Water is a valuable and finite resource on earth, and water quality (WQ) is becoming a sensitive issue in many countries. WQ is influenced by natural (e.g., changes in precipitation trends, erosion) and anthropogenic (e.g., urban, industrial, agricultural activities) factors whose influence changes with time and space and affects water use. Thus, a monitoring program that will provide empirical evidence to support decision-making on surface WQ issues is vital. Scientists face the challenge of providing stakeholders with accurate indicators of surface WQ monitoring that help guide the water management decision process. This PhD study uses different statistical approaches to solving questions linked to surface WQ of the Paute river basin (PRB) (south of Ecuador). As such, an extensive database collected in the PRB throughout a 5-year monitoring program (the year 2008 and period 2010-2013) was analysed. Statistical approaches involve hierarchical and non-hierarchical cluster methods, non-parametric classification algorithms, mathematical optimisation techniques, measures of similarity of clustering outputs, ordination methods, and regression models. The overall research goal of this doctoral project was to mine relevant information about the different components of surface WQ in the PRB streams using adequate statistical procedures. Firstly, a spatial WQ characterisation of the studied basin was performed using only chemical and microbiological information. As a result, two WQ classes were established, representing low and high pollution levels. Further, the key chemical and microbiological parameters that most explain these groups and their spatial variability were identified. Then, the biotic/ecological component (i.e., benthic macroinvertebrates) was used as a biological response variable regarding descriptive parameters (i.e., physical, chemical, microbiological, geomorphological and integrity habitat quality parameters). Different statistical algorithms were implemented using the macroinvertebrate community's taxonomic/biotic and functional approaches to find the most accurate indices/metrics that best discriminated among less, moderately and highly polluted sites in the PRB streams. Also, the relevant descriptive variables to explain the spatial variability of both taxonomical/biotic and functional macroinvertebrates aspects were identified. Likewise, using the benthic macroinvertebrates, questions about using fine (i.e., genera) or coarse (i.e., families) taxonomic resolutions were answered, giving a new perspective on using these bioindicators in the PRB. This research showed degraded streams are characterised by high levels of faecal coliforms, electric conductivity, chlorides, total hardness, nitrogen, total alkalinity, biochemical oxygen demand, turbidity, water temperature, low levels of oxygen and poor macroinvertebrate functional and taxonomic attributes, as well as fluvial habitat degradation. Contrarily, clean sites are associated with good land use plans, adequate habitat conditions (i.e., riparian vegetation and streambed heterogeneity), and high levels of taxonomic/biotic and functional diversity of macroinvertebrates. These ecological integrity features were congruent with permissible records of chemical and microbiological WQ parameters. Also, the geomorphological parameters such as elevation, slope and river order were relevant to explain the macroinvertebrate's responses. The results show that these outputs of different surface WQ components can vary in magnitude regarding the sub-basins of the PRB hydrological system, and one of the primary factors that affect their trends is land use. The macroinvertebrate community and their taxonomic and functional mechanisms are affected by stress linked with anthropization. Thus, a remarkable result of the current research is that these mechanisms were reasonably parameterised to help assess the impact of stress on the macroinvertebrate community, providing validated tools for future WQ monitoring in the PRB.