Spatial smoothing techniques for the assessment of habitat suitability

Precise knowledge about factors influencing the habitat suitability of a certain species forms the basis for the implementation of effective programs to conserve biological diversity. Such knowledge is frequently gathered from studies relating abundance data to a set of influential variables in a regression setup. In particular, generalised linear models are used to analyse binary presence/absence data or counts of a certain species at locations within an observation area. However, one of the key assumptions of generalised linear models, the independence of observations is often violated in practice since the points at which the observations are collected are spatially aligned. In this paper, we describe a general framework for semiparametric spatial generalised linear models that allows for the routine analysis of non-normal spatially aligned regression data. The approach is utilised for the analysis of a data set of synthetic bird species in beech forests, revealing that ignorance of spatial dependence actually may lead to false conclusions in a number of situations.

Geostatistical Assessment of the Impact of World War I on the Spatial Occurrence of Soil Heavy Metals

Previous research showed a regional Cu enrichment of 6 mg kg⁻¹ in the top soil of the Ypres war zone (Belgium), caused by corrosion of WWI shell fragments. Further research was required since in addition to Cu, also As, Pb, and Zn were used during the manufacturing of ammunition. Therefore, an additional data collection was conducted in which the initial Cu data set was tripled to 731 data points and extended to eight heavy metals (As, Cd, Cr, Cu, Hg, Ni, Pb, and Zn) which permitted (1) to evaluate the environmental impact of the heavy metals at a regional scale and (2) to assess their regional spatial occurrence by performing an optimized geostatistical modeling. The results showed no pollution at a regional scale, but sometimes locally concentrations exceeded the soil sanitation threshold, especially for Cu, Pb, and Zn. The spatial patterns of Ni and Cr were related to variations in soil texture whereas the occurrences of Cu and Pb were clearly linked to WWI activities. This difference in spatial behavior was confirmed by an analysis of coregionalization.     

Identification of trace element sources and associated risk assessment in vegetable soils of the urban-rural transitional area of Hangzhou, China

In the urban-rural transitional area of Hangzhou, China, 74 topsoil samples were collected from vegetable fields to measure the contents of arsenic (As), copper (Cu), cobalt (Co), cadmium (Cd), chromium (Cr), mercury (Hg), manganese (Mn), nickel (Ni), lead (Pb) and zinc (Zn). The combination of multivariate statistical and geostatistical methods successfully separated the contaminating elements (As, Cd, Cu, Hg, Pb and Zn) from uncontaminated elements (Co, Cr, Ni and Mn). A significant correlation was found between these uncontaminated elements and total Al2O3, Fe2O3, and SiO2 of the soils, indicating that the source of these elements was mainly controlled by soil-forming factors. On the other hand, these contaminating elements showed relatively weaker correlation and higher spatial variability, indicating that their enrichment and spatial heterogeneity were mostly affected by anthropic inputs. Through the pollution evaluation, it was found that only 30.8% of the study area did not suffer from moderate or severe pollution.     

Metal accumulation in mosses across national boundaries: uncovering and ranking causes of spatial variation

This study aimed at cross-border mapping metal loads in mosses in eight European countries in 1990, 1995, and 2000 and at investigating confounding factors. Geostatistics was used for mapping, indicating high local variances but clear spatial autocorrelations. Inference statistics identified differences of metal concentrations in mosses on both sides of the national borders. However, geostatistical analyses did not ascertain discontinuities of metal concentrations in mosses at national borders due to sample analysis in different laboratories applying a range of analytical techniques. Applying Classification and Regression Trees (CART) to the German moss data as an example, the local variation in metal concentrations in mosses were proved to depend mostly on different moss species, potential local emission sources, canopy drip and precipitation.     

Heavy metal sources identification and sampling uncertainty analysis in a field-scale vegetable soil of Hangzhou, China

At a field-scale (6.7 ha), 100 surface soil samples were collected from a vegetable field to determine total concentrations of Cd, Co, Cu, Hg, Mn, Ni and Zn. To identify possible sources of these metals and characterize their spatial variation, classic statistic and geostatistic techniques were applied. Through correlation and geostatistical analysis, it was found that the primary inputs of Co, Mn and Ni were due to pedogenic sources, whereas the sources of Hg and Cd were mainly due to human activities. Because of their different sources, their variations followed: Hg > Cd ≈ Cu > Zn ≈ Co ≈ Mn ≈ Ni. Based on their relationships with other soil properties, co-kriging was used to minimize sampling density. Sampling numbers for Cd, Cu, Zn, Mn, Co and Ni can be reduced from 100 to 90, 80, 70, 60, 60 and 60, respectively, without losing accuracy relative to ordinary kriging.     

Soil salinity evolution and its relationship with dynamics of groundwater in the oasis of inland river basins: case study from the Fubei region of Xinjiang Province, China

Soil salinization is an important worldwide environmental problem, especially in arid and semi-arid regions. Knowledge of its temporal and spatial variability is crucial for the management of oasis agriculture. The study area has experienced dramatic change in the shallow groundwater table and soil salinization during the 20th century, especially in the past two decades. Classical statistics, geostatistics and geographic information system (GIS) were applied to estimate the spatial variability of the soil salt content in relation to the shallow groundwater table and land use from 1983 to 2005. Consumption of reservoir water for agricultural irrigation was the main cause of a rise in the shallow groundwater table under intense evapotranspiration conditions, and this led indirectly to soil salinization. The area of soil salt accumulation was greater in irrigated than in non-irrigated landscape types with an increasing of 40.04% from 1983 to 2005 in cropland at ∼0.43 t ha⁻¹ year⁻¹, and an increase at ∼0.68 t ha⁻¹ year⁻¹ in saline alkaline land. Maps of the shallow groundwater table in 1985 and 2000 were used to deduce maps for 1983 and 1999, respectively, and the registration accuracy was 99%.     

Comparing proximity measures of exposure to geostatistical estimates in environmental justice research

Abstract

Recent hazard literature frequently refers to sustainability and resilience as the guiding principles behind effective hazard planning. Certainly, structurally organizing communities to minimize effects of disasters and to recover quickly by restoring socio-economic vitality are laudable goals. However, while anticipating such outcomes is relatively easy from a theoretical standpoint, practical implementation of comprehensive plans is much more elusive. Indeed, relationships between community sustainability/resilience and hazards are complex involving many social, economic, political and physical factors. A conceptual framework for analysis of sustainability and resilience, then, is described based on three theoretical models, a mitigation model, a recovery model, and a structural-cognitive model. This framework is examined using data from Florida, USA, where local context, social and political activities, and economic concerns present difficulties in application. The question remains, therefore, to what extent can communities truly develop sustainable and resilient characteristics?     

Heavy metals in top core sediments from the southern Campania shelf (Italy): Hints to define large-scale geochemical backgrounds

The results of a geochemical survey carried out on the southern Campania shelf (southern Italy) are illustrated, offering a tempting opportunity to discuss the statement and definition of regional geochemical backgrounds for selected heavy metals. A total of 104 top core sediments, collected on average 1.5 km from one another and covering a shelf area of about 1300 km², were analysed for grain size, organic carbon content (OC), and heavy-metal (As, Al, Fe, Cd, V, Cr, Zn, Cu, Ni, Hg, and Pb) concentrations. The characteristics of the area, scarcely influenced by industrial activities, and the extensive data set, make this study suitable for an appropriate multivariate and spatial geostatistical methodologies used for i) a reliable definition of large-scale marine sediments' geochemical backgrounds and ii) an accurate discrimination of different geological controls on trace-element distribution patterns. Enrichment factors estimated with respect to both the average shale composition and the weighted average metal concentrations measured in inland soils rule out significant anthropogenic inputs. A spatial-geostatistics approach (kriging of spatial component based on an accurate implementation of variographic surveying) was applied, and this is proposed as an appropriate quantitative methodology to define the geochemical backgrounds for the different heavy metals. Two nested spatial structures were identified for some of the heavy metals, thus allowing clear identification of local and regional geochemical controls related to the different spatial variability of grain size.     

Spatial distribution of lindane in topsoil of Northern France

Lindane is a persistent organochlorine insecticide and the use of this insecticide in agriculture was banned in France in 1998. In this study we investigated the concentrations of lindane in topsoil in Northern France and used robust geostatistics to map the geographical distribution of lindane. The study was based on a 16 km × 16 km grid covering an area of ca 25 000 km². Lindane was found in all soils, even those from non-agricultural-application areas. Very low ratios of α-/γ-HCH and δ-/γ-HCH suggested that a long time had passed since technical HCH was used in the studied area, or that emission sources of lindane were still present. A strong gradient in lindane concentration was observed, with the highest lindane concentrations in an area located in the northern region. Results suggested that some of the lindane observed in the high concentration area may have come from volatilization of old lindane applied to intensively cultivated areas, which was then transported by prevailing winds coming from the south-west and deposited in a densely inhabited depression.     

Spatial methods for plot-based sampling of wildlife populations

Classical sampling methods can be used to estimate the mean of a finite or infinite population. Block kriging also estimates the mean, but of an infinite population in a continuous spatial domain. In this paper, I consider a finite population version of block kriging (FPBK) for plot-based sampling. The data are assumed to come from a spatial stochastic process. Minimizing mean-squared-prediction errors yields best linear unbiased predictions that are a finite population version of block kriging. FPBK has versions comparable to simple random sampling and stratified sampling, and includes the general linear model. This method has been tested for several years for moose surveys in Alaska, and an example is given where results are compared to stratified random sampling. In general, assuming a spatial model gives three main advantages over classical sampling: (1) FPBK is usually more precise than simple or stratified random sampling, (2) FPBK allows small area estimation, and (3) FPBK allows nonrandom sampling designs.