber of EPT families (EPTf) Decrease Yes Yes Yes ValidNumber of EPTsBH families (EPTsBHf) Decrease Yes Yes Yes ValidShannon-Wiener index families (Shanf) Decrease Yes Yes Yes ValidSimpson index genera Decrease Yes Yes No –Simpson index families Decrease Yes Yes No –Evenness index genera Decrease Yes No – –Evenness index families Decrease Yes No – –Margalef index genera (Margg) Decrease Yes Yes Yes ValidMargalef index families (Margf) Decrease Yes Yes Yes ValidEquitability index genera Decrease Yes No – –Equitability index families (Equif) Decrease Yes Yes Yes ValidDominance index families (Domf) Decrease Yes Yes Yes Valid

Composition % Coleoptera (%Coleo) Decrease Yes Yes Yes Valid% Odonata Variable Yes No – –% Ephemeroptera Decrease Yes Yes No –% Plecoptera (%Ple) Decrease Yes Yes Yes Valid% Trichoptera Decrease Yes Yes No –% EPT Decrease Yes Yes Yes Valid% EPTsBH Decrease Yes Yes Yes Valid% EPT/%Chironomidae (%EPT/%Chi) Decrease Yes Yes Yes Valid% Chironomidae Increase Yes No – –% Diptera Increase Yes Yes No –MOLD Increase Yes No – –% MOLD Increase Yes Yes No -

Tolerance IBE-IOC Decrease Yes Yes Yes ValidBMWP-CETEC Decrease Yes Yes Yes ValidHydropsychidae/Trichoptera Increase Yes No – –Chironomidae Increase Yes No – –Chironomidae/Diptera Increase Yes No – –Baetidae/Ephemeroptera Decrease Yes No – –

Trophic % Scrapers Decrease Yes No – –% Shredders Decrease Yes Yes No –% Filterers Decrease Yes No – –% Predators Variable Yes Yes No –%Collector-gatherer Variable Yes Yes No –

Table 3Stepwise forward regression between each metric and environmental variables. Bio04 – temperature seasonality (standard deviation × 100), Bio12 – annual precipitation.Residual indicate if residuals of regressions were used instead of raw metric values.

Metrics r2 P Intercept Slope Environmental Variable Status

Margf – – – – – –BMWP re 0.286 <0.05 −49.109 0.078 Bio04 ResidualEPTsBHf re 0.4058 <0.01 −9.349 0.009 Bio04 Residual%EPTsBH – – – – – –EPTf re 0.4888 <0.001 −17.280 0.013 Bio04 Residual%EPT – – – – – –Equif – – – – – –%EPT/%Chi re 0.2389 <0.05 −4.707 0.006 Bio12 Residual%Coleo re 0.3092 <0.05 −28.058 0.041 Bio12 ResidualShanf – – – – – –IBE-IOC re 0.3154 <0.01 −3.209 0.006 Bio04 Residual%Ple – – – – – –Domf – – – – – –

Margf – Margalef index in family-level; BMWP re-Residuals for natural variation of metric adaptation of the Biological Monitoring Working Party; EPTsBHf re – Residuals fornatural variation of metric number of Ephemeroptera, Plecoptera and Trichoptera excluding Baetidae and Hydropsychidae families; %EPTsBH – Percentage of Ephemeroptera,Plecoptera and Trichoptera excluding Baetidae and Hydropsychidae families; EPTf – Number of Ephemeroptera, Plecoptera and Trichoptera families; Equif-Equitabilityindex; %EPT/%Chi re – Residuals for natural variation of metric ratio of Percentage of Ephemeroptera, Plecoptera and Trichoptera and %Chironomidae; %Coleo – Percentage ofColeoptera; Shanf – Shannon-Wiener index families; IBE-IOC re – Residuals for natural variation of metric Índice Biótico Estendido Instituto Oswaldo Cruz; %Ple – Percentageof Plecoptera and Domf – Dominance index (1-Simpson).

612 P.S. Pereira et al. / Ecological Indicators 69 (2016) 606–616

Fig. 2. Cluster analysis (Pearson correlation as similarity index, Ward’s method). (a) MMI-1 – including adjusted metrics; (b) MMI-3 – only unadjusted metrics. The term“ re” refers to adjusted metrics (residual values). Margf – Margalef index in family-level; BMWP-adaptation of the Biological Monitoring Working Party; EPTsBHf – Numberof Ephemeroptera, Plecoptera and Trichoptera excluding Baetidae and Hydropsychidae families; %EPTsBH – Percentage of Ephemeroptera, Plecoptera and Trichopteraexcluding Baetidae and Hydropsychidae families; EPTf – Number of Ephemeroptera, Plecoptera and Trichoptera families; Equif – Equitability index families; %EPT/%Chi –ratio of Percentage of Ephemeroptera, Plecoptera and Trichoptera and %Chironomidae; %Coleo – Percentage of Coleoptera; Shanf-Shannon-Wiener index families; IBE-IOC –Índice Biótico Estendido Instituto Oswaldo Cruz; %Ple – Percentage of Plecoptera and Domf – Dominance index (1-Simpson).

Table 4Mean correlation (and standard deviation) between metrics and the three criteria used to evaluate the performance of the MMIs (precision, responsiveness and sensitivity).Precision (standard deviation among reference site values), responsiveness (Student’s t-value, F-value and p-value for reference and impaired site index values), sensitivity(r2 of regression between MMI and anthropogenic stressors). HAP = Habitat Assessment Protocol, NH3 = ammonia, DO = dissolved oxygen. Values in bold represent MMI withbest performance in each criterion.

MMIs Mean correlation Precision Responsiveness Sensitivity

(±std. dev.) t-score F p r2 F Variables p

MMI-1 (adjusted, cluster) 0.36 (±0.26) 14.44 11.91 1.04 <0.0001 0.89 32.64 HAP + NH3 <0.0001MMI-2 (adjusted, one of each category) 0.39 (±0.16) 13.83 14.04 1.2 <0.0001 0.92 123.3 HAP + NH3 <0.0001MMI-3 (unadjusted, cluster) 0.28 (±0.17) 10.85 13.12 1.17 <0.0001 0.91 85.1 HAP + NH3 <0.0001MMI-4 (unadjusted, one of each category) 0.34 (±0.28) 13.83 9.32 1.45 <0.0001 0.81 44.06 HAP + NH3 + DO <0.0001

Table 5Metrics selected for MMI-2. Minimum and maximum values and quartiles 25% and 75% (Q25 and Q75, respectively) of reference and impaired sites.

Metrics Minimum Maximum Q25 Reference Q75 Reference Q25 Impaired Q75 Impaired

Margf 1.466 4.73 3.0925 3.99 1.97575 2.8245EPTsBHf re −9.6843 3.7168 −0.8732 0.7461 −6.4726 −3.1836% Ple 0 20.39356 2.799944 11.12323 0 0.181848IBE-IOC re −7.2086 1.9977 −0.8107 1.0888 6.0098 4.2168

M er of Ec ral var

3cs1ttibfa

4

cH

argf – Margalef index, EPTsBHf re – residuals for natural variation of metric Numbhidae families, %Ple – Percentage of Plecoptera and IBE-IOC re – residuals for natu

3 sites). Only two out of 23 reference sites were not classifiedorrectly (one as “impaired”, one as “regular”). Six intermediateites were classified as “impaired”. Regarding the PCAs, only axis

was significant according to the broken-stick model (56.7% ofhe variance explained for the development dataset and 49% forhe validation dataset). In both cases, axis 1 displayed sites accord-ng to the impairment gradient. Significant correlations were foundetween MMI-2 scores and their corresponding PCA axis 1 valuesor both datasets (r = −0.61, p < 0.001 for the development datasetnd r = −0.59, p < 0.001 for the validation dataset).

. Discussion

Accounting for naturally occurring variation is an ideal and logi-al approach for MMIs (Stoddard et al., 2008; Hawkins et al., 2000).owever, this is a major difficulty in metric selection and MMI

PT (Ephemeroptera, Plecoptera and Trichoptera) excluding Baetidae and Hydropsy-iation of metric Índice Biótico Estendido-Instituto Oswaldo Cruz.

development (Cao et al., 2007; Buss et al., 2015). Other studiesshowed that calibrating metrics for natural gradients improvedMMI performance and decreased type I and type II errors of infer-ence (Hawkins et al., 2010; Chen et al., 2014). This was only partiallytrue in our study. MMI-3 – built only with unadjusted metrics – hadthe best precision. In addition, in our “adjusted” MMI-2 only halfthe metrics had to be adjusted because of their response to naturalvariability. Still, as shown in Table 4, MMI-2 increased its respon-siveness and sensitivity to detect impairment, which is the ultimategoal of bioassessment tools (Stoddard et al., 2008).

Other studies also reported an apparent low response ofmacroinvertebrate to natural variation. According to Stoddard et al.

(2008) calibrating metrics appears to be more important for aquaticvertebrate MMIs than for macroinvertebrate MMIs. Macedo et al.(2016) found that only one in four metrics included in their MMIneeded to be adjusted. They also found – as well as Chen et al. (2014)

P.S. Pereira et al. / Ecological Indicators 69 (2016) 606–616 613

Fig. 3. (a) Box-and-whisker plots showing values at the development of the MMIs between different classes of impact. Ref – Reference sites and Imp – Impaired sites. (b)Box-and-whisker plots showing values at the validation of the MMIs between different classes of impact. Ref – Reference sites and Int – Intermediate sites. Rectanglesdelineate the 1st and 3rd quartiles; small squares are medians; circles are extreme points; bars are maxima and minima.

F Refer2 ngles

p

asa

tweF2reeti

ig. 4. Box-and-whisker plots with MMI-2 values for the two stream types. R-Ref – and S-Imp – Impaired stream type 1. SD – Sedimentary deposits ecoregion. Rectaoints; bars are maxima and minima.

nd our results – that all indices, adjusted or unadjusted, performedignificantly well in the three criteria (precision, responsivenessnd sensitivity).

Some authors argue that selecting one metric from each clus-er (or each axis of a principal component analysis) is an effectiveay of identifying groups of statistically covarying metrics, which

nable to select non-redundant metrics for MMIs (Cao et al., 2007).ollowing this procedure improved MMI performance (Chen et al.,014). Stoddard et al. (2008) argue that redundancy between met-ics should not be avoided per se, but that it could occur because

ach metric could be responding to a single environmental param-ter that may covariate with each other in the database used forhe study. One proposed solution to circumvent that is eliminat-ng the stressor gradient from this step. They state that this way

ence stream type 2, S-Ref – Reference stream type 1, R-Imp – Impaired stream typedelineate the 1st and 3rd quartiles; small squares are medians; circles are extreme

it is possible to avoid eliminating metrics that respond to similarstressor gradients, but that reflect different taxonomic informa-tion. In our study, using only reference sites to reduce the effectof stressors, MMI-1 and MMI-3 – which used cluster analysis – didnot have lower correlations between metrics than the ones wheremetrics were selected from each category (Table 4). Also, althoughMMI-1 had no pair of metrics with correlation r > 0.70, MMI-3 hada pair with very high correlation (Domf × Shanf r = −0.91). Chenet al. (2014) have also reported high correlations (r > 0.8) betweenmetrics when using the cluster approach. They have decided to

exclude one of the correlated metrics, choosing to miss one poten-tial unique response (since this metric was in a different clusterbranch), in a way to avoid redundancy. An alternative would beselecting the second-best metric in that cluster, provided it was not

6 al Indi

rM

mcpofobmcMtcEevpcaistq

4

waoacaatcBHatSMeeBaatw

fsaldIuM

4

re

14 P.S. Pereira et al. / Ecologic

edundant with the other selected metrics, but that could decreaseMI discrimination power – as we showed in our study.We had two hypotheses in our study: (1) testing and adjusting

etrics to landscape parameters and (2) selecting metrics using aluster analysis to avoid metrics redundancy, would increase MMIerformance. Our results do not support it. In fact, MMI-1 did notutperform other MMIs in any criteria. However, the two best per-ormance MMIs used one of this approaches each: MMI-2, with twout of four adjusted metrics, one selected from each category, hadest responsiveness and sensitivity; MMI-3, only with unadjustedetrics using a cluster approach for metric selection, had lower

orrelation among metrics and better precision. We believe thatMI’s development should benefit from incorporating responses

o natural variation, and also by using a system less subjective thanhoosing one from each category. For example, many metrics (e.g.PT richness) could be classified either as a “richness” or as a “tol-rance” metric. We support that further studies are necessary toalidate those methods. In addition, we notice that some recentublications failed to report the full set of information to allow MMIalculation. We believe these tools should be made readily avail-ble for managers and we urge studies to provide the following:ntercept and slope for the residuals of each environmental variableignificantly associated with the selected metrics; the transforma-ion used for each metric to reach normalization (if any); and theuartiles 25% and 75% of reference and impaired sites.

.1. Considerations on selected metrics

EPT-derived metrics had been used in several MMIs around theorld (e.g. Helson and Williams, 2013; Lakew and Moog, 2015),

nd also in Brazil (e.g. Ferreira et al., 2011; Couceiro et al., 2012). Inur study, two metrics were selected representing this group: %Plend the number of EPT families, excluding Baetidae and Hydropsy-hidae (EPTsBHf). Plecoptera is often classified as sensitive taxa,nd have been used in many MMIs around the world (Rosenbergnd Resh, 1993; Törnblom et al., 2011). Recent studies indicatehis group is also a good surrogate for the assessment of climatehange effects (Tierno de Figueroa et al., 2010). The metric EPTs-Hf performed better than EPT richness because Baetidae andydropsychidae had higher relative abundance in intermediatelynd impaired sites (46% and 69%, respectively), thus decreasingheir discriminating power. Among Hydropsychidae, the genusmicridea was responsible for 98% abundance of the family. OtherMIs around the world report that (Barbour et al., 1999; Bellucci

t al., 2013). This is also supported by other studies in Brazil. Forxample, Buss and Salles (2007) described the habitat and habits ofaetidae species, and classified the genus Americabaetis as “toler-nt” and the genera Baetodes, Camelobaetidius and Cryptonymphas “somewhat sensitive”. We recommend other studies to considerhis metric for screening while developing MMIs in Brazil and else-here.

The other selected metrics for MMI-2 were the Margalef index inamily level and the IBE-IOC index. Other recent multimetric indiceselected the Margalef index (Nguyen et al., 2014), including Helsonnd Williams (2013) that developed a multimetric index in low-and streams in Panama, an ecological condition similar to the oneescribed in the present study. The better performance of the IBE-

OC is likely to have occurred because the index was developedsing information from a similar study area (Rio de Janeiro state;ugnai et al., 2008).

.2. Further considerations for MMIs development

No trophic metrics were retained for our final MMI. Trophic met-ics are often used in MMIs (e.g. Stoddard et al., 2008; Couceirot al., 2012) because they can serve as a proxy for ecosystem func-

cators 69 (2016) 606–616

tioning (Petchey and Gaston, 2006). Two metrics were discardedbecause they did not discriminate the three impairment conditions(%Scrapers and %Filterers) and the three other metrics (%Shred-ders, %Predators and %Gatherers-collector) were excluded becausethey did not respond linearly to the impairment gradient. Moyaet al. (2011), studying Bolivian streams, also reported that trophicmeasures were not approved while screening for a MMI. In theneotropical region only a few studies aimed to determine macroin-vertebrate functional feeding groups (FFGs) either by analyzing gutcontents (Tomanova et al., 2006; Miserendino, 2007; Chará-Sernaet al., 2012) or ultra-structure of mouthparts (Baptista et al., 2007).As a result, most MMIs developed in the region simply apply theFFG attributed to a taxon based on temperate species, which maylead to wrong classifications. Some studies in neotropical regionsuggest most genera have omnivorous feeding habits (Tomanovaet al., 2006), adding difficulties to correctly applying this approach.Furthermore, in many cases MMIs are calculated in family-level,which make FFG assignment increasingly prone to errors consid-ering there is considerable within-family variation in the feedinghabits of macroinvertebrate species in neotropical streams (Moyaet al., 2007). Another limitation for using these metrics is that sincethey express the energy flow through the compartments of the foodchain and the ecosystem, it is necessary to evaluate the biomass ofeach FFG, and not their relative abundance (a common approachfor calculating other metrics). Using data on relative abundanceas a proxy clearly change the relative importance of FFGs. Forexample, Masese et al. (2014) showed that shredder relative abun-dance in closed-canopy forested streams was ∼20%, but the samegroups yielded corresponded to ∼80% of total macroinvertebratewet mass in Kenya. Similarly, Chara et al. (2007) found that shred-ders represented 13% of the abundance and 68% of the biomassof invertebrates colonizing leaf litter bags in Andean Colombianstreams. The easiest method for calculating the biomass is usinglenth-biomass regressions based on total length, head width orinterocular distance (Benke et al., 1999). That requires a previouseffort to build the regressions and then measure organisms individ-ually to convert length to biomass. The use of FFG biomass may beused as a tool for assessing ecosystem functioning. However, thereare few length-biomass regression curves available for neotropi-cal macroinvertebrates, not covering most groups (Cressa, 1999;in Venezuela; Miserendino, 2001; in Argentina; Chará-Serna et al.,2012, for leaf litter fauna in Colombia; and Becker et al., 2009; onlyfor the shredder genus Phylloicus – family Calamoceratidae, orderTrichoptera – in Brazil). More studies are necessary in this region,which will allow not only the application of FFG-based metrics inbiomonitoring programs, but also on studies of ecosystem func-tioning and to provide information on the supposed scarcity ofshredders in the tropics (Masese et al., 2014; Boyero et al., 2011).

Substrate availability is one of the main drivers for the colo-nization and the structure of macroinvertebrate assemblages instreams. Several papers have demonstrated the importance of thesubstrate for macroinvertebrate assemblages (Jiang et al., 2010;Demars et al., 2012). In Brazil, Buss et al. (2004) reported thattaxa occurrence was highly dependent on substrate type (82 outof 86 taxa had a >40% preference for one substrate), and althoughdegradation influence significantly the macroinvertebrate fauna oneach substrate, they were responding to different physicochemicalparameters. Many biomonitoring programs recognize the impor-tance of substrate types for macroinvertebrates, and recommenda multi-habitat sampling approach, usually in proportion to theiroccurrence (e.g. Barbour et al., 1999; AQEM, 2002). In New Zealand,Stark et al. (2001) showed significant differences in the structure

and composition of the community of macroinvertebrates betweenhard- and soft-bottom stream types. They reported the latter ashaving less productive habitats and lower diversity and numberof invertebrates. In our study, however, we did not find signifi-

al Indi

ctsi

5

ra(sarais

A

ma

R

A

A

A

B

B

B

B

B

B

B

B

B

B

B

B

P.S. Pereira et al. / Ecologic

ant differences for any metric values between reference sites ofhe two stream types (“transitional/sedimentary areas” and “rockyubstrates”), and reference sites were significantly different thanmpaired sites, for both stream types.

. Conclusions

Although all MMIs performed well in all three decision crite-ia, adjusting metrics to natural variation increased MMI responsend sensitivity to impairment. In addition, the selected MMI-2adjusted metrics, one representing each category) was able to clas-ify sites of two stream types and two ecoregions. The use of clusternalysis, however, did not avoid high redundancy between met-ics of different branches. The MMI developed using the traditionalpproach had the poorest performance among all tested MMIs andt was not able to distinguish adequately reference and impairedites from ecoregion 2.

cknowledgments

This work was supported by Conselho Nacional de Desenvolvi-ento Científico e Tecnológico (CNPq/PROEP 400107/2011-2). We

re grateful to LAPSA team members for field assistance.

eferences

lvares, C.A., Stape, J.L., Sentelhas, P.C., Gonc alves, J.L.M., Sparovek, G., 2013.Köppen’s climate classification map for Brazil. Meteorol. Z. 22 (6), 711–728.

PHA, AWWA, WPCF, 2000. Standard methods for the examination of water andwastewater, 20th ed. American Public Health Association/American. WaterWorks Association/Water Environment Federation. Washington, DC.

QEM Consortium, 2002. Manual for the application of the AQEM system. Acomprehensive method to assess European streams using benthicmacroinvertebrates, developed stream macroinvertebrates. J. N. Am. Benthol.Soc. 25, 977–997.

aptista, D.F., Buss, D.F., Egler, M., Giovanelli, A., Silveira, M.P., Nessimian, J.L., 2007.A multimetric index based on benthic macroinvertebrates for evaluation ofAtlantic Forest streams at Rio de Janeiro state, Brazil. Hydrobiologia 575, 83–94.

arbour, M.T., Gerritsen, J., Griffith, G.E., Frydenborg, R., McCarron, E., White, J.S.,Bastian, M.L., 1996. A framework for biological criteria for Florida streamsusing benthic macroinvertebrates. J. N. Am. Benthol. Soc. 15, 185–211.

arbour, M.T., Gerritsen, J., Snyder, B.D., Stribling, J.B., 1999. Rapid BioassessmentProtocols for use in streams and wadeable rivers: Periphyton, BenthicMacroinvertebrates and Fish, 3rd ed. Washington: U.S. EnvironmentalProtection Agency; Office of Water, EPA 841–B–99–002.

ecker, B., Moretti, M.S., Callisto, M., 2009. Length-dry mass relationships for atypical shredder in Brazilian streams (Trichoptera: calamoceratidae). Aquat.Insects 31 (3), 227–234.

ellucci, C.J., Becker, M.E., Beauchane, M., Dunbar, L., 2013. Classifying the health ofConnecticut streams using benthic macroinvertebrates with implications forwater management. Environ. Manag. 51, 1274–1283.

enke, A., Huryn, A., Smock, L., Wallace, J., 1999. Length-mass relationships forfreshwater macroinvertebrates in North America with particular reference tothe southeastern United States. J. N. Am. Benthol. Soc. 18 (3), 308–343.

locksom, K.A., 2003. A performance comparison of metric scoring methods for amultimetric index for Mid-Atlantic highlands streams. Environ. Manag. 31 (5),416–432.

oyero, L., Pearson, R.G., Dudgeon, D., Grac a, M.A.S., Gessner, M.O., Albarino, R.J.,Ferreira, V., Yule, C.M., Boulton, A.J., Arunachalam, M., Callisto, M., Chauvet, E.,Ramírez, A., Chará, J., Moretti, M.S., Gonc alves, J.F., Helson, J.E., Chará- Serna,A.M., Encalada, A.C., Davies, J.N., Lamothe, S., Cornejo, A., Castela, J., Li, A.O.Y.,Buria, L.M., Villanueva, V.D., Zúniga, M.C., Pringle, C.M., 2011. Globaldistribution of a key trophic guild contrasts with common latitudinal diversitypatterns. Ecology 92, 1839–1848.

rasil. Ministério das Minas e Energia. Secretaria Geral., 1983. ProjetoRADAMBRASIL: Folha SD. 23. Rio de Janeiro, (Levantamento de RecursosNaturais, v. 29), pp. 660.

uss, D.F., Salles, F.F., 2007. Using baetidae species as biological indicators ofenvironmental degradation in a Brazilian River Basin. Environ. Monit. Assess.130, 365–372.

uss, D.F., Baptista, D.F., Nessimian, J.L., Mariana, M., 2004. Substrate specificity:environmental degradation and disturbance structuring macroinvertebrate

assemblages in neotropical streams. Hydrobiologia 518, 179–188.

uss, D.F., Carlisle, D.M., Chon, T.S., Culp, J., Harding, J.S., Keizer-Vlek, H.E.,Robinson, W.A., Strachan, S., Thirion, C., Hughes, R.M., 2015. Streambiomonitoring using macroinvertebrates around the globe: a comparison oflarge scale programs. Environ. Monit. Assess. 187, 1–21.

cators 69 (2016) 606–616 615

Cao, Y., Hawkins, C.P., Olson, J.R., Kosterman, M.A., 2007. Modeling naturalenvironmental gradients improves the accuracy and precision of diatom-basedindicators. J. N. Am. Benthol. Soc. 26, 566–585.

Carter, J.L., Resh, V.H., 2001. After site selection and before data analysis: sampling,sorting, and laboratory procedures used in stream benthic macroinvertebratemonitoring programs by USA state agencies. J. N. Am. Benthol. Soc. 20,658–682.

Chará-Serna, A.M., Chará, J.D., Zúniga, M.D.C., Pearson, R.G., Boyero, L., 2012. Dietsof leaf litter-associated invertebrates in three tropical streams. Int. J. Limnol.48, 139–144.

Chara, J., Baird, D., Telfer, T., Giraldo, L., 2007. A comparative study of leafbreakdown of three native tree species in a slowly-flowing headwater streamin the Colombian Andes. Int. Ver. Hydrobiol. 92 (2), 183–198.

Chen, K., Hughes, R.M., Xu, S., Zhang, J., Cai, D., Wang, B., 2014. Evaluatingperformance of macroinvertebrate-based adjusted and unadjustedmulti-metric indices (MMI) using multi-season and multi-year samples. Ecol.Indic. 36, 142–151.

Costa, S.S., Melo, A.S., 2008. Beta diversity in stream macroinvertebrateassemblages: among-site and among microhabitat components. Hydrobiologia598, 131–138.

Couceiro, S.R.M., Hamada, N.B.R., Pimentel, T.P., Luz, S.L.B., 2012. Amacroinvertebrate multimetric index to evaluate the biological condition ofstreams in the central Amazon region of Brazil. Ecol. Indic. 18, 118–125.

Cressa, C., 1999. Dry mass estimates of some tropical aquatic insects. Rev. Biol.Trop. 47 (1–2), 133–141.

Demars, B.O., Kemp, J.L., Friberg, N., Usseglio-Polatera, P., Harper, D.M., 2012.Linking biotopes to invertebrates in rivers: biological traits: taxonomiccomposition and diversity. Ecol. Indic. 23, 301–311.

Ferreira, W.R., Paiva, L.T., Callisto, M., 2011. Development of a benthic multimetricindex for biomonitoring of a neotropical watershed. Braz. J. Biol. 71 (1), 15–25.

Gerth, W.J., Herlihy, A.T., 2006. Effect of sampling different habitat types in regionalmacroinvertebrate bioassessment surveys. J. N. Am. Benthol. Soc. 25, 501–512.

Hawkins, C.P., Norris, R.H., Gerritsen, J., Hughes, R.M., Jackson, S.K., Johnson, R.K.,Stevenson, R.J., 2000. Evaluation of the use of landscape classifications for theprediction of freshwater biota: synthesis and recommendations. J. N. Am.Benthol. Soc. 19 (3), 541–556.

Hawkins, C.P., Cao, Y., Roper, B., 2010. Method of predicting reference conditionbiota affects the performance and interpretation of ecological indices. Freshw.Biol. 55 (5), 1066–1085.

Helson, J.E., Williams, D.D., 2013. Development of a macroinvertebrate multimetricindex for the assessment of low-land streams in the neotropics. Ecol. Indic. 29,167–178.

Jiang, X.M., Xiong, J., Qiu, J.W., Wu, J.M., Wang, J.W., Xie, Z.C., 2010. Structure ofmacroinvertebrate communities in relation to environmental variables in asubtropical Asian river system. Int. Rev. Hydrobiol. 95, 42–57.

Jun, Y.C., Won, D.H., Lee, S.H., Kong, D.S., Hwang, J.S., 2012. A multimetric benthicmacroinvertebrate index for the assessment of stream biotic integrity in Korea.Int. J. Environ. Res. Publ. Health 9, 3599–3628.

Junqueira, V.M., Campos, S.C.M., 1998. Adaptation of the BMWP for water qualityevaluation to Rio das Velhas watershed (Minas Gerais, Brazil). Acta Limnol.Bras. 10 (2), 125–135.

Kong, W.J., Meng, W., Zhang, Y., Christopher, G., Qu, X.D., 2013. A freshwaterecoregion delineation approach based on freshwater macroinvertebratecommunity features and spatial environmental data in Taizi River Basinnortheastern China. Ecol. Res. 28, 581–592.

Lakew, A., Moog, O., 2015. A multimetric index based on benthicmacroinvertebrates for assessing the ecological status of streams and rivers incentral and southeast highlands of thiopia. Hydrobiologia 751 (1), 229–242.

Ligeiro, R., Hughes, R.M., Kaufmann, P.R., Macedo, D.R., Firmiano, K.R., Ferreira,W.R., 2013. Defining quantitative stream disturbance gradients and theadditive role of habitat variation to explain macroinvertebrate taxa richness.Ecol. Indic. 25, 45–57.

Macedo, D.R., Hughes, R.M., Ferreira, W.R., Firmiano, K.R., Silva, D.R., Ligeiro, R.,Callisto, M., 2016. Development of a benthic macroinvertebrate multimetricindex (MMI) for Neotropical Savanna headwater streams. Ecol. Indic. 64,132–141.

Masese, F.O., Kitaka, N., Kipkemboi, J., Gettel, G.M., Irvine, K., McClain, M.E., 2014.Macroinvertebrate functional feeding groups in Kenyan highland streams:evidence for a diverse shredder guild. Freshw. Sci. 33 (2), 435–450.

Mereta, S.T., Boets, P., Meester, L.D., Goethals, P.L.M., 2013. Development of amultimetric index based on benthic macroinvertebrates for the assessment ofnatural wetlands in Southwest Ethiopia. Ecol. Indic. 29, 510–521.

Miserendino, M.L., 2001. Macroinvertebrate assemblages in Andean Patagonianrivers and streams. Hydrobiologia 444, 147–158.

Miserendino, M.L., 2007. Macroinvertebrate functional organization and waterquality in a large arid river from Patagonia (Argentina). Int. J. Limnol. 43 (3),133–145.

Moya, N., Tomanova, S., Oberdorff, T., 2007. Index based on aquaticmacroinvertebrates to assess streams condition in the Upper Isiboro-SécureBasin, Bolivian Amazon. Hydrobiologia 589 (1), 107–116.

Moya, N., Hughes, R.M., Dominguez, E., Gibon, F., Goittia, E., Oberdorff, T., 2011.

Macroinvertebrate-based multimetric predictive models for evaluating thehuman impact on biotic condition of Bolivian streams. Ecol. Indic. 11, 840–847.

Mugnai, R., Oliveira, R.B.S.O., Carvalho, A.L., Baptista, D.F., 2008. Adaptation of theIndice Biotico Esteso (IBE) for water quality assessment in rivers of Serra doMar Rio de Janeiro state, Brazil. Trop. Zool. 21, 57–74.

6 al Indi

N

O

O

P

R

R

R

16 P.S. Pereira et al. / Ecologic

guyen, H.H., Everaert, G., Gabriels, W., Hoang, T.H., Goethals, P.L.M., 2014. Amultimetric macroinvertebrate index for assessing the water quality of the Cauriver basin in Vietnam. Limnologica 45, 16–23.

liveira, R.B., Mugnai, R., Castro, C.M., Baptista, D.F., 2011. Determiningsubsampling effort for the development of a rapid bioassessment protocolusing benthic macroinvertebrates in streams of Southeastern Brazil. Environ.Monit. Assess. 175, 75–85.

mernik, J.M., 1995. Ecoregions: a spatial framework for environmentalmanagement. In: Davis, W.S., Simon, T.P. (Eds.), Biological Assessment andCriteria-tools for Water Resource Planning and Decision Making. CRC Press,Boca Raton, Florida, pp. 49–62.

etchey, O.L., Gaston, K.J., 2006. Functional diversity: back to basics and lookingforward. Ecol. Lett. 9 (6), 741–758.

Core Team, 2015. R: A Language and Environment for Statistical Computing. RFoundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/.

eynoldson, T.B., Norris, R.H., Resh, V.H., Day, K.E., Rosenberg, D.M., 1997. Thereference condition approach: a comparison of multimetric and multivariate

approaches to assess water quality impairment using benthicmacroinvertebrates. J. N. Am. Benthol. Soc. 16, 833–852.

ibeiro, M.C., Martensen, A.C., Metzger, J.P., Tabarelli, M., Scarano, F., Fortin, M.J.,2011. The Brazilian Atlantic Forest: a shrinking biodiversity hotspot. In:Zachos, F.E., Habel, J.C. (Eds.), Biodiversity Hotspots: Distribution and

cators 69 (2016) 606–616

Protection of Conservation Priority Areas. Springer, Heidelberg, Berlin,pp. 405–434.

Rosenberg, D.M., Resh, V.H., (Org.), 1993. Freshwater Biomonitoring and BenthicMacroinvertebrates. Chapman & Hall, New York (NY), 488 p.

Stark, J.D., Boothroyd, I.K.G, Harding, J.S., Maxted, J.R., Scarsbrook, M.R., 2001.Protocols for sampling macroinvertebrates in wadeable streams. New ZealandMacroinvertebrate Working Group Report No. 1. Prepared for the Ministry forthe Environment. Sustainable Management Fund Project No. 5103, pp. 57.

Stoddard, J.L., Larsen, D.P., Hawkins, C.P., Johnson, R.K., Norris, R.H., 2006. Settingexpectations for the ecological condition of streams: the concept for referencecondition. Ecol. Appl. 16 (4), 1267–1276.

Stoddard, J.L., Herlihy, A.T., Peck, D.V., Hughes, R.M., Whittier, T.R., Tarquinio, E.,2008. A process for creating multimetric indices for large-scale aquaticsurveys. J. N. Am. Benthol. Soc. 27, 878–891.

Törnblom, J., Degerman, E., Angelstam, P., 2011. Forest proportion as indicator ofecological integrity in streams using Plecoptera as a proxy. Ecol. Indic. 11 (5),1366–1374.

Tierno de Figueroa, J.M., López-Rodríguez, M.J., Lorenz, A., Graf, W.,

Schmidt-Kloiber, A., 2010. Vulnerable taxa of European Plecoptera (Insecta) inthe context of climate change. Biodivers. Conserv. 19, 1269–1277.

Tomanova, S., Goittia, E., Helezic, J., 2006. Trophic levels and functional feedinggroups of macroinvertebrates in neotropical streams. Hydrobiologia 556,251–264.