Stream Health Assessment Research Articles

Extreme Learning Machine model for assessment of stream health using the Qualitative Habitat Evaluation Index

Abstract The Extreme Learning Machine (ELM) approach was used to predict stream health with a Qualitative Habitat Evaluation Index (QHEI), and watershed metrics. A dataset of 112 sites in Ontario, Canada with their Hilsenhoff Biotic Index (HBI) and richness values was used in the development of two ELM models. Each model used 70 and 30% of the dataset for training and testing respectively. The models show a great fit with Root Mean Square Error (RMSE)=0.12 and 0.33 for HBI and richness test models, respectively. Then, features elimination based on ELM coefficients and coefficient of variation showed a slight increase in the models' RMSE to reach 0.09 and 0.33 correspondingly. Accordingly, this high predictability of the models in this research provide better insights into which factors influence HBI or richness, and suggests that ELM has a better architecture than other machine learning models and ANN to learn complex non-linear relationships. Also, sensitivity analysis expressed channel slope as the most affecting stream-health parameter for stream health.

Effect of substrate type on diatom-based water quality assessments in the Mount Lofty Ranges, South Australia

Diatoms have been widely used in stream health assessments as they are highly sensitive to water quality. There is no standardised method for diatom-based stream health assessments causing many substrates, both natural and artificial, to be used. Few studies have investigated the impact of substrate type on diatom assemblages in streams with highly variable water levels. To determine if the substrate type affects these assessments, diatoms were analysed from an artificial substrate (rope), rock and mud, from 17 sites in the Mount Lofty Ranges, South Australia. Rope has been advocated as a substrate as it mimics aquatic plants, while having advantages of artificial substrates (e.g. standardised habitat). We found that substrate type did not affect the stream health assessment based on the Diatom Species Index for Australian Rivers and hence the additional cost of deploying ropes in such studies is unlikely to be warranted. However, substrate type influenced diatom–nutrient relationships, with no relationship between nutrients in the stream water and diatoms in the mud substrate—possibly due to the nutrient subsidy provided by mud. Therefore, by contrast, the absence of nutrients provided by the rope substrate indicates it is suited to studies of nutrient status and sensitivity in streams.

Stream health assessment using chemical and biological multi-metric models and their relationships with fish trophic and tolerance indicators

Abstract Nonsan Stream is a tributary of the Geum River that is strongly affected by urbanization and agricultural activities in the watershed. The present study was performed to diagnose the chemical and biological health of the stream from 2010 to 2016 using the multi-metric water pollution index (WPI) and index of biotic integrity (IBI) models. Monthly and annual variations in stream water chemistry were directly influenced by the monsoon season. Inputs of total phosphorus, suspended solids, and ionic substances, as well as dilution of total nitrogen, occurred during July and August. Chlorophyll increase was largely influenced by total phosphorus (R2 = 0.43, p

The Stream Quality Index: A multi-indicator tool for enhancing environmental management

Assessment of stream health is a function of the physical, chemical, and biological integrity of the water body. While monitoring of all three indicator types is common, combining them into a unified assessment of stream quality is rare. In this study, a unified index was developed that compares biological response to physical and chemical stressors for southern California wadeable streams using a scientifically rigorous, easy-to-understand tool intended to facilitate stream management. The Stream Quality Index (SQI) is based on a stressor-response empirical model that quantifies the expected likelihood that chemical and physical stressors will impact multiple components of biological condition. While the individual stressor and response components are quantitative and have similar meaning across a variety of environmental settings, the final SQI narrative assessment is categorical and designed to be directly actionable within a management context. The four narrative assessment categories are: (1) “healthy and unstressed” (i.e., unimpacted biology, no stressors); (2) “healthy and resilient” (i.e., stressed, but biological communities are healthy); (3) “impacted and stressed” (i.e., impacted biology from observed stressors); and (4) “impacted by unknown stress” (i.e., biology is impacted, but stressors are low). To facilitate adoption by managers, a web-based application was developed that not only maps overall SQI results, but also enables users to readily access underlying quantitative information for stressors and biological responses. This transparent design was intended; high-level output and foundational components of the SQI are relevant for different audiences and details are not sacrificed for accessibility.

Sign, strength and shape of stream fish-based metric responses to geo-climatic and human pressure gradients

Abstract Efficient community indices and indicators are crucial for the adequate management and design of measures ensuring the ecosystem integrity. In this study we analyse the shape, sign and strength of the response of some biotic integrity indices and indicators of structure and function of fish communities along geo-climatic and human pressures gradients at catchment scale. To that purpose, >300 sites all over the Iberian Peninsula were characterized at the catchment scale by means of two anthropogenic drivers (agricultural and urban land proportion) and seven natural environmental descriptors covering geographical and climatic aspects. Regarding to fish-based metrics, a set of the most frequently used in stream health assessment studies have been selected, including taxonomic classic indicators, size related indicators and also recent multimetric indices created in the European context (WFD). We applied boosted regression trees that allow estimating the sign and strength of the response as well as considering non-linearity and impact thresholds. Our results show that the jointly contribution of anthropic drivers was lower than geo-climatic drivers. For most of indices and indicators, one single land cover contributed more markedly to the total deviance explained than the other, and they responded rather consistently to land-use variables, i.e., most of them responded negatively to the increase of anthropic use in the catchment. Size diversity, Fish Region Index (FRI) and maximum weight were those more sensible to agriculture land, while EFI+, mean weight, distinctness and FRI were those more sensitive to urban land. Regarding the shape of the response, urban land proportion affects normally at extremely low values, while agriculture land proportion induces smoother changes on a wider range. Our results may have practical implications, such as the selection of an efficient array of fish-based metrics to be included in ecological assessment and monitoring programs.

Assessing the relative importance of parameter estimation in stream health based environmental justice modeling

Performance of environmental justice models depends on the level of accuracy in measuring or estimating the health of the environment. In the past decades, and especially in the area of stream health modeling, significant improvement has been observed. However, the impacts of these improvements on the robustness of environmental justice models have not been evaluated. Therefore in this study, the relative importance of parameter estimation in stream health based environmental justice modeling was evaluated. The Saginaw River Basin in Michigan was considered as the study area, and four major ecological indices evaluating the response of fish and macroinvertebrates to instream stressors were used for stream health assessment. Seventeen socioeconomic and physiographic indices were evaluated at three census levels of county, census tract, and block group. Then the ecological, socioeconomic, and physiographic indices were used in the development of stream health based environmental justice models. Results showed that incorporating ecologically relevant indices and a using two-phase modeling approach not only improved the performance of stream health predictive models but also reduced the sensitivity of environmental justice models to aggregation at different census levels. In addition, using improved stream health indices reduced the redundancy of the independent variables (socioeconomic and physiographic indices), where the total number of significant parameters was reduced from 171 to 115. Besides that, more robust and meaningful spatial dependencies were observed among stream health measures and environmental justice parameters at different spatial levels. In summary having a reliable stream health information is the key for development of robust environmental justice models as evidence by improving model predictability and eliminating contradictory results compared to previous studies.

Stream health, topography, and land use influences on the distribution of the Eurasian otter Lutra lutra in the Nakdong River basin, South Korea

In South Korea, Eurasian otter (Lutra lutra) habitat has increased since the 1990 s because of legal protection, forest development, and increased public awareness. Although otters have been detected throughout the country, a precise investigation of habitat conditions has not previously been undertaken. We hypothesized that higher values of stream health assessment programme (SHAP) indices would indicate better habitat for otters. Otter distribution based on spraint density was investigated along 600 m riverside transects at 250 sites in the Nakdong River basin. We defined four otter population categories (absent, rare, normal, and abundant) and compared the results with presence and absence data. Use of the four population categories provided more details about habitat conditions than presence and absence analysis. The highest spraint densities were found at higher altitudes, and densities declined in areas with high human density and urbanization. The benthic macro-invertebrate community index was highly positively related to the otter population indices. In terms of fish as prey items, the proportion of species tolerant to environmental change were negatively related to otter indices, otherwise number of species sensitive to environmental change positively related. Overall, stream health assessment indices provided the highest correlations with otter population numbers and distribution.

Effective Stream Health Assessment: Soil Microbes Versus Aquatic Insects

Effective Stream Health Assessment: Soil Microbes Versus Aquatic Insects

Large-scale climate change vulnerability assessment of stream health

Freshwater streams are critical resources that provide multiple benefits to humans and aquatic biota alike. As climate changes, it is projected that changes to the hydrological cycle and water temperatures will affect individual biota and aquatic ecosystems as a whole. The goal of this study was to determine the extent of climate change impacts on stream ecosystems as represented by four commonly used stream health indicators (Ephemeroptera, Plecoptera, and Trichoptera taxa (EPT), Family Index of Biotic Integrity (FIBI), Hilsenhoff Biotic Index (HBI), and fish Index of Biotic Integrity (IBI)). Seven watersheds in Michigan were selected based on stream thermal regimes. The Soil and Water Assessment Tool was used to simulate streamflow and pollutant loads. Important variables for each thermal class were selected using a Bayesian variable selection method and used as inputs to adaptive neuro-fuzzy inference systems models of EPT, FIBI, HBI, and IBI. Finally, an ensemble of climate models from the Coupled Model Intercomparison Project Phase 5 were used to determine the impacts of climate on the stream health in 2020–2040 compared to 1980–2000. The risk of declining stream health was determined using cumulative distribution functions. A stream temperature regression model was also developed to assess potential changes in stream thermal regimes, which could cause shifts in composition of aquatic communities. Several flow regime variables, including those related to flow variability, duration of extreme events, and timing were mainly affected by changing climate. At the watershed scale, most indicators were relatively insensitive to changing climate and the magnitude of stream health decline was low. However, at the reach scale, there are many instances of high risk and large magnitude of declines in the stream health indicators. At the same time, several streams experienced changes in thermal class, mostly transitioning from cold-transitional and cool streams to warm streams. This research demonstrated the applicability of the stream health modeling process in performing a climate change impacts assessment.

Are bioassessments based on the Reference Condition Approach (RCA) affected by 'rapid' approaches to sample collection and processing?

<p>Benthic invertebrates are used by a number of agencies worldwide as indicators for assessing stream health, which has resulted in the development of a variety of protocols for collecting and processing benthic samples. The large number of methods used means that calibration of data collection is not always possible, but if different methods produce similar estimates of community composition and metric values, then sharing of data can make bioassessments more efficient. This study explored the effect of two approaches to subsampling and sorting of benthic invertebrates on community composition, calculation of metrics, and assessment of stream health. We compared two commonly used sampling methods: a rapid approach, employing live, unaided sorting and a standard approach using microscope sorting of preserved samples, through a comparison of replicate samples collected from 61 streams. This study found that both methods resulted in similar estimates of community composition at a site, as determined by the Bray-Curtis similarity index. However, the live sorting methodology resulted in greater family richness and higher estimates of metrics that reflect large taxa (<em>i.e</em>., %EPT). Despite differences in a number of metrics, both methods performed equally well at identifying impairment in the test sites, with live-sorting samples slightly more sensitive.</p>

Ecological health assessments using multiple parameters of fish blood tissues to community along with water chemistry in urban streams

The objectives of this study were to identify multi-level stressors from blood biomarkers to community-level bioindicators and diagnose the stream ecosystem health in polluted streams. Blood chemistry such as total protein (TPro), blood urea nitrogen (BUN), total cholesterol (TCho) and albumin (Alb) were analyzed from sentinel fish tissues; the functions of kidney, gill and liver were significantly decreased in the impacted zone (I z), compared to the control zone (Cz). Histopathological analysis showed that fish liver tissues were normal in the C z. Fish liver tissues in the Iz, however, showed large cell necrosis and degeneration and also had moderate lobular inflammation and inflammatory cell infiltration of lymphocytic histocytes. Species biotic index (SBI) at species level and stream health assessment (SHA) at community level indicated that chemical impacts were evident in the Iz (ecological health; poor - very poor), and this was matched with the blood tissue analysis and histopathological analysis. The impairments of the streams were supported by water chemistry analysis (nitrogen, phosphorus). Tolerance guild analysis and trophic guild analysis of fish were showed sig nificant differences ( P < 0.01) between Cz and Iz. Overall, multiple parameter analysis from biomarker level (blood tissues) to bioindicator level (community health) showed significantly greater impacts in the I z than Cz. This approach may be effective as a monitoring tool in identifying the multilateral and forthcoming problems related to chemical pollution and habitat degradation of stream ecosystems.

Volunteer and professional macroinvertebrate monitoring provide concordant assessments of stream health

Whilst volunteer monitoring has many benefits for both volunteers and professionals, volunteer data must be validated to understand the value and potential applications of information from volunteer monitoring programmes. Our study aimed to assess the concordance between volunteer and professional data, including state and trend assessments. We compared macroinvertebrate data collected by volunteers using a simplified identification protocol to data collected by professionals following standard national protocols for collection and identification. We found that volunteer and professional macroinvertebrate data expressed as summary indices of ecological health were significantly correlated. However, the coarser level of taxonomic identification in the volunteer dataset limited the use of taxon richness as a biodiversity measure. We also demonstrated that the ability of volunteer data to detect long-term trends in ecological health is comparable to professional data. Overall, stream monitoring data collected by volunteers provided an assessment of stream health that was concordant with assessments based on data collected by professionals, indicating that volunteer data could be used to support professional monitoring programmes.

Assessment of Stream Integrity in Relation to Neighboring Land Use Coverage

Landscape morphology of a watershed plays an important role on determination of the ecological health in the stream or river ecosystems. A habitat-riparian index system (HIS) is a newly developed metric to estimate stream naturalness in Korea. In this study, we focus on a predictive assessment of the HIS associated with stream integrity, and then aim to evaluate essential features of the stream integrity in conjunction with ambient land use patterns obtained from geographical information system (GIS). Concerning the stream health assessment, we applied a decision-tree model to the HIS prediction. As a results, the best models of the decision-trees exhibited a 72.2% accuracy in prediction of the stream health index, and also a 65.8% accuracy in classifying river catchments. A hierarchical structure of the decision tree models in effect, was able to identify and classify more influential land use patterns that could reflex stream integrity. We highlight that GIS-involved analyses and approaches accommodates a wide variety of benefits in a rapid assessment of ecological data associated with field measurements. Furthermore, such methodological applications will be able to become an effective alternative for efficient ecological health assessment in streams and rivers.

Comparing alternatives for combining invertebrate and diatom assessment in stream quality classification

The present study aimed to determine if a predictive model integrating freshwater assemblages from different trophic levels (macroinvertebrates and diatoms) produces a more sensitive assessment of stream health than single-assemblage assessments combined a posteriori. For this purpose, individual and combined models based on two approaches (BEAST and RIVPACS) were developed for Portuguese streams: two for diatoms; two for macroinvertebrates; and two combining diatoms and macroinvertebrates as a single community. Twenty-three sites affected by organic contamination, industrial effluents and mine drainage were evaluated with the predictive models and also by the official biotic indices used in Portugal. The sensitivity of the RIVPACS assessment to disturbance was improved by the a priori combination of diatoms and macroinvertebrates, whereas for BEAST the a posteriori approaches were slightly more sensitive. Diatom and invertebrate indices combined a posteriori performed better than single-assemblage indices but with lower sensitivity than combined models. We conclude that the a priori combination of the two biological assemblages is valuable (more sensitive to disturbance) for the RIVPACS approach and that the a posteriori combination of assessments for individual biological elements may not always provide the most realistic indication of stream health.

Ecological Health Assessments of an Urban Lotic Ecosystem Using a Multimetric Model along with Physical Habitat and Chemical Water Quality Assessments

The objective of this study was to diagnose integrative ecological health of an urban stream,which is located in the Asian temperate region. The research approach was primarily based on the Index ofBiological Integrity (IBI) using fish assemblage and a Qualitative Habitat Evaluation Index (QHEI). And these indexes were compared with long-term conventional chemical dataset during 1996-2005. For the experiment, four sampling sites were chosen from the stream and wading method was used for fish collection during 2004-2005. We developed a stream health assessment model (SHA model) for the regional application. Stream health conditions, based on the SHA model, averaged 23 indicating a fair-poor condition, and varied from 18 to 40 depending on the sampling stations. Values of QHEI varied from 55 (fair-poor condition) to 112 (good-fair condition) and the values were significantly lower in the down stream than the headwater. Water quality, based on COD, BOD, TN and TP declined from the upstream to downstream reach. The proportion of sensitive species showed a negative linear function with BOD and the tolerant was vise versa. Values of SHA model reflected the chemical and physical habitat conditions and this result was clearly evident in the downstream reach. These outcomes were supported byprincipal component analysis (PCA) of IBI vs. other factors. Overall these results suggest that the impacts to the health of this stream are due to the combined effects of both chemical and habitat degradation.

Stream Health Assessment on Hoeya River Basin and Other Streams Based on Fish Community and Land Use in the Surrounding Watersheds

Stream Health Assessment on Hoeya River Basin and Other Streams Based on Fish Community and Land Use in the Surrounding Watersheds

Assessing Stream Health in the Chesapeake Basin Using the SHARP Model

An assessment of stream health within the Chesapeake Bay Basin can be made using the Stream Health and Runoff Potential (SHARP) model, which is based solely on the relationship between land cover and stream constituents: Total phosphorus (TP), total nitrogen (TN), and total suspended sediment (TSS). While not intended to compete with more complex models that utilize a range of specific input data, SHARP’s advantage is that it requires little input, is easily applied, and can show whether a stream or watershed is likely to be impacted (impaired). The model allows the user to define a watershed boundary on screen within which a stream health index (SHI), concentrations of TP, TN and TSS, percentages of five land cover types, a color-coded land cover snapshot, impervious surface area and fractional vegetation cover are output. The paper describes SHARP, its output and an overview of how it can be used.

标准化方法筛选参照点构建大型底栖动物生物完整性指数

Benthic macroinvertebrate Index of Biological Integrity(B-IBI) was developed of relationships among habitat variables,water quality variables and benthic macroinvertebrate data from 340 sampling sites in the Huntai River for assessment of stream health and standardized reference and impaired sites selection.In this study,9 reference sites and 6 impaired sites were selected to construct the B-IBI index.Five major benthic macroinvertebrate metric parameters(i.e.taxonomic composition,relative abundance,pollution tolerance characteristics,dominant groups,and functional feeding groups) were assessed by applying the B-IBI.The results showed that the Huntai River is highly degraded and loss of biological integrity in which 24.8% of sampling sites were in excellent and good conditions,27.8% of sampling sites were in normal conditions,and 47.4% of sampling sites were in or very poor conditions.The box-plot distributions of B-IBI were highly clustered within the references sites,impaired sites and test sites indicated that the assessment was reliable.Moreover,the resulting B-IBI was highly correlated with the data set of macroinverterbates across 70 sampling sites in the Taizi River in May 2009,with 58% of the total sampling sites graded in the same stream health categories and 40.6% of total sampling sites with one grade difference in stream health.This correlation was employed to check the accuracy of the assessment result between standardized sites selection and expert option reference sites selection and showed strong consistency between two methods.The regression between the B-IBI and habitat assessment was significant in linear regression(R2=0.25,P0.01).And the regressions between the B-IBI and conductivity(Cond.),permanganate index(CODMn) and ammonia(NH3-N) were also significant in curved regressions(R2=0.23—0.34,P0.01).According to the regressions,the inflexion values of three water quality variables were around 1,000 μs/cm(Cond.),5.0mg/L(CODMn) and 1.0mg/L(NH3-N),respectively.The B-IBI decrease sharply when these water quality variables are lower than the inflexion value and decrease minimally when the concentration is higher than the inflexion value.The CODMn and NH3-N would have significant effect on the stream health and macroinvertebrates communities when the concentration is higher than the III grade of Chinese surface water standards.The sensitivity of the B-IBI and its component metrics to environmental stressors supports the use of this index for monitoring ecological conditions in streams in the Huntai River.

APPLIED ISSUES: Exploring the response of functional indicators of stream health to land‐use gradients

Summary1. Broad‐scale assessment of stream health is often based on correlative relationships between catchment land‐use categories and measurements of stream biota or water chemistry. Few studies have attempted to characterise the response curves that describe how measures of ecosystem function change along gradients of catchment land use, or explored how these responses vary at broad spatial scales.2. In autumn 2008, we conducted a survey of 84 streams in three bioregions of New Zealand to assess the sensitivity of functional indicators to three land‐use gradients: percentage of native vegetation cover, percentage of impervious cover (IC) and predicted nitrogen (N) concentration. We examined these relationships using general linear models and boosted regression trees to explore monotonic, non‐monotonic and potential threshold components of the response curves.3. When viewing the responses to individual land‐use gradients, four of five functional indicators were positively correlated with the removal of native vegetation cover and N. In general, weaker and less responsive models were observed for the IC gradient. An analysis of the response to multiple stressors showed δ15N of primary consumers and gross primary productivity (GPP) to be the most responsive functional indicators to land‐use gradients. The multivariate models identified thresholds for change in the relationship between the functional indicators and all three land‐use gradients. Apparent thresholds were &lt;10%IC, between 40 and 80% loss of native vegetation cover and at 0.5 and 3.2 mg L−1 N.4. The strength of regression models and the nature of the response curves suggest that measures of ecosystem function exhibit predictable relationships with land use. Furthermore, the responses of functional indicators varied little among three bioregions. This information provides a strong argument for the inclusion of functional indicators in a holistic assessment of stream health.

Stream ecosystem health in rural mixed land-use watersheds

A paired stream approach was used to assess ecosystem health in two rural Nova Scotia streams (Thomas Brook and Sharpe Brook) with varying land-use practices. The objectives of the study were to assess stream health within an agricultural catchment, the Thomas Brook Watershed, and to identify parameters that could be used to characterize the impacts of agricultural land use on stream ecosystem health within intensively farmed watersheds in Nova Scotia. General water quality (nutrient concentrations, turbidity, pH, temperature) and the hydrology of both watersheds were monitored from May to October 2006. In addition, continuous dissolved oxygen (DO) and chlorophyll a data were collected in both streams, and benthic invertebrate populations were characterized during the study period. Diurnal DO data were analyzed to determine photosynthesis and respiration ratios. Macroinvertebrate data provided information on productivity, and on a number of other relevant metrics. Findings determined that agricultural land-use generally led to high nutrient concentrations, large dissolved oxygen variability, turbid waters, high chlorophyll a content, and impacted macroinvertebrate populations in streams. Forested land-use demonstrated typically unimpacted conditions. It was concluded that DO dynamics and macroinvertebrate metrics would be very useful for providing a generalized assessment of stream health in agricultural watersheds.