Search

Excess phosphorus loading to waterbodies has led to increasing frequency and severity of harmful algal blooms, negatively impacting economic activity and human health. While interventions to improve water quality can create large societal benefits, these investments are costly and the value of benefits is often unknown. Understanding the social and economic impacts of reduced phosphorus loading is critical for developing effective land use policies and for generating public and political support for these initiatives. Here, we quantify the social benefits and costs of improving water quality in Lake Champlain under a range of phosphorus reduction and climate change scenarios between 2016 and 2050. We use statistical models to link water quality outputs from an established integrated assessment model with three categories of benefits: tourism expenditures, property sales, and avoided human health impacts. We estimate the costs of reducing phosphorus loading using data reported by the State of Vermont. We find that under the most aggressive phosphorus reduction scenario, the total benefits of improved water quality are $55 to $60 million between 2016 and 2050. Over this 35 year time horizon, the combined benefits do not outweigh the costs under any scenario. If the time horizon is extended to 2100 or beyond, however, the benefits may exceed the costs if the applied discount rate is less than 3%. Importantly, we almost certainly underestimate the value of clean water, due to the omission of other types of benefits. Despite this uncertainty, our study provides a tractable framework for disentangling the complex relationships between water quality and human well-being, and illuminates the value of reductions in phosphorus loading to society.

Conversation has been a primary means for the exchange of information since ancient times. Understanding patterns of information flow in conversations is a critical step in assessing and improving communication quality. In this paper, we describe COnversational DYnamics Model (CODYM) analysis, a novel approach for studying patterns of information flow in conversations. CODYMs are Markov Models that capture sequential dependencies in the lengths of speaker turns. The proposed method is automated and scalable, and preserves the privacy of the conversational participants. The primary function of CODYM analysis is to quantify and visualize patterns of information flow, concisely summarized over sequential turns from one or more conversations. Our approach is general and complements existing methods, providing a new tool for use in the analysis of any type of conversation. As an important first application, we demonstrate the model on transcribed conversations between palliative care clinicians and seriously ill patients. These conversations are dynamic and complex, taking place amidst heavy emotions, and include difficult topics such as end-of-life preferences and patient values. We perform a versatile set of CODYM analyses that (a) establish the validity of the model by confirming known patterns of conversational turn-taking and word usage, (b) identify normative patterns of information flow in serious illness conversations, and (c) show how these patterns vary across narrative time and differ under expressions of anger, fear and sadness. Potential applications of CODYMs range from assessment and training of effective healthcare communication to comparing conversational dynamics across language and culture, with the prospect of identifying universal similarities and unique "fingerprints" of information flow., Comment: 34 pages, 20 figures, submitted to PLOS One (in review)

Given the limited resources available for managing erosion hazards and addressing water quality impairment along rivers, stakeholders engaged in water resource management would benefit from tools to identify those river reaches most prone to adjustment and which disproportionately load sediment to receiving waters. The extent and rate of vertical and lateral channel adjustments in response to natural and human disturbances vary considerably across space and time; and this complexity and nonlinearity introduce challenges for classification or modeling of river reaches using conventional statistical techniques. The Self-Organizing Map (SOM) is a data-driven computational tool with advantages for clustering or classifying multivariate observations and for exploratory data analysis and visualization of complex, nonlinear systems. We applied a SOM to cluster multivariate stream geomorphic assessment data into reach-scale sediment process domains for 193 river reaches in glacially-conditioned catchments of northeastern US using field- and GIS-derived hydraulic and geomorphic parameters. The reaches comprised a range of channel types from confined to unconfined, steep- to shallow-gradient, mid-to-high order, and bedrock to alluvial channels. Fifteen variables were identified that meaningfully separated reaches into seven sediment regimes, following a two-stage application of the SOM. A coarse-tune SOM identified sediment regime classes at the supply-limited and transport-limited extremes of a continuum, including bedrock channels and confined, steep-gradient reaches as well as braided, depositional channels at alluvial fan or delta settings. A second-stage, fine-tune SOM nuanced differences in sediment production and transport for unconfined reaches with varying degrees of floodplain disconnection resulting from natural or human stressors. This classification framework is transferable to other hydroclimatic regions, with consideration of additional or alternate independent variables unique to those regions, and can provide valuable insights for river management to promote flood resiliency, restore water quality and improve instream and riparian habitats.

It is well established that cyanobacteria populations in shallow lakes exhibit dramatic fluctuations on both interannual and intraannual timescales; however, despite extensive research, disentangling the drivers of interannual variability in bloom severity has proved challenging. Critical thresholds of abiotic drivers such as wind, irradiance, air temperature, and tributary inputs may control the development and collapse of blooms, but these thresholds are difficult to identify in large and complex datasets. In this study, we compared high-frequency estimates of oxygen metabolism in a shallow bay of Lake Champlain to concurrent measurements of physical and chemical parameters over 3 years with very different bloom dynamics. We clustered the data using supervised and unsupervised self-organizing maps to identify the environmental drivers associated with key stages of bloom development. We then used threshold analysis to identify subtle yet important thresholds of thermal stratification that drive t...

The leading cause of bridge failure has often been identified as bridge scour, which is generally defined as the erosion or removal of streambed and/or bank material around bridge foundations due to flowing water. These scour critical bridges are particularly vulnerable during extreme flood events, and pose a major risk to human life, transportation infrastructure, and economic sustainability. Retrofitting the thousands of undersized and scour critical bridges to more rigorous standards is prohibitively expensive requiring effective yet economical countermeasures. This research tested the efficacy of using approach embankments as intentional sacrificial “fuses” to protect the bridge integrity and minimize damage during large flow events by allowing the streams to access their natural floodplain and reduce channel velocities. This countermeasure concept was evaluated using the Hydrologic Engineering Center’s River Analysis System models. Steady flow models were developed for three specific bridges on two river reaches. Streamflow return period estimators for both river reaches were developed using Bayesian analysis and available United States Geological Survey stream gauge data to evaluate sacrificial embankments under non-stationary climatic conditions. The use of sacrificial embankments was determined to be a cost-effective scour mitigation strategy for bridges with suboptimal hydraulic capacity and unknown or shallow foundations. Additional benefits of sacrificial embankments include reductions in upstream flood stage and velocity.

Structural alterations to bridges may result in unintended consequences up- or downstream, such as changes in discharge, velocity, stream power, and water levels. This work presents a framework and methodology to analyze interconnected bridge-stream interactions under flood conditions. Such analysis may help prioritize limited resources available for bridge and river rehabilitations, facilitate holistic design of bridges and better-informed cost/benefit analyses, and address stakeholder concerns raised in response to planned bridge and infrastructure alterations. A two-dimensional unsteady HEC-RAS hydraulic model of the Otter Creek between Rutland and Middlebury, VT is used to simulate the impact of individual structures on the bridge-stream network, as well as potential sensitivity to those impacts, during extreme flood events. The presence of a bridge and approach roadway may induce measurable changes in peak discharge throughout the entire 46 miles of modeled river. These may be by as much as 10% at adjacent structures, down to 1% at structures as far as six miles upstream and nine miles downstream. Depending on their characteristics, bridges and roadways may either attenuate or amplify peak flood flows up- and downstream, or have little to no effect at all, suggesting that there is no easily predictable impact, and that hydraulic modeling is necessary for such analysis on rivers. Alterations to structures that develop substantial backwaters may result in the most dramatic impacts to the network, which can be both positive and negative. Affected bridges may or may not be sensitive to these changes in discharge; structures that do not experience relief (e.g., roadway overtopping) may be most sensitive to any distant perturbations.

We propose a new evolutionary approach for discovering causal rules in complex classification problems from batch data. Key aspects include (a) the use of a hypergeometric probability mass function as a principled statistic for assessing fitness that quantifies the probability that the observed association between a given clause and target class is due to chance, taking into account the size of the dataset, the amount of missing data, and the distribution of outcome categories, (b) tandem age-layered evolutionary algorithms for evolving parsimonious archives of conjunctive clauses, and disjunctions of these conjunctions, each of which have probabilistically significant associations with outcome classes, and (c) separate archive bins for clauses of different orders, with dynamically adjusted order-specific thresholds. The method is validated on majority-on and multiplexer benchmark problems exhibiting various combinations of heterogeneity, epistasis, overlap, noise in class associations, missing data, extraneous features, and imbalanced classes. We also validate on a more realistic synthetic genome dataset with heterogeneity, epistasis, extraneous features, and noise. In all synthetic epistatic benchmarks, we consistently recover the true causal rule sets used to generate the data. Finally, we discuss an application to a complex real-world survey dataset designed to inform possible ecohealth interventions for Chagas disease.

Hydrological storm events are a primary driver for transporting water quality constituents such as suspended sediments and nutrients. Analyzing the concentration (C) of these water quality constituents in response to river discharge (Q), particularly when monitored at high temporal resolution during a hydrological event, helps to characterize the dynamics and flux of such constituents. A conventional approach to storm event analysis is to reduce C-Q time series to two-dimensional (2-D) hysteresis loops and analyze these 2-D patterns. While informative, this hysteresis loop approach has limitations because projecting the C-Q time series onto a 2-D plane obscures detail (e.g., temporal variation) associated with the C-Q relationships. In this paper, we address this limitation using a multivariate event time series (METS) clustering approach that is validated using synthetically generated event times series. The METS clustering is then applied to river discharge and suspended sediment data (acquired through turbidity-based monitoring) from six watersheds in the Lake Champlain Basin located in the northeastern United States, and results in identifying four common types of hydrological water quality events. Statistical analysis on the events partitioned by both methods (METS clustering and 2-D hysteresis classification) helped identify hydrometeorlogical features of common event types. In addition, the METS and hysteresis analysis were simultaneously applied to a regional Vermont dataset to highlight the complimentary nature of using them in tandem for hydrological event analysis.

Characterization of surface gas permeability measurements on a variety of natural and engineered building materials using two relatively new, non-destructive surface permeameters is presented. Surface gas permeability measurements were consistent for both laboratory and field applications and correlated well with bulk gas permeability measurements. This research indicates that surface permeability measurements could provide reliable estimates of bulk gas permeability; and due to the non-destructive nature and relative sampling ease of both surface gas permeability tools, it is possible to quantify the range of the spatial autocorrelation, heterogeneity, and anisotropy in porous building materials and their degree of degradation from weathering.

Floodplain restoration offers an opportunity to enhance communities’ resilience to flooding. However, the degree to which these interventions mitigate damages is often unknown, and identifying the best locations for implementation is a challenge. Further, the extent to which the benefits of flood mitigation are equitably distributed within communities is rarely considered in restoration projects. Here, we develop a novel framework to optimize investments in floodplain restoration that maximizes the utility of avoided damages from flood inundation for a range of budgetary constraints. We estimate the expected reduction in flood damages from restoration interventions by integrating a hydraulic flood model and an economic damage cost model. Using equity-weighted utility functions, we explicitly evaluate how the value of reduction in flood damages varies for different property owners. We demonstrate the potential of this approach in the Lewis Creek watershed, located in Vermont, USA. Under all optimal scenarios, the benefits of avoided flood damages over a 100-year time period outweigh the costs of restoration by at least 5-to-1. Floodplain restoration has the potential to reduce the present value of damages by up to $400,000, a 5% decrease from the baseline, at a cost of only $75,000. We also show that the equity-weighted utility of flood mitigation increase when restoration interventions protect the lowest-income property owners, particularly those who live in mobile homes. Together, our results illuminate the importance of evaluating the distribution of benefits and costs associated with alternative restoration strategies, as well as underscore the capacity for floodplain restoration to build resilience to flooding.

Chagas disease is caused by the protozoan parasite Trypanosoma cruzi and transmitted by triatomine insect vectors. In Guatemala, insecticide spraying is an integral part of management of the main vector, Triatoma dimidiata. Spraying typically has low efficacy, which may be due to incomplete elimination from infested houses, within-village dispersal, or influx from other villages or sylvan environments. To evaluate how these mechanisms contribute to reinfestation, we conducted a time-course analysis of T. dimidiata infestation, abundance and household genetic structure in two nearby villages in Jutiapa, Guatemala; houses in the first village were surveyed, treated with insecticide if infested and then re-surveyed at eight and 22 months following spraying, while the second village served as an untreated control to quantify changes associated with seasonal dispersal. Insects were genotyped at 2–3000 SNP loci for kinship and population genetic analyses. Insecticide application reduced overall infestation and abundance, while the untreated village was stable over time. Nevertheless, within two years 35.5% of treated houses were reinfested and genetic diversity had largely recovered. Insects collected from reinfested houses post-spraying were most closely related to pre-spray collections from the same house, suggesting that infestations had not been fully eliminated. Immigration by unrelated insects was also detected within a year of spraying; when it occurred, dispersal was primarily local from neighboring houses. Similar dispersal patterns were observed following the annual dispersal season in the untreated village, with high-infestation houses serving as sources for neighboring homes. Our findings suggest that the efficacy of pyrethroid application is rapidly diminished by both within-house breeding by survivors and annual cycles of among-house movement. Given these patterns, we conclude that house structural improvements, an integral part of the Ecohealth approach that makes houses refractory to vector colonization and persistence, are critical for long-term reduction of T. dimidiata infestation.

Chagas disease, considered a neglected disease by the World Health Organization, is caused by the protozoan parasite Trypanosoma cruzi, and transmitted by >140 triatomine species across the Americas. In Central America, the main vector is Triatoma dimidiata, an opportunistic blood meal feeder inhabiting both domestic and sylvatic ecotopes. Given the diversity of interacting biological agents involved in the epidemiology of Chagas disease, having simultaneous information on the dynamics of the parasite, vector, the gut microbiome of the vector, and the blood meal source would facilitate identifying key biotic factors associated with the risk of T. cruzi transmission. In this study, we developed a RADseq-based analysis pipeline to study mixed-species DNA extracted from T. dimidiata abdomens. To evaluate the efficacy of the method across spatial scales, we used a nested spatial sampling design that spanned from individual villages within Guatemala to major biogeographic regions of Central America. Information from each biotic source was distinguished with bioinformatics tools and used to evaluate the prevalence of T. cruzi infection and predominant Discrete Typing Units (DTUs) in the region, the population genetic structure of T. dimidiata, gut microbial diversity, and the blood meal history. An average of 3.25 million reads per specimen were obtained, with approximately 1% assigned to the parasite, 20% to the vector, 11% to bacteria, and 4% to putative blood meals. Using a total of 6,405 T. cruzi SNPs, we detected nine infected vectors harboring two distinct DTUs: TcI and a second unidentified strain, possibly TcIV. Vector specimens were sufficiently variable for population genomic analyses, with a total of 25,710 T. dimidiata SNPs across all samples that were sufficient to detect geographic genetic structure at both local and regional scales. We observed a diverse microbiotic community, with significantly higher bacterial species richness in infected T. dimidiata abdomens than those that were not infected. Unifrac analysis suggests a common assemblage of bacteria associated with infection, which co-occurs with the typical gut microbial community derived from the local environment. We identified vertebrate blood meals from five T. dimidiata abdomens, including chicken, dog, duck and human; however, additional detection methods would be necessary to confidently identify blood meal sources from most specimens. Overall, our study shows this method is effective for simultaneously generating genetic data on vectors and their associated parasites, along with ecological information on feeding patterns and microbial interactions that may be followed up with complementary approaches such as PCR-based parasite detection, 18S eukaryotic and 16S bacterial barcoding., Author summary Chagas disease is caused by the parasite Trypanosoma cruzi, which is spread by triatomine kissing bugs. There are many biotic factors that influence the risk of disease transmission, including the strain of the parasite, the vector movement patterns, the community of microbes interacting with the parasite inside the vector's gut, and the availability of suitable vertebrate hosts. DNA from all of these species can be found in the gut of an infected bug, providing an opportunity to investigate all of them simultaneously by genetically analyzing this single tissue. In this study, we developed a DNA-based method to retrieve, separate, and analyze genetic information from the abdomens of 32 T. dimidiata kissing bug vectors collected across Central America. We found two distinct strains of T. cruzi, and four T. dimidiata genetic clusters associated with environmental and geographical characteristics. These populations harbored different bacterial gut communities that were augmented by specifically infection-associated bacteria when the vector was infected by the parasite. In some cases, we could identify what the insect had recently fed on, including chicken, duck, dog and human. Having simultaneous information on all of these organisms may help to fine-tune control strategies that influence the risk of T. cruzi transmission.

Genetic studies are recognized increasingly as important for understanding naturally occurring disease dynamics and are used to predict host genetic diversity and coevolutionary processes and to identify species composition in ecological communities. Tubifex tubifex, the definitive host of the whirling disease parasite Myxobolus cerebralis, comprises 6 known lineages that vary widely in parasite susceptibility. We used 16S ribosomal DNA (16S rDNA) to identify relationships among genetic variability of 3 oligochaete genera (T. tubifex, Rhyacodrilus spp., and Ilyodrilus spp.; Oligochaeta:Tubificidae), oligochaete assemblage composition, and the presence of whirling disease in 9 locations across 4 watersheds in Montana, USA. We assessed genetic variability among 183 tubificid worms from locations classified as positive or negative for whirling disease based on 5 to 8 y of monitoring by the Montana Department of Fish, Wildlife, and Parks. Within genera, we found 2 groups of T. tubifex (lineages I and ...

The increasing incidence of thyroid cancer has resulted in the rate tripling over the past 30 years. Reasons for this increase have not been established. Geostatistics and geographic information system (GIS) tools have emerged as powerful geospatial technologies to identify disease clusters, map patterns and trends, and assess the impact of ecological and socioeconomic factors (SES) on the spatial distribution of diseases. In this study, these tools were used to analyze thyroid cancer incidence in a rural population.Thyroid cancer incidence and socio-demographic factors in Vermont (VT), United States, between 1994 and 2007 were analyzed by logistic regression and geospatial and temporal analyses.The thyroid cancer age-adjusted incidence in Vermont (8.0 per 100,000) was comparable to the national level (8.4 per 100,000), as were the ratio of the incidence of females to males (3.1:1) and the mortality rate (0.5 per 100,000). However, the estimated annual percentage change was higher (8.3 VT; 5.7 U.S.). Incidence among females peaked at 30-59 years of age, reflecting a significant rise from 1994 to 2007, while incidence trends for males did not vary significantly by age. For both females and males, the distribution of tumors by size did not vary over time; ≤1.0 cm, 1.1-2.0 cm, and2.0 cm represented 38%, 22%, and 40%, respectively. In females, papillary thyroid cancer (PTC) accounted for 89% of cases, follicular (FTC) 8%, medullary (MTC) 2%, and anaplastic (ATC) 0.6%, while in males PTC accounted for 77% of cases, FTC 15%, MTC 1%, and ATC 3%. Geospatial analysis revealed locations and spatial patterns that, when combined with multivariate incidence analyses, indicated that factors other than increased surveillance and access to healthcare (physician density or insurance) contributed to the increased thyroid cancer incidence. Nine thyroid cancer incidence hot spots, areas with very high normalized incidence, were identified based on zip code data. Those locations did not correlate with urban areas or healthcare centers.These data provide evidence of increased thyroid cancer incidence in a rural population likely due to environmental drivers and SES. Geospatial modeling can provide an important framework for evaluation of additional associative risk factors.

Summary Although commonly used by those tasked with lake management, the statistical approach of data averaging (DA) followed by ordinary least-squares regression (OLSR) to generate nutrient limitation models is outdated and may impede the understanding and successful management of lake eutrophication. Using a 21-year data set from Lake Champlain as a case study, the traditional DA-OLSR-coupled approach was re-evaluated and improved to quantify the cause–effect relationships between chlorophyll (Chl) and total nitrogen (TN) or total phosphorus (TP). We confirmed that the commonly used DA-OLSR approach results in misleading cause–effect nutrient limitation inferences by illustrating how the process of DA reduces the range of data distribution considered and masks meaningful temporal variation observed within a given period. Our model comparisons demonstrate that using quantile regression (QR) to fit the upper boundary of the response distribution (99th quantile model) is more robust than the OLSR analysis for generating eutrophication models and developing nutrient management targets, as this method reduces the effects of unmeasured factors that plague the OLSR-derived model. Because our approach is statistically in line with the ecological ‘law of the minimum’, it is particularly powerful for inferring resource limitation with broad potential utility to the ecological research community. By integrating percentile selection (PS) with QR-derived model output, we developed a PS-QR-coupled approach to quantify the relative importance of TN and TP reductions in a eutrophic system. Utilising this approach, we determined that the reduction in TP to meet a specific Chl target should be the first priority to mitigate eutrophication in Lake Champlain. The structure of this statistically robust and straightforward approach for developing nutrient reduction targets can be easily adopted as an individual lake-specific tool for the research and management of other lakes and reservoirs with similar water quality data sets. Moreover, the PS-QR-coupled approach developed here is also of theoretical importance to understanding and modelling the interacting effects of multiple limiting factors on ecological processes (e.g. eutrophication) with broad application to aquatic research.

Over the past 30+ years, researchers and water resource managers have often relied on a set of regression-based equations to describe the relationships between secchi depth (SD), chlorophyll (Chl) and total phosphorous (TP) and quantitatively assess lake trophic status after Carlson (1977). Here, we develop a revised framework for eutrophication assessment that incorporates recent statistical advances in ecology and leverages the increasing availability of lake-specific datasets in the 21st Century. Long-term (1992–2012) water quality data from Lake Champlain (LC) are used to revisit and revise classic equations of tropic state indices (TSIChl/TP). The upper boundaries of SD–ln(Chl) and ln(Chl)–ln(TP) distributions within this dataset fit well with quantile regression (99th, QR) to generate LC-specific TSIChl/TP equations. Our results illustrate that Carlson (1977)'s original TSIChl/TP equations overestimate the trophic status of LC relative to LC-specific equations, and highlight the power of the QR-derived TSIChl/TP metric. We combine TSISD and TSIChl into one metric to indicate pseudoeutrophication and pseudomesotrophication of oligotrophic waters as well as pseudoeutrophication of mesotrophic waters to identify waters threatened by potential trophic shift. Additionally, TSIChl and TSITP were coupled as a complimentary dual metric to indicate potential risks of excessive phosphorus loading to oligotrophic and mesotrophic waters. With these dual metric schemes, we performed cluster analysis of 15 locations to spatially assess trophic status and phosphorous risks across LC. This study describes a relatively simple and robust approach for lake-specific status assessment, the structure of which can be broadly utilized within monitoring and research communities.

We present evidence of increasing persistence in daily precipitation in the northeastern United States that suggests that global circulation changes are affecting regional precipitation patterns. Meteorological data from 222 stations in 10 northeastern states are analyzed using Markov chain parameter estimates to demonstrate that a significant mode of precipitation variability is the persistence of precipitation events. We find that the largest region-wide trend in wet persistence (i.e., the probability of precipitation in 1 day and given precipitation in the preceding day) occurs in June (+0.9% probability per decade over all stations). We also find that the study region is experiencing an increase in the magnitude of high-intensity precipitation events. The largest increases in the 95th percentile of daily precipitation occurred in April with a trend of +0.7 mm/d/decade. We discuss the implications of the observed precipitation signals for watershed hydrology and flood risk.

Plug-in hybrid electric vehicles (PHEVs) offer the potential to significantly reduce greenhouse gas emissions, if vehicle consumers are willing to adopt this new technology. Consequently, there is much interest in exploring PHEV market penetration models. In prior work, we developed an agent-based model (ABM) of potential PHEV consumer adoption that incorporated several spatial, social, and media influences to identify nonlinear interactions among potential leverage points that may impact PHEV market penetration. In developing that model, the need for additional data to properly inform both the decision-making rules and agent initialization became apparent. To address these issues, we recently conducted and analyzed an extensive consumer survey; in this paper, we modify the ABM to reflect the survey findings. A unique aspect is a one-to-one correspondence between agents in the model and survey respondents, and thus yielding distributions and cross correlations in agent attributes that accurately reflect the survey population. We also implement a used-PHEV market, and allow agents to purchase new or used compact PHEVs or vehicles of their current type. Based on our prior survey response analysis, our modified model includes a PHEV-technology threshold component, a multinomial logistic prediction of willingness to consider a compact PHEV based on dynamically changing attitudes, and agent-specific delay discounting functions that predict the amount agents are willing to pay up front for greater fuel savings. We thus independently account for agents’ discomfort with the new PHEV technology, their desire to drive a more environmentally friendly vehicle, and their willingness to pay a higher sticker price for a PHEV. Results of ten survey-based ABM scenarios are reported with implications for policy-makers and manufacturers. We believe close integration of the design of consumer surveys and the development of ABMs is a key step in developing useful decision-support models; this paper serves as an example of one way to achieve that.

The East Antarctic Ice Sheet (EAIS) is the largest potential contributor to sea-level rise. However, efforts to predict the future evolution of the EAIS are hindered by uncertainty in how it responded to past warm periods, for example, during the Pliocene epoch (5.3 to 2.6 million years ago), when atmospheric carbon dioxide concentrations were last higher than 400 parts per million. Geological evidence indicates that some marine-based portions of the EAIS and the West Antarctic Ice Sheet retreated during parts of the Pliocene1,2, but it remains unclear whether ice grounded above sea level also experienced retreat. This uncertainty persists because global sea-level estimates for the Pliocene have large uncertainties and cannot be used to rule out substantial terrestrial ice loss 3 , and also because direct geological evidence bearing on past ice retreat on land is lacking. Here we show that land-based sectors of the EAIS that drain into the Ross Sea have been stable throughout the past eight million years. We base this conclusion on the extremely low concentrations of cosmogenic 10Be and 26Al isotopes found in quartz sand extracted from a land-proximal marine sediment core. This sediment had been eroded from the continent, and its low levels of cosmogenic nuclides indicate that it experienced only minimal exposure to cosmic radiation, suggesting that the sediment source regions were covered in ice. These findings indicate that atmospheric warming during the past eight million years was insufficient to cause widespread or long-lasting meltback of the EAIS margin onto land. We suggest that variations in Antarctic ice volume in response to the range of global temperatures experienced over this period—up to 2–3 degrees Celsius above preindustrial temperatures 4 , corresponding to future scenarios involving carbon dioxide concentrations of between 400 and 500 parts per million—were instead driven mostly by the retreat of marine ice margins, in agreement with the latest models5,6. Analysis of cosmogenic isotopes from a marine sediment core shows that much of the land-based East Antarctic Ice Sheet has remained stable for the past eight million years, including during the warm Pliocene epoch.

On August 28, 2011, Tropical Storm Irene hit the state of Vermont with a severity that deposited 100–200 mm (4–8 in.) of rain across the state and resulted in damage or failure of over 300 ...

Concentrations of in situ 10Be measured in detrital fluvial sediment are frequently used to estimate long-term erosion rates of drainage basins. In many regions, basin-averaged erosion rates are positively correlated with basin average slope. The slope dependence of erosion allows model-based erosion rate estimation for unsampled basins and basins where human disturbance may have biased cosmogenic nuclide concentrations in sediment. Using samples collected from southeastern North America, we demonstrate an approach that explicitly considers the relationship between average basin slope and erosion rate. Because dams and reservoirs are ubiquitous on larger channels in the field area, we selected 36 undammed headwater subbasins (average area 10.6 km2) from which we collected river sand samples and measured 10Be concentrations. We used these data to train a predictive model that relates average basin slope and 10Be-inferred erosion rate. Applying our model to 28 basins in the same region previously studied with 10Be, we find that the model successfully predicts erosion rates for basins of different sizes if they are undammed or if samples were collected >25 km downstream of dams. For samples collected closer to dams, measured erosion rates exceed modeled erosion rates for two-thirds of the samples. In three of four cases where paired samples were collected upstream of reservoirs and downstream of the impounding dam, 10Be concentrations were lower downstream. This finding has implications for detrital cosmogenic studies, whether or not samples were collected directly downstream of dams, because dams obstruct most major rivers around the world, effectively trapping sediment that originated upstream.

Phylogenetic relationships of insect vectors of parasitic diseases are important for understanding the evolution of epidemiologically relevant traits, and may be useful in vector control. The sub-family Triatominae (Hemiptera:Reduviidae) includes ∼140 extant species arranged in five tribes comprised of 15 genera. The genus Triatoma is the most species-rich and contains important vectors of Trypanosoma cruzi, the causative agent of Chagas disease. Triatoma species were grouped into complexes originally by morphology and more recently with the addition of information from molecular phylogenetics (the four-complex hypothesis); however, without a strict adherence to monophyly. To date, the validity of proposed species complexes has not been tested by statistical tests of topology. The goal of this study was to clarify the systematics of 19 Triatoma species from North and Central America. We inferred their evolutionary relatedness using two independent data sets: the complete nuclear internal transcribed spacer-2 ribosomal DNA (ITS-2 rDNA) and head morphometrics. In addition, we used the Shimodaira-Hasegawa statistical test of topology to assess the fit of the data to a set of competing systematic hypotheses (topologies). An unconstrained topology inferred from the ITS-2 data was compared to topologies constrained based on the four-complex hypothesis or one inferred from our morphometry results. The unconstrained topology represents a statistically significant better fit of the molecular data than either the four-complex or the morphometric topology. We propose an update to the composition of species complexes in the North and Central American Triatoma, based on a phylogeny inferred from ITS-2 as a first step towards updating the phylogeny of the complexes based on monophyly and statistical tests of topologies.

Plug-in Hybrid Electric Vehicles (PHEVs) show potential to reduce greenhouse gas (GHG) emissions, increase fuel efficiency, and offer driving ranges that are not limited by battery capacity. However, these benefits will not be realized if consumers do not adopt this new technology. Several agent-based models have been developed to model potential market penetration of PHEVs, but gaps in the available data limit the usefulness of these models. To address this, we administered a survey to 1000 stated US residents, using Amazon Mechanical Turk, to better understand factors influencing the potential for PHEV market penetration. Our analysis of the survey results reveals quantitative patterns and correlations that extend the existing literature. For example, respondents who felt most strongly about reducing US transportation energy consumption and cutting greenhouse gas emissions had, respectively, 71 and 44 times greater odds of saying they would consider purchasing a compact PHEV than those who felt least strongly about these issues. However, even the most inclined to consider a compact PHEV were not generally willing to pay more than a few thousand US dollars extra for the sticker price. Consistent with prior research, we found that financial and battery-related concerns remain major obstacles to widespread PHEV market penetration. We discuss how our results help to inform agent-based models of PHEV market penetration, governmental policies, and manufacturer pricing and marketing strategies to promote consumer adoption of PHEVs.

Excessive phosphorus (P) export to aquatic ecosystems can lead to impaired water quality. There is a growing interest among watershed managers in using restored wetlands to retain P from agricultural landscapes and improve water quality. We develop a novel framework for prioritizing wetland restoration at a regional scale. The framework uses an ecosystem service model and an optimization algorithm that maximizes P reduction for given levels of restoration cost. Applying our framework in the Lake Champlain Basin, we find that wetland restoration can reduce P export by 2.6% for a budget of $50 M and 5.1% for a budget of $200 M. Sensitivity analysis shows that using finer spatial resolution data for P sources results in twice the P reduction benefits at a similar cost by capturing hot-spots on the landscape. We identify 890 wetlands that occur in more than 75% of all optimal scenarios and represent priorities for restoration. Most of these wetlands are smaller than 7 ha with contributing area less than 100 ha and are located within 200 m of streams. Our approach provides a simple yet robust tool for targeting restoration efforts at regional scales and is readily adaptable to other restoration strategies.

Knowledge of microscopic geomorphic structures is critical to understanding transport processes in porous building materials. X-ray scans were obtained of a variety of commonly used porous building materials to both qualitatively and quantitatively evaluate their pore structures. The specimens included natural materials (two sandstones and a limestone) and engineered materials (three types of concretes and a brick). Scanned images were processed to reconstruct the geomorphic structures of these materials. Random walk analyses were performed on the reconstructed pore structures to estimate macroscopic transport properties (including tortuosity, specific surface, and permeability). The effective porosity and permeability of these materials were also experimentally determined and compared to computed values. Calibration of the threshold pixel value used in the postprocessing of X-ray images against measured effective porosity appears to be a more appropriate method of selecting this value than the typical approach, which employs selection based solely on observed histograms. The resulting permeabilities computed by using a calibrated threshold pixel value compare better with the measured permeabilities. This study also demonstrates that the relatively homogeneous and heterogeneous pore structures associated with the natural and engineered building materials under investigation can be captured by X-ray tomography.

[1] Organizing or clustering data into natural groups is one of the most fundamental aspects of understanding and mining information. The recent explosion in sensor networks and data storage associated with hydrological monitoring has created a huge potential for automating data analysis and classification of large, high-dimensional data sets. In this work, we develop a new classification tool that couples a Naive Bayesian classifier with a neural network clustering algorithm (i.e., Kohonen Self-Organizing Map (SOM)). The combined Bayesian-SOM algorithm reduces classification error by leveraging the Bayesian's ability to accommodate parameter uncertainty with the SOM's ability to reduce high-dimensional data to lower dimensions. The resulting algorithm is data-driven, nonparametric and is as computationally efficient as a Naive Bayesian classifier due to its parallel architecture. We apply, evaluate and test the Bayesian-SOM network using two real-world hydrological data sets. The first uses genetic data to classify the state of disease in native fish populations in the upper Madison River, MT, USA. The second uses stream geomorphic and water quality data measured at ∼2500 Vermont stream reaches to predict habitat conditions. The new classification tool has substantial benefits over traditional classification methods due to its ability to dynamically update prior information, assess the uncertainty/confidence of the posterior probability values, and visualize both the input data and resulting probabilistic clusters onto two-dimensional maps to better assess nonlinear mappings between the two.

Disease transmission is difficult to model because most vectors and hosts have different generation times. Chagas disease is such a situation, where insect vectors have 1-2 generations annually and mammalian hosts, including humans, can live for decades. The hemataphagous triatominae vectors (Hemiptera: Reduviidae) of the causative parasite Trypanosoma cruzi (Kinetoplastida: Trypanosomatidae) usually feed on sleeping hosts, making vector infestation of houses, peridomestic areas, and wild animal burrows a central factor in transmission. Because of difficulties with different generation times, we developed a model considering the dwelling as the unit of infection, changing the dynamics from an indirect to a direct transmission model. In some regions, vectors only infest houses; in others, they infest corrals; and in some regions, they also infest wild animal burrows. We examined the effect of sylvatic and peridomestic vector populations on household infestation rates. Both sylvatic and peridomestic vectors increase house infestation rates, sylvatic much more than peridomestic, as measured by the reproductive number R0. The efficacy of manipulating parameters in the model to control vector populations was examined. When R01, the number of infested houses increases. The presence of sylvatic vectors increases R0 by at least an order of magnitude. When there are no sylvatic vectors, spraying rate is the most influential parameter. Spraying rate is relatively unimportant when there are sylvatic vectors; in this case, community size, especially the ratio of houses to sylvatic burrows, is most important. The application of this modeling approach to other parasites and enhancements of the model are discussed.

Watershed managers often use physical geomorphic and habitat assessments in making decisions about the biological integrity of a stream, and to reduce the cost and time for identifying stream stressors and developing mitigation strategies. Such analysis is difficult since the complex linkages between reach-scale geomorphic and habitat conditions, and biological integrity are not fully understood. We evaluate the effectiveness of a generalized regression neural network (GRNN) to predict biological integrity using physical (i.e., geomorphic and habitat) stream-reach assessment data. The method is first tested using geomorphic assessments to predict habitat condition for 1,292 stream reaches from the Vermont Agency of Natural Resources. The GRNN methodology outperforms linear regression (69% vs. 40% classified correctly) and improves slightly (70% correct) with additional data on channel evolution. Analysis of a subset of the reaches where physical assessments are used to predict biological integrity shows no significant linear correlation, however the GRNN predicted 48% of the fish health data and 23% of macroinvertebrate health. Although the GRNN is superior to linear regression, these results show linking physical and biological health remains challenging. Reasons for lack of agreement, including spatial and temporal scale differences, are discussed. We show the GRNN to be a data-driven tool that can assist watershed managers with large quantities of complex, nonlinear data.

Stakeholders are often required to make judgments and decisions about the tradeoffs between multiple competing objectives inherent in any engineering design. Design optimization can provide decision support for such situations, but often prescribes that only a single design solution be selected for a given set of preferences. The purpose of this study is to frame an objective function for assessing how the sensitivity of one objective relative to another varies in space and to demonstrate the method using a real site, with spatially-dependent floodplain access and bridge scour as the objective tradeoffs. Bridge scour is a widespread and expensive infrastructure problem, and the proposed methodology provides the ability to assess how the sensitivity of bridge scour to floodplain access varies at different locations in a river reach. The site chosen for demonstration purposes was the Lewis Creek in the vicinity of the Quinlan Covered Bridge in Charlotte, VT. Differential evolution (DE) was wrapped around an existing HEC-RAS model. The decision variables corresponded to floodplain access at locations up and downstream of the bridge; the objective function was constructed so that optimal solutions may be interpreted as relative salience of floodplain access to bridge scour. Multiple weightings of the objectives were used to verify that the rank-order of locations was robust. The optimal DE solutions for all weightings resulted in the same sensitivity ranking of locations, providing evidence that the analysis is not dependent on a particular choice of stakeholder objective weightings. For systems with spatially dependent variables that impact a constraint or objective of interest to stakeholders, a tool for identifying locations where that variable has a particularly strong or weak impact (e.g. where floodplain access is more or less important for bridge scour) has obvious advantages. This study demonstrates a method for conducting such a sensitivity analysis using a numerical optimization scheme. On the real test site, the sensitivity ranking was consistent across multiple stakeholder weightings, providing evidence that the technique is robust, and one that can be used at multiple stages of design. This work demonstrates the utility of a novel interpretation of optimization results in which locations are ranked according to the relative sensitivity of competing objectives.

We develop an algorithm to evolve sets of probabilistically significant multivariate feature interactions, with co-evolved feature ranges, for classification in large, complex datasets. The datasets may include nominal, ordinal, and/or continuous features, missing data, imbalanced classes, and other complexities. Our age-layered evolutionary algorithm generates conjunctive clauses to model multivariate interactions in datasets that are too large to be analyzed using traditional methods such as logistic regression. Using a novel hypergeometric probability mass function for fitness evaluation, the algorithm automatically archives conjunctive clauses that are probabilistically significant at a given threshold, thus identifying strong complex multivariate interactions. The method is validated on two synthetic epistatic datasets and applied to a complex real-world survey dataset aimed at determining the drivers of household infestation for an insect that transmits Chagas disease. We identify a set of 178,719 predictive feature interactions that are associated with household infestation, thus dramatically reducing the size of the search space for future analysis.

Bedrock outcrops are common on central Appalachian Mountain ridgelines. Because these ridgelines define watersheds, the rate at which they erode influences the pace of landscape evolution. To estimate ridgeline erosion rates, we sampled 72 quartz-bearing outcrops from the Potomac and Susquehanna River Basins and measured in situ–produced 10 Be. Ridgeline erosion rates average 9 ± 1 m m.y. −1 (median = 6 m m.y. −1 ), similar to 10 Be-derived rates previously reported for the region. The range of erosion rates we calculated reflects the wide distribution of samples we collected and the likely inclusion of outcrops affected by episodic loss of thick slabs and periglacial activity. Outcrops on main ridgelines erode slower than those on mountainside spur ridges because ridgelines are less likely to be covered by soil, which reduces the production rate of 10 Be and increases the erosion rate of rock. Ridgeline outcrops erode slower than drainage basins in the Susquehanna and Potomac River watersheds, suggesting a landscape in disequilibrium. Erosion rates are more similar for outcrops meters to tens of meters apart than those at greater distances, yet semivariogram analysis suggests that outcrop erosion rates in the same physiographic province are similar even though they are hundreds of kilometers apart. This similarity may reflect underlying lithological and/or structural properties common to each physiographic province. Average 10 Be-derived outcrop erosion rates are similar to denudation rates determined by other means (sediment flux, fission-track thermochronology, [U-Th]/He dating), indicating that the pace of landscape evolution in the central Appalachian Mountains is slow, and has been since post-Triassic rifting events.

We developed a methodology to predict brook trout (Salvelinus fontinalis) distribution using summer temperature metrics as predictor variables. Our analysis used long-term fish and hourly water temperature data from the Dog River, Vermont (USA). Commonly used metrics (e.g., mean, maximum, maximum 7-day maximum) tend to smooth the data so information on temperature variation is lost. Therefore, we developed a new set of metrics (called event metrics) to capture temperature variation by describing the frequency, area, duration, and magnitude of events that exceeded a user-defined temperature threshold. We used 16, 18, 20, and 22°C. We built linear discriminant models and tested and compared the event metrics against the commonly used metrics. Correct classification of the observations was 66% with event metrics and 87% with commonly used metrics. However, combined event and commonly used metrics correctly classified 92%. Of the four individual temperature thresholds, it was difficult to assess which threshold had the “best” accuracy. The 16°C threshold had slightly fewer misclassifications; however, the 20°C threshold had the fewest extreme misclassifications. Our method leveraged the volumes of existing long-term data and provided a simple, systematic, and adaptable framework for monitoring changes in fish distribution, specifically in the case of irregular, extreme temperature events.

A spatially explicit agent-based vehicle consumer choice model is developed to explore sensitivities and nonlinear interactions between various potential influences on plug-in hybrid vehicle (PHEV) market penetration. The model accounts for spatial and social effects (including threshold effects, homophily, and conformity) and media influences. Preliminary simulations demonstrate how such a model could be used to identify nonlinear interactions among potential leverage points, inform policies affecting PHEV market penetration, and help identify future data collection necessary to more accurately model the system. We examine sensitivity of the model to gasoline prices, to accuracy in estimation of fuel costs, to agent willingness to adopt the PHEV technology, to PHEV purchase price and rebates, to PHEV battery range, and to heuristic values related to gasoline usage. Our simulations indicate that PHEV market penetration could be enhanced significantly by providing consumers with ready estimates of expected lifetime fuel costs associated with different vehicles (e.g., on vehicle stickers), and that increases in gasoline prices could nonlinearly magnify the impact on fleet efficiency. We also infer that a potential synergy from a gasoline tax with proceeds is used to fund research into longer-range lower-cost PHEV batteries.

The performance and temporal variation of three hybrid and three integrated, saturated flow, pilot-scale constructed wetlands (CWs) were tested for treating dairy farm effluent. The three hybrid systems each consisted of two CWs in-series, with horizontal and vertical flow. Integrated systems consisted of a CW (horizontal and vertical flow) followed by a steel slag filter for removing phosphorus. Time series temporal semivariogram analyses of measured water parameters illustrated different treatment efficiencies existed over the course of one season. As a result, data were then divided into separate time period groups and CW systems were compared using ANOVA for parameter measurements within each distinct time period group. Both hybrid and integrated CWs were efficient in removing organics; however, hybrid systems had significantly higher performance ( p < 0.05) during peak vegetation growth. Compared to hybrid CWs, integrated CWs achieved significantly higher DRP reduction ( p< 0.05) throughout the period of investigation and higher ammonia reduction ( p < 0.05) in integrated CWs was observed in late summer. Geochemical modeling demonstrates hydroxyapatite and vivianite minerals forming on steel slag likely control the fate of phosphate ions given the reducing conditions prevalent in the system. The model also demonstrates how the wastewater:slag ratio can be adjusted to maximize phosphorus removal while staying at a near-neutral pH.

At the University of Vermont, an interdisciplinary faculty team developed an introductory watershed science field course. This course honed field skills and catalyzed communication across water-rel...

Summary In this work, we develop and test two artificial neural networks (ANNs) to forecast streamflow in ungauged basins. The model inputs include time-lagged records of precipitation and temperature. In addition, recurrent feedback loops allow the ANN streamflow estimates to be used as model inputs. Publically available climate and US Geological Survey streamflow records from sub-basins in Northern Vermont are used to train and test the methods. Time-series analysis of the climate-flow data provides a transferable and systematic methodology to determine the appropriate number of time-lagged input data. To predict streamflow in an ungauged basin, the recurrent ANNs are trained on climate-flow data from one basin and used to forecast streamflow in a nearby basin with different (more representative) climate inputs. One of the key results of this work, and the reason why time-lagged predictions of steamflow improve forecasts, is these recurrent flow predictions are being driven by time-lagged locally-measured climate data. The successful demonstration of these flow prediction methods with publicly available USGS flow and NCDC climate datasets shows that the ANNs, trained on a climate-discharge record from one basin, prove capable of predicting streamflow in a nearby basin as accurately as in the basin on which they were trained. This suggests that the proposed methods are widely applicable, at least in the humid, temperate climate zones shown in this work. A scaling ratio, based on a relationship between bankfull discharge and basin drainage area, accounts for the change in drainage area from one basin to another. Hourly streamflow predictions were superior to those using daily data for the small streams tested due the loss of critical lag times through upscaling. The ANNs selected in this work always converge, avoid stochastic training algorithms, and are applicable in small ungauged basins.

Repeated measurements of two small streams in northeastern Vermont document change in channel width and suggest variable rates of widening because of passive reforestation over four decades. Historic data on channel width are available for several tributaries to Sleepers River in Danville, VT, USA from the 1960s. In 2004 and 2008, we re-measured channel dimensions in two of these tributaries, in two reaches of upper Pope Brook and along seven reaches of an unnamed tributary (W12). Four reaches had reforested since 1966; two reaches remained nonforested. The other three reaches have been forested since at least the 1940s. Comparisons between 1966 and 2004 showed that reforested reaches widened significantly, and comparisons between 2004 and 2008 showed continued widening, but at a greater rate. Between 1966 and 2004, reforested reaches widened at an average rate of 4.1 cm/year, while the rate more than doubled for the last four years (8.7 cm/year). Additionally, turbulence data collected during five peak flows in the spring of 2005 showed significantly greater turbulent kinetic energy (TKE) in the reforested reach than in either the forested or nonforested reach. Our data add supporting information to the conceptual model of stream W12 that describes a process of incision, widening, and recovery of a stream reach transitioning from nonforested to forested riparian vegetation.

Summary Watershed managers and planners have long sought decision-making tools for forecasting changes in stream-channels over large spatial and temporal scales. In this research, we apply non-parametric, clustering and classification artificial neural networks to assimilate large amounts of disparate data types for use in fluvial hazard management decision-making. Two types of artificial neural networks (a counterpropagation algorithm and a Kohonen self-organizing map) are used in hierarchy to predict reach-scale stream geomorphic condition, inherent vulnerability and sensitivity to adjustments using expert knowledge in combination with a variety of geomorphic assessment field data. Seven hundred and eighty-nine Vermont stream reaches (+7500 km) have been assessed by the Vermont Agency of Natural Resources’ geomorphic assessment protocols, and are used in the development of this work. More than 85% of the reach-scale stream geomorphic condition and inherent vulnerability predictions match expert evaluations. The method’s usefulness as a QA/QC tool is discussed. The Kohonen self-organizing map clusters the 789 reaches into groupings of stream sensitivity (or instability). By adjusting the weight of input variables, experts can fine-tune the classification system to better understand and document similarities/differences among expert opinions. The use of artificial neural networks allows for an adaptive watershed management approach, does not require the development of site-specific, physics-based, stream models (i.e., is data-driven), and provides a standardized approach for classifying river network sensitivity in various contexts.

ARTICLE I NFO Measurements of two small streams in northeastern Vermont, collected in 1966 and 2004-2005, document considerable change in channel width following a period of passive reforestation. Channel widths of several tributaries to Sleepers River in Danville, VT, USA, were previously measured in 1966 when the area had a diverse patchwork of forested and nonforested riparian vegetation. Nearly 40 years later, we remeasured bed widths and surveyed large woody debris (LWD) in two of these tributaries, along 500 m of upper Pope Brook and along nearly the entire length (3 km) of an unnamed tributary (W12). Following the longitudinal survey, we collected detailed channel and riparian information for nine reaches along the same two streams. Four reaches had reforested since 1966; two reaches remained nonforested. The other three reaches have been forested since at least the 1940s. Results show that reforested reaches were significantly wider than as measured in 1966, and they are more incised than all other forested and nonforested reaches. Visual observations, cross-sectional surveys, and LWD characteristics indicate that reforested reaches continue to change in response to riparian reforestation. The three reaches with the oldest forest were widest for a given drainage area, and the nonforested reaches were substantially narrower. Our observations culminated in a conceptual model that describes a multiphase process of incision, widening, and recovery following riparian reforestation of nonforested areas. Results from this case study may help inform stream restoration efforts by providing insight into potentially unanticipated changes in channel size associated with the replanting of forested riparian buffers adjacent to small streams.

Reach-scale physical habitat assessment scores are increasingly used to make decisions about management. We characterized the spatial distribution of hydraulic habitat characteristics at the reach and sub-reach scales for four fish species using detailed two-dimensional hydraulic models and spatial analysis techniques (semi-variogram analyses). We next explored whether these hydraulic characteristics were correlated with commonly used reach-scale geomorphic assessment (RGA) scores, rapid habitat assessment (RHA) scores, or indices of fish biodiversity and abundance. River2D was used to calculate weighted usable areas (WUAs) at median flows, Q50, for six Vermont streams using modelled velocity, depth estimates, channel bed data and habitat suitability curves for blacknose dace (Rhinichthys atratulus), brown trout (Salmo trutta), common shiner (Notropis cornutus) and white sucker (Catostomus commersoni) at both the adult and spawn stages. All stream reaches exhibited different spatial distributions of WUA ranging from uniform distribution of patches of high WUA to irregular distribution of more isolated patches. Streams with discontinuous, distinct patches of high score WUA had lower fish biotic integrity measured with the State of Vermont's Mixed Water Index of Biotic Integrity (MWIBI) than streams with a more uniform distribution of high WUA. In fact, the distribution of usable habitats may be a determining factor for fish communities. A relationship between predicted WUAs averaged at the reach scale and RGA or RHA scores was not found. Future research is needed to identify the appropriate spatial scales to capture the connections between usable patches of stream channel habitat. Copyright © 2008 John Wiley & Sons, Ltd.

Artificial neural networks (ANNs) are used to estimate vertical ground surface movement when soils expand and contract due to changes in soil moisture content caused by changing climate conditions. Several counterpropagation ANN test cases were investigated to map climate data (i.e. temperature and rainfall) to vertical ground surface movement at field sites in Texas and Australia. Three of the four ANN test cases use a historical time series of climate data to forecast ground surface elevation relative to a specified datum. The fourth ANN test case predicts the rate of ground surface movement, and requires post-processing of the predicted rates to calculate ground surface elevation relative to a specified datum. The counterpropagation network has demonstrated a successful mapping of temperature and rainfall data to vertical ground surface movement at a field site when it is trained with a subset of data from the same field site (test cases 1 and 2). The results of training an ANN on one field site and testing it on another field site (test cases 3 and 4) demonstrate the ability of the ANN to capture trends in vertical ground surface movement. When compared with the predictions from a physics-based method (shrink test-water content method) that requires measurements/estimates of changes in soil water content, the ANN-based predictions (based on climatic changes) captured the trends in the field measurements of shrinking–swelling soil surface movements equally well. These findings are promising and merit further investigation with data from additional field sites. Copyright © 2007 John Wiley & Sons, Ltd.