Range Of Spatial Autocorrelation Research Articles

Spatial kernel models capturing field heterogeneity for accurate estimation of genetic potential.

According to Fisher’s principles, an experimental field is typically divided into multiple blocks for local control. Although homogeneity is supposed within a block, this assumption may not be practical for large blocks, such as those including hundreds of plots. In line evaluation trials, which are essential in plant breeding, field heterogeneity must be carefully treated, because it can cause bias in the estimation of genetic potential. To more accurately estimate genotypic values in a large field trial, we developed spatial kernel models incorporating genome-wide markers, which consider continuous heterogeneity within a block and over the field. In the simulation study, the spatial kernel models were robust under various conditions. Although heritability, spatial autocorrelation range, replication number, and missing plots directly affected the estimation accuracy of genotypic values, the spatial kernel models always showed superior performance over the classical block model. We also employed these spatial kernel models for quantitative trait locus mapping. Finally, using field experimental data of bioenergy sorghum lines, we validated the performance of the spatial kernel models. The results suggested that a spatial kernel model is effective for evaluating the genetic potential of lines in a heterogeneous field.

The spatial statistic trinity: A generic framework for spatial sampling and inference

Geospatial referenced environmental data are extensively used in environmental assessment, prediction, and management. Data are commonly obtained by nonrandom surveys or monitoring networks, whereas spatial sampling and inference affect the accuracy of subsequent applications. Design-based and model-based procedures (DB and MB for short) both allow one to address the gap between statistical inference and spatial data. Creating independence by sampling implies that DB may neglect spatial autocorrelation (SAC) if the sampling interval is beyond the SAC range. In MB, however, a particular sampling design can be irrelevant for inferential results. Empirical studies further showed that MSE (mean squared error) values for both DB and MB are affected by SAC and spatial stratified heterogeneity (SSH). We propose a novel framework for integrating SAC and SSH into DB and MB. We do so by distinguishing the spatial population from the spatial sample. We show that spatial independence in a spatial population results in independence in a spatial sample, whereas SAC in a spatial population is reflected in a spatial sample if sampling distances are within the range of dependence; otherwise, SAC is absent in the spatial sample. Similarly, SSH in a population may or may not be inherited in data, and this depends on the sampling method. Thus, the population, sample, and inference constitute a so-called spatial statistic trinity (SST), providing a new framework for spatial statistics, including sampling and inference. This paper shows that it greatly simplifies the choice of method in spatial sampling and inferences. Two empirical examples and various citations illustrate the theory.

Role of Sampling Design When Predicting Spatially Dependent Ecological Data With Remote Sensing

Remote sensing opens opportunities to assess spatial patterns on ecological data for a wide range of ecosystems. This information can be used to more effectively design sampling strategies for fieldwork, either to capture the maximum spatial dependence related to the ecological data or to completely avoid it. The sampling design and the autocorrelation observed in the field will determine whether there is a need to use a spatial model to predict ecological data accurately. In this article, we show the effects of different sampling designs on predictions of a plant trait, as an example of an ecological variable, using a set of simulated hyperspectral data with an increasing range of spatial autocorrelation. Our findings show that when the sample is designed to estimate population parameters such as mean and variance, a random design is appropriate even where there is strong spatial autocorrelation. However, in remote sensing applications, the aim is usually to predict characteristics of unsampled locations using spectral information. In this case, regular sampling is a more appropriated design. Sampling based on close pairs of points and clustered over a regular design may improve the accuracy of the training model, but this design generalizes poorly. The use of spatially explicit models improves the prediction accuracy significantly in landscapes with strong spatial dependence. However, such models have low generalization capacities to extrapolate to other landscapes with different spatial patterns. When the combination of design and size results in sample distances similar to the range of the spatial dependence in the field, it increases predictions uncertainty.

Effect of DEM Interpolation Neighbourhood on Terrain Factors

Topographic factors such as slope and aspect are essential parameters in depicting the structure and morphology of a terrain surface. We study the effect of the number of points in the neighbourhood of a digital elevation model (DEM) interpolation method on mean slope, mean aspect, and RMSEs of slope and aspect from the interpolated DEM. As the moving least squares (MLS) method can maintain the inherent properties and other characteristics of a surface, this method is chosen for DEM interpolation. Three areas containing different types of topographic features are selected for study. Simulated data from a Gauss surface is also used for comparison. First, the impact of the number of points on the DEM root mean square error (RMSE) is analysed. The DEM RMSE in the three study areas decreases gradually with the number of points in the neighbourhood. In addition, the effect of the number of points in the neighbourhood on mean slope and mean aspect was studied across varying topographies through regression analysis. The two variables respond differently to changes in terrain. However, the RMSEs of the slope and aspect in all study areas are logarithmically related to the number of points in the neighbourhood and the values decrease uniformly as the number of points in the neighbourhood increases. With more points in the neighbourhood, the RMSEs of the slope and aspect are not sensitive to topography differences and the same trends are observed for the three studied quantities. Results for the Gauss surface are similar. Finally, this study analyses the spatial distribution of slope and aspect errors. The slope error is concentrated in ridges, valleys, steep-slope areas, and ditch edges while the aspect error is concentrated in ridges, valleys, and flat regions. With more points in the neighbourhood, the number of grid cells in which the slope error is greater than 15° is gradually reduced. With similar terrain types and data sources, if the calculation efficiency is not a concern, sufficient points in the spatial autocorrelation range should be analysed in the neighbourhood to maximize the accuracy of the slope and aspect. However, selecting between 10 and 12 points in the neighbourhood is economical.

Using geographical semi-variogram method to quantify the difference between NO2 and PM2.5 spatial distribution characteristics in urban areas

Urban air pollutant distribution is a concern in environmental and health studies. Particularly, the spatial distribution of NO2 and PM2.5, which represent photochemical smog and haze pollution in urban areas, is of concern. This paper presents a study quantifying the seasonal differences between urban NO2 and PM2.5 distributions in Foshan, China. A geographical semi-variogram analysis was conducted to delineate the spatial variation in daily NO2 and PM2.5 concentrations. The data were collected from 38 sites in the government-operated monitoring network. The results showed that the total spatial variance of NO2 is 38.5% higher than that of PM2.5. The random spatial variance of NO2 was 1.6 times than that of PM2.5. The nugget effect (i.e., random to total spatial variance ratio) values of NO2 and PM2.5 were 29.7 and 20.9%, respectively. This indicates that urban NO2 distribution was affected by both local and regional influencing factors, while urban PM2.5 distribution was dominated by regional influencing factors. NO2 had a larger seasonally averaged spatial autocorrelation distance (48km) than that of PM2.5 (33km). The spatial range of NO2 autocorrelation was larger in winter than the other seasons, and PM2.5 has a smaller range of spatial autocorrelation in winter than the other seasons. Overall, the geographical semi-variogram analysis is a very effective method to enrich the understanding of NO2 and PM2.5 distributions. It can provide scientific evidences for the buffering radius selection of spatial predictors for land use regression models. It will also be beneficial for developing the targeted policies and measures to reduce NO2 and PM2.5 pollution levels.

Spatiotemporal variability of road weather conditions and optimal RWIS density — an empirical investigation

This paper presents a study aiming at understanding the relationship between the spatiotemporal characteristics of road weather conditions and a number of road weather information systems (RWIS) stations using real-world case studies. Semivariogram models are constructed to determine the spatial variability of road weather conditions, especially, autocorrelation range which describes a separation distance at which the measurements are no longer correlated to each other. An optimal RWIS density is then determined through an optimization process that minimizes the total inference errors across the underlying road network. The findings suggest that the regions with less varied topography tend to have a longer spatial correlation range than the regions with more varied topography. The study further reveals that the range of spatial autocorrelation is related to the optimal density of RWIS network — the region with a longer range requires fewer RWIS stations, than the region having a shorter range.

Spatio-Temporal Variability of Soil Beta-glucosidase Activity at the Arable Field Scale

Knowledge about the spatio-temporal variability of soil enzymatic properties is crucial in evaluating their structure-function relationship and their impact on ecosystem functions. The aim of this study was to assess the spatio-temporal variability of soil β-glucosidase (BG) activity against selected physico-chemical properties at the arable fild scale. A grid soil sampling (10×10 m) was used to measure the spatial variation of soil properties across a 0.4-ha fild. The BG activity was analyzed in soil samples that were collected from the upper 20 cm of Luvisols at 50 locations in April and August 2007 and 2008. Additionally, total organic carbon (CORG), total nitrogen (NTOT) concentrations, soil pHKCl as well as texture and water content were determined. The dataset was analyzed using classical statistical and geostatistical methods. Based on the analysis of variance, it was found that the BG activity, CORG, and water content showed signifiant differences between the four sampling dates. The BG activity showed a high contribution of nugget effect in sill (over 50%) and revealed a moderate spatial structure. The range of spatial autocorrelation calculated for the BG activity was between 15.0 and 61.0 m. Less intensive sampling grid should be recommended for soil BG activity in further studies concerning spatial variability in arable fild scale. In turn, more frequent sampling must be included in the sampling strategy in order to better understand whether the BG activity always shows permanent spatial patterns in soil or whether it is more randomized.

Spatio-temporal variability of soil sulfur content and arylsulfatase activity at a conventionally managed arable field

Sulfur (S) has long been recognized as essential for plant metabolism, but over the last 30years the increasing deficit of sulfur has become a problem all over the world. For this reason, more research should be devoted to determining the content and spatio-temporal variability of sulfur in relation to an improved management of this element. The aims of this study were to assess the spatio-temporal variability of soil sulfur forms and arylsulfatase activity against selected physicochemical properties on a plot scale.A grid soil sampling (10×10m) was used to measure the spatial variation of soil properties across a 0.4ha field. Soil samples were collected from the upper 27cm of Luvisols at 50 location in April and August 2007 and 2008. The total (STOT), organic (SORG) and available (SO42−-S) sulfur concentrations and arylsulfatase (AS) activity were analyzed. Additionally, total organic carbon (CORG), total nitrogen (NTOT) concentrations, soil pHKCl as well as texture and water content were determined. The dataset was analyzed using classical statistical and geostatistical methods.Based on the analysis of variance, it was found that soil properties, except for CORG, NTOT and pHKCl, showed significant differences between the four sampling dates (Tukey's test, p<0.05). Additionally, it was found that the temporal stability of the data set showed no clear time-associated shift based on Spearman's rank correlations coefficients. Relatively high degree of variability of the soil properties over time was also indicated by the mean relative differences (MRD). Generally, higher level of temporal stability was noted for the S forms and CORG and NTOT concentrations as compared to the AS activity. Among the studied properties, only AS activity at all sampling times and all of the S forms measured in August (except STOT in 2008 and SO42-S in 2007) revealed a spatial autocorrelation (Moran's I). All the studied properties revealed a high contribution of nugget effect in sill (over 50%) or pure nugget effect. Some properties revealed a moderate spatial structure (a nugget/sill ratio between 59.2 and 70.3%), whereas the concentration of SORG measured in August 2008 and AS activity analyzed in 2008 were in the weak variability class. Only AS activity in 2007 showed a pure nugget effect. The range of spatial autocorrelation calculated for the analyzed properties was between 20.0 and 34.0m.Results of this study are essential to understand and predict the contribution of internal factors in the total variability of soil properties that can interfere with the influence of soil management practices. The spatio-temporal variability showed that the studied area was temporally unstable and more frequent sampling must be included in the sampling strategy in order to better understand whether the S forms and AS activity always show permanent spatial patterns in soil or whether they are more randomized.

Wolf population genetics in Europe: a systematic review, meta-analysis and suggestions for conservation and management.

The grey wolf (Canis lupus) is an iconic large carnivore that has increasingly been recognized as an apex predator with intrinsic value and a keystone species. However, wolves have also long represented a primary source of human-carnivore conflict, which has led to long-term persecution of wolves, resulting in a significant decrease in their numbers, genetic diversity and gene flow between populations. For more effective protection and management of wolf populations in Europe, robust scientific evidence is crucial. This review serves as an analytical summary of the main findings from wolf population genetic studies in Europe, covering major studies from the 'pre-genomic era' and the first insights of the 'genomics era'. We analyse, summarize and discuss findings derived from analyses of three compartments of the mammalian genome with different inheritance modes: maternal (mitochondrial DNA), paternal (Y chromosome) and biparental [autosomal microsatellites and single nucleotide polymorphisms (SNPs)]. To describe large-scale trends and patterns of genetic variation in European wolf populations, we conducted a meta-analysis based on the results of previous microsatellite studies and also included new data, covering all 19 European countries for which wolf genetic information is available: Norway, Sweden, Finland, Estonia, Latvia, Lithuania, Poland, Czech Republic, Slovakia, Germany, Belarus, Russia, Italy, Croatia, Bulgaria, Bosnia and Herzegovina, Greece, Spain and Portugal. We compared different indices of genetic diversity in wolf populations and found a significant spatial trend in heterozygosity across Europe from south-west (lowest genetic diversity) to north-east (highest). The range of spatial autocorrelation calculated on the basis of three characteristics of genetic diversity was 650-850 km, suggesting that the genetic diversity of a given wolf population can be influenced by populations up to 850 km away. As an important outcome of this synthesis, we discuss the most pressing issues threatening wolf populations in Europe, highlight important gaps in current knowledge, suggest solutions to overcome these limitations, and provide recommendations for science-based wolf conservation and management at regional and Europe-wide scales.

Application of Area to Point Kriging to Low Birth Weight Incidence in Ghana

Background: The study examines the spatial distribution of low birth weight by the ten administrative Regions in Ghana using Area to Point Kriging method. Low birth weight babies, defined by World Health Organization as babies born at term who weigh less than 2.5 kg is an important indicator of reproductive health and general health status of population. The incidence of LBW is quite high in the sub region which has a public health concern. Methods: The study used a data set based on a Multiple Indicators Cluster Survey conducted by Ghana Statistical Service in 2011 with a sample of 10,963 women within the reproductive age. The geostatistical analysis applied in this study consists of three steps: filtering of noise in the data based on Poisson kriging, mapping of the corresponding risk at a fine scale and estimating geographical clustering of the low birth weight at the administrative units Results: This study has demonstrated how geostatistical method can be used to model low birth weight incidence by administrative units. The Area to Point method employed has given an insight Original Research Article Fosu et al.; BJAST, 15(5): 1-9, 2016; Article no.BJAST.25752 2 into a more localized potential “hot spots” for low birth weight incidence. The research showed a large range of spatial autocorrelation in the northern part than in the south in the incidence of low birth weight. The risk associated with low birth weight is centred broadly in the northern districts, districts in Central region and districts in the southern part of Ashanti region in the country which coincidentally are dominated by people of Sissala, Kassena, Mamprusi, Mole Dagbani, Wassa and Akan descends. The least affected areas are those settlements along the Volta lake who are predominantly Ewes. This suggests that low birth weight incidence in Ghana is more of an ethnic problem with some cultural undertones and parity as a main contributing factor than any other factor. Conclusion: The geostatistical method adopted has been able to identify a more localized potential “hot spots” for low birth weight incidence that may not be evident using other non geostatistical methods. The results further show that low birth weight incidence in Ghana is more of an ethno-cultural problem with parity as a driving factor.

ESTIMACION DEL HORIZONTE DE AUTOCORRELACION ESPACIAL: EL CASO DE LAS VARIABLES FISICO - QUIMICAS ESTUDIADAS EN LA CIENAGA GRANDE DE SANTA MARTA, COLOMBIA

The spatial autocorrelation range: estimations of physical-chemical variables studied at the Ciénaga Grande de Santa Marta, Northern Colombia. The spatial autocorrelation range, understood here as the minimum distance between points at which the spatial correlation becomes zero, is a very useful parameter to design samples that have as goal the study of the spatial behavior of a variable. In this paper the estimation of the autocorrelation range by means of the correlograms of confidence bands are discussed in detail, and estimations for a set of physical-chemical variables studied at the Ciénaga Grande de Santa Marta, Northern Colombia, are made. The results of the application show that for four of the variables included in the study (Nitrite, Nitrate, the addition of Nitrite and Nitrate, and Oxygen concentration) it is not possible to analyze their spatial behavior because the autocorrelation range is less than the minimum distance between the points that constituted the sample. The spatial autocorrelation range for the other variables were, 4900 m for ammonia, 4600 m for salinity, 3900 m for silicate, 2700 m for orthophosphate and 2700 m for chlorophyll. Through the application it can be further noticed the distortion that may result in estimating the ranges if the spatial trend is not previously removed.

Development of “best practices” for sampling of an important surface-dwelling soil mite in pastoral landscapes

In this study, we analyzed 1145 vacuum samples of redlegged earth mites (RLEM) [Halotydeus destructor (Tucker) (Acari: Penthaleidae)] from 18 sampling events at six locations in pastoral landscapes of Western Australia during three growing seasons (2012-2014) (total of 228,299 RLEM individuals). The specific objectives were to determine: (1) presence/absence effects of a range of vegetation characteristics, (2) possible factors influencing RLEM sampling performance during the course of the season and day, (3) effects of size of area sampled and duration of sampling, (4) the spatial structure of RLEM counts in uniform pastoral vegetation, and (5) develop "best practices" regarding field-based vacuum sampling of surface dwelling soil mites in pastoral landscapes. We found that sampling of completely bare ground will lead to very low RLEM counts but spots with sparse vegetation (presence of bare ground) probably increases the presence of microhabitats for mites to shelter in and therefore lead to higher RLEM counts. RLEM counts were positively associated with the height of vegetation, at least up to about 15 cm in height. In early season (May-August), highest RLEM counts will be obtained in the afternoon hours (2-4 pm), whereas in late season sampling (August-November), highest RLEM counts will be obtained around noon. Higher RLEM counts should be expected from spots with grazed/mowed vegetation including cape weed and without presence of grasses and stubble. Variogram analyses of high-resolution data sets suggested that considerable range of spatial autocorrelation should be expected from fields with fairly uniform vegetation, especially if RLEM population densities are high. We are therefore recommending that samples are collected at least 30 m apart, if the objective is to obtain independent (spatially non-correlated) counts. The results from this study may be used to develop effective sampling protocols deployed in field ecology studies of soil surface dwelling mesofauna in pastoral landscapes and other ecosystems.

Habitat vs. dispersal constraints on bryophyte diversity in the Mojave Desert, USA

Preserving desert plant diversity requires an understanding of the degree to which plant communities are constrained either by suitable habitat or by dispersal limitations. Using a spatially explicit approach, we surveyed diversity in an undisturbed xeric bryophyte community within a regular sampling grid of 1000 adjacent cells. Eigenvector mapping and variation partitioning revealed that spatially independent habitat processes accounted for little of the variation in community composition (2.0%); most variation was attributed either to habitats that were spatially structured (25.6%) or to dispersal independent of habitat (28.0%). At three nested scales, the most influential habitat factors were topography (broad-scale >5 m2), shrub “islands” (meso-scale 1−5 m2), and physical substrate characteristics (fine-scale < 1 m2). Dispersal limitation was less obviously scale-dependent at the community level, yet was important when considering individual species, which showed a wide range of spatial autocorrelation (Moran's I = −0.003–0.601). Our findings suggest that bryophyte diversity is the joint result of dispersal limitations and the distinctive spatially patterned environment of desert shrublands. Therefore, conservation efforts must not only account for dispersal limitations of individual species of concern, but should also seek to retain native shrubs and consequent landscape physiognomy that create bryophyte habitat.

Application of area to point Kriging to breast cancer incidence in Ashanti Region of Ghana

This paper provides a spatial analysis of breast cancer incidence in the Ashanti region area during the period of 2010 to 2011. Breast cancer disease has prevalence in Ghana particularly in the major cities including Ashanti region. Geographical units vary in shape and size and incidence count is non homogeneous in nature. For this reason, assigned area to point kriging approach is adopted as methodology. There is a large range of spatial autocorrelation in ages above 40 years than that of below 40 years in the various administrative units. The surrounding administrative units in the regional capital are less endemic for women whose ages are above 40 years. However, for those whose ages are below 40 years in all the surrounding administrative units are endemic but the capital is not. Most of the endemic districts share boundaries with Kumasi metropolis, the regional capital, where the only Teaching hospital is located. Most of the districts do not have good health facilities where women report for early treatment. Key words: Area to point kriging, breast cancer, area to area, kriging.

Spatial heterogeneity of fine root biomass of Pinus massoniana forests in the Three Gorges Reservoir Area, China

Environmental heterogeneity is a constant presence in the natural world that significantly affects plant behavior at a variety of levels of complexity. In order to estimate the spatial pattern of fine root biomass in the Three Gorges Reservoir Area, the spatial heterogeneity of fine root biomass in the upper layer of soils (0–10 cm) in three Masson pine (Pinus massoniana) stands in the Three Gorges Reservoir Area, China, was studied in 30 m × 30 m plots with geostatistical analysis. The results indicate that 1) both the live and dead fine root biomass of stand 2 were less than those of other stands, 2) the spatial variation of fine roots in the three stands was caused together by structural and random factors with moderate spatial dependence and 3) the magnitude of spatial heterogeneity of live fine roots ranked as: stand 3 > stand 1 > stand 2, while that of dead fine roots was similar in the three stands. These findings suggested that the range of spatial autocorrelation for fine root biomass varied considerably in the Three Gorges Reservoir Area, while soil properties, such as soil bulk density, organic matter and total nitrogen, may exhibit great effect on the spatial distribution of fine roots. Finally, we express our hope to be able to carry out further research on the quantitative relationship between the spatial heterogeneous patterns of plant and soil properties.

Individual‐based approach to the detection of disturbance history through spatial scales in a natural beech‐dominated forest

AbstractQuestionWe connected tree‐census and dendrochronological research data (74.2 ha) in order to answer the following questions: How do we apply an individual‐based approach, which allows us to abandon the traditional patch model, in the research of disturbance history through spatial scales? What is the disturbance history of the natural forest? How do we understand the influence of the Kyrill storm, which affected the reserve on 18 January 2007? How does the disturbance history in dendrochronological records correspond to real disturbances?LocationŽofínský Prales (48 °40′ N, 14 °42′ E, 735–830 m a.s.l., granite, Czech Republic)MethodsRepeated tree‐censuses (&gt;20 000 trees) were carried out in 1975, 1997 and 2008. These surveys served as the basis of dendrochronological research, during which 3 020 trees were sampled. A boundary line (BL) approach and geostatistical techniques were used to detect disturbance intensity and its spatial pattern.ResultsThe mean disturbance rate was 11.0% canopy loss per decade (maximum 33.7%) in the core zone. Bordering forests were directly affected by humans after 1800. A maximum of nine disturbance events were recorded per tree, but there were also 3% of non‐suppressed trees without records of disturbance events. The Kyrill storm represented a singular event in the forest history from the point of view of its exceptional spatial pattern (ranges up to 320.0 m), but not from the perspective of disturbance intensity. The relationship between disturbance intensity (y) and the range of spatial autocorrelation (x) can be explained by the equation y = 10.6863 + 0.0783*x (R2 = 0.546, P = 0.009). The spatial variability of releases ≥50% of BL corresponds to real disturbance events, while the insertion of weak releases led to the overestimation of range. The frequency of releases was affected by a decrease in air pollution at the beginning of the 1990s.ConclusionsEffects of small‐scale disturbances predominate in forest dynamics. Disturbance intensity was often not directly in accordance with the range of spatial autocorrelation. The ecological role of disturbance is therefore biased if disturbance intensity is used as the sole criterion. The results suggest the importance of historical and geographical contingency in ecosystem development.

Spatial Autocorrelation in Farmland Grasshopper Assemblages (Orthoptera: Acrididae) in Western France

Agricultural intensification in western Europe has caused a dramatic loss of grassland surfaces in farmlands, which have resulted in strong declines in grassland invertebrates, leading to cascade effects at higher trophic levels among consumers of invertebrates. Grasshoppers are important components of grassland invertebrate assemblages in European agricultural ecosystems, particularly as prey for bird species. Understanding how grasshopper populations are distributed in fragmented landscapes with low grassland availability is critical for both studies in biodiversity conservation and insect management. We assessed the range and strength of spatial autocorrelation for two grasshopper taxa (Gomphocerinae subfamily and Calliptamus italicus L.) across an intensive farmland in western France. Data from surveys carried out over 8 yr in 1,715 grassland fields were analyzed using geostatistics. Weak spatial patterns were observed at small spatial scales, suggesting important local effects of management practices on grasshopper densities. Spatial autocorrelation patterns for both grasshopper taxa were only detected at intermediate scales. For Gomphocerinae, the range of spatial autocorrelation varied from 802 to 2,613 m according to the year, depending both on grasshopper density and on grassland surfaces in the study site, whereas spatial patterns for the Italian locust were more variable and not related to grasshopper density or grassland surfaces. Spatial patterns in the distribution of Gomphocerinae supported our hypothesis that habitat availability was a major driver of grasshopper distribution in the landscape, and suggested it was related to density-dependent processes such as dispersal.

Deriving the optimal scale for relating topographic attributes and cover crop plant biomass

The use of cover crops generates a number of agro-ecological benefits for sustainable row-crop agriculture. However, their performance across agricultural fields is often highly spatially variable and there is insufficient information on factors affecting this variability and on tools to manage it. Topography is one of the main factors affecting spatial patterns of plant growth in the American Midwest. Digital elevation models are readily available for deriving topographic attributes; also sensor digital data can be used to indirectly assess cover crop biomass. However, processing procedures for identifying the proper scale of topographic and biomass representations are not well defined. The objectives of this study are to examine how relationships between cover crop biomass, assessed using the normalized difference vegetation index (NDVI), and topography depend on the neighborhood size used for deriving topographic attributes and creating NDVI maps; and identify the optimal neighborhood size for correlation and regression analyses. Slope, relative elevation and the potential solar radiation index were the variables that contributed the most to explaining variability in NDVI for raw data. However, other topographic attributes became significant predictors of NDVI at larger neighborhood sizes. We demonstrated that neighborhood size greatly affects some topographic attributes, i.e. curvature, flow accumulation, flow length and the wetness index; and changing the neighborhood size in both topography and NDVI considerably changes the strength of the prediction performance in multiple regression models. We studied six neighborhood sizes from 1 to 40m and the original raw data. On average, across all studied fields the best performance of multiple regression, as determined by the adjusted-R2, was obtained at neighborhood sizes 20 and 40m. Parameters of semivariogram models for terrain slope, such as the spatial autocorrelation range and the nugget/sill ratio, were found to be good indicators of prediction performance and optimum neighborhood size for filtering the raw data. The results demonstrate that topographic effects on growth and biomass production of cover crops are most pronounced at certain spatial scales, and topographic model predictions will be most accurate when used at the optimal scales.

The distribution of Biomphalaria (Gastropoda: Planorbidae) in Lake Victoria with ecological and spatial predictions using Bayesian modelling

Intestinal schistosomiasis is a global disease of enormous public health importance. Assessing the local risk of transmission of the parasite causing the disease requires an appraisal of the distribution of its intermediate snail hosts, Biomphalaria spp.. In East Africa, the Lake Victorian Basin is a major freshwater ecosystem and highly endemic for intestinal schistosomiasis, although detailed distribution data for Biomphalaria have not been collected. We used on-the-ground malacological surveys and conjoint measurement of environmental determinants to develop models for analysis of variables associated with distribution and predictive snail mapping into data-deficient areas. Four expeditions were undertaken to collect snails along the Lake Victoria shoreline, visiting 223 sites overall. Environmental measurements were recorded at the time of inspection and water samples taken for quantification of anion and cation concentrations. The spatial distributions of Biomphalaria choanomphala and Biomphalaria sudanica were modelled in two Bayesian multivariate models: one non-spatial and one spatial with random effects. The results showed that chloride, nitrate, sulphate and the number of sympatric snail species were significant predictors of B. choanomphala whereas habitat, water depth, pH and sulphate were significant predictors for B. sudanica. The range of spatial autocorrelation was large (572.9 km for B. choanomphala and 175.3 km for B. sudanica). Interpolating snail abundance data by kriging revealed two ‘hot-spots’ of high abundance of Biomphalaria. These areas should be targeted in future expeditions to ground-truth the model’s predictions. Our study is the first to use Bayesian methods for determination of biomedically important snail distributions and sets crucial limnological baseline data for assessment of future changes in snail biodiversity in the Lake Victoria Basin.

Spatial autocorrelation in predictors reduces the impact of positional uncertainty in occurrence data on species distribution modelling

Aim To investigate the impact of positional uncertainty in species occurrences on the predictions of seven commonly used species distribution models (SDMs), and explore its interaction with spatial autocorrelation in predictors. Methods A series of artificial datasets covering 155 scenarios including different combinations of five positional uncertainty scenarios and 31 spatial autocorrelation scenarios were simulated. The level of positional uncertainty was defined by the standard deviation of a normally distributed zero-mean random variable. Each dataset included two environmental gradients (predictor variables) and one set of species occurrence sample points (response variable). Seven commonly used models were selected to develop SDMs: generalized linear models, generalized additive models, boosted regression trees, multivariate adaptive regression spline, random forests, genetic algorithm for rule-set production and maximum entropy. A probabilistic approach was employed to model and simulate five levels of error in the species locations. To analyse the propagation of positional uncertainty, Monte Carlo simulation was applied to each scenario for each SDM. The models were evaluated for performance using simulated independent test data with Cohen's Kappa and the area under the receiver operating characteristic curve. Results Positional uncertainty in species location led to a reduction in prediction accuracy for all SDMs, although the magnitude of the reduction varied between SDMs. In all cases the magnitude of this impact varied according to the degree of spatial autocorrelation in predictors and the levels of positional uncertainty. It was shown that when the range of spatial autocorrelation in the predictors was less than or equal to three times the standard deviation of the positional error, the models were less affected by error and, consequently, had smaller decreases in prediction accuracy. When the range of spatial autocorrelation in predictors was larger than three times the standard deviation of positional error, the prediction accuracy was low for all scenarios. Main conclusions The potential impact of positional uncertainty in species occurrences on the predictions of SDMs can be understood by comparing it with the spatial autocorrelation range in predictor variables.