Assessment of urban stream morphology: an integrated index and modelling system

Physical morphology is an important attribute of a stream system. The morphological state of a natural pristine stream often reflects its biological condition because of their close links. In contrast, the morphology of an urban stream may exhibit different behaviours due to serious human disturbances. For an urban stream system, the morphological condition not only determines the in-stream habitat quality but also provides the physical basis for the stream’s municipal functionalities. By comparing the morphological characters of urban and natural streams, this paper develops an integrated index system and model for the assessment of urban stream morphology. The model is applied to the Ancient Canal (Zhenjiang, China) with the aim of comparing the morphological conditions of reaches with and without ongoing restoration programs and further of assessing the effectiveness of the restoration methods and techniques. The results indicate that the water security and the landscape functionality of the canal have been upgraded. However, the quality of the in-stream habitat has been degraded as a result of the restoration. Based on the modelled results, recommendations are given for improving the effects of the next-phase restoration. The assessment system and findings from the application presented here are expected to have important implications for the restoration of disturbed urban streams in many other cities in China and elsewhere in the world.     

Correspondence of biological condition models of California streams at statewide and regional scales

We used boosted regression trees (BRT) to model stream biological condition as measured by benthic macroinvertebrate taxonomic completeness, the ratio of observed to expected (O/E) taxa. Models were developed with and without exclusion of rare taxa at a site. BRT models are robust, requiring few assumptions compared with traditional modeling techniques such as multiple linear regression. The BRT models were constructed to provide baseline support to stressor delineation by identifying natural physiographic and human land use gradients affecting stream biological condition statewide and for eight ecological regions within the state, as part of the development of numerical biological objectives for California’s wadeable streams. Regions were defined on the basis of ecological, hydrologic, and jurisdictional factors and roughly corresponded with ecoregions. Physiographic and land use variables were derived from geographic information system coverages. The model for the entire state (n?=?1,386) identified a composite measure of anthropogenic disturbance (the sum of urban, agricultural, and unmanaged roadside vegetation land cover) within the local watershed as the most important variable, explaining 56 % of the variance in O/E values. Models for individual regions explained between 51 and 84 % of the variance in O/E values. Measures of human disturbance were important in the three coastal regions. In the South Coast and Coastal Chaparral, local watershed measures of urbanization were the most important variables related to biological condition, while in the North Coast the composite measure of human disturbance at the watershed scale was most important. In the two mountain regions, natural gradients were most important, including slope, precipitation, and temperature. The remaining three regions had relatively small sample sizes (n?≤?75 sites) and had models that gave mixed results. Understanding the spatial scale at which land use and land cover affect taxonomic completeness is imperative for sound management. Our results suggest that invertebrate taxonomic completeness is affected by human disturbance at the statewide and regional levels, with some differences among regions in the importance of natural gradients and types of human disturbance. The construction and application of models similar to the ones presented here could be useful in the planning and prioritization of actions for protection and conservation of biodiversity in California streams.     

The effects of landscape-level disturbance on the composition of Minnesota caddisfly (Insecta: Trichoptera) trophic functional groups: evidence for ecosystem homogenization

Over 300,000 caddisfly specimens representing 249 species were collected from nearly 250 sites throughout Minnesota during 2000 and 2001 to determine the effects of human disturbance on the composition of caddisfly trophic functional groups at the landscape level. Canonical correspondence analysis determined that stream width was the most important variable influencing functional group composition in regions of the state with relatively low disturbance, and that differences in the caddisfly fauna between sizes of streams generally followed trends predicted by the river continuum concept. In regions of the state with moderate disturbance, both stream width and the percentage of disturbed habitat upstream of a site were important variables influencing functional group composition. In highly disturbed regions, no variables corresponded to changes in the composition of caddisfly functional groups. Instead, ecosystems were homogeneous: fine-particle filtering collectors dominated in all sizes of streams. The observed aquatic ecosystem homogenization is attributed mostly to input of fine-particle organic and inorganic sediment from extensive agriculture.     

城镇化对西苕溪大型底栖动物群落的影响

于2004—2010年5月对西苕溪流域20个参照点,5个中度干扰点以及2个重度干扰点的底栖动物进行了调查,共鉴定出74科190属226个种;受干扰后,底栖动物群落物种丰富度、EPT物种丰富度、Shannon-wiener多样性指数和B-IBI指数显著下降,BI指数显著升高。CCA分析结果表明,参照、中度和重点干扰样点的底栖动物群落差异明显。城镇化引起的溪流水温上升、营养盐升高、泥沙输入量增多和堤岸固化是导致溪流底栖动物群落退化的主要原因。     

A native species-based index of biological integrity for Hawaiian stream environments

Based upon ecological data provided by a 6-year study of native species assemblage structure and function in near-pristine Limahuli Stream (Kauai), The Hawaii Stream Index of Biological Integrity (HS-IBI) incorporates 11 metrics covering five ecological categories (taxonomic richness, sensitive species, reproductive capacity, trophic–habitat capacity, and tolerance capacity). The HS-IBI was shown to effectively distinguish stream biological condition on a continuum from undisturbed (near-pristine) to severely impaired in sampling of 39 sites (6 estuarine reaches) on 18 Hawaiian streams located on all major islands. A significant relationship was validated between relative levels of human impact occurring within-watersheds (determined through use of a landscape indicator) and IBI ratings with metrics responding predictably to gradients of human influence. For management interpretation of HS-IBI results, a framework comprised of five “integrity classes” (excellent–good–fair–poor–impaired) is provided which can be used to operationalize HS-IBI results obtained through standardized sampling of stream sites that “…translates into a verbal and visual portrait of biological condition.” Through its focus on native species, the HS-IBI incorporates evolutionary and biogeographic variation for the region with biological expectations based upon reference condition benchmarks established in near-pristine stream environments where ecological functioning is naturally self-sustaining and resilient to normal environmental variation. The methods and tools described in this study are appropriate for application in all perennial streams in Hawaii and may be adapted for use in streams on other tropical Pacific islands where native species assemblages persist in near-pristine stream environments.     

Hydrologic disturbance reduces biological integrity in urban streams

The impact of urbanization on stream ecosystems is linked by land cover changes to the alteration of the natural hydrology and subsequent physical disruption of stream biota and habitat. Seasonal floods are part of the natural disturbance regime of many streams, but urbanization increases their frequency and magnitude. This study evaluated the impact of hydrologic disturbance on fish and aquatic macroinvertebrates in 81 (56 urban/25 reference) Ohio streams. Hydrologic variables included annual and monthly 24-h rainfall maxima and computed annual peak discharge, with computation supported by GIS-based drainage area delineation and land cover characterization. Ohio biological criteria for fish and macroinvertebrates measured during the late spring and summer were negatively impacted by annual peak discharge in urban streams as compared to reference streams. Results support the application of stormwater best management practices as part of stream restoration efforts to mitigate urbanization impacts to fish and macroinvertebrates.     

Quantifying the sensitivity of ephemeral streams to land disturbance activities in arid ecosystems at the watershed scale

Large areas of public lands administered by the Bureau of Land Management and located in arid regions of the southwestern United States are being considered for the development of utility-scale solar energy facilities. Land-disturbing activities in these desert, alluvium-filled valleys have the potential to adversely affect the hydrologic and ecologic functions of ephemeral streams. Regulation and management of ephemeral streams typically falls under a spectrum of federal, state, and local programs, but scientifically based guidelines for protecting ephemeral streams with respect to land-development activities are largely nonexistent. This study developed an assessment approach for quantifying the sensitivity to land disturbance of ephemeral stream reaches located in proposed solar energy zones (SEZs). The ephemeral stream assessment approach used publicly-available geospatial data on hydrology, topography, surficial geology, and soil characteristics, as well as high-resolution aerial imagery. These datasets were used to inform a professional judgment-based score index of potential land disturbance impacts on selected critical functions of ephemeral streams, including flow and sediment conveyance, ecological habitat value, and groundwater recharge. The total sensitivity scores (sum of scores for the critical stream functions of flow and sediment conveyance, ecological habitats, and groundwater recharge) were used to identify highly sensitive stream reaches to inform decisions on developable areas in SEZs. Total sensitivity scores typically reflected the scores of the individual stream functions; some exceptions pertain to groundwater recharge and ecological habitats. The primary limitations of this assessment approach were the lack of high-resolution identification of ephemeral stream channels in the existing National Hydrography Dataset, and the lack of mechanistic processes describing potential impacts on ephemeral stream functions at the watershed scale. The primary strength of this assessment approach is that it allows watershed-scale planning for low-impact development in arid ecosystems; the qualitative scoring of potential impacts can also be adjusted to accommodate new geospatial data, and to allow for expert and stakeholder input into decisions regarding the identification and potential avoidance of highly sensitive stream reaches.     

Comparative Water Quality of Lightly- and Moderately-Impacted Streams in the Southern Blue Ridge Mountains, USA

For less-developed regions like the Blue Ridge Mountains, data are limited that link basin-scale land use with stream quality. Two pairs of lightly-impacted (90–100% forested) and moderately-impacted (70–80% forested) sub-basins of the upper Little Tennessee River basin in the southern Blue Ridge were identified for comparison. The pairs contain physically similar stream reaches, chosen for the purpose of isolating forest conversion as a potential driver of any detected differences in water quality. Streams were sampled during baseflow conditions twice monthly over a six-month period from September 2003 through February 2004. Parametric t-tests were run for each parameter measured between the lightly-and moderately-impacted streams within each pair. Statistically significantly higher mean values of suspended and dissolved solids, nitrate, specific conductivity, turbidity, and temperature were observed in the moderately impacted streams versus the lightly impacted streams in both pairs, while dissolved oxygen levels were lower in the moderately-impacted streams. No significant differences were demonstrated in orthophosphate or ammonium concentration. A near-bankfull runoff event on February 6, 2004, was sampled for stormflow values, and the results support baseflow findings. The water quality of these streams is very good when compared with lower relief areas like the Piedmont, and none of the parameters measured in this study exceeds levels of known threat to stream biota. However, the demonstration that moderate reductions in forest cover are associated with stream water quality degradation carries important implications for stream management in this rapidly developing mountainous region.     

Factors affecting river health and its assessment over broad geographic ranges: the Western Australian experience

AusRivAS is an Australia-wide program that measures river condition using predictive models to compare the macroinvertebrate families occurring at a river site with those expected if the site were in natural condition. Results of assessment of 685 sites across all major rivers in Western Australia are presented. Most rivers were in relatively natural condition in the northern half of the state where the human population is low and pastoralism is the major land use. In the south, where the human population is higher and agriculture is more intensive, rivers were mostly more disturbed. AusRivAS assessment produced some erroneous results in rivers of the south-west cropping zone because of the lack of appropriate reference site groups and biased distribution of sampling sites. Collecting low numbers of animals from many forested streams, because of low stream productivity and samples that were difficult to sort, also affected assessments. Overall, however, AusRivAs assessment identified catchment processes that were inimical to river health. These processes included salinisation, high nutrient and organic loads, erosion and loss of riparian vegetation. River regulation, channel modification and fire were also associated with river degradation. As is the case with other assessment methods, one-off sampling at individual sites using AusRivAS may be misleading. Seasonal drought, in particular, may make it difficult to relate conditions at the time of sampling to longer-term river health. AusRivAS has shown river condition in Western Australia is not markedly different from other parts of Australia which, as a whole, lacks the substantial segments of severely degraded river systems reported in England.     

Nutrient concentrations in Maryland non-tidal streams

Using a spatially extensive database from the Maryland Biological Stream Survey (MBSS), we describe nutrient relationships of small-order, non-tidal streams to Maryland watershed basins, Maryland Tributary Strategy basins, and stream order. In addition, we estimate the number of stream km affected by nutrient loading, using derived nutrient criteria. Based on the MBSS spring water quality sampling, we determined several important factors relating to nutrient levels in non-tidal streams. There are strong linear relationships of nutrients to the percentage of agriculture and forested land present within MBSS sampling strata. Both mean total nitrogen (TN) and mean total phosphorus (TP) levels for watershed basins by stream order show exceedances of derived nutrient reference criteria for Maryland. Four Maryland basins have over 85% of their stream kilometers exceeding the TN criterion, with three basins over 90% of the TP criterion. To protect small stream integrity in Maryland, we recommend an upper stream TN criterion between 1.34 and 1.68 mg/L and an upper stream TP criterion between 0.025 and 0.037 mg/L, based on quantile analyses. Elevated levels of both TN and TP are present in non-tidal streams, with subsequent nutrient inputs into the upper freshwater tidal reaches of the Chesapeake Bay.     

Assessing the soil quality of alpine grasslands in the Qinghai-Tibetan Plateau using a modified soil quality index

Soil degradation has caused various problems on the planet. Human disturbance and land use changes always negatively affect soil quality. In this study, we used a modified soil quality index (SQI) to assess soil quality under differing degrees of human disturbance and land use. The alpine grasslands were studied at different levels of degradation [i.e., severely degraded grassland, heavily degraded grassland, moderately degraded grassland, and non-degraded grassland (NDG)] in a case study conducted in Qinghai-Tibetan Plateau (QTP) to test the feasibility of using the SQI. Fifteen chemical, physical, and biological soil parameters were measured in each type of grassland. Significant variations in SQI were found across the different types of grasslands according to severity of human disturbance and changes in land use. Urease, the ratio of microbial biomass nitrogen to total nitrogen, proteinase, and soil organic carbon were found to be the most important indicators for assessing soil quality. NDG had a higher SQI than the other three types of grasslands. It was concluded that SQI is effective for assessing the soil quality of alpine grasslands in the QTP. The intensity of human disturbance had a negative effect on soil quality in the QTP.     

An evaluation of a bed instability index as an indicator of habitat quality in mountain streams of the northwestern United States

Managers of aquatic resources benefit from indices of habitat quality that are reproducible and easy to measure, demonstrate a link between habitat quality and biota health, and differ between human-impacted (i.e., managed) and reference (i.e., nonimpacted or minimally impacted) conditions. The instability index (ISI) is an easily measured index that describes the instability of a streambed by relating the tractive force of a stream at bankfull discharge to the median substrate size. Previous studies have linked ISI to biological condition but have been limited to comparisons of sites within a single stream or among a small number of streams. We tested ISI as an indicator of human impact to habitat and biota in mountain streams of the northwestern USA. Among 1428 sites in six northwestern states, ISI was correlated with other habitat measures (e.g., residual pool depth, percent fine sediment) and indices of biotic health (e.g., number of intolerant macroinvertebrate taxa, fine sediment biotic index) and differed between managed and reference sites across a range of stream types and ecoregions. While ISI could be useful in mountain streams throughout the world, this index may be of particular interest to aquatic resource managers in the northwestern USA where a large dataset, from which ISI can be calculated, exists.     

Estimating reference nutrient criteria for Maryland ecoregions

Management of stream nutrients is becoming increasingly important in order to protect both water quality and aquatic resources throughout the USA. Using an extensive water quality database from the long-term Maryland Biological Stream Survey (MBSS), we describe nutrient relationships to landscape characteristics as total nitrogen (TN) and total phosphorus (TP) of small-order, non-tidal streams in USEPA L2 and L3 ecoregions in Maryland and by MBSS stream order at the L2 and L3 ecoregion levels. To protect stream ecosystem integrity, preliminary reference nutrient estimates (TN and TP) as percentiles (25th of all stream reaches and 75th of stream reference reaches) for the six Maryland L3 ecoregions are: Blue Ridge TN 0.29 and 0.64 mg/L, TP 0.0065 and 0.0090 mg/L; Central Appalachians TN 0.40 and 1.0 mg/L, TP 0.0060 and 0.015 mg/L; Middle Atlantic Coastal Plains TN 0.93 and 2.5 mg/L, TP 0.094 and 0.065 mg/L; Northern Piedmont TN 1.6 and 1.8 mg/L, TP 0.010 and 0.015 mg/L; Ridge and Valley TN 0.40 and 0.98 mg/L, TP 0.0063 and 0.012 mg/L; and Southeastern Plains TN 0.33 and 0.82 mg/L, TP 0.016 and 0.042 mg/L. High levels of both TN and TP are present in many streams found in non-tidal watersheds associated with all Maryland ecoregions, but are especially elevated in the Northern Piedmont and Middle Atlantic Coastal Plain ecoregions, with the latter second-order streams (average TN?>?2.9 mg/L) significantly higher than all other ecoregion–order combinations. Across all six ecoregions, mean nutrient loading for both TN and TP was generally equivalent in first-order streams to nutrient concentrations seen in both second- and third-order streams, indicating a definite need to increase efforts in preventing nutrients from entering first-order streams. Small-order stream nutrient levels are the drivers for subsequent TN and TP inputs into the upper freshwater tidal reaches of the Chesapeake Bay, resulting in a potential risk for altered estuarine ecosystems.     

Relationships among nekton assemblage structure and abiotic conditions in three Texas tidal streams

The objective of this study was to assess the effects of spatially and temporally varying abiotic conditions on the nekton residing in tidally influenced streams and determine the comparability of the assemblage response across the various gears needed to fully characterize the biota. Three tidal streams were sampled seasonally for 2 years for chemical (physiochemical profiles), physical (in-stream and riparian habitat classification), and biological (nekton sampled with bag seines, trawls, and gill nets) components of ecosystem integrity. Multiple sampling stations on each stream encompassed the transitional character of the tidally influenced ecosystem, from the freshwater of the river to the saltwater of the bay. Instead of characterizing the biota with traditional fish-based indexes of biotic integrity, analysis methodologies relying heavily on multivariate ordination techniques were used. This study showed that the temporal and spatial relationships among nekton assemblages and abiotic environmental conditions were quite gear dependent. The greatest degree of difference in indicators of ecosystem health all involved upstream–downstream gradients that appear to be driven by salinity structure. Based on the results of this study, dissolved oxygen concentration does not appear to be a major structuring factor in the physical, chemical, or biological component of ecosystem integrity.     

Comparing the natural variation of oribatid mite communities with their changes associated with anthropogenic disturbance

Several organism communities serve as ecological and environmental indicators to detect changes in human-impacted habitats. However, the composition of indicator communities may vary because of natural variation in addition to the changes associated with human disturbances. This meta-analysis compared the natural variation of oribatid mite assemblages, a good indicator model group in soil ecosystems, with their deviations associated with disturbance using diversity and dissimilarity indices and three human disturbance types. Literature data were collected about oribatid mite assemblages from natural and disturbed habitats. Human disturbances consisted of agriculture, heavy metal pollution and forest management. Biodiversity indices (Shannon and Berger-Parker) and dissimilarity indices (Jaccard and Bray-Curtis) were calculated among natural habitats and between disturbed and control habitats at the species and genus level. We considered oribatid mite assemblages as effective community-level indicators when the methods separated the differences of assemblages between disturbed and control habitats from their varieties among natural habitats. In addition, the study analysed the correlation between these indices and environmental variables of the study sites. Oribatid mite assemblages performed high indication strength with Bray-Curtis index for agricultural disturbances since dissimilarity values between disturbed and control habitats were as high as between different natural habitats and higher than among the same types of natural habitats. Genus-level values showed similar results to the species level. This approach may be useful to test the effectiveness of other indicator groups and methods.     

The Stream and Its Altered Valley: Integrating Landscape Ecology into Environmental Assessments of Agro-Ecosystems

Little is known about the importance of landscape and land cover to the implementation and performance of agricultural conservation projects designed to improve stream quality. In our study, we addressed the potential importance of landscape and land cover to conservation projects by measuring variation across 191 μ-basins (100–2400 ha) and integrating the observed variation into a study design aimed at determining the effectiveness of conservation projects. Our findings indicate that there are strong gradients across which landscape and land cover attributes vary. Land cover varied along a gradient of agricultural intensity, basin morphometry across gradients of stream closure and basin size, basin substrate was described by variation in drumlin formation, glacial landform type, and soil drainage, while agricultural conservation projects varied according to the level of project implementation. Correlation of these gradients found several associations between landscape and land cover, indicating that agricultural intensity was being constrained predominantly by drumlin formation and glacial landform type. Landscape and land cover did not appear to be determining factors in the implementation of conservation projects by land owners. Based on these findings we chose 32 μ-basins which represented the variability along each of the defined gradients for further study. We conclude that landscape scale variables demonstrate important variation and covariation that can and should be integrated into study designs for the assessment of streams and human activities affecting streams.     

Live diatoms as indicators of urban stormwater runoff

Diatom bioassessment of streams/rivers does not distinguish between live (cells with intact chloroplasts) and dead (empty cells) individuals, even though most diatom samples collected from the field will be composed of a mixture of both. This study aimed to evaluate whether percentage of live diatoms (PLD), live diatom density and chlorophyll a, and diatom species compositions can be used as indicators of hydrologic disturbance in an urban stream. We deployed artificial substrates on a monthly basis and collected periphyton samples weekly over the course of one calendar year (n = 182) in three tributaries of urbanized Ruddiman Creek (Michigan, USA). We also collected samples before and after six major storm events (>0.5 cm rain). We found no temporal patterns in PLD (Mann-Kendall test p > 0.05) or species composition (non-metric multidimensional scaling (NMDS) ordination), which may be explained by a diatom composition already tolerant to frequent disturbance. There was no difference in PLD before and after storm events, which might partially be explained by their disturbance resistance due to different assemblage ages (1, 2, and 4 weeks old) before the storms. High flow had differential effects on diatom species; loosely attached Navicula and Nitzschia species were more easily removed compared to stalk-forming Gomphonema parvulum. The most important environmental variable that was found to affect live diatom density and chlorophyll was stream width, which has an indirect effect (as a measure of discharge) on periphyton assemblages. In conclusion, PLD was found to be unsuitable metric for assessing stormwater runoff in urban streams where periphyton may not have enough time to form mature communities.     

Methods development and use of macroinvertebrates as indicators of ecological conditions for streams in the Mid-Atlantic Highlands Region

The Mid-Atlantic Highlands Assessment (MAHA) included the sampling of macroinvertebrates from 424 wadeable stream sites to determine status and trends, biological conditions, and water quality in first through third order streams in the Mid-Atlantic Highlands Region (MAHR) of the United States in 1993–1995. We identified reference and impaired sites using water chemistry and habitat criteria and evaluated a set of candidate macroinvertebrate metrics using a stepwise process. This process examined several metric characteristics, including ability of metrics to discriminate reference and impaired sites, relative scope of impairment, correlations with chemical and habitat indicators of stream disturbance, redundancy with other metrics, and within-year variability. Metrics that performed well were compared with metrics currently being used by three states in the region: Pennsylvania, Virginia, and West Virginia. Some of the metrics used by these states did not perform well when evaluated using regional data, while other metrics used by all three states in some form, specifically number of taxa, number of EPT taxa, and Hilsenhoff Biotic Index, performed well overall. Reasons for discrepancies between state and regional evaluations of metrics are explored. We also provide a set of metrics that, when used in combination, may provide a useful assessment of stream conditions in the MAHR.     

Using Bacterial Growth on Insects to Assess Nutrient Impacts in Streams

A combination field and laboratorystudy was conducted to evaluate the ability of arecently developed bioindicator to detect detrimentalnutrient conditions in streams. The method utilizesbacterial growth on aquatic insects to determinenutrient impacts. Field investigations indicated thatelevated concentrations of nitrate and phosphate wereassociated with growth of filamentous bacteria oninsect body surfaces, and that there was a significantreduction in the density of major insect taxa in thenutrient-enriched stream reaches. Laboratoryinvestigations confirmed a strong linkage betweenbacterial growth and reduced survival of insects. Survival was examined for insects with bacterialinfestation ranging from 0% to greater than 50%coverage of the body surface. A threshold forcatastrophic mortality occurred at about 25% bodycoverage; there were few survivors above that amount. Based on these findings, the diagnostic endpoint forthe bioindicator is 25% body coverage by bacterialgrowth, a level that signifies major impacts and isalso easy to detect visually. This study providesadditional evidence that the insect-bacteriabioindicator is a reliable tool for assessing nutrientimpacts on stream macroinvertebrate communities. Thebioindicator should prove useful for identifyingnutrient-impacted sites as well as monitoring thesuccess of management actions to improve water quality.     

Effects of subsample size on seasonal and spatial comparisons of stream macroinvertebrate communities

We examined the effect of subsample size on the accuracy of information obtained from aquatic macroinvertebrate assemblage samples. Subsamples containing 100 organisms or 300 organisms were compared on the bases of processing time and the ability to discern ecological differences among samples. Independently of subsample size, assemblages differed between study streams, primarily reflecting an intermittent vs. permanent stream difference, and between seasons at most streams. It required, on average, two additional hours to process the larger subsamples. Larger subsamples gave significantly higher estimates of total richness and Ephemeroptera, Plecoptera, and Trichoptera (EPT) richness, but the relative abundances of many assemblage subsets (e.g., EPT organisms and most functional feeding groups) were similar using both subsample sizes. Larger subsamples did not typically enhance the ability to discriminate between samples from different seasons, but did more accurately distinguish among streams when differences were subtle. They also appeared to avoid Type I error in comparisons of compositionally similar reaches within a study stream.