The use of species traits to characterize the functional composition of benthic invertebrate communities has become well established in the ecological literature. This approach holds much potential for predicting changes of both species and species assemblages along environmental gradients in terms of traits that are sensitive to local environmental conditions. Further, in the burgeoning field of biomonitoring, a functional approach provides a predictive basis for understanding community-level responses along gradients of environmental alteration caused by humans. Despite much progress in recent years, the full potential of the functional traits-based approach is currently limited by several factors, both conceptual and methodological. Most notably, we lack adequate understanding of how individual traits are intercorrelated and how this lack of independence among traits reflects phylogenetic (evolutionary) constraint. A better understanding is needed if we are to make the transition from a largely univariate approach that considers single-trait responses along single environmental gradients to a multivariate one that more realistically accounts for the responses of many traits across multiple environmental gradients characteristic of most human-dominated landscapes. Our primary objective in this paper is to explore the issue of inter-trait correlations for lotic insects and to identify opportunities and challenges for advancing the theory and application of traits-based approaches in stream community ecology. We created a new database on species-trait composition of North American lotic insects. Using published accounts and expert opinion, we collected information on 20 species traits (in 59 trait states) that fell into 4 broad categories: life-history, morphological, mobility, and ecological. First, we demonstrate the importance of considering how the linkage of specific trait states within a taxon is critical to developing a more-robust traits-based community ecology. Second, we examine the statistical correlations among traits and trait states for the 311 taxa to identify trait syndromes and specify which traits provide unique (uncorrelated) information that can be used to guide trait selection in ecological studies. Third, we examine the evolutionary associations among traits by mapping trait states onto a phylogentic tree derived from morphological and molecular analyses and classifications from the literature. We examine the evolutionary lability of individual traits by assessing the extent to which they are unconstrained by phylogenic relationships across the taxa. By focusing on the lability of traits within lotic genera of Ephemeroptera, Plecoptera, and Trichoptera, taxa often used as water-quality indicators, we show how a traits-based approach can allow a priori expectations of the differential response of these taxa to specific environmental gradients. We conclude with some ideas about how specific trait linkages, statistical correlations among traits, and evolutionary lability of traits can be used in combination with a mechanistic understanding of trait response along environmental gradients to select robust traits useful for a more predictive community ecology. We indicate how these new insights can direct the research in statistical modeling that is necessary to achieve the full potential of models that can predict how multiple traits will respond along multiple environmental gradients.

Mesoscale spatial (centimeters–meters) and temporal (days–weeks) flow variability plays an important role in stream ecosystem structure and function. The structural effects of flow variability have been widely tested but its influence on ecosystem processes is less well understood. Our study examined the influence of mesoscale flow variability on cobble biofilm metabolism in an Australian regulated upland stream that had been subjected to severe catchment disturbance as a result of extensive fire 10 mo before the study. Rates of primary production and respiration were calculated from changes in dissolved O₂ over 24 h in benthic chambers containing stream cobbles sampled before and after high-discharge events in areas with differing surface-flow types. Overall, cobble biofilm communities were strongly autotrophic (primary production » respiration), probably because of reduced shading and increased nutrient influx caused by the recent fire. Differences in production and respiration before and after individual high-discharge events were inconsistent and, therefore, not statistically significant. The effect of high discharge on respiration was greater when high-discharge events were preceded by long periods of low discharge, but the number of events sampled was limited and this result could not be tested statistically. Rates of respiration and concentrations of organic material, chlorophyll a, and pheophytin a also varied spatially, and these variables were affected more by mesoscale variation in hydraulic conditions than by reach-scale variation in hydraulic conditions. Our results suggest that flow management has the potential to alter benthic metabolism.

Spatial patterns in macroinvertebrate communities were examined in 4 to 5 rheocrene springbrooks and 4 to 5 nearby rhithral streams in 4 different regions of New Zealand. Physicochemical attributes of springbrooks and rhithral streams were similar, but springbrooks were more stable. Standing crops of periphyton biomass, epilithic C, and organic matter were greater and more variable in springbrooks than in rhithral streams. The number of macroinvertebrate taxa and the total number of individuals were greater at more stable sites. Altitude, habitat stability, and food resource levels were the best predictors of the number of macroinvertebrate taxa, whereas pH, PO₄ concentration, habitat stability, and food resource levels were the best predictors of the total number of individuals at a site. Differences in the number of macroinvertebrate taxa between Northern Hemisphere and New Zealand springbrooks may be related to the larger diversity of macroinvertebrate predators in New Zealand springbrooks and to the fact that many New Zealand invertebrates do not use temperature-mediated life-history cues. Altitude, which limits invertebrate dispersal, also may play an important role in determining the maximum number of macroinvertebrate taxa in a given stream. A dispersal–stability framework is proposed to explain the observed patterns of macroinvertebrate diversity in these streams. The general pattern of noninsect dominance in lowland springs may be the result of the interaction between dispersal ability and the effects of the last glaciation.

We studied variation in C and energy flow in stream food webs by examining primary consumer diets and potential food sources at 8 sites of different drainage areas in the South Fork Eel River drainage. Both heptageniid mayfly nymphs and Glossosoma caddisfly larvae are considered scrapers in traditional functional feeding group classification, but past studies suggested that they differed in their relative use of terrestrial and algal C in some streams. In our study, microscopic examination and stable C isotope ratios (δ¹³C) suggested an increasing contribution of algae to both epilithic biofilms and fine particulate organic matter as stream drainage area and productivity increased. The proportion of algal cells in biofilms of small, unproductive streams was low, and biofilm δ¹³C values were similar to those of terrestrial detritus, suggesting that biofilms were composed primarily of heterotrophic microorganisms. Glossosoma larvae fed selectively on algae where it was scarce within the biofilms of small forested streams. In contrast, heptageniid mayfly nymphs did not appear to feed selectively on algae, but consumed algae and other materials in proportion to their abundance in the environment. These feeding patterns may have consequences for energy flow through food webs. Heptageniid mayflies feeding on biofilms in unproductive streams may augment the flow of dissolved organic C from terrestrial sources through food webs. In contrast, selective feeding by abundant Glossosoma larvae may reduce the flow of algal C through food webs because they are resistant to aquatic predators.

The gut contents of a guild of invertebrate grazers inhabiting the moss Fontinalis and feeding on epiphytic diatoms in a small Québec stream were analyzed to characterize resource partitioning and food selection. A multivariate approach (RLQ analysis coupled with a revised version of 4^th-corner analysis) identified distinct diet patterns among co-occurring grazers. These patterns were mainly explained by differential ingestion of diatoms that differed in their spatial positions within the multilayered periphyton mat. When the size range of available diatoms was large, diet differences were partly explained by diatom size. Comparison of diatoms in grazer guts with diatoms available in the environment indicated selective feeding in different levels of the periphyton mat by grazers. Some grazers (scrapers) fed preferentially on tightly attached diatoms, whereas others (surfers) favored overstory diatoms. Spatial segregation of feeding within the periphyton mat by members of the grazer guild was more evident in a period of potential resource limitation (July) than when food was abundant (May). Our results suggest that all layers/growth forms in the diatom mat are used, resulting in spatial partitioning of the resource when considering the entire grazer community. Therefore, foraging theories already established for other ecosystems are confirmed in the unique context of stream benthos.

Regional acidification of aquatic habitats has caused severe reductions in biodiversity. Reduced sulfur dioxide emissions over the past several decades have resulted in increased pH and alkalinity in some areas of North America and Europe. However, biological recovery has lagged behind increases in lake pH. We propose that acidification-induced changes in predator assemblages can provide biological resistance to recovery of assemblages in lower trophic levels in lakes. Many recovering lakes remain fishless because of low colonization rates and, therefore, support a high abundance of macroinvertebrate predators that may have a large impact on zooplankton community structure. We assessed the distribution of water beetles in relation to pH and presence/absence of fish in 29 lakes on the Canadian Shield. We found that water beetle assemblage composition was not related to pH. However, the occurrence of fish was significantly negatively correlated with water beetle distribution, particularly for the predaceous diving beetle, Graphoderus liberus. Mesocosm experiments in Swan Lake, a fishless lake recovering from acidification, revealed that larval G. liberus can reduce total zooplankton abundance, species richness, and species diversity. In particular, the densities of 4 taxa (Leptodiaptomus minutus, Diaphanosoma birgei, Bosmina [Bosmina] spp., and calanoid copepodids) were reduced by larval G. liberus predation. The high abundance of G. liberus in the absence of fish and the impact of larval G. liberus predation on crustacean zooplankton in Swan Lake suggest that biological resistance may be an important impediment to the recovery of aquatic food webs, despite increasing pH.

In June 1995, heavy rains caused severe flooding and massive debris flows on the Staunton River, a 3^rd-order stream in the Blue Ridge Mountains (Virginia, USA). Scouring caused the loss of the riparian zone and repositioned the stream channel of the lower 2.1 km of the stream. Between 1998 and 2001, we conducted seasonal macroinvertebrate surveys at sites on the Staunton River and on White Oak Canyon Run, a reference stream of similar size and geology that was relatively unaffected by the flood. Our study was designed to determine the extent to which flood-induced changes to the stream channel and riparian habitats caused long-term changes to macroinvertebrate community structure and composition. Sites within the impacted zone of the Staunton River supported diverse stable benthic macroinvertebrate assemblages 3 y after the flood despite dramatic and persistent changes in environmental factors known to be important controls on stream ecosystem function. However, significant differences in total macroinvertebrate density and trophic structure could be attributed to the flood. In autumn, densities of most feeding guilds, including shredders, were higher at impacted-zone sites than at all other sites, suggesting higher overall productivity in the impacted zone. Higher shredder density in the impacted zone was surprising in light of expected decreases in leaf-litter inputs because of removal of riparian forests. In contrast, in spring, we observed density differences in only one feeding guild, scrapers, which showed higher densities at impacted-zone sites than at all other sites. This result conformed to a priori expectations that reduced shading in the impacted zone would lead to increased light and higher instream primary production. We attribute the seasonal differences in trophic structure to the effects of increased temperatures on food quality and to the relationship between the timing of our sampling and the emergence patterns of important taxa.

River InVertebrate Prediction And Classification System (RIVPACS)-type predictive models assess the biological conditions of streams by comparing observed assemblage composition (O) with an expected assemblage composition derived from reference site observations (E). The ratio of the 2 values (O/E) can be interpreted as a measure of taxonomic completeness. Values of O/E that are near 1 at a test site suggest that the site is comparable to reference sites, whereas values that differ substantially from 1 suggest that the site is degraded. I analyzed the computations involved in estimating E mathematically. The analysis reveals that many predictive models produce values of O/E that are systematically <1, even when the models are applied to sites that have the same average assemblage composition as reference sites. The amount by which O/E values are <1 depends upon the number of reference sites used to build the model, the manner in which these reference sites are distributed among discrete clusters, the distribution of estimated capture probabilities of the taxa in the reference sites, and the value used to screen the estimated capture probabilities for inclusion of taxa in the model. Based on this analysis, a theoretical adjustment to O/E values is proposed, and its application is demonstrated on a predictive model developed for the Mid-Atlantic Highlands, USA.

An enhanced water-quality monitoring project was established in 2000 for streams providing drinking water to New York City (NYC). The project's design considered the history of the NYC source watersheds, and some of the broader issues facing freshwater supply systems in general. NYC's relationship with its watershed has historically been acrimonious and filled with mistrust, a situation that became critical in 1989 when the US Environmental Protection Agency (EPA) issued the Surface Water Treatment Rule (SWTR), which required all unfiltered public water-supply systems either to provide filtration or to comply with a stringent set of water-quality, operational, and watershed-control standards. Plans to implement this rule caused further mistrust and lawsuits, which led in 1997 to the NYC Watershed Memorandum of Agreement (MOA), a compromise that was accepted by all the stakeholders. The MOA addressed fundamental issues about: 1) the protection, allocation, and ownership of water resources, 2) the identification and valuation of ecosystem services, 3) the compatibility of environmental protection and economic development, and 4) strategies for bringing together diverse stakeholders in the watershed. One of the provisions of the MOA was to enhance the existing city, state, and federal monitoring programs for NYC's source watersheds. The monitoring project described in this series, which is part of that enhancement, recognizes philosophically that source watersheds and their ecosystems are: 1) the ultimate source of the water, 2) the major source of anthropogenic contaminants in the water, and 3) the primary natural processors of water-borne contaminants. Protecting NYC's source-water areas requires an integrated approach that ties historical and contemporary land use into the design of a large-scale, enhanced, water-quality monitoring project (the Project). The Project set forth 4 primary objectives: 1) to create a quantitative baseline of selected physical, chemical, and biological characteristics of source-water streams and reservoirs for use in assessing future changes in the quality of NYC's drinking water and the integrity of the associated aquatic ecosystems, 2) to include in the baseline factors that are sensitive to temporal variability, are reproducible, and lend themselves to unconfounded analyses among sampling sites and times, 3) to integrate temporal and spatial change in both the level of selected contaminants and the structure and function of biological communities and ecosystems to assess whether impairment impacts the ability of the streams to provide ecosystem services related to water quality, and 4) to provide additional direction and perspective to the overall watershed management plan for the NYC source-water area. All papers in this series cover Phase I of the monitoring project, which involved physical, chemical, and biological measurements made during 2000 to 2002 at 60 stream sites distributed across a 5066-km² study area.

New York City (NYC) receives >99% of its drinking-water supply from streams, rivers, and reservoirs north and northwest of the city (east or west of Hudson River [EOH or WOH, respectively]). As part of a large-scale enhanced water-quality monitoring project (the Project) in NYC's drinking-water-supply watersheds, 60 stream and 8 reservoir sampling sites were established in the water-supply area (30 WOH and 30 EOH) and sampled from 2000 to 2002. Our study describes watershed characteristics (including climate and hydrology, land use, human population, and known point-source discharges) at each study site and provides an analysis of differences in land use quantified at 3 scales: 1) watershed, 2) riparian (30 m on each side of entire stream network upstream of a site), and 3) reach (same as riparian, but truncated 1 km upstream of the study site). Regression analysis was used to determine relationships among scales, and principal components analysis was used to describe spatial differences in watershed characteristics across the study region. EOH sites are on smaller streams than WOH sites because the WOH region is much larger than the EOH region. EOH sites had smaller mean annual area-specific discharges than WOH sites, reflecting differences in precipitation and in watershed hydrologic retention that were related to surficial geology and the presence of wetlands, lakes, and reservoirs. Population densities, point-source discharges, and flows from those discharges were higher in EOH watersheds than in WOH watersheds. Landuse values in the EOH watersheds ranged from 87% forest to 57% urban. Agricultural land use exceeded 16% in only one watershed. Landuse values in WOH watersheds indicated either largely forest (several sites near 98%) or agriculture and grassland (many near 25%, largely in pasture). Urban landuse values were never >11%. Values for most landuse categories were strongly correlated (most R² > 0.75) between the watershed and riparian scales. In WOH watersheds, values for categories indicating human land use (e.g., agriculture, urban) were greater at the riparian than at the watershed scale, indicating that human land use was concentrated along the stream network. In EOH watersheds, values for categories indicating human land use were lower at the riparian than at the watershed scale. Values for most landuse categories were not correlated (typically R² < 0.50 or not significant) between the reach and watershed scales, indicating that local landuse values described statistically different conditions than watershed- or riparian-scale landuse values.

Stream inorganic chemistry was sampled under summer baseflow conditions from 2000 to 2002 at 60 sites as part of a large-scale, enhanced water-quality monitoring project (the Project) across New York City's drinking-water-supply watersheds. The 60 stream sites were evenly divided between regions east and west of the Hudson River (EOH and WOH, respectively). EOH sites had generally higher ionic concentrations than WOH sites, reflecting differences in land use and geology. Within each region, variability in inorganic chemistry data between sites was far greater than annual variability within sites. Geology was an important factor controlling underlying baseflow chemistry differences within and between regions. However, after taking into account geological controls, anthropogenic land uses primarily defined ion and nutrient baseflow chemistry patterns at regional and watershed levels. In general, watershed-scale landscape attributes had either the strongest relationships with analytes or had relationships with analytes that did not differ fundamentally from relationships of riparian- or reach-scale landscape attributes. Individual analyses indicated no dominant watershed-scale landscape attribute that could be used to predict instream inorganic chemistry concentrations, and no single ion or nutrient was identified as the best indicator of a given anthropogenic land use. Our results provide a comprehensive baseline of information for future water-quality assessments in the region and will aid in examining other components of the Project.

Organic matter (OM) in streams that provide drinking water is a potential energy source for bacterial regrowth in distribution systems and a precursor for disinfection byproducts. Baseflow concentrations of OM were measured over a 3-y period in 60 streams divided evenly between water-supply regions east and west of the Hudson River (EOH or WOH) in New York State. A baseline of OM concentrations in the 2 regions was generated, and land use/cover variables were related to those baseline concentrations. Dissolved organic C (DOC), biodegradable DOC (BDOC), and particulate OM (POM) reflected regional differences in land use and point-source discharges. Three-year mean concentrations for these variables and for total organic C (TOC) were significantly lower in the WOH than in the EOH by factors of 1.5 to 2.3. Size fractionation of POM showed similarities between regions with >70% of particles in the 0.5- to 10-μm size class. DOC made up most of the TOC in both regions, and DOC:POC ratios ranged from 1.7 to 54.4. Stepwise multiple linear regressions revealed that agriculture and forest land uses explained most of the variation in OM concentrations in the WOH, whereas wetlands and point-source discharges, primarily associated with wastewater treatment plants, explained most of the variation in OM concentrations in the EOH. Despite the potential problems from OM for drinking water quality, OM is a natural and important component of stream ecosystems, so its total elimination from watersheds is neither advisable nor possible. Our data from watersheds in the WOH region with high percentages (>97%) of forested land use and from small to mid-sized watersheds in the EOH with no point-source discharges provide lower limits for OM concentrations and targets for best management practices.

A molecular tracer method was used to assess the extent and sources of pollution to 60 stream sites that were distributed across the watersheds that supply drinking water to the greater New York City area. Samples were collected from each site annually from 2000 to 2002 during summer baseflow conditions. Twelve polycyclic aromatic hydrocarbons (PAH), 2 fragrance materials (FM), caffeine (CAF), and 7 fecal steroids (FS) were measured using a modification of EPA method 8270, which quantified concentrations to laboratory reporting levels ranging from 0.00009 to 0.016 μg/L or 3 to 5 orders of magnitude lower than method detection levels (MDL) given by EPA 8270. In 54 of 180 stream samples, concentrations of ≥1 PAHs exceeded suggested, nonregulatory EPA guidance values for water supplies (0.0038 μg/L for the 5 most toxic PAHs), and PAH signatures (ratios) and spatial patterns suggested that soot from local urban/suburban combustion was the primary source. CAF, FM, and FS all showed their highest concentrations at the 3 sites with large, failing sewage treatment plants, but more complex relationships to landscape variables at remaining sites suggested a variety of anthropogenic point and nonpoint sources. Concentrations of all molecular tracers measured were strongly negatively correlated with % forest cover (= all forest variables used) in the watershed.

Macroinvertebrate communities were examined in conjunction with landuse and water-chemistry variables at 60 sites in the NYC drinking-water-supply watersheds over a 3-y period. The watersheds are in 2 adjacent regions of New York State (east of Hudson River [EOH] and west of Hudson River [WOH]) that are geographically distinct and have unique macroinvertebrate communities. Nonforested land use at EOH sites was mostly urban (4–57%), whereas land use at sites in the rural WOH region was more agricultural (up to 26%) and forested (60–97%). Land use accounted for 47% of among-site variability in macroinvertebrate communities in the EOH region and was largely independent of geological effects. Land use accounted for 40% of among-site variability in macroinvertebrate communities in the WOH region but was correlated with underlying geology. Comparisons among 3 landuse scales emphasized the importance of watershed- and riparian-scale land use to macroinvertebrate communities in both regions. Multivariate and bivariate taxa–environment relationships in the EOH and WOH regions identified specific landuse and water-chemistry gradients and, in general, showed a continuum in conditions across the watersheds. WOH macroinvertebrate communities varied primarily with specific conductance, population density, and agricultural and urban land use, but communities were not classified as impaired along these gradients. EOH macroinvertebrate communities were associated with a wider range of watershed conditions than WOH communities. Conditions ranged from forested to urban, and distinctive communities were associated with point-source discharges, road density, and lake outlets. The severity of the impact gradient in the EOH region resulted in impaired macroinvertebrate communities with decreased total and Ephemeroptera, Plecoptera, and Trichoptera (EPT) taxon richness and increased densities of oligochaetes and chironomids.

Quantitative sampling of benthic macroinvertebrate communities from 60 sites in New York City's drinking-water-supply watersheds was undertaken in 2000, 2001, and 2002 as part of a large-scale enhanced water-quality monitoring project (the Project). Sampling yielded 543 macroinvertebrate taxonomic units, most of which (including Chironomidae) were identified to the genus/species level. Our goals were to investigate the effect of level of taxonomic resolution on statistical macroinvertebrate–environment relationships, the effect of including rare taxa on among-site similarity and macroinvertebrate–environment relationships, and the correlations between the common and rare components of the total community at each site. Mean site richness ranged from 90.8 to 101.2 taxa for sites west of Hudson River (WOH) and 62.2 to 78.5 taxa for sites east of Hudson River (EOH). Species-level identifications provided the greatest separation of sites in multivariate space, but genus- and family-level identifications discriminated between most- and least-impacted sites, particularly in the EOH region where anthropogenic impact was greatest. Of the 543 taxa, 175 (32%) were found at ≤3 sites, and nearly ½ of the taxa within a given site were occasional (found once in 3 y). Numerically rare taxa (defined as either <1% or <0.3% relative abundance within each site) accounted for 42 to 75% (WOH) and 37 to 73% (EOH) of mean site richness. Ordinations of data sets including or excluding rare taxa revealed similar impact gradients, and the % of spatial variance explained by environmental factors was similar with and without rare taxa included. Common taxa contributed noise to site-similarity patterns in the WOH region, and rare taxa provided information that was redundant with information provided by common taxa in the EOH region.

Uptake of nutrients and organic C was measured once annually between 2000 and 2002 in each of 10 streams within the water-supply source areas for New York City. Nutrients (PO₄^3– and NH₄ ) and organic C (glucose and arabinose) were injected into the streams for 1 to 2 h, and uptake lengths were estimated from the longitudinal declines in downstream concentration relative to that of a conservative tracer. Uptake lengths increased with stream size and were converted to uptake velocities, V_f, to remove scaling effects. V_f s of PO₄^3– and NH₄ varied inversely with the ambient concentration of total dissolved P (TDP) and total dissolved N (TDN), respectively, and were described by a model based on Michaelis–Menten kinetics. However, V_f s of glucose and arabinose were unrelated to the concentrations of any solute. V_f s of PO₄^3–, NH₄ , arabinose, and (with less certainty) glucose varied positively with measures of ecosystem metabolism (24-h community respiration and gross primary productivity). Uptake flux (U) of NH₄ also varied positively with ecosystem metabolism, but Us of PO₄^3–, glucose, and arabinose did not. The V_f s of PO₄^3– and NH₄ were positively related to invertebrate species richness and % forest cover, and negatively related to molecular tracer concentrations (polyaromatic hydrocarbons and fecal steroids [PO₄^3–-V_f], fragrance materials [NH₄ -V_f]) and population density. Spiraling, as a measure of ecosystem function, was sensitive to human impacts, most clearly through responses to nutrient loadings, but very probably through responses to other impacts as well.

Ecosystem metabolism was measured in 10 streams flowing into New York City drinking-water-supply reservoirs. Six of the streams were located west of Hudson River (WOH) in the Catskill Mountains and 4 were in the Croton River watershed east of Hudson River (EOH). Measurements were made for 3-d periods between June and November in each of 3 y using an open-system O₂ technique with reaeration determined from propane evasion. Chlorophyll a concentrations, algal cover types, and nutrient uptake were measured concurrently. Gross primary productivity ranged from 2.02 to 4.32 g O₂ m⁻² d⁻¹ in the WOH streams and from 0.23 to 1.13 g O₂ m⁻² d⁻¹ in the EOH streams. Community respiration ranged from 3.94 to 8.30 g O₂ m⁻² d⁻¹ in the WOH streams and from 1.39 to 6.12 g O₂ m⁻² d⁻¹ in the EOH streams. All streams were heterotrophic. The WOH streams were larger and more open than the EOH streams. Metabolism was strongly correlated with instream environmental and water-chemistry variables and riparian shade. Land use was largely forested with some agriculture in the WOH watersheds, and it was forested or urbanized in EOH watersheds. Landuse impacts were confounded by the smaller size and denser shade along EOH streams than along WOH streams.

Primary productivity and chlorophyll a concentrations were measured in 8 reservoirs in New York City drinking-water-supply watersheds. The light-and-dark bottle O₂-change procedure was used to measure gross primary productivity (GPP) once each summer from 2000 to 2002. GPP normalized for photosynthetically active radiation (PAR) in the Neversink and Schoharie averaged only 0.025 and 0.035 g O₂/mol quanta, respectively. Values for New Croton and Cannonsville averaged 0.118 and 0.125 g O₂/mol quanta, respectively. Values in the other reservoirs (west basin Ashokan, Pepacton, Rondout, and Kensico) were intermediate. Chlorophyll a concentrations in reservoir photic zones ranged from mean values of <10 to 100 mg/m², with highest values in New Croton and Cannonsville and lowest concentrations in Neversink, Pepacton, and Schoharie. Cannonsville was eutrophic, and New Croton was at the mesotrophic–eutrophic boundary. Neversink, Schoharie, and Pepacton were at the oligotrophic–mesotrophic boundary, and the remaining reservoirs (Kensico, Rondout, and west basin Ashokan) were mesotrophic. Reservoir conditions were related to watershed-scale land use. Gradients within reservoirs in chlorophyll a, depth of photic zone, and primary productivity indicated an influence of the major tributary on reservoir conditions in several of the reservoirs.

Distributing 4.5 billion liters of clean fresh water every day to >9 million New York City (NYC) and suburban residents and countless other users is an enormous task that is made even more difficult by the aspiration to supply that water without filtration. To accomplish that task with a sense of confidence requires adequate data to: 1) gauge the quality of water in the source streams, 2) measure changes in that quality over time, and 3) assess the factors that might contribute to future degradation. The primary goal of the Stroud Water Research Center's large-scale enhanced water-quality monitoring project (the Project) described in the papers in this special series was to create a baseline of water quality and ecosystem health for the streams and reservoirs that provide drinking water to NYC and to relate current conditions to land use/cover. The results show that streams and rivers located west of Hudson River (WOH) deliver good to very good water to most of the receiving reservoirs. The project confirmed the eutrophic condition of the WOH Cannonsville reservoir and further linked that condition to nutrient inputs from the West Branch Delaware River. The project also confirmed that many streams located east of Hudson River (EOH) had fair to poor water quality and that streams in the Croton and Kensico watersheds were biologically and functionally degraded. Streams in some parts of the WOH region appear to be on a trajectory toward conditions already present in streams in the EOH region. Anthropogenic changes in land use from forested to agricultural in the WOH region have affected water chemistry, macroinvertebrate community structure, and stream function, although the impact is less than that caused by changes in land use from forested to urban in the EOH region. Understanding the processes of change in both regions should improve conservation, restoration, and best management practices by revealing the causes of problems, the extent and nature of those problems, and the type of landuse conditions that lead to water-quality degradation. The addition of novel parameters such as nutrient spiraling and whole-stream metabolism to traditional biomonitoring tools has established a new bridge between basic and applied research at the ecosystem level.