Assessment and analysis of ecological quality, macroinvertebrate communities and diversity in rivers of a multifunctional tropical island

Ecological Indicators 77 (2017) 228–238

Contents lists available at ScienceDirect

Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind

Assessment and analysis of ecological quality, macroinvertebrate communities and diversity in rivers of a multifunctional tropical island Marie Anne Eurie Forio a,∗ , Koen Lock a , Eve Daphne Radam a , Marlito Bande b , Victor Asio b,c , Peter L.M. Goethals a a

Laboratory of Environmental Toxicology and Aquatic Ecology, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium Institute of Tropical Ecology and Environmental Management, Visayas State University, 6521 Baybay, Leyte, Philippines c Department of Soil Science, Visayas State University, 6521 Baybay, Leyte, Philippines b

a r t i c l e

i n f o

Article history: Received 23 February 2016 Received in revised form 8 February 2017 Accepted 10 February 2017 Keywords: BMWP-Viet Shannon-Wiener diversity index Margalef index Stream ecology Leyte Philippines Lotic ecosystem

a b s t r a c t This study aims to assess the macroinvertebrate communities, diversity and ecological quality (expressed as BMWP-Viet) of rivers on a tropical island. Environmental factors associated with macroinvertebrate communities, diversity and ecological quality were identified to assist conservation planning and management. Biological (macroinvertebrates), chemical, physical and hydromorphological characteristics of 85 river reaches were assessed in Leyte island, Philippines. Canonical Correspondence Analysis (CCA) and multivariable linear regression (LRM) were performed to relate environmental variables and macroinvertebrates. Several taxa of snails, shrimps, dragonflies, beetles, bugs and caddisflies were found on the island. Although many sites had good to very good ecological quality and high diversity, about 41% had moderate to very bad ecological quality and low diversity. Based on CCA, we can conclude that macroinvertebrate communities were associated with velocity, sediment, conductivity and dissolved oxygen. Particularly, sensitive and tolerant taxa were encountered at high and low flow velocities, respectively. LRM indicated that macroinvertebrate diversity and ecological quality were associated with physical (turbidity), chemical (chlorophyll), hydromorphological characteristics (bank slope & pool/riffle class), habitat degradation (gravel/sand quarrying, erosion) and the presence of logs and twigs. Consequently, this study gives support to the use of invertebrates as indicators of certain environmental conditions and the results of this investigation can serve as a basis to set up dedicated experiments to further prove the causality of these discovered relations. Strikingly, organic pollution, as reflected by biological oxygen demand and chemical oxygen demand, was in general weakly related to invertebrate composition, diversity and ecological quality. This might be linked to the low input in most sites and the relatively short rivers which are closely connected to the marine system. Thus, typical midstream and downstream systems were not encountered and the accumulation of these pollutants along the river is less likely. Although the island is situated in the western Pacific Ocean and encounters intensive natural disturbances (severe typhoons), the taxa (families) are similar to other tropical systems and the effects of the environmental conditions are comparable. Findings of this study are valuable in understanding tropical island systems and provide insights into the effects of environmental conditions on stream invertebrates, which aids in protecting and conserving tropical insular systems. © 2017 Elsevier Ltd. All rights reserved.

1. Introduction Globally, biodiverse ecosystems are severely threatened due to rapid population growth and increasing anthropogenic activities

∗ Corresponding author. E-mail addresses: anne [email protected], [email protected] (M.A.E. Forio). http://dx.doi.org/10.1016/j.ecolind.2017.02.013 1470-160X/© 2017 Elsevier Ltd. All rights reserved.

(Hill et al., 2015). Conserving and protecting a biodiverse ecosystem is essential as biodiversity is connected to the functioning of ecosystems and is therefore related to the society (Cardinale, 2012). Biodiversity supports ecosystem services (Francis et al., 2014). For instance, a biodiverse forest supplies pollination services to agriculture, which increase crop yield and quality (Ricketts et al., 2004). Thus, a biodiverse ecosystem is more productive as they comprise key species that largely influence productivity (Cardinale, 2012;

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

Helm and Hepburn, 2012). Biodiversity is also strongly associated with certain regulating services (Cardinale, 2012). Creating a biodiverse set of habitats in non-production agricultural areas increases multiple ecosystem functions (e.g. water regulation and carbon storage) of agricultural lands (Smukler et al., 2010). Lastly, biodiverse natural environments stimulate better health and well-being (Lovell et al., 2014). Biodiversity hotspots are defined as areas featuring exceptional concentrations of endemic species (plant and vertebrates) which experience exceptional loss of habitat (Myers et al., 2000). These hotspots are identified as priorities for protection, conservation and management. The Philippines, composed of 7100 islands, is identified as a biodiversity hotspot and one of the world’s biologically richest countries (Myers et al., 2000; Proches et al., 2015). Endemic invertebrates are numerous in the country. However, information on spatial distribution of invertebrate communities in the Philippines is limited (David, 2003; Dyer et al., 2003; Freitag, 2004). Aquatic invertebrates inhabit different types of aquatic systems. Most of them live part or most of their life cycle attached to submerged rocks, logs and vegetation (Gabriels et al., 2010). They are critical in the stream’s food web (USEPA, 1997). Thus, macroinvertebrates may be crucial in the overall biodiversity as the extinction of one species within a food web can result in secondary extinctions due to bottom-up effects (Calizza et al., 2015). They are also commonly used in environmental monitoring and assessment (Lock and Goethals, 2014; Resh and Rosenberg, 2010) because they reflect the quality of an aquatic system due to their varying degrees of tolerances towards disturbances (De Pauw et al., 2001; Morse et al., 2007) and integrate environmental stresses that have occurred over an extended period (Rosenberg and Resh, 1996). They reveal not only the cumulative impacts of pollution but also the impacts from habitat loss not detected by traditional water quality assessments (USEPA, 1997). Although tropical systems are reported to support greater species diversity than temperate systems (Myers et al., 2000; Pianka, 1966; Proches et al., 2015), knowledge of tropical stream invertebrates is generally limited (Jardine, 2014). In tropical island systems, it is believed that macroinvertebrate faunas are sparse (Bass, 2003). However, according to Covich (2006), diversity of species in a tropical island may depend on the age, location, height and size of an island. Anthropogenic activities are increasingly growing in tropical islands (Fomba et al., 2013; Kura et al., 2015; Somboonna et al., 2014). Investigation of the impacts of anthropogenic activities on the ecology and biodiversity in tropical island systems is limited (Ramirez et al., 2009). Janse et al. (2015) reported that there is a negative relation between biodiversity and environmental stressors (land use changes, climate change, nutrient emissions) in all types of freshwater ecosystems. Thus, knowledge of the impact of human activities on stream invertebrates in the tropical islands facilitates conservation planning and management of these systems (Covich, 2006). The framework and objectives of this study are reflected in Fig. 1. We aim to assess the macroinvertebrate communities based on two indices (BMWP-Viet and diversity) and use this information as a proxy for the river quality status. For this, biological (macroinvertebrates), chemical, physical and hydromorphological characteristics of river reaches were monitored in Leyte island, Philippines. Subsequently, Canonical Correspondence Analysis (CCA) and multivariable linear regression (LRM) were performed to relate environmental variables and macroinvertebrates (phase 1 in Fig. 1). Although the results do not guarantee a proof for causality, they can contribute to determine key variables and indicators that serve as a basis to identify relevant management and policy actions, such as water quality standards, habitat restoration

229

Fig. 1. The applied framework for water restoration and management, of which phase 1 was practically implemented in this study.

and regulations related to sand quarrying. Cause deduction (phase 2 in Fig. 1), as indicated in the framework, is not covered in our study since this would require additional experiments in labs or in (artificial/actual) rivers. 2. Material and methods 2.1. River quality assessment 2.1.1. Study area The Leyte island is the eighth largest island in the Philippines and has a surface area of 7368 km2 (Fig. 2). The island is irregular in shape and has mountains in the centre. The highest mountain reaches 1,349 m. A complex system of short streams drains from the mountains to the coasts. Plains are found in the coastal areas, mainly in the north (Pletcher, 2015). The climate of the island is characterized by a relatively high temperature (24–33 ◦ C), a high humidity and abundant rainfall. Average annual rainfall is 2100–4500 mm. Typhoons occur every year, usually in the period of June-December. Human activities within the island include crop cultivation, industry, quarrying, urbanization and aquaculture. Rice and coconut are the main crops. Other crops include corn (maize), abaca, tobacco, bananas, pineapple and sugarcane. Manganese deposits, sandstone and limestone are quarried in the northwest. Coconut oil mills, a copper smelting plant, a phosphate fertilizer and ethyl alcohol production plants and a geothermal production field are located on the island. 2.1.2. Data collection Aquatic macroinvertebrates were sampled on the island at 85 different locations with varying degrees of disturbances. The sites were selected to ensure safety and accessibility of all locations. The sampling campaign was conducted during the dry season (April–May 2015). Macroinvertebrates were monitored through kick sampling with a standard handnet (conical net with a frame size of 20 × 30 cm and a mesh size of 500 ␮m, attached to a stick)

230

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

Fig. 2. The sampling sites on Leyte island.

as prescribed by Gabriels et al. (2010). For each sampling site, a 10–20 m stretch was sampled during 5 min. Sampling effort was proportionally distributed across all aquatic habitats present at the sampling site, including bed substrates (stones, sand or mud), macrophytes (floating, submerged, emerging) and other floating or submerged natural and artificial substrates. All the collected material was transferred to buckets with covers. Afterwards, samples were sieved and organisms were sorted alive in the laboratory. Macroinvertebrates were identified to the family level. For each sampling site, physical-chemical water quality characteristics were measured (Table 1). Additionally, elevation was measured. Hydromorphological characteristics (Table 1) and site descriptions, such as river morphology, water flow, the presence or absence of macrophytes, substrate characteristics, main land use, shading, presence of macroalgae (i.e. filamentous algae), presence of twigs, branches and logs, presence of carabao (domesticated water buffalos) and presence of gravel/sand quarrying were determined through field inspection. For a brief definition of each hydromorphological variables and site description, we refer to SI Table 1.

2.1.3. Ecological indices The ecological quality of each site was assessed by Biological Monitoring Working Party adapted for Vietnam (BMWP-Viet) (Nguyen et al., 2004). BMWP index is globally known in ecological quality assessment. We used BMWP-Viet index because Vietnam is close to the Philippines and climatic conditions between countries and therefore also macroinvertebrate communities are similar. Representative macroinvertebrate taxa have assigned sensitivity scores; the higher is the sensitivity score, the more sensitive is the taxa towards disturbance. The sensitivity scores of the taxa present at each site were summed. The higher the score, the better is the ecological quality. BMWP-Viet was divided into five quality classes: very good (>100 score), good (71–100), moderate (41–70), poor (11–40) and bad (<11). The diversity of macroinvertebrates was assessed by family richness, Margalef diversity index and Shannon-Wiener diversity

index. Family richness was calculated by the number of different families at each site. Margalef diversity index (d) (Margalef, 1958) is often used to measure taxa richness which can be calculated by d=

S−1 lnN

where S is the number of taxa, and N is the total number of individuals in the sample. The Shannon-Wiener diversity index (H) (Shannon and Weaver, 1949) is a diversity index commonly used to characterize species diversity in a community, which accounts for both abundance and evenness dimensions of diversity. It is calculated by s 

H=−

Pi ∗ lnPi

i=s

where Pi represents the relative abundance of the ith taxon in the sample, s is the total number of taxa in the sample. The index identifies major changes in community structure of taxa (Pettersson, 1998).The higher the calculated value, the more diverse is the given site. 2.2. Determination of key environmental variables Prior to the analysis, total N was removed, because most of the data were below detection limit. All records in BDL (below detection limit) were set to the detection limit. The data were explored as described in Zuur et al. (2010). Subsequently, the data were divided into two groups. The first and second set comprises of sites with an elevation lower than 30 m a.s.l. (39 sites) and sites with an elevation higher than 30 m a.s.l. (46 sites), respectively. The first set contained all the sites with brackish waters. 2.2.1. Canonical correspondence analysis To test whether a linear method or unimodal method was needed, a Detrended Correspondence Analysis (DCA) was performed on each set of data. If the Length of Gradient (LoG) is higher

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

231

Table 1 All measured chemical, physical and hydromorphological variables considered in the study. Chemical variables (units)

Measuring Device

Physical variables (units)

Measuring device

Hydromorphological variables

Specific conductivity (␮S/cm) Total dissolved solids (g/L) Water pH Dissolved oxygen (mg/L) Dissolved oxygen saturation (%) Chlorophyll a concentration (␮g/L) Biological oxygen demand (mg/L) Chemical oxygen demand (mg/L) Orthophosphate-P (mg/L) Total phosphorus (mg/L) Nitrate-N (mg/L) Nitrite-N (mg/L) Ammonium-N (mg/L) Total Nitrogen (mg/L)

Multiprobea Multiprobea Multiprobea Multiprobea Multiprobea Multiprobeb Spectrophotometerc Spectrophotometerc Spectrophotometerc Spectrophotometerc Spectrophotometerc Spectrophotometerc Spectrophotometerc Spectrophotometerc

Water temperature (◦ C) Average flow velocity (m/s) Turbidity (NTU)

Multiprobea Current meterd Multiprobeb

Valley form Channel form Variation in width Extent of erosion Bank profile Variation of flow Depth of sludge layer Pool/Riffle class Bank shape Bank slope Bed compaction Bank material Sediment matrix Sediment angularity Main sediment type Average stream depth Average stream width Floodplain width Flood prone width Maximum depth Entrenchment depth

a b c d

model Three-Multi 3430 IDS, WTW GmbH. model YSI 6600 V2, YSI manufacturer. Hach Lange GmbH spectrophotometric method. model höntzsch HFA, Höntzsch GmbH manufacturer.

than 3, a unimodal method is needed (i.e. Canonical Correspondence Analysis), whereas if the LoG is smaller than 3, a linear method is designated (i.e. Redundancy Analysis) (Ter Braak and Prentice, 2004). As the LoG was higher than 3, a Canonical Correspondence Analysis (CCA) was implemented on each set of data. CCA (Ter Braak and Verdonschot, 1995) was performed to explore the relationships between macroinvertebrate communities and the environmental conditions and to identify the corresponding patterns in the composition of the macroinvertebrate. Initially, there were 40 environmental variables for analysis. To minimize the number of variables prior to analysis, one of each pair of correlated continuous/ordinal categorical variables was removed. Correlation analysis was performed with the Spearman rank test at p = 0.01 (SI Table 2). As the remaining variables depended on which variables were removed, two sets of CCA analysis were performed (SI Table 3). Rare taxa (occurred in less than 5% of the samples, SI Table 4) were also removed prior to analysis to avoid skewedness of the CCA plots. The macroinvertebrate abundances and all variables were log transformed. All analysis were performed in the ade4 package in R (Dray and Dufour, 2007).

2.2.2. Multivariable linear regression models Based on data exploration, one observation was believed to be an outlier and was removed. Thus, 84 observations were left for analysis. Collinear variables were removed based on variance inflation factor (VIF), wherein variables with VIF values higher than three were regarded as collinear. The variable with the highest VIF value was removed. Subsequently, the VIF values of the remaining variables were recalculated. The process was repeated until all VIF values were less than three (SI Table 5). This process resulted in 21 variables that were included in further analysis. Multivariable linear regression (LRM) was used to determine significant environmental variables related with BMWP-Viet, Margalef diversity index and Shannon-Wiener diversity index in each data set. A stepbackward selection procedure was employed. The method starts from a model which includes all non-collinear variables and then reduces this model by removing variables one by one. A Drop1 function in R was implemented. Each individual explanatory variable

was dropped one by one based on the hypothesis testing procedures (lowest F-test or highest p-value). Each time, the model was refitted. The process was repeated until the p-values of the variables were lower than 0.05. This implies that any variable contained in the last model cannot be further dropped. When the optimal model was found, model validation was applied by plotting residuals against fitted values to assess homogeneity, QQ plot was checked to verify normality and residuals were plotted against each explanatory variable to check the independence assumption. All analyses were implemented with R software (R-Core-Team, 2013) and as described in Zuur et al. (2009).

3. Results 3.1. River quality assessment 3.1.1. Macroinvertebrate composition and indices A total of 26,110 macroinvertebrate specimens belonging to 82 macroinvertebrate families were encountered at the sampled sites (SI Table 4). The taxonomic classes found in the sites were Bivalvia, Gastropoda, Hirudinea, Insecta, Malacostraca, Turbellaria, Oligochaeta and Polychaeta. The most common families that occurred in more than 80% of the sites were Baetidae, Caenidae and Chironomidae. The families Atyidae, Coenagrionidae, Elmidae, Grapsidae, Gerridae, Hydropsychidae, Leptophlebiidae, Thiaridae and Veliidae occurred in more than 50% of the sites. Nineteen EPT (Ephemeroptera, Plecoptera, Trichoptera) families were encountered in the sampled sites. Six taxa thrived in brackish water. These families were Ischyroceridae, Mysidae, Ocypodidae, Paguridae, Sphaeromatidae and Trochidae. Five sites were considered as an estuary or had an influence of brackish water. Diptera and Trichoptera had the highest number of families, which were 11. Ten families of Hemiptera and Coleoptera were collected. All ecological quality classes were represented on Leyte Island (Fig. 3a). The most diverse sites were located in the southern part of the island or in the mountainous regions, while a low diversity was observed along the west coast (Fig. 3b–e).

232

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

Fig. 3. (a) The island map representing the BMWP quality classes (b) Margalef diversity indices (c) Shannon-Wiener diversity indices (d) number of families and (e) number of EPT taxa of each site.

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

233

Table 2 Mean, median, minimum and maximum values of physical and chemical environmental variables. Variable

units

Mean

Median

Minimum

Maximum

Elevation Average water velocity Water temperature Specific conductivity pH Dissolved oxygen (DO) Dissolved oxygen saturation Total dissolved solids (TDS) Chlorophyll Turbidity Chemical oxygen demand (COD) Biological oxygen demand (5days) BOD5 Nitrate-N Nitrite-N Ammonium-N Total nitrogen Orthophosphate-P Total phosphorus Flood plain width Flood prone width Entrenchment depth Average stream width Average stream depth Maximum depth

m a.s.l. m/s (◦ C) ␮S/cm – mg/L % g/L ␮g/L NTU mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L m m m m m m

65.2 0.5 29.3 339.8 7.7 7.6 99.0 0.2 2.1 3.2 6.1 1.0 0.3 0.003 0.06 1.0 0.06 0.08 25.7 22.7 1.7 15.0 0.2 0.5

32.9 0.5 29.7 196.5 7.9 7.8 99.9 0.1 1.4 1.1 5 0.8 0.2 0.002 0.04 1.0 0.06 0.07 14.4 12.2 1.5 8.7 0.2 0.3

5.12 0.02 21.8 34.3 6.0 3.3 43.7 0.02 0.0 0.0 5* 0.5* 0.2* 0.002* 0.01* 1.0* 0.01* 0.01* 2.4 0.3 0.5 1.1 0.04 0.1

777.6 1.4 34.1 7862.0 8.8 12.3 174.4 5.1 21.0 34.0 43.4 6.4 1.1 0.03 0.6 1.3 0.3 0.4 200.0 150.0 6 100 0.6 7.0

*Records are below the detection limit. Values are expressed as the detection limits of the kits.

3.1.2. Physical and chemical variables Table 2 presents the physical and chemical measurements performed on the 85 locations. Six freshwater sites had a conductivity higher than 500 ␮S/cm, which may have an impact on aquatic life according to Behar (1997). Both water pH and ammonium-N for all locations were within the normal range of concentrations in rivers which are 6–9 and 0.03–0.77, respectively. One site had a nitrite-N concentration higher than the standard limit (0.02 mg/L) as reported by Bartram and Ballance (1996). Nitrate concentrations were in general low and about half of the sites had phosphate concentrations that may impact aquatic life in rivers (2.25 mg/L for Nitrate-N and 0.05 mg/L for phosphate-P) as documented by Behar (1997) and DES (2015). Ten sites had chlorophyll concentrations greater than the natural levels (3 ␮g/L) according to the standards at DES (2015) and USEPA (2015). Likewise, fifteen sites had turbidity higher than typical turbidity (5 NTU) of most rivers (DES, 2015). Except for 2 sites, DO, COD and BOD5 are within the natural concentrations in rivers according to the standards reported by Bartram and Ballance (1996), Behar (1997) and USEPA (2015). 3.2. Determination of key environmental variables by CCA and LRM In this step, we determined the relation between macroinvertebrates and environmental conditions. As this relation might be affected by altitude (cf. material and methods), the results are presented for conditions above and below 30 m a.s.l. CCA results revealed that at higher elevation (>30 m a.s.l.), macroinvertebrate communities were associated with sediment matrix, sludge layer, mean depth, variation in flow, dissolved oxygen, elevation and flow velocity as depicted in their long arrow lengths (Fig. 4a–b, SI Fig. 1a–b). The arrow length is proportional to the maximum rate of change in the value of the associated variable. Variables with long arrows vary much across the diagram (Ter Braak and Verdonschot, 1995). The lengths and positions of the arrows provide information about the relationship between the environmental variables and the derived axes. Arrows that are parallel to an axis (e.g. DO and axis 2) indicate a correlation. Likewise, Viviparidae had a high score of axis 2, however, the correlation between DO and axis 2 is negative (also shown the direction of the DO arrow); large positive

scores on axis 2 should have low values for DO, whilst large negative values on axis 2 should have high values for DO. Thus, it is expected that Viviparidae are abundant at sites with low DO, while Lepidosmatidae and Ephemerellidae are abundant at sites with high DO. Correspondingly, black flies (Simuliidae), Lepidosmatidae and Pleidae preferred high flow velocities, while the molluscs (Corbiculidae, Physidae, Viviparidae), Tubificidae and Corixidae were abundant at low flow velocities and low dissolved oxygen. Sensitive taxa such as Perlidae and Lepidostomatidae were more abundant at higher elevations, whereas Neritidae, Physidae, Tubificidae and Mesoveliidae were encountered at lower elevations. At lower elevation (<30 m a.s.l.), macroinvertebrate communities were related with variation in width, macrophytes, velocity, mineral substrate, the presence of logs, pool/riffle class and conductivity as reflected in their arrow lengths (Fig. 4c–d, SI Fig. 1c–d). BMWP-Viet index, Shannon-Wiener diversity index and Margalef diversity index were successfully modelled with LRM as confirmed in the validation results (SI Fig. 2–7). Based on the LRM models, the main variables related with BMWP-Viet at higher elevation (>30 m a.s.l.) were elevation, chlorophyll, pool/riffle class and the presence of macroalgae. The results revealed that BMWPViet score decreased with increasing chlorophyll concentration, while the BMWP-Viet score increased with increasing elevation, the presence of a variety of pool/riffle sequences and the presence of macroalgae. At lower elevation (<30 m a.s.l.), BMWP-Viet score was negatively associated with stream width and erosion, while it was positively related to mean depth, riffle class and the presence of macroalgae (Table 3). Both the final LRM models developed for Margalef and Shannon-Wiener diversity indices were positively related with bank slope (Tables 4 and 5). Both Margalef and Shannon-Wiener diversity indices were negatively associated with turbidity and quarrying at lower elevation. Margalef diversity index decreased with increasing chlorophyll and the presence of erosion at higher and lower elevation, respectively. Margalef diversity index increased with increasing elevation and abundance of twigs at higher elevation. Shannon-Wiener index was positively associated with the presence of a variety of pool/riffle sequences and presence of logs at higher elevation, whereas this index was negatively related to the abundance of twigs at lower elevation.

234

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

Fig. 4. Canonical Correspondence Analysis (CCA) plots for macroinvertebrate families responses to environmental gradients. Plots are separated for clarity with a & b representing macroinvertebrate families and the first set of environmental variables (SI Table 3) for sites at higher elevation c & d representing macroinvertebrate families and the second set of environmental variables (SI Table 3) for sites at lower elevation. Arrow length is proportional to the relative importance of environmental variables. Abbreviations: DO (Dissolved oxygen), OP (Orthophosphate P), VF (Variation in flow), Ac (Acari), Bae (Baetidae), Chi (Chironomidae), Coe (Coenagrionidae), Dy (Dytiscidae), Ecn (Ecnomidae), El (Elmidae), Hep (Heptageniidae), Hy (Hydroptilidae), Hyd (Hydropsychidae), Ge (Gerridae), Glo (Glossosomatidae), Gra(Grapsidae), Gyr (Gyrinidae), Le (Leptoceridae), Lep (Leptophlebiidae), Lim (Limoniidae), Lib (Libellulidae), Pal (Palaemonidae), Ph (Philopotamidae), Pse (Psephenidae), Poly (Polycentropodidae), Pyralidae (Py), Psy (Psychomiidae), Sci (Scirtidae), Vel (Veliidae).

Table 4 Estimates and p-values of variables for Margalef diversity index.

Table 3 Estimates and p-values of variables for BMWP-Viet scores. Explanatory variable

Estimate

Model developed on data with elevation higher than 30 m a.s.l. 67.73 Intercept 0.057 Elevation −3.287 Chlorophyll 19.78 Riffleclass3 Riffleclass4 28.93 18.57 Presence of macroalgae Model developed on data with elevation lower than 30 m a.s.l. Intercept 55.8320 −0.7866 Average stream width 93.8511 Mean depth −30.5667 Erosion 1 21.7605 Riffle class 3 9.9249 Riffle class 4 16.359 Presence of macroalgae

p-value

5.31e − 09 0.01813 0.01806 0.04617 0.00613 0.01778

0.00139 0.00110 0.03231 0.00021 0.01518 0.39470 0.02571

*Pool/Riffle class 2–4 represents pool-riffle pattern is absent, uniform pool-riffle pattern; pool-riffle pattern is poorly developed, low variety in pools and riffles; and pool-riffle pattern is moderately developed, variety in pools and riffles but locally, respectively. Extent of erosion variable 1 represents limited erosion. Pool/Riffle class 2, absence of erosion and absence of macroalgae is used as a baseline for the remaining categories of pool/riffle class, erosion and presence of macroalgae, respectively.

Explanatory variable

Estimate

Model developed on data with elevation higher than 30 m a.s.l. 1.1372 Intercept 0.0030 Elevation −0.1241 Chlorophyll 0.7552 Twigs 1 Twigs 2 2.4383 2.0346 Bank slope 2 1.6233 Bank slope 3 1.8032 Bank slope 4 Model developed on data with elevation lower than 30 m a.s.l. 3.0424 Intercept −0.0366 Turbidity −0.7375 Erosion 1 0.9161 Bank slope 3 0.7965 Bank slope 4 −1.4017 Quarrying

p-value

0.08989 0.00093 0.00874 0.02793 0.00073 0.00092 0.00456 0.00478

2.04e − 14 0.04646 0.00716 0.00095 0.01221 7.55e − 05

*Twigs 0–2 represents absent, limited and abundant respectively. Bank slope 1–4 represents 80–90◦ , 60–80◦ , 30–60◦ , and 10–30◦ , respectively. Extent of erosion 1 represents limited erosion. Absence of twigs, bank slope class 1 and absence of erosion is used as a baseline for the remaining categories of twigs, bank slope and erosion, respectively.

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238 Table 5 Estimates and p-values of variables for Shannon-Wiener diversity index. Explanatory variable

Estimate

Model developed on data with elevation higher than 30 m a.s.l. Intercept 0.7602 0.9703 Logs 1 0.9107 Bank slope 2 0.9619 Bank slope 3 0.8476 Bank slope 4 0.3795 Riffle class 3 Riffle class 4 0.4662 Model developed on data with elevation lower than 30 m a.s.l. Intercept 2.1451 −0.0324 Turbidity −0.3122 Twigs1 −1.1747 Twigs2 0.3468 Bank slope 3 0.3073 Bank slope 4 −0.8296 Quarrying 1

p-value

0.008692 0.000520 0.002503 0.000791 0.006319 0.023197 0.009969

6.76e − 15 0.000313 0.031431 0.002324 0.009597 0.047644 1.30e − 06

*Logs or Twigs 0–2 represents absent, limited and abundant respectively. Bank slope 1–4 represents 80–90◦ , 60–80◦ , 30–60◦ , and 10–30◦ , respectively. Pool/Riffle class 2–4 represents pool-riffle pattern is absent, uniform pool-riffle pattern; pool-riffle pattern is poorly developed, low variety in pools and riffles; and pool-riffle pattern is moderately developed, variety in pools and riffles but locally, respectively. Absence of logs or twigs, bank slope class 1, and Pool/Riffle class 2 is used as a baseline for the remaining categories of twigs, bank slope and Pool/Riffle class, respectively.

4. Discussion 4.1. Macroinvertebrate communities Various shell-forming invertebrates such as gastropods and bivalves were found on the island. These invertebrates were located in sites near the coast (at low elevation). Snail abundance is influenced by the available Ca in the water (Lewis and Magnuson, 1999; Skeldon et al., 2007). Sites near the coast have high salt concentration and therefore have high Ca concentrations. Snails may also be physiologically susceptible at low pH (Lonergan and Rasmussen, 1996). However, most locations were alkaline and therefore, water pH on the island may not be a concern for the abundance of these snails. Sediment structure had a significant role in macroinvertebrate communities. This is expected as sediments serve as habitat for some benthic macroinvertebrates (Kondolf and Wolman, 1993). Furthermore, different types of substrate support different fauna. Organic-rich sediments provide food for the filter- and depositfeeder invertebrates such as molluscs (Pan et al., 2015). In our study, molluscs (Viviparidae, Corbiculidae, Thiaridae, Neritidae) were generally found at sites with fine bed substrate. Likewise, Aqrawi and Evans (1994) observed molluscs dominate the macrofauna at surface waters with bed substrate of predominantly clay or silt. Few taxa indicated the conductivity of the water. For instance, Spionidae were strikingly found in both brackish and freshwater. Usually, this taxon is found in marine or estuarine systems (Essink and Kleef, 1988). The specimen found on the island probably adapted to freshwater conditions. According to the CCA, this taxon occurs at relatively high conductivities, particularly at waters with conductivities higher than 200 ␮S/cm. Therefore, we can infer that the presence of this taxon indicates that the water has a conductivity higher than 200 ␮S/cm. At lower elevation (<30 m a.s.l.), some invertebrates (Pleidae, Naucoridae, Lampyridae, Dytiscidae, Athericeridae, Gomphidae, Pyralidae) were found at relatively low conductivities (>120 ␮S/cm). However, at higher elevation (>30 m a.s.l.), these taxa were also found at higher conductivities (of up to 520 ␮S/cm). Perhaps, these taxa coincidentally occurred at relatively low conductivities at lower elevation.

235

Tubificidae, some molluscs (Corbiculidae, Physidae, Viviparidae) and Corixidae were abundant at sites with low flow velocities and low oxygen concentrations. Tubificidae is a widely known taxon dominating at waters with low oxygen concentration (Giere et al., 1999) and tolerant to organic pollution (Roldán, 2003). Similar to the study of Jewsbury (1985), the molluscs Viviparidae and Corbiculidae were encountered in rivers with low flow velocities. Some molluscs are sensitive to pollution while others are tolerant (Elder and Collins, 1991). In our study, it is apparent that the molluscs Corbiculidae, Physidae, and Viviparidae are tolerant to pollution as they are abundant at sites with low dissolved oxygen. Few families were associated with velocity. Simuliidae usually prefer sites with high flow velocities, which is also illustrated in the study of Fenoglio et al. (2013). In our study, the presence of Simuliidae indicated water with high flow velocity. Aside from Simuliidae, the families Lepidosmatidae, Ephemerellidae, Philopotamidae, Psephenidae, Scirtidae, Gyrinidae, Pleidae favoured waters with high flow velocity. The first four families are known to be sensitive to pollution (Chesters, 1980; Nguyen et al., 2004; Roldán, 2003). In general, sensitive families (Lepidosmatidae, Ephemerellidae, Philopotamidae) are abundant in waters with high flow velocities, whereas the tolerant families (mostly molluscs) dominate at sites with low flow velocities, low dissolved oxygen and dominant silt or clay bed substrate. Bass (2003) reported that typical tropical island systems have low macroinvertebrate fauna diversity due to their small size, frequent disturbances of their freshwater environments and most likely due to their oceanic origin. However, our investigation showed a high richness of macroinvertebrate families. The tropical island we investigated is about 10 times larger than the Caribbean islands investigated by Bass (2003). More diverse habitats can be found on the island of Leyte. The diverse habitats support a lot of different macroinvertebrate taxa, which confirms the hypothesis of Covich (2006), in which the patterns of the presence of macroinvertebrate faunas can be related to habitat characteristics (i.e. island age, size, location). Aside from the habitat characteristics, natural disturbances (i.e. hurricanes, drought) and habitat degradation also influence the occurrence of these taxa (Bass, 2003; Covich, 2006). 4.2. Determination of key environmental variables Natural disturbance such as heavy rainfall and severe tropical storms, causing rising water levels and increased flow, may scour the substrate and destroy many individuals (Bass, 2003). Aside from natural disturbances, anthropogenic activities can affect the occurrence of aquatic invertebrates (Damanik-Ambarita et al., 2016; Forio et al., 2016, 2017; Li et al., 2012; Rogers et al., 2002). Locations at higher altitude tend to be more diverse and of better quality than locations at lower altitude. According to Rezende et al. (2014) and Feio et al. (2015), higher altitudes likely increased taxonomic richness and density of macroinvertebrate communities. However, such effects are often confounded with anthropogenic disturbance, which is often most extensive at lowest altitudes (Montano-Centellas and Garitano-Zavala, 2015; Vasquez et al., 2015). Chlorophyll a concentration appeared to be an important variable associated with diversity and ecological quality of sites. Measurement of chlorophyll a is used as an indicator of phytoplankton abundance (Kiefer et al., 2015). Phytoplankton concentration increases with an increase of external nutrient loading (Deng et al., 2014; Wiik et al., 2015). Even with small additions of nutrients, phytoplankton growth may increase. This changes the abundance and diversity of algal consumers (Bowman et al., 2005). Thus, alterations in chlorophyll concentration affect the food chain and may affect macroinvertebrate diversity and composition (Bowman et al., 2007; Forio et al., 2015; Smith et al., 1999).

236

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

Turbidity was associated with richness and evenness of macroinvertebrate communities based on Margalef and ShannonWiener diversity indices, respectively. In a similar study, macroinvertebrates responded strongest to turbidity, which is likely influenced by anthropogenic impacts (Backus-Freer and Pyron, 2015). Turbidity in streams could be due to erosion, waste discharge, urban runoff, fine bottom sediments, eroding stream banks, excessive algal growth and heavy precipitation causing water flow to carry large volumes of particulate matter (Munyika et al., 2014; USEPA, 1997). Despite the growing population on the island (approximately 2.5 million inhabitants), organic pollution was less likely occurring in the streams, as most sites have low COD and BOD5 and high DO concentration (Table 2). These could be due to low input of organic pollution in most sites as septic tanks are commonly used by the inhabitants for sewage treatment. The rivers are also relatively short and are closely connected to the marine system. Thus, typical midstream and downstream systems were not encountered and the accumulation of these pollutants along the river is less likely. Hydromorphological variables were significantly related to ecological quality based on macroinvertebrates. Numerous sequences of pools and riffles were positively associated with ecological quality and macroinvertebrate diversity. Buffagni et al. (2004) observed dissimilarity in the structure of benthic communities between pool and riffle microhabitats. Sequences of these sub-habitats enhance diverse macroinvertebrate communities and ecological quality. Our results indicate that a moderately steep bank is significantly associated with diversity more than a steep bank. Likewise, Heatherly et al. (2005) found that macroinvertebrate diversity was negatively related to bank slope. The steepness of littoral zone can limit macrophyte abundance. Sequentially, macroinvertebrate diversity increases as the presence of macrophytes influences macroinvertebrate communities (Nguyen et al., 2015). The presence of vegetation at the banks was not quantified in our study although it was observed that locations with steep banks had less riparian vegetation than those characterized by moderately steep banks. We recommend further investigation to understand the associations between bank slope and the presence of riparian vegetation and their effect on biodiversity. Bank erosion is negatively related with ecological quality and diversity at lower elevation. Our results are in line with the study of Simpson et al. (2014), wherein bank erosion negatively influenced macroinvertebrate richness. The risk of bank erosion increases with the absence of floodplain and riparian vegetation and result in the reduction of habitat complexity and diversity (O’donnell et al., 2015; Peters et al., 2016). In some sites, the absence of riparian vegetation is a result of sand/gravel mining in rivers and alteration of land use into crop production, wherein the vegetation cover is systematically removed. Diversity increased with the presence of logs (diameter of >30 cm) and twigs (diameter of <3 cm) at higher elevation. Investigation of the effect of the presence of logs in rivers is limited. A study, however, indicates that the presence of logs (woody debris) increases habitat heterogeneity and causes variable water flow velocities. As a result, macroinvertebrate diversity increases and stabilizes (Brugmans et al., 2015). Aside from habitat complexity, dead woods may also serve as food for some aquatic invertebrates. However, at lower elevation (<30 m a.s.l.), the presence of abundant twigs negatively affect evenness which is expressed in ShannonWiener diversity index. This finding could be an indirect effect, for instance, the presence of twigs at lower elevation might increase the productivity, which may lead to a few highly competitive taxa to dominate the system, as was concluded in other studies (Drobner et al., 1998) and therefore the number of families became unbalanced.

On the island, quarrying/mining of sand and gravel in rivers is common, although the activity has been recently regulated in some rivers. Sand and gravel are used in the construction of roads, highways and buildings. It is reported that quarrying causes lowering of water level, creates a pit which traps sediment and interrupts transport of sediments through the river (Peng, 1999), leads to coarsening of bed material, results to loss of spawning gravels for salmon and trout and erosion of channel bed and banks which produce channel incision (Kondolf, 1997). Investigation of the impact of sand/gravel quarrying on macroinvertebrate communities is limited. However, based on our model results, sand/gravel quarrying was related with macroinvertebrate diversity. Gravel and cobbles serve as habitat for some benthic macroinvertebrates and severe habitat disturbances can affect the life cycles of these animals (Kondolf and Wolman, 1993). In contrast, most quarried locations are located at lower altitude. These locations are also affected by the cumulative impacts of human activities (i.e. agriculture, household), which may have an effect on macroinvertebrate communities. Unexpectedly, the LRM model suggests that presence of macroalgae was positively related with ecological water quality. There were 31 sites containing macroalgae. However, macroalgae in most of these sites are few and are growing on some stones or floating on limited parts of the river reaches. Algae could naturally grow with the amount of sunlight available, but their growth is limited by nutrients (Townsend and Padovan, 2005). Algae could have varying degrees of tolerances towards pollution and therefore can be used as an indicator of the ecological state of water systems (Palmer, 1969). In our study, these algae were not identified. Thus, the sensitivity of these algae is unknown. However, these algae were mostly present at sites with good ecological quality. Based on our findings, further work is needed to define the possible application of macroalgae as ecological indicators in flowing waters on tropical islands. The key drivers associated with macroinvertebrate diversity and ecological quality are related to various types of environmental variables, such as chemical (chlorophyll), physical (turbidity) hydromorphological (bank slope, pool/riffle class), the presence of logs and twigs and habitat degradation (sand quarrying, bank erosion). These variables can be used to indicate ecological quality and diversity degradation in flowing waters on tropical islands. However, further in-depth study is needed for the possible application of these indicators to describe the ecological status of these systems. The findings of this study could serve as a baseline to develop useful indicators for flowing waters on larger tropical islands. Furthermore, the discovered relations can serve as a basis to set up dedicated experiments to further prove the causal relations. Few studies have documented the impacts of habitat degradation and loss of physical complexity in lotic ecosystems (Friberg, 2014). However, this study illustrates the added value of the ecological index BMWP-Viet and diversity indices, Margalef and Shannon-Wiener. They were able to indicate not only the chemical quality of the water but also hydromorphological and habitat degradation. These indices incorporate the impact of a multitude of stressors on the ecological status of lotic systems in the tropics. 4.3. Implications for management Although many sites have good to very good ecological quality and high diversity, about 41% had moderate to bad ecological quality and low diversity. Maintenance and improvement of the ecological quality of the rivers need proper management. Strict regulation of sand/gravel quarrying in the rivers is necessary. Minimizing quarrying not only reduces degradation of banks which allows regrowth of diverse riparian vegetation but also preserves various sequences of pool/riffle successions. Though nutrient input

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

in some rivers is low, some sites showed some evidence of low to moderate nutrient enrichment. This nutrient input most likely came from nutrient runoff of fertilizers applied in agricultural fields, specifically in rice field areas. Proper amount and sustainable application of fertilizer in agricultural areas to minimize nutrient runoff are recommended. Regulation of fertilizer application is not only economical but also aids in conserving aquatic ecosystems. Limiting nutrient input in the rivers also limits the growth of phytoplankton, which contributes to better ecological quality. Furthermore, a buffer zone is suggested between the river banks and crop production and/or residential land use to conserve the riparian vegetation. Consequently, bank erosion will be minimized, habitat heterogeneity will be increased and nutrient runoff will be reduced. The impact of other toxic substances such as pesticides/pharmaceuticals is not covered in this investigation; thus we recommend future studies on the impacts of these substances on these systems. Moreover, further investigation can be elaborated in the causal relationship between diversity and ecological quality and these key environmental variables. The results of this study contribute inputs in the literature on the characteristics of insular stream systems and provide insights into the responses of aquatic macroinvertebrate communities on the environmental conditions in a tropical island system. This allows a better understanding of river managers and conservationists on tropical stream ecology. Acknowledgements Marie Anne Eurie Forio is funded by the special research fund of Ghent University to support the VLIR Ecuador biodiversity network. The authors would like to thank DA-BFAR (Department of Agriculture-Bureau of Fisheries and Aquatic Resources) for facilitating and providing the sampling permit; Hannah Rissah Abad and Jik Abad for assisting the logistics and preparation of the sampling campaign; Department of Soil Science, Visayas State University for accommodating the laboratory activities; Cesar Yap, John Paul Poliquit, Jose Talavera, Anthony Sinahon and all the people involved during the sampling campaign. The authors would like to thank the anonymous reviewers for the valuable suggestions, which improved the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ecolind.2017. 02.013. References Aqrawi, A.A.M., Evans, G., 1994. Sedimentation in the lakes and marshes (Ahwar) of the Tigris-Euphrates Delta southern Mesopotamia. Sedimentology 41, 755–776. Backus-Freer, J., Pyron, M., 2015. Concordance among fish and macroinvertebrate assemblages in streams of Indiana, USA. Hydrobiologia 758, 141–150. Bartram, J., Ballance, R., 1996. Water quality monitoring. In: A Practical Guide to the Design and Implementation of Freshwater Quality Studies and Monitoring Programmes. Chapman & Hall, London, UK. Bass, D., 2003. A comparison of freshwater macroinvertebrate communities on small caribbean islands. Bioscience 53, 1094–1100. Behar, S., 1997. Testing the Waters: Chemical and Physical Vital Signs of a River, Montpelier, VT. River Watch Network. Bowman, M.F., Chambers, P.A., Schindler, D.W., 2005. Epilithic algal abundance in relation to anthropogenic changes in phosphorus bioavailability and limitation in mountain rivers. Can. J. Fish. Aquat. Sci. 62, 174–184. Bowman, M.F., Chambers, P.A., Schindler, D.W., 2007. Constraints on benthic algal response to nutrient addition in oligotrophic mountain rivers. River Res. Appl. 23, 858–876. Brugmans, B., Moeleker, M., Weerman, E., Lapperre, L., 2015. Woody debris increases macroinvertebrates communities in stream construction works. In: Angelopoulos, N., Buijse, T., Vanoorschot, M., Kampa, E. (Eds.), Proceedings of the International Conference on River and Stream Restoration Novel

237

Approaches to Assess and Rehabilitate Modified Rivers. Wageningen, The Netherlands, pp. 232–423. Buffagni, A., Erba, S., Cazzola, M., Kemp, J.L., 2004. The AQEM multimetric system for the southern Italian Apennines: assessing the impact of water quality and habitat degradation on pool macroinvertebrates in Mediterranean rivers. Hydrobiologia 516, 313–329. Calizza, E., Costantini, M.L., Rossi, L., 2015. Effect of multiple disturbances on food web vulnerability to biodiversity loss in detritus-based systems. Ecosphere 6. Cardinale, B.J., 2012. Biodiversity loss and its impact on humanity. Nature 489, 326. Chesters, R.K., 1980. Biological Monitoring Working Party. The 1978 National Testing Exercise. Department of the Environment, UK (Water data unit technical memorandum 19). Covich, A.P., 2006. Dispersal − limited biodiversity of tropical insular streams. Polish J. Ecol. 54, 523–547. De Pauw, N., Beyst, B., Heylen, S., 2001. Development of a biological assessment method for river sediments in Flanders, Belgium. Int. Assoc. Theor. Appl. Limnol. 27 (Pt. 5), 2703–2708 (Proceedings). DES, 2015. Description of River Water Quality Parameters. http://des.nh.gov/. [Accessed 15 December 2015]. Damanik-Ambarita, M.N., Lock, K., Boets, P., Everaert, G., Nguyen, T.H.T., Forio, M.A.E., Musonge, P.L.S., Suhareva, N., Bennetsen, E., Landuyt, D., Dominguez-Granda, L., Goethals, P.L.M., 2016. Ecological water quality analysis of the Guayas river basin (Ecuador) based on macroinvertebrates indices. Limnologica 57, 27–59. David, C.P.C., 2003. Establishing the impact of acid mine drainage through metal bioaccumulation and taxa richness of benthic insects in a tropical Asian stream (The Philippines). Environ. Toxicol. Chem. 22, 2952–2959. Deng, J., Qin, B., Paerl, H.W., Zhang, Y., Zhang, Y., Chen, Zhang, Y., 2014. Effects of nutrients, temperature and their interactions on spring phytoplankton community succession in Lake Taihu, China. PLoS One 9, e113960. Dray, S., Dufour, A.B., 2007. The ade4 package: implementing the duality diagram for ecologists. J. Stat. Softw. 22, 1–20. Drobner, U., Bibby, J., Smith, B., Wilson, J.B., 1998. The relation between community biomass and evenness: what does community theory predict, and can these predictions be tested? Oikos 82, 295–302. Dyer, S.D., Peng, C., Mcavoy, D.C., Fendinger, N.J., Masscheleyn, P., Castillo, L.V., Lim, J.M.U., 2003. The influence of untreated wastewater to aquatic communities in the Balatuin River. The Philippines. Chemosphere 52, 43–53. Elder, J.F., Collins, J.J., 1991. Freshwater molluscs as indicators of bioavailability and toxicity of metals in surface-water systems. In: Waregw, G.W. (Ed.), Reviews of Environmental Contamination and Toxicology: Continuation of Residue Reviews. Springer, New York, New York, NY. Essink, K., Kleef, H.L., 1988. Marenzelleria viridis (Verril, 1873) (Polychaeta: Spionidae): a new record from the ems estuary (The Netherlands/Federal republic of Germany). Zoologische Bijdragen 38, 3–13. Friberg, N., 2014. Impacts and indicators of change in lotic ecosystems. Wiley Interdiscipl. Rev.: Water 1, 513–531. Feio, M.J., Ferreira, W.R., Macedo, D.R., Eller, A.P., Alves, C.B.M., Franca, J.S., Callisto, M., 2015. Defining and testing targets for the recovery of tropical streams based on macroinvertebrate communities and abiotic conditions. River Res. Appl. 31, 70–84. Fenoglio, S., Boano, F., Bo, T., Revelli, T., Ridolfi, L., 2013. The impacts of increasing current velocity on the drift of Simulium monticola (Diptera: Simuliidae): a laboratory approach. Ital. J. Zool. 80, 443–448. Fomba, K.W., Mueller, K., Van Pinxteren, D., Herrmann, H., 2013. Aerosol size-resolved trace metal composition in remote northern tropical Atlantic marine environment: case study Cape Verde islands. Atmos. Chem. Phys. 13, 4801–4814. Forio, M.A.E., Landuyt, D., Bennetsen, E., Lock, K., Nguyen, T.H.T., Ambarita, M.N.D., Musonge, P.L.S., Boets, P., Everaert, G., Dominguez-Granda, L., Goethals, P.L.M., 2015. Bayesian belief network models to analyse and predict ecological water quality in rivers. Ecol. Modell. 312, 222–238. Forio, M.A.E., Van Echelpoel, W., Dominguez-Granda, L., Mereta, S.T., Ambelu, A., Hoang, T.H., Boets, P., Goethals, P.L.M., 2016. Analysing the effects of water quality on the occurrence of freshwater macroinvertebrate taxa among tropical river basins from different continents. Ai Commun. 29, 665–685. Forio, M.A.E., Mouton, A., Lock, K., Boets, P., Nguyen, T.H.T., Damanik Ambarita, M.N., Musonge, P.L.S., Dominguez-Granda, L., Goethals, P.L.M., 2017. Fuzzy modelling to identify key drivers of ecological water quality to support decision and policy making. Environ. Sci. Policy 68, 58–68. Francis, C., Van Wart, J., Johnson, B., 2014. How to regenerate rural community and ecoservices: reversing the tragedy of the commons. Agron. J. 106, 95–99. Freitag, H., 2004. Composition and longitudinal patterns of aquatic insect emergence in small rivers of Palawan Island, the Philippines. Int. Rev. Hydrobiol. 89, 375–391. Gabriels, W., Lock, K., DE Pauw, N., Goethals, P.L.M., 2010. Multimetric Macroinvertebrate Index Flanders (MMIF) for biological assessment of rivers and lakes in Flanders (Belgium). Limnologica 40, 199–207. Giere, O., Preusse, J.-H., Dubilier, N., 1999. Tubificoides benedii (Tubificidae: oligochaeta) − a pioneer in hypoxic and sulfidic environments. An overview of adaptive pathways. Hydrobiologia 406, 235–241. Heatherly, T., Whiles, M.R., Knuth, D., Garvey, J.E., 2005. Diversity and community structure of littoral zone macroinvertebrates in southern Illinois reclaimed surface mine lakes. Am. Midland Nat. 154, 67–77. Helm, D., Hepburn, C., 2012. The economic analysis of biodiversity: an assessment. Oxford Rev. Econ. Policy 28, 1–21.

238

M.A.E. Forio et al. / Ecological Indicators 77 (2017) 228–238

Hill, M.J., Mathers, K.L., Wood, P.J., 2015. The aquatic macroinvertebrate biodiversity of urban ponds in a medium-sized European town (Loughborough, UK). Hydrobiologia 760, 225–238. Janse, J.H., Kuiper, J.J., Weijters, M.J., Westerbeek, E.P., Jeuken, M.H.J.L., Bakkenes, M., Alkemade, R., Mooij, W.M., Verhoeven, J.T.A., 2015. GLOBIO-aquatic: a global model of human impact on the biodiversity of inland aquatic ecosystems. Environ. Sci. Policy 48, 99–114. Jardine, T.D., 2014. Organic matter sources and size structuring in stream invertebrate food webs across a tropical to temperate gradient. Freshw. Biol. 59, 1509–1521. Jewsbury, J.M., 1985. Effects of water velocity on snails and cercariae. Parasitol. Today 1, 116–117. Kiefer, I., Odermatt, D., Anneville, O., Wueest, A., Bouffard, D., 2015. Application of remote sensing for the optimization of in-situ sampling for monitoring of phytoplankton abundance in a large lake. Sci. Total Environ. 527, 493–506. Kondolf, G.M., Wolman, M.G., 1993. The sizes of salmonid spawning gravels. Water Resour. Res. 29, 2275–2285. Kondolf, G.M., 1997. Hungry water: effects of dams and gravel mining on river channels. Environ. Manag. 21, 533–551. Kura, N.U., Ramli, M.F., Ibrahim, S., Sulaiman, W.N.A., Aris, A.Z., Tanko, A.I., Zaudi, M.A., 2015. Assessment of groundwater vulnerability to anthropogenic pollution and seawater intrusion in a small tropical island using index-based methods. Environ. Sci. Pollut. Res. 22, 1512–1533. Lewis, D.B., Magnuson, J.J., 1999. Intraspecific gastropod shell strength variation among north temperate lakes. Can. J. Fish. Aquat. Sci. 56, 1687–1695. Li, F., Chung, N., Bae, M.J., Kwon, Y.S., Park, Y.S., 2012. Relationships between stream macroinvertebrates and environmental variables at multiple spatial scales. Freshw. Biol. 57, 2107–2124. Lock, K., Goethals, P.L.M., 2014. Predicting the occurrence of stoneflies (Plecoptera) on the basis of water characteristics: river morphology and land use. J. Hydroinf. 16, 812–821. Lonergan, S.P., Rasmussen, J.B., 1996. A multi-taxonomic indicator of acidification: isolating the effects of pH from other water-chemistry variables. Can. J. Fish. Aquat.Sci. 53, 1778–1787. Lovell, R., Wheeler, B.W., Higgins, S.L., Irvine, K.N., Depledge, M.H., 2014. A systematic review of the health and well-being benefits of biodiverse environments. J. Toxicol. Environ. Health-Part B-Crit. Rev. 17, 1–20. Margalef, R., 1958. Information theory in ecology. Int. J. Gen. Syst. 3, 36–71. Montano-Centellas, F.A., Garitano-Zavala, A., 2015. Andean bird responses to human disturbances along an elevational gradient. Acta Oecolog.-Int. J. Ecol. 65–66, 51–60. Morse, J.C., Bae, Y.J., Munkhjargal, G., Sangpradub, N., Tanida, K., Vshivkova, T.S., Wang, B.X., Yang, L.F., Yule, C.M., 2007. Freshwater biomonitoring with macroinvertebrates in East Asia. Front. Ecol. Environ. 5, 33–42. Munyika, S., Kongo, V., Kimwaga, R., 2014. River health assessment using macroinvertebrates and water quality parameters: a case of the Orange River in Namibia. Phys. Chem. Earth A/B/C 76–78, 140–148. Myers, N., Mittermeier, R.A., Mittermeier, C.G., DA Fonseca, G.A.B., Kent, J., 2000. Biodiversity hotspots for conservation priorities. Nature 403, 853–858. Nguyen, X.Q., Mai, D.Y., Pinder, C., Tillings, S., 2004. Biological Surveillance of Freshwaters: A Practical Manual and Identification Key for Use in Vietnam, Hanoi, Vietnam. National University Publishers. Nguyen, T.H.T., Boets, P., Lock, K., Ambarita, M.N.D., Forio, M.A.E., Sasha, P., Dominguez-Granda, L.E., Hoang, T.H.T., Everaert, G., Goethals, P.L.M., 2015. Habitat suitability of the invasive water hyacinth and its relation to water quality and macroinvertebrate diversity in a tropical reservoir. Limnologica 52, 67–74. O’donnell, J., Fryirs, K., Leishman, M.R., 2015. Can the regeneration of vegetation from riparian seed banks support biogeomorphic succession and the geomorphic recovery of degraded river channels? River Res. Appl. 31, 834–846. Palmer, C.M., 1969. A composite rating of algae tolerating organic pollution. J. Phycol. 5, 78. Pan, B.-Z., Wang, Z.-Y., Li, Z.-W., Lu, Y.-J., Yang, W.-J., Li, Y.-P., 2015. Macroinvertebrate assemblages in relation to environments in the West River, with implications for management of rivers affected by channel regulation projects. Quat. Int. 384, 180–185. Peng, J.S., 1999. Impact of Quarrying Coarse Gravel from River Channel. A a Balkema Publishers, Leiden. Peters, D.L., Caissie, D., Monk, W.A., Rood, S.B., St-Hilaire, A., 2016. An ecological perspective on floods in Canada. Canad. Water Resour. J. 41, 288–306. Pettersson, M., 1998. Monitoring a freshwater fish population: statistical surveillance of biodiversity. Environmetrics 9, 139–150. Pianka, E.R., 1966. Latitudinal gradients in species diversity: a review of concepts. Am. Nat. 100, 33–46.

Pletcher, K., 2015. Leyte. Encyclopædia Britannica Online. http://www.britannica. com/place/Leyte. [Accessed 2 December 2015]. Proches, S., Ramdhani, S., Perera, S.J., Ali, J.R., Gairola, S., 2015. Global hotspots in the present-day distribution of ancient animal and plant lineages. Sci. Rep. 5, 6. R-CORE-TEAM. 2013. R: A language and environment for statistical computing, R foundation for statistical computing. Available: http://www.R-project.org/ [Accessed 1 April 2014]. Roldán, 2003. Bioindicacíon de la calidad del agua en Colombia. Uso del método BMWP/Col. Universidad de Antioquia, Medellín. Ramirez, A., DE Jesus-Crespo, R., Martino-Cardona, D.M., Martinez-Rivera, N., Burgos-Caraballo, S., 2009. Urban streams in Puerto Rico: what can we learn from the tropics? J. North Am. Benthol. Soc. 28, 1070–1079. Resh, V.H., Rosenberg, D.M., 2010. Recent trends in life-history research on benthic macroinvertebrates. J. North Am. Benthol. Soc. 29, 207–219. Rezende, R.S., Santos, A.M., Henke-Oliveira, C., Goncalves, J.F., 2014. Effects of spatial and environmental factors on benthic a macroinvertebrate community. Zoologia 31, 426–434. Ricketts, T.H., Daily, G.C., Ehrlich, P.R., Michener, C.D., 2004. Economic value of tropical forest to coffee production. Proc. Natl. Acad. Sci. U. S. A. 101, 12579–12582. Rogers, C.E., Brabander, D.J., Barbour, M.T., Hemond, H.F., 2002. Use of physical, chemical, and biological indices to assess impacts of contaminants and physical habitat alteration in urban streams. Environ. Toxicol. Chem. 21, 1156–1167. Rosenberg, D.M., Resh, V.H., 1996. Use of aquatic insects in biomonitoring. In: Merritt, R.W., Cummins, K.W. (Eds.), An Introduction to the Aquatic Insects of North America. Kendall/Hunt Publishing Company, Dubuque, IA, USA, pp. 87–96. Shannon, C.E., Weaver, W., 1949. The Mathematical Theory of Communication. University of Illinois Press, Urbana, Illinois, USA. Simpson, A., Turner, I., Brantley, E., Helms, B., 2014. Bank erosion hazard index as an indicator of near-bank aquatic habitat and community structure in a southeastern Piedmont stream. Ecol. Indic. 43, 19–28. Skeldon, M.A., Vadeboncoeur, M.A., Hamburg, S.P., Slum, J.D., 2007. Terrestrial gastropod responses to ecosystem-level calcium manipulation a northern hardwood forest. Canad. J. Zool.-Revue Canad. De Zool. 85, 994–1007. Smith, V.H., Tilman, G.D., Nekola, J.C., 1999. Eutrophication: impacts of excess nutrient inputs on freshwater, marine, and terrestrial ecosystems. Environ. Pollut. 100, 179–196. Smukler, S.M., Sánchez-Moreno, S., Fonte, S.J., Ferris, H., Klonsky, K., O’geen, A.T., Scow, K.M., Steenwerth, K.L., Jackson, L.E., 2010. Biodiversity and multiple ecosystem functions in an organic farmscape. Agric. Ecosyst. Environ. 139, 80–97. Somboonna, N., Wilantho, A., Assawamakin, A., Monanunsap, S., Sangsrakru, D., Tangphatsornruang, S., Tongsima, S., 2014. Structural and functional diversity of free-living microorganisms in reef surface, Kra island, Thailand. Bmc Genom. 15. Ter Braak, C.F., Verdonschot, P.M., 1995. Canonical correspondence analysis and related multivariate methods in aquatic ecology. Aquat. Sci. 57, 255–289. Ter Braak, C.J.F., Prentice, C., 2004. A theory of gradient analysis. In: Caswell, H. (Ed.), Advances in Ecological Research, vol. 34. Elsiever Academic Press., U.S.A. Townsend, S.A., Padovan, A.V., 2005. The seasonal accrual and loss of benthic algae (Spirogyra) in the Daly River: an oligotrophic river in tropical Australia. Mar. Freshwater Res. 56, 317–327. USEPA, 1997. Volunteer Stream Monitoring: A Methods Manual [Online]. USA. Available: http://www.water-research.net/Waterlibrary/Lake/STREAM.PDF [Accessed Nov. 19 2015]. USEPA, 2015. National Recommended Water Quality Criteria − Aquatic Life Criteria Table. http://www.epa.gov/wqc/national-recommended-waterquality-criteria-aquatic-life-criteria-table. [Accessed 15 December 2015]. Vasquez, D.L.A., Balslev, H., Sklenar, P., 2015. Human impact on tropical-alpine plant diversity in the northern Andes. Biodivers. Conserv. 24, 2673–2683. Wiik, E., Bennion, H., Sayer, C.D., Davidson, T.A., Mcgowan, S., Patmore, I.R., Clarke, S.J., 2015. Ecological sensitivity of marl lakes to nutrient enrichment: evidence from Hawes Water, UK. Freshw. Biol. 60, 2226–2247. Zuur, A.F., Ieno, E.N., Walker, N.J., Saveliev, A.A., Smith, G.M., 2009. Mixed Effects Models and Extensions Ion Ecology with R. Springer Science + Business Media LLC, NY, USA, pp. 2009. Zuur, A.F., Ieno, E.N., Elphick, C.S., 2010. A protocol for data exploration to avoid common statistical problems. Methods Ecol. Evol. 1, 3–14.