Accuracy and uncertainty assessment on geostatistical simulation of soil salinity in a coastal farmland using auxiliary variable

Understanding the spatial soil salinity aids farmers and researchers in identifying areas in the field where special management practices are required. Apparent electrical conductivity measured by electromagnetic induction instrument in a fairly quick manner has been widely used to estimate spatial soil salinity. However, methods used for this purpose are mostly a series of interpolation algorithms. In this study, sequential Gaussian simulation (SGS) and sequential Gaussian co-simulation (SGCS) algorithms were applied for assessing the prediction accuracy and uncertainty of soil salinity with apparent electrical conductivity as auxiliary variable. Results showed that the spatial patterns of soil salinity generated by SGS and SGCS algorithms showed consistency with the measured values. The profile distribution of soil salinity was characterized by increasing with depth with medium salinization (EC_e 4–8 dS/m) as the predominant salinization class. SGCS algorithm privileged SGS algorithm with smaller root mean square error according to the generated realizations. In addition, SGCS algorithm had larger proportions of true values falling within probability intervals and narrower range of probability intervals than SGS algorithm. We concluded that SGCS algorithm had better performance in modeling local uncertainty and propagating spatial uncertainty. The inclusion of auxiliary variable contributed to prediction capability and uncertainty modeling when using densely auxiliary variable as the covariate to predict the sparse target variable.     

Arsenic and other trace elements in two catfish species from Paranaguá Estuarine Complex, Paraná, Brazil

Concentrations of arsenic and four additional trace elements (Cu, Cr, Ni, and Zn) were determined by inductively coupled plasma–optical emission spectrometry in the muscular tissue of the yellow catfish (Cathorops spixii) and the urutu catfish (Genidens genidens) from Paranaguá Estuarine Complex, Brazil (PEC). The PEC can be characterized by an environment of high ecological and economic importance in which preserved areas of rainforest and mangroves coexist with urban activities as ports and industries. The average concentrations (in milligram per kilogram dry weight) of elements in the muscle tissue of C. spixii are as follows: Zn (31), As (17), Cu (1.17), Cr (0.62), and Ni (0.28). Similar concentrations could be found in G. genidens with exception of As: Zn (36), As (4.78), Cu (1.14), Cr (0.51), and Ni (0.14). Fish from the geographic northern rural region (Guaraqueçaba–Benito) display higher As concentrations in the muscle tissues than fish found in the south-western (urban) part of the PEC. An international comparison of muscle tissue concentrations of trace elements in fish was made. Except for Ni in C. spixii, a tendency of decrease in element concentration with increasing size (age) of the fish could be observed. According to the National Health Surveillance Agency of Brazil, levels of Cr and As exceeded the permissible limits for seafood. An estimation of the provisional tolerable weekly intake of As was calculated with 109 % for C. spixii and with 29 % for G. genidens.     

Fluctuating asymmetry rather than oxidative stress in <Emphasis Type="Italic">Bufo raddei</Emphasis> can be an accurate indicator of environmental pollution induced by heavy metals

Oxidative stress (OS) and fluctuating asymmetry (FA) as risk markers for environmental stress are widely used to predict changes in the health and fitness of many animals exposed to pollutants. However, from the perspective of protecting declining amphibians, it remains to be verified which one would be a reliable indicator for amphibians exposed to long-term heavy metal pollution under natural conditions. In this study, the OS and FA of Bufo raddei exposed to natural heavy metal pollution were analyzed to determine which marker is more accurate for indicating heavy metal-induced stress. Three years of data were collected during the breeding season of B. raddei from Baiyin (BY), which has been mainly contaminated with Cu, Zn, Pb, and Cd compounds for a long period, and from Liujiaxia (LJX), which is a relatively unpolluted area. Unexpectedly, although significant accumulation of the four heavy metals was found in the kidney and liver of B. raddei from BY, the levels of superoxide dismutase, glutathione peroxidase, and malondialdehyde in these two organs were found to be irregular, with low repeatability in both BY and LJX. However, significant differences in the levels of FA were observed in B. raddei populations from these two areas over the past 3 years (P < 0.01). The degrees of FA in B. raddei populations from BY and LJX were assessed as degree 4 and 1, respectively. In short, this study suggested that FA was a more reliable and effective indicator than OS to monitor and predict long-term environmental stress on anuran amphibians.     

Passive sampling of glycol ethers and their acetates in indoor air

This study examined the performances of a thermal desorbable radial diffusive sampler for the weekly measurement of eight glycol ethers in indoor air and described the results of an application of this method carried out as part of HABIT'AIR Nord - Pas de Calais program for the air monitoring of these compounds in sixty homes located in northern France. The target compounds were the four glycol ethers banned from sale to the public in France since the 1990s (i.e. 2-methoxy ethanol, 2-ethoxy ethanol and their acetates) and four other glycol ethers derivatives of which the use have increased considerably (i.e. 1-methoxy-2-propanol, 2-butoxy ethanol and their acetates).A test program was carried out with the aim of validating the passive sampling method. It allowed the estimation of all the parameters of a method for each compound (calibration, analytical precision, desorption efficiency, sampling rate in standard conditions, detection limit and stability of sample before and after exposure), the examination of the influence of environmental factors on the sampling rate by some exposure chamber experiments and the assessment of the uncertainty of the measurements.The results of this evaluation demonstrated that the method has turned out to be suitable for six out of eight glycol ethers tested. The effect of the environmental factors on the sampling rates was the main source of measurement uncertainty. The measurements done in sixty homes revealed a relative abundance of 1-methoxy-2-propanol that was found in more than two thirds of homes at concentration levels of 4.5 microg m(-3) on average (a maximum value of 28 microg m(-3)). 1-methoxy-2-propanol acetate and 2-butoxy ethanol were also detected, but less frequently (in 19% of homes) and with the concentrations below 12 microg m(-3). The highest levels of these glycol ethers appear to be in relation to the emissions occurring at the time of cleaning tasks.     

Spatial variation in soil carbon in the organic layer of managed boreal forest soil—implications for sampling design

We studied within-site spatial variation of the carbon stock in the organic layer of boreal forest soil. A total of 1,006 soil samples were taken in ten forest stands (five Scots pine stands and five Norway spruce stands). Our results indicate that the spatial autocorrelation disappears at a distance of 75-225 cm. This spatial autocorrelation should be taken into account in the sampling design by locating the sampling points at adequate intervals. With a sample size of over 20-30 samples per site, additional soil samples do not notably improve the precision of the site mean estimate. An adequate sample size is dependent on the purpose of sampling and on the site-specific soil variation. Our results on the dependence between sample size and precision of the mean estimates can be applied in designing efficient soil monitoring in boreal coniferous forests.     

Utilization of statistics based classification approach to investigate water supply profile of Turkey

In the paper water supply profile of Turkey was examined. In this scope, the questionnaire survey conducted by Turkish Statistical Institute in 2004 to investigate annual amount of water abstracted to drinking water networks by type of resources in 81 provinces was evaluated. In the questionnaire, sources were grouped under five categories as spring, (artificial) lake, river, reservoir and well. Due to the complex and multivariate characteristics of the data sets, to replace a large collection of variables with a smaller number of factors the statistical method "factor analysis" was performed. Results revealed that, water supply systems in the country were mainly governed by groundwater sources (well and/or spring). However, in the northeastern part of the country, rivers were allocated for drinking water supply. On the other hand, reservoir dependent cities were densely located in Marmara, Central Anatolia and Southeast Anatolia Regions. This study showed that statistics based classification methods assist decision makers to extract information from multidimensional complex data sets representing environmental conditions.