Metrics of benthic communities and habitat quality associated to different types of land use

Eng Sanit Ambient | v.24 n.4 | jul/ago 2019 | 737-746 ABSTRACT Aquatic ecosystems are under severe anthropogenic modifications. Thus, the dependent biological communities in these environments are also changed. The objective of this study was to investigate the effects of urban development in a highly impacted ecosystem. We selected 15 sampling points along the stream Ribeirão Vermelho, in which were sampled benthic macroinvertebrates and assessed the water and habitat diversity. It was found an impact gradient, with some the reference points classified as natural and others as impacted. There was a significant difference in all biological indicators used. The total number of taxa, the wealth of Diptera, the taxa Ephemeroptera, Plecoptera and Trichoptera, and the diversity indices, the Water Quality Indices, and the percentage of herbivores crushers and predators were significantly higher in points classified as natural and changed. The relative abundances of collectors, filter feeding, chironomids and parasites were significantly lower in sites classified as natural in relation to impacted ones. The metrics of the macroinvertebrate community benthic and Habitat Diversity Protocol were influenced by environmental degradation, being a useful tool for planning and development actions for the preservation of watersheds and the prioritization of high-value transmission systems for protection and rehabilitation of aquatic ecosystems


INTRODUCTION
Aquatic ecosystems are subject to strong modification, especially the rivers and streams that are polluting environments receivers of the atmosphere and the terrestrial environment (ALLAN, 2004;RAMACHANDRA;BHARATH;BHARATH, 2014;MENEZES et al., 2016). Environmental change encompasses many anthropogenic disturbances; land use and their interaction are expected to have the most significant effects on biodiversity.
Land use patterns in river basins influence fundamental processes over hierarchically distributed spatiotemporal scales (ALLAN, 2004). DOI: 10.1590 For example, agricultural activities can provoke erosion and runoff of sediments, nutrients, and pesticides (DUPAS et al., 2015). The anthropogenic sealing of soils in urban areas causes increased runoff, channel erosion, and threatens water quality from diverse pollutants such as metals, oils, and road salts (WALSH et al., 2005;PARR et al., 2015;MENEZES et al., 2016). Riparian clearance and subsequent increased solar radiation elevate water temperature and alters fundamental physicochemical processes (BOOTH et al., 2014). Combined effects of increased concentration of contaminants, temperature elevation and hydrologic changes modify dependent communities of aquatic ecosystems (PARR et al., 2015).
Amongst the communities of aquatic ecosystem, the benthic macroinvertebrate fauna is by far the most widely studied and well known in freshwater biomonitoring studies, a result of their ease of sampling, well studies taxonomic and functional diversity, ubiquity, and comparatively rapid response to a variety of stressors and life-cycle patterns that are compatible with timescales associated with the persistence and effect of known stressors such as organic pollution (FRIBERG et al., 2011).
The bioindicators concept for assessing the quality of freshwater systems, such as rivers and streams is not new. As Birk et al. (2012) succinctly put it, "aquatic bioindicators are organisms accumulating toxic substances or responding to environmental stress, such as pollution, nutrient enrichment, habitat loss or overexploitation".
Several methods have been developed to assess the ecological status of surface waters in Brazil, among they, the use of metrics has been an alternative to national indices for assessing the effects of changes in land use. Metrics synthesise biological and ecological information of macroinvertebrate assemblages (HAYBACH et al., 2004) and are frequently used to identify change in relation to natural and anthropogenic gradients (VARANDAS & CORTES, 2010;FEIO & DOLÉDEC, 2012;TUPINAMBÁS et al., 2014). Some studies have associated changes in human disturbance with an integrated index of biotic integrity, subsuming changes in individual metrics in the larger index (GUIMARÃES et al., 2009;VARANDAS & CORTES, 2010;CORTES et al., 2011;TUPINAMBÁS et al., 2014). Other studies have demonstrated associations between stressors and particular biological metrics, but these associations were expressed only in terms of whether a statistically significant positive or negative relationship existed and, thus, did not investigate the relative responses of the metrics.
Developing biomonitoring tools requires evaluation and definition of the efficient ways to detect human impacts. In this way, the aim of this study was to evaluate how different types of land use influence the metrics of the benthic macroinvertebrate communities. It was tested the hypothesis that the benthic community using metrics will answer in differentiated way to the impacts caused for the different types of land use activities in comparison with the regions with absence of anthropogenic disturbances, in order to implement appropriate policies and management practices to minimize the impacts of urbanization on aquatic ecosystems.

MATERIALS AND METHODS
The methodological steps are described in the next sub-items, including the study area, physicochemical and hydrological parameters, land use, benthic macroinvertebrates, data treatment and statistical analyses.
Water velocity was measured with a jctm-hidromet® portable flow meter (four measurements per site). Inorganic substrate was collected at each site with a corer (a composite sample comprising four cores per site) to approximately 10 cm depth. The granulometric composition of the collected substrate (%) was determined after drying by sieving through a set of Unite States Standard sieves that separated the material into pebble, gravel, coarse sand, fine sand, and silt/clay fractions (BLOTT & PYE, 2001). Stones and boulders were not observed in the study area. Organic content (%) was determined using the gravimetric ash-free dry weight method. Aliquots (0.3±0.1 g) were ashed (550°C for 4 h) and weighed; the difference between the initial weight of sample and weight after ashing gave the percentage of content of organic sediment samples.

Land-use
Land-use data were obtained from Google Earth satellite images. Screen scanning of features was made at the scale of 1:1,500 (ArcGIS 10, ESRI enterprise) using image interpretation techniques (supervised classification) to obtain information of proximal land use type, percent occupation and distribution at each sample station (SANTOS; LOUZADA; EUGÊNIO, 2007). A land use classification scheme (agriculture, forest, pasture, soil, urban area and water body) was developed based on In each study site, the Rapid Assessment Protocol (RAP) for habitat diversity (HANNAFORD; BARBOUR; RESH, 1997) modified by Callisto et al. (2002) was used to analyze 15 parameters (see CALLISTO et al., 2002). In the first group, the score assigned to each parameter was 0, 2 and 4, and in the second, 0, 2, 3 or 5 points. The final result was the sum of the scores assigned to each parameter. Final scores indicate the level of ecological preservation of the sites under study, scores from 0 to 20 indicate that the sites have been poor (impacted); from 42 to 60, have been altered or suboptimal; and from 61 to 88, are in optimal condition (natural).

Benthic macroinvertebrates
Three replicate samples of benthic macroinvertebrates were taken in riffle/run sections at each sampling site (total of 180 sub-samples) on each sampling occasion (pools comprised less than 10% of total stream channel area at all sites) using a Surber sampler (250 μm mesh, 0.1 m 2 ).

Data treatment and statistical analyses
Redundant modalities and environmental data were removed using Spearman Rank Correlation (threshold value r≥ or r≤0.65) and draftsman plots (HUGHES et al., 2009) to avoid multicollinearity.  We derived three biotic data sets to compare the effect of ecological conditions, water quality, habitat diversity and land-use on taxonomic composition (family level identification, relative abundance log10 (x+1) transformed), metrics describing community structure and composition (standardized). A Principal Components Analysis (PCA) was carried out on environmental data to reduce dimensionality, identify principal environmental gradients and construct orthogonal stressor gradients.

RESULTS AND DISCUSSION
The Excerpts from the watercourse that were considered to have altered next score impacted condition, showing that changes are taking place around the Ribeirão tending to homogeneity of habitats.
According GALDEAN et al. (2000), evaluation of habitat diversity is an important tool in health approach to aquatic ecosystems due to the strong relationship between availability of habitats and aquatic  Urban streams are subject to a "cocktail" of anthropogenic stressors, impacts and contaminants. Macroinvertebrate assemblages respond to combinations of environmental (e.g., abiotic factors such as flow and substratum) and pressure drivers that may act together or in isolation (WIBERG-LARSEN et al., 2000). The "habitat template" theory proposes that temporal and spatial variability in physical habitats influence the evolution of physical, behavioural, and physiological species profiles (SOUTHWOOD, 1977). The results of this study on a partly urban stream show shifts in macroinvertebrates assemblage structure expressed via taxonomic groups and macroinvertebrate metrics response to urban impacts.
According to Ormerod et al. (2010), in freshwater management, these problems almost always involve simultaneous challenges because human pressure usually change more than an environmental factor (for example, urbanization affects the amount of flow, water quality, habitat availability and others), making it difficult to establish direct associations between drivers and change in assemblages (STEWART et al., 2001). Results from our study on macroinvertebrate assemblage structure and function support this statement.
A total of 54 taxa and 68,808 individuals were identified and counted in the study. Mean total density was a lower at most "natural" site than Modifications to the structure and composition of macroinvertebrate community can occur due to changes in the viability of trophic resources (TUPINAMBÁS et al., 2014). Selected metrics were able to discriminate different degrees of biotic integrity. Species richness, EPT richness and Diversity Index (Shannon-Wiener and Simpson) were lower at all altered and impacted sites than natural sites (Figure 3).
All impacted sites had lower BMWP and DSFI values than their respective natural sites. The most dramatic changes were recorded at impacted sites. Percentage of Hirudinea and Oligochaeta density did not differ between natural and impacted sites, but at altered site Hirudinea was different, show more number of individuals. Percentage of Diptera increased from natural to impacted and altered sites, and was most pronounced in altered sites, whereas the percentages of Ephemeroptera, Plecoptera and Trichoptera was higher at natural sites than at alteredimpacted sites in Ribeirão Vermelho stream (Figure 4).
Macroinvertebrate density patterns varied considerably among natural and impacted-altered sites. Thus, mean macroinvertebrates density was higher at impacted sites.
In this study, they were selected several metrics commonly used to quantify the ecological degradation in general and organic contamination. In many cases organism groups and metrics responded predominantly to specific stressors. Our results indicate that metrics based on macroinvertebrate communities were useful to identify different degrees of pollution and disturbance in the studied urban streams.
The impacted-altered site was shown to be the most disturbed reach by the majority of metrics. The results suggest that the ratios tested produce very consistent results with each other, and similarly broadly reflect variations in the structure of the communities and their sensitivity to natural environmental factors.
It was expected an improvement in water quality owing to presence of the wastewater treatment plant at Lavras, but none of the physicchemical and biological parameters suggested stream water recovery. However, in the last few years, Lavras has experienced a growth in population, and the ability of the plant to cope with the volume of waste produced now appears to be inadequate. Also, the impervious surfaces of the town have increased as more streets and neighborhood areas have been paved. In some areas, the domestic sewage system is frequently flooded by storm-water, thus sewage and pluvial drainage systems are not always working as separate units high discharge events associated with rains. This result is a serious alarm for local managers to consider stricter control measures on the aquaculture activities.   (VANNOTE et al., 1980). Also, factors biotic, such as predation and competition, can influence the structure location of invertebrate communities (ALLAN, 2004).
Except for percentage nPlecoptera and nDiptera, all metrics showed at least one significant correlation with land use variables ( Table 1).
The Diversity index (Shannon-Weaver, Margaref and Simpson), nEphemeroptera, nPlecoptera, EPT, number of taxa (s) and number of families showed positive and significant relationships with forest areas and agriculture areas, all of these metrics are related to good ecological integrity. In contrast, these rates had a negative correlation with urban areas. Urban areas showed positive and significant relationships with nOligochaeta, nHirudinea and abundance of macroinvertebrate (N).
Benthic invertebrate metrics and biotic indices differed between types of land use, which indicates an increasing gradient of impairment from upstream to downstream Ribeirão Vermelho river. High on the watershed of the Ribeirão Vermelho is still possible to find green areas, while in the downstream direction the stream goes through city of Lavras to the following section for agricultural areas (Figure 1).
The degree to which current assemblage structure in these streams reflects land-use history is unknown. In summary, the modification land use gradient was reflected in the changes in the benthic macroinvertebrate communities, which indicated ecological impairment at structural level.

CONCLUSIONS
In conclusion, a trend in the benthic community observed in predicting changes and alteration of habitat quality caused by different land uses and alteration of habitat quality, especially when the source point pollution, such as the case of the urban area. Sampling points located near the urban area showed significant changes in macroinvertebrate density due to predominance tolerant species.
The response of the taxonomic composition and metrics of benthic macroinvertebrates assemblages to different stressors varied in the basin of the Ribeirão Vermelho stream. Land use and parameters of water quality were important variables that affect the answers of the data groups (taxonomic composition and metrics) of macroinvertebrates.
It was proposed to implement an integrated approach in the assessment of the health of urban aquatic ecosystems involving the analysis of the processes at the ecosystem level, as well as structural and functional biological indicators. These results are important for local managers of the studied river as well as those of other rivers in Minas Gerais, which are under stress of the same land uses.