Robust Regression with Data-Dependent Regularization Parameters and Autoregressive Temporal Correlations

We introduce robust procedures for analyzing water quality data collected over time. One challenging task in analyzing such data is how to achieve robustness in presence of outliers while maintaining high estimation efficiency so that we can draw valid conclusions and provide useful advices in water management. The robust approach requires specification of a loss function such as the Huber, Tukey’s bisquare and the exponential loss function, and an associated tuning parameter determining the extent of robustness needed. High robustness is at the cost of efficiency loss in parameter loss. To this end, we propose a data-driven method which leads to more efficient parameter estimation. This data-dependent approach allows us to choose a regularization (tuning) parameter that depends on the proportion of “outliers” in the data so that estimation efficiency is maximized. We illustrate the proposed methods using a study on ammonium nitrogen concentrations from two sites in the Huaihe River in China, where the interest is in quantifying the trend in the most recent years while accounting for possible temporal correlations and “irregular” observations in earlier years.     

The Impact of Succession Stage on Share and Traits of Clonal Plants in Abandoned <Emphasis Type="Italic">Molinion caeruleae</Emphasis> Meadows

The impact of succession stage on the share of clonal plants in the plant cover, seed rain and seedling pool has been studied in the years 2014 and 2015. The investigations were carried out in abandoned Molinion caeruleae meadows dominated by small-statured species (Patch I), tall growing large-tussock grasses (Patch II) and shrub-willows (Patch III). Although the clonal taxa dominated in the plant cover, seed rain and seedling pool, their share in subsequent study sites was quite different. The spectra of life-history traits relevant to clonality in the plant cover and seed rain were similar and did not vary significantly among subsequent locations, while they mostly differed as observed in the seedling pool. The various spectra observed in seedling pool in subsequent study sites might be due to fact that the seed germination and recruitment of offspring are strongly influenced by habitat conditions which varied remarkably in the successive Patches.     

Periphyton Developed on Artificial Substrates: Effect of Substrate Type and Incubation Depth

The aim of this study was to assess the effect of substrate type and incubation depth on periphyton that had developed on artificial substrates. Uniform rectangular tiles made out of artificial substrates: glass, ceramic, willow tree and yew tree, were fixed on a floating buoy and deployed at three different depths in a photic zone of the Sava Lake (Belgrade, Serbia). Non-taxonomic attributes in the developed biofilm were estimated week-by-week from the start of the experiment in July, until its end in September 2014. Through assessment of substrate type and depth of incubation effect we concluded that these parameters for the fact influence periphyton development and composition. Glass was preferred by autotrophic component over ceramic and wooden substrates. In general, substrate type effect was diminished by increasing incubation depth. When non-taxonomic parameters are to be used in biomonitoring studies, our results suggest that glass substrate and shallow layer of water column (up to 50 cm) for incubation should be preferred.     

Annual Nutritional Organ Mass Cycle of <Emphasis Type="Italic">Bufo gargarizans</Emphasis> (Cantor, 1842) in Response to Extreme High Temperature

The annual relative mass of storage organs cycle of Bufo gargarizans in southwestern China was studied from December 2005 to November 2006. We described how those organs changed in response to extreme temperatures. Energy reservation dropped to the lowest levels after dormancy for both sexes, and then increased for period of 4.25 month. However, the mass of these organs dropped again rapidly until August (liver) and September (fat bodies) due to the constraint from extreme high temperature occurred during the summer, and then non-significantly increased before hibernation. For ovaries, varied with a different profile to those of liver and fat bodies have a resting period of four months. In addition, the toads’ ovaries kept growing when the mass of liver and fat bodies declining accordingly during hibernation. Therefore, ovarian mass increased during the hibernation and non-hibernation period.     

Plant Species Richness and Developmental Morphology Stage Influence Mycorrhizal Patagonia Plants Root Colonization

The objectives of this study were to determine the percentage of root colonization by arbuscular mycorrhizal (AM) fungi at various levels of plant species richness and developmental morphology stages in various perennial grass, and herbaceous and woody dicots species using experimental plots during 2013 and 2014. An auger was used to obtain six replicate root + soil samples at each sampling time on each of the study parameters. Roots were washed free of soil, and percentage AM was determined. The shrub Larrea divaricata was the species which showed the lowest percentage of colonization by AM at the vegetative developmental morphology stage at the monocultures and six-species-mixtures on the experimental plots. Dicots, but not grass, species showed a greater percentage colonization by AM fungi at the greatest (i.e., six-species-mixtures) than lowest (i.e., monocultures) species richness. Although at different degrees of species richness and developmental morphology stages, the perennial grasses Nassella longiglumis and N. tenuis, the herbaceous dicot Atriplex semibaccata, and the shrubs L. divaricata and Schinus fasciculatus showed a greater (p < 0.050) percentage colonization by AM fungi during the second than the first study year. Even though it was species- and sampling time-dependent, percentage colonization by AM fungi increased as species richness also increased most of the times. Our results demonstrated that the plant species differences in percentage colonization by AM fungi in the experimental plots were species richness-, developmental morphology stage-, and sampling-time dependents.     

Identifying Key Habitats to Conserve the Threatened Brown Bear in Northern Iran

Many factors such as human activities threat brown bears (Ursus arctos) in Southern Asia, and limit it to small populations in remote and rocky mountainous regions. Brown bears are generally studied in North America and Europe, but there is little information about its conditions and requirements for survival in Asia. During the recent years, brown bear populations in Iran have decreased sharply. Therefore, they are now officially listed as a threatened species in local scale. Therefore, we tried to recognize brown bear habitat relationships in northern Iran using species distribution model (SDM). Maxent method was applied using multiscale approach to predict suitable habitats and habitat relationships of the species. Our results revealed that the predictive ability of environment variables and species distribution maps varied across scales strongly. Also, our findings showed that identifying a proper scale is important issue to improve habitat modeling accuracy. Only 17% of the protected areas was found suitable for brown bear and divided to 5.1% poor, 8.8% suitable, and 3.1% high-quality habitat. Consequently, it is suggested that the protected area of northern Iran reconsiders with interventions aimed at maintaining suitable habitats of brown bear.     

Larval and Juvenile Fish Assemblage Structure of Inshore Habitats in the Middle Reaches of Li River,China: Spatial and Temporal Patterns in Relation to Abiotic Factors

To evaluate the effect of abiotic factors on larval and juvenile fish assemblage in shallow waters of middle reaches of Li River, Guilin, China. Samplings were taken monthly at 7 stations in the mainstream and 2 stations in 3 tributaries from May 2014 to April 2015. Synchronously, temperature, pH, turbidity, conductivity, dissolved oxygen rainfall and discharge were obtained. In all, 11.886 larval and juvenile fish from 18 species belonging to 8 families, 16 genera was captured. Principal Component Analysis (PCA) showed obvious significant difference between the environmental variables across sampling areas and months. The results of Canonical Correspondence Analysis (CCA) indicated that water temperature, turbidity, discharge and rainfall were dominant factors on larval and juvenile fish assemblage.     

A Mathematical Model for the Effects of Fertilization on Nitrogen Concentrations in Unsaturated Soil on Blueberry Farms in Southern New Jersey

Nitrogen is commonly known as a food source for crops. However, the nitrogen compounds used in crop fertilizers, most commonly nitrate (NO₃) and ammonium (NH₄), are not widely understood. Blueberry plants do not take up these compounds as efficiently as organic nitrogen so varying amounts of leaching into the soil and groundwater will occur. A biogeochemical model consisting of ordinary and partial differential equations is implemented to computationally predict the concentrations of nitrate and ammonium in unsaturated soil of blueberry plants, specifically in the southern region of New Jersey. The model takes into account the type of soil of the region, the nitrate uptake of the plant, the water content in the roots region, the pressure heads in the soil pores, and the application rates of fertilizers containing nitrate, ammonium, and organic nitrogen. Computational simulations demonstrate that the model accounts for natural processes and, in addition, show that commonly used fertilizer application rates cause unnecessarily high concentrations of both nitrate and ammonium in the unsaturated soil level. Further, simulations show that decreasing nitrate fertilizer applications by 85.7% in annual and 91.8% in bi-annual schedules provides an optimal system for safe reapplication.     

GPU Accelerated Lattice Boltzmann Simulation of a Cyanide-Release Accident in the Danjiangkou Reservoir

A graphics processing unit (GPU) accelerated two-dimensional contamination transport model is developed using the lattice Boltzmann method (LBM). The intrinsic parallel features of LBM make it particularly suitable for implementation on GPUs. After validation with two benchmarks, the model was applied to a cyanide-release accident in the Danjiangkou Reservoir, using a numerical combination of hydrodynamics and solute transport. The results show good agreement with an Environmental Fluid Dynamics Code (EFDC) model. A single GPU workstation speedup reaches 1.58 times that of a central processing unit (CPU) calculation alone. This speedup provides the opportunity for applying the model as a pollution accident information and decision support system for the Middle Route of the South to North Water Diversion Project in China.