Striving for consistent bioassessment across diverse landscapes: Using land use matters for classifying reference and disturbed sites for index development.

Consistency in ecological assessments is challenging across large diverse landscapes because natural geological, climatic, and hydrological factors vary greatly. As a result, large landscapes are often subdivided into ecoregions and assessments are based on ecoregion specific indices. In the present study, we developed and compared multimetric indices (MMIs) using benthic diatom data from the 2008-2009 dataset from the United States (US) National Rivers and Streams Assessment. Nationwide and separate ecoregion specific MMIs were developed with reference, moderately disturbed, and highly disturbed sites selected using criteria based on physicochemical condition of the habitat or based on watershed land use (% agriculture and % urban). Metrics were adjusted to account for variation in natural conditions when needed. We found only land use criteria selected reference sites with consistently low median % watershed disturbance (%WD) and large differences in %WD between reference and highly disturbed sites. <38 % of sites were identified as reference or highly disturbed by both physicochemical and land use criteria. All MMIs displayed substantial discrimination ability between reference and highly disturbed sites. At the national scale, MMIs based on land use outperformed MMIs based on physicochemical conditions for all performance attributes tested. When national scale MMIs were applied to ecoregions, MMIs based on land use were again better than MMIs based on physicochemical conditions for most performance attributes and even had better or comparable performance to the land use MMIs developed separately for each ecoregion. Our findings show that developing MMIs using land use criteria and adjusting metrics for natural variation could improve assessment consistency without losing MMI performance across large, diverse landscapes as in the US National Rivers and Streams Assessment.

Getting off the ladder: Disentangling water quality indices to enhance the valuation of divergent ecosystem services.

Many water quality valuation studies and Federal cost-benefit analyses build from pioneering work using a "water quality ladder" or a single water quality index (WQI) to characterize both current conditions and effects of policies. When policies lead to contrasting changes in valued ecosystem services like recreational fishing and swimming, analyses using a single ladder or index might obscure important underlying service trade-offs. We test for this effect using alternative approaches that separate water quality indices and value changes in distinct ecosystem services stemming from policies with small to moderate changes in water quality. The indices we test relate to nutrient loadings in Michigan's rivers, lakes, and Great Lakes. Our split-sample experiment compares economic values for treatments with two versus three quality metrics. The key distinction is that the two-index survey, like many existing studies, aggregates subindices for water contact (for swimming and boating) and fish biomass scores (for fishing) into a single WQI, whereas the three-index survey separately utilizes both. We find that changes in our index reflecting changes in fecal bacteria and water clarity are valued differently from changes in our recreational fishing index. Aggregating changes in these two distinct recreational services using a single WQI yields consistently lower benefit estimates across a range of underlying changes in our experiment. In valuation scenarios with small changes in overall water quality, the WQI-based benefit estimates can differ substantially from benefits measured by decomposing the index and valuing the disparate subindices, differences which might change balance of benefits and costs in regulatory evaluations.

Model-based assessment and mapping of total phosphorus enrichment in rivers with sparse reference data.

Water nutrient management efforts are frequently coordinated across thousands of water bodies, leading to a need for spatially extensive information to facilitate decision making. Here we explore potential applications of a machine learning model of river low-flow total phosphorus (TP) concentrations to support landscape nutrient management. The model was trained, validated, and then applied for all rivers of Michigan, USA to identify potential drivers of nutrient variation, predict alteration in nutrient concentrations from minimally disturbed conditions, and explore reach-specific sensitivity to riparian agricultural change. A boosted regression tree model of low-flow TP concentrations trained on natural and anthropogenic landscape predictors accounted for 53 % of variation in cross-validation data, had good accuracy, little bias, and plausible relationships between predictors and response. Percent riparian agricultural cover accounted for the greatest root mean square error reduction in the modeled response (33.2 %), followed by riparian soil permeability (12.9 %), watershed slope (9.6 %), and percent urban cover (9.6 %). An apparent non-linear relationship between TP concentrations and percent riparian agricultural cover suggested steep positive increases in stream TP concentrations between 10 and 30 % upstream riparian agricultural cover. Predicted minimally disturbed TP concentrations were spatially variable and ranged from 7.0 to 48.5 µg/L, with the highest concentrations in watersheds draining low-permeability lake plain soils. Comparison of minimally disturbed predictions to those from the early 2000s suggested that much of northern Michigan existed close to the reference condition, while southern Michigan streams were often substantially enriched. Our predicted values of minimally disturbed condition generally agreed with previous studies but offer greater geographic specificity. Expanded application of machine learning modeling with landscape predictor data have great potential to inform stream nutrient strategy development in settings with sparse reference data.

The importance of natural versus human factors for ecological conditions of streams and rivers.

Streams are influenced by watershed-scale factors, such as climate, geology, topography, hydrology, and soils, which mostly vary naturally among sites, as well as human factors, agriculture and urban development. Thus, natural factors could complicate assessment of human disturbance. In the present study, we use structural equation modeling and data from the 2008-2009 United States National Rivers and Streams Assessment to quantify the relative importance of watershed-scale natural and human factors for in-stream conditions. We hypothesized that biological condition, represented using a diatom multimetric index (MMI), is directly affected by in-stream physicochemical environment, which in turn is regulated by natural and human factors. We evaluated this hypothesis at both national and ecoregion scales to understand how influences vary among regions. We found that direct influences of in-stream environment on diatom MMIs were greater than natural and human factors at the national scale and in all but one ecoregion. Meanwhile, in-stream environments were jointly explained by natural variations in precipitation, base flow index, hydrological stability, % volcanic rock, soil water table depth, and soil depth and by human factors measured as % crops, % other agriculture, and % urban land use. The explained variance of in-stream environment by natural and human factors ranged from 0.30 to 0.75, for which natural factors independently accounted for the largest proportion of explained variance at the national scale and in seven ecoregions. Covariation between natural and human factors accounted for a higher proportion of explained variance of in-stream environment than unique effects of human factors in most ecoregions. Ecoregions with relatively weak effects by human factors had relatively high levels of covariance, high levels of human disturbance, or small ranges in human disturbance. We conclude that accounting for effects of natural factors and their covariation with human factors will be important for accurate ecological assessments.

Relationships between diatom metrics based on species nutrient traits and agricultural land use.

We assessed how diatom metrics were related to different ranges of agricultural land use. Diatom assemblage composition, nutrients, and landscape characteristics were determined at 232 sites in eight agriculturally dominated study areas of the continental United States. Two regional groups based on differences in diatom relations to human disturbance were determined. Changes in diatom species composition were related to nutrients, pH, and conductivity in the eastern study areas (due to more wetlands) and more exclusively to nutrients in the west-central study areas. Homogenization of diatom flora among streams was related to high agricultural disturbance at this transcontinental scale. Species traits were developed separately for the east and west-central study groups and calculated two ways: indicator species analysis for taxa in low and high TN or TP conditions and weighted average partial least squares models of TN and TP concentration. These diatom metrics were significantly related to many indicators of agricultural land use in watersheds, especially percent row crops. Further analysis was conducted on only the west-central region due to its larger sample size. Overall, diatom metrics using species responses to N gradients were better related to agricultural land use than were species responses to P gradients. Most nutrient-based diatom metrics changed greatly in response to low ranges of percent row crops, but only a few high N diatom metrics responded to high row crop conditions. The greater response of diatoms to changes in low agriculture conditions may be due to past diatom evolution occurring when most waters had low nutrient conditions.

Advancing evaluation of bioassessment methods: A reply to Liu and Cao.

A series of three papers was written about the development of multimetric indices (MMIs) using diatoms in rivers, streams and lakes for transcontinental surveys conducted by the United States Environmental Protection Agency. Stevenson et al. (2013) used the surface sediment diatom data from the 2007 National Lake Assessment to develop national scale site specific models for MMIs to account for natural variation in condition among sites. Liu and Stevenson (2017) also used the 2007 lakes data to evaluate performance of MMIs by grouping sites by ecoregions or typologies (naturally similar types of lakes defined by similarity in diatom species composition) with site specific metric models (SSMMs) that adjust metrics for natural variability among sites. Tang et al. (2016) used benthic diatom data from the 2008-2009 National River and Stream Assessment to develop SSMMs and MMIs by ecoregion and typology. All three studies showed that SSMMs improved performance of diatom MMIs by accounting for natural variation among sites. None of the studies provided consistent evidence that grouping sites by typologies produced better MMI performance than grouping sites by ecoregions. Liu and Cao (2018) criticized the Tang et al. (2016) paper for using means and standard errors to evaluate relative performance of MMI calculation methods at the site group scale, however, their criticism is incorrect. Actually, Tang et al. (2016) only used means to summarize and report relative performance of MMI calculation methods in the body of the paper. Tang et al. (2016) appropriately used non-parametric rank sum approaches to evaluate the probability that the multiple MMI calculations for separate site groups were the same for ecoregion (n = 9) and typology (n = 7) site groups. Liu and Stevenson (2017) used this same non-parametric approach for tests of lake diatom MMIs. Liu and Cao's (2018) concerns can be addressed by distinguishing between the goals and methods used for testing and evaluation of MMI calculation methods at the national and site-group scales. Tang et al. (2016) did not aggregate data across site groups to test MMI performance at the national scale because they were following standard EPA methods that develop separate MMIs for each site group. In conclusion, Liu and Cao (2018) misunderstood the MMI evaluation in Tang et al. (2016) and added no new information to this body of work, because all the concerns they raised were discussed in Liu and Stevenson (2017).

Eutrophication and recovery of a Lake inferred from sedimentary diatoms originating from different habitats.

Eutrophication and recovery of Muskegon Lake resulted from a complex set of interacting factors according to diatom-inferred total phosphorus (TP), geochemical proxies, detailed modeling of land use land cover change in the watershed, and accounts of past point source management and non-native species invasions. Benthic and planktonic diatoms responded to phosphorus environments differently in this lake that receives 95% of its input from one river and has only a 23days average retention time. Planktonic diatoms reflected river conditions more than benthic diatoms, and benthos reflected lake conditions more than plankton. Inferred TP from planktonic diatoms indicated the Muskegon River was relatively nutrient rich compared to inferred TP for Muskegon Lake based on benthic diatoms before Europeans settled the watershed. Early European settlement and logging caused no changes in phosphorus condition in the Muskegon River, but modest increases in phosphorus were indicated in Muskegon Lake during the middle and late thirds of the 19th century. Extensive watershed-scale agricultural activity in the early 20th century apparently had little effect on trophic status of the lake, perhaps because it preceded high fertilizer use on farms. During the industrial and population boom in the watershed during the early half of the 20th century, river conditions changed little, but eutrophication of Muskegon Lake increased greatly. Reduction in river phosphorus by dams occurred during the first half of the 20th century. Phosphorus reduction in the lake was indicated after advanced wastewater treatment for Muskegon Township was implemented in 1973. Current diatom inferred phosphorus concentration in the lake is the same as before European settlement, however many attributes of the lake still differ because other stressors persist.

Isothermal amplification of environmental DNA (eDNA) for direct field-based monitoring and laboratory confirmation of Dreissena sp.

Loop-mediated isothermal amplification (LAMP) of aquatic invasive species environmental DNA (AIS eDNA) was used for rapid, sensitive, and specific detection of Dreissena sp. relevant to the Great Lakes (USA) basin. The method was validated for two uses including i) direct amplification of eDNA using a hand filtration system and ii) confirmation of the results after DNA extraction using a conventional thermal cycler run at isothermal temperatures. Direct amplification eliminated the need for DNA extraction and purification and allowed detection of target invasive species in grab or concentrated surface water samples, containing both free DNA as well as larger cells and particulates, such as veligers, eggs, or seeds. The direct amplification method validation was conducted using Dreissena polymorpha and Dreissena bugensis and uses up to 1 L grab water samples for high target abundance (e.g., greater than 10 veligers (larval mussels) per L for Dreissena sp.) or 20 L samples concentrated through 35 µm nylon screens for low target abundance, at less than 10 veligers per liter water. Surface water concentrate samples were collected over a period of three years, mostly from inland lakes in Michigan with the help of a network of volunteers. Field samples collected from 318 surface water locations included i) filtered concentrate for direct amplification validation and ii) 1 L grab water sample for eDNA extraction and confirmation. Though the extraction-based protocol was more sensitive (resulting in more positive detections than direct amplification), direct amplification could be used for rapid screening, allowing for quicker action times. For samples collected between May and August, results of eDNA direct amplification were consistent with known presence/absence of selected invasive species. A cross-platform smartphone application was also developed to disseminate the analyzed results to volunteers. Field tests of the direct amplification protocol using a portable device (Gene-Z) showed the method could be used in the field to obtain results within one hr (from sample to result). Overall, the direct amplification has the potential to simplify the eDNA-based monitoring of multiple aquatic invasive species. Additional studies are warranted to establish quantitative correlation between eDNA copy number, veliger, biomass or organismal abundance in the field.

Improving assessment accuracy for lake biological condition by classifying lakes with diatom typology, varying metrics and modeling multimetric indices.

Site grouping by regions or typologies, site-specific modeling and varying metrics among site groups are four approaches that account for natural variation, which can be a major source of error in ecological assessments. Using a data set from the 2007 National Lakes Assessment project of the USEPA, we compared performances of multimetric indices (MMI) of biological condition that were developed: (1) with different lake grouping methods, ecoregions or diatom typologies; (2) by varying or not varying metrics among site groups; and (3) with different statistical techniques for modeling diatom metric values expected for minimally disturbed condition for each lake. Hierarchical modeling of MMIs, i.e. grouping sites by ecoregions or typologies and then modeling natural variability in metrics among lakes within groups, substantially improved MMI performance compared to using either ecoregions or site-specific modeling alone. Compared with MMIs based on ecoregion site groups, MMI precision and sensitivity to human disturbance were better when sites were grouped by diatom typologies and assessing performance nationwide. However, when MMI performance was evaluated at site group levels, as some government agencies often do, there was little difference in MMI performance between the two site grouping methods. Low numbers of reference and highly impacted sites in some typology groups likely limited MMI performance at the group level of analysis. Varying metrics among site groups did not improve MMI performance. Random forest models for site-specific expected metric values performed better than classification and regression tree and multiple linear regression, except when numbers of reference sites were small in site groups. Then classification and regression tree models were most precise. Based on our results, we recommend hierarchical modeling in future large scale lake assessments where lakes are grouped by ecoregions or diatom typologies and site-specific metric models are used to establish expected metric values.

Effect of phosphorus stress on Microcystis aeruginosa growth and phosphorus uptake.

This study was designed to advance understanding of phosphorus regulation of Microcystis aeruginosa growth, phosphorus uptake and storage in changing phosphorus (P) conditions as would occur in lakes. We hypothesized that Microcystis growth and nutrient uptake would fit classic models by Monod, Droop, and Michaelis-Menten in these changing conditions. Microcystis grown in luxury nutrient concentrations was transferred to treatments with phosphorus concentrations ranging from 0-256 µg P∙L-1 and luxury nitrogen. Dissolved phosphorus concentration, cell phosphorus quota, P uptake rate and cell densities were measured at day 3 and 6. Results showed little relationship to predicted models. Microcystis growth was asymptotically related to P treatment from day 0-3, fitting Monod model well, but negatively related to P treatment and cell quota from day 3-6. From day 0-3, cell quota was negatively related to P treatments at <2 µg∙L-1, but increased slightly at higher P. Cell quota decreased greatly in low P treatments from day 3-6, which may have enabled high growths in low P treatments. P uptake was positively and linearly related to P treatment during both periods. Negative uptake rates and increases in measured culture phosphorus concentrations to 5 µg∙L-1 in the lowest P treatments indicated P leaked from cells into culture medium. This leakage during early stages of the experiment may have been sufficient to stimulate metabolism and use of intracellular P stores in low P treatments for rapid growth. Our study shows P regulation of Microcystis growth can be complex as a result of changing P concentrations, and this complexity may be important for modeling Microcystis for nutrient and ecosystem management.

Accounting for regional variation in both natural environment and human disturbance to improve performance of multimetric indices of lotic benthic diatoms.

Regional variation in both natural environment and human disturbance can influence performance of ecological assessments. In this study we calculated 5 types of benthic diatom multimetric indices (MMIs) with 3 different approaches to account for variation in ecological assessments. We used: site groups defined by ecoregions or diatom typologies; the same or different sets of metrics among site groups; and unmodeled or modeled MMIs, where models accounted for natural variation in metrics within site groups by calculating an expected reference condition for each metric and each site. We used data from the USEPA's National Rivers and Streams Assessment to calculate the MMIs and evaluate changes in MMI performance. MMI performance was evaluated with indices of precision, bias, responsiveness, sensitivity and relevancy which were respectively measured as MMI variation among reference sites, effects of natural variables on MMIs, difference between MMIs at reference and highly disturbed sites, percent of highly disturbed sites properly classified, and relation of MMIs to human disturbance and stressors. All 5 types of MMIs showed considerable discrimination ability. Using different metrics among ecoregions sometimes reduced precision, but it consistently increased responsiveness, sensitivity, and relevancy. Site specific metric modeling reduced bias and increased responsiveness. Combined use of different metrics among site groups and site specific modeling significantly improved MMI performance irrespective of site grouping approach. Compared to ecoregion site classification, grouping sites based on diatom typologies improved precision, but did not improve overall performance of MMIs if we accounted for natural variation in metrics with site specific models. We conclude that using different metrics among ecoregions and site specific metric modeling improve MMI performance, particularly when used together. Applications of these MMI approaches in ecological assessments introduced a tradeoff with assessment consistency when metrics differed across site groups, but they justified the convenient and consistent use of ecoregions.

The role of light availability and herbivory on algal responses to nutrient enrichment in a riparian wetland, Alaska.

We investigated how the relative availability of solar radiation in the presence or absence of grazing alters the ability of benthic algae to respond to nutrient enrichment in an Alaskan marsh. We used a factorial mesocosm experiment that included nutrient enrichment (enriched or control), grazing (grazed or ungrazed), and light (unshaded or shaded) to simulate shading by macrophytes early and late in the growing season, respectively. We found stronger effects of grazers and nutrients compared to light on benthic algal biomass and taxonomic composition. Algal biomass increased in nutrient-enriched treatments and was reduced by grazing. Shading did not have an effect on algal biomass or taxonomic composition, but the concentration of chl a per algal biovolume increased with shading, demonstrating the ability of algae to compensate for changes in light availability. Algal taxonomic composition was more affected by grazer presence than nutrients or light. Grazer-resistant taxa (basal filaments of Stigeoclonium) were replaced by diatoms (Nitzschia) and filamentous green algae (Ulothrix) when herbivores were removed. The interacting and opposing influences of nutrients and grazing indicate that the algal community is under dual control from the bottom-up (nutrient limitation) and from the top-down (consumption by herbivores), although grazers had a stronger influence on algal biomass and taxonomic composition than nutrient enrichment. Our results suggest that low light availability will not inhibit the algal response to elevated nutrient concentrations expected with ongoing climate change, but grazers rapidly consume algae following enrichment, masking the effects of elevated nutrients on algal production.

Contributions of algae to GPP and DOC production in an Alaskan fen: effects of historical water table manipulations on ecosystem responses to a natural flood.

The role of algae in the metabolism of northern peatlands is largely unknown, as is how algae will respond to the rapid climate change being experienced in this region. In this study, we examined patterns in algal productivity, nutrients, and dissolved organic carbon (DOC) during an uncharacteristically wet summer in an Alaskan rich fen. Our sampling was conducted in three large-scale experimental plots where water table position had been manipulated (including both drying and wetting plots and a control) for the previous 4 years. This study allowed us to explore how much ecosystem memory of the antecedent water table manipulations governed algal responses to natural flooding. Despite no differences in water table position between the manipulated plots at the time of sampling, algal primary productivity was consistently higher in the lowered water table plot compared to the control or raised water table plots. In all plots, algal productivity peaked immediately following seasonal maxima in nutrient concentrations. We found a positive relationship between algal productivity and water-column DOC concentrations (r (2) = 0.85, P < 0.001). Using these data, we estimate that algae released approximately 19% of fixed carbon into the water column. Algal exudates were extremely labile in biodegradability assays, decreasing by more than 55% within the first 24 h of incubation. We suggest that algae can be an important component of the photosynthetic community in boreal peatlands and may become increasingly important for energy flow in a more variable climate with more intense droughts and flooding.

Grazer control of nutrient availability in the periphyton.

Benthic algal assemblages are regulated by both abiotic (e.g., nutrient) and biotic (e.g., grazing) constraint. The objective of this study was to determine how changes in these two factors affected the structure of an algal assemblage in an ephemeral stream. Coverslips were incubated for 21 days in enclosures containing one of three nutrient environments (ambient, phosphorus-enriched, or phosphorus and nitrogen enriched) and one of four densities of the snail Gonibasis (0, 40, 80, or 120 snails/m2) and examined directly to enumerate the algal assemblage. The effect of grazing on algal biomass was dependent on the nutrient environment. An overstory of diatoms was susceptible to removal by grazing and was not strongly affected by nutrient enrichment. An understory of Stigeoclonium was more resistant to grazing and responded strongly to nutrient enrichment only in the presence of grazers. Snail grazers may mediate nutrient availability to the understory indirectly by removing overlying cells or by direct excretion of nutrients. Multiple interactions occur between benthic herbivores and algae, and, as shown here, some of them are positive and involve modifications of the nutrient environment.