Can trail spatial attributes predict trail use level in urban forest park? An examination integrating GPS data and space syntax theory

Understanding users’ spatial distribution in forest park is crucial for providing visitors with quality recreation experiences and for park planning and management. Utilizing users’ spatial distribution data, this study aims at investigating associations between trail use level and trail spatial attributes, through examining two large urban forest parks (Gongqing forest park and Paotaiwan forest park) in Shanghai, China. Users’ spatial distribution was measured utilizing GPS trackers with the interval of 10 seconds. This study conceptualizes trail spatial attributes as trail metric attributes and trail configurational attributes. Trail metric attributes include trail mean distance to gates, length, width and level of enclosure, which are calculated based on park map and on-site observation. We computed trail configurational attributes utilizing space syntax theory, which comprise measures of global integration, control, and connectivity. Trail connection with features/facilities, visual connection with water and shading are included as covariate variables. In total, we obtained 134 valid samples in Gongqing forest park and 210 valid samples in Paotaiwan forest park for analysis. Multivariate regression analyses indicate that when involving covariate variables, consistently in both parks, a greater trail usage is significantly related to trail shorter mean distance to park gate, trail width wider than 3 meters, higher global integration and higher control values. Collectively, these four trail spatial attributes explained 31.7 % (p < 0.001) and 27.3 % (p < 0.001) of the variances in trail use level in Gongqing forest park and Paotaiwan forest park. These findings provide direct implications to park designers and managers for providing visitors with different desirable social conditions, and ultimately improve users’ experiences and satisfactions.     

基于Web电子商务平台的食用菌产品共词聚类算法

针对食用菌产品电子商务平台的数据挖掘问题,采用了共词聚类算法进行主题聚类,在密度聚类算法的基础上,使用了移动非密集单元的处理方法实现了共词聚类。该算法克服了传统聚类算法裁剪造成的精度误差,提高了聚类结果的精确度,应用在电子商务平台的效果良好。     

牧场土壤含水率与坚实度空间变异与相关性分

利用土壤水分/圆锥指数复合测量装置,应用精细农作技术体系网格定点测量与GPS定位,在一块面积约1.27 hm2的草地上获取了土壤含水率与坚实度空间分布基础数据,并针对采样过程中出现的数据缺失,用偏最小二乘法对数据进行修补.然后运用克里格插值法进行数字化成图,并在此基础上分别对含水率与坚实度的空间变异性及两者的相关性做了分析.     

Spatial epidemiology of the Asian honey bee mite (Varroa destructor) in the North Island of New Zealand

We describe the spatial epidemiology of Varroa destructor infestation among honey bee apiaries in the greater Auckland area of the North Island of New Zealand. The study population was comprised of 641 apiaries located within the boundaries of the study area on 11 April 2000. Cases were those members of the study population declared Varroa-infested on the basis of testing conducted between April and June 2000. The odds of Varroa was highest in apiaries in the area surrounding transport and storage facilities in the vicinity of Auckland International Airport. A mixed-effects geostatistical model, accounting for spatial extra-binomial variation in Varroa prevalence, showed a 17% reduction in the odds of an apiary being Varroa infested for each kilometre increase in the squared distance from the likely site of incursion (95% Bayesian credible interval 7–28%). The pattern of spatially autocorrelated risk that remained after controlling for the effect of distance from the likely incursion site identified areas thought to be ‘secondary’ foci of Varroa infestation initiated by beekeeper-assisted movement of infested bees. Targeted investigations within these identified areas indicated that the maximum rate of local spread of Varroa was in the order of 12 km/year (interquartile range 10–15 km/year).     

The influence of sampling scheme and interpolation method on the power to detect spatial effects of forest birds in Ontario (Canada)

Spatial ecology is becoming an increasingly important component of resource management, and the general monitoring of how human activities affect the distribution and abundance of wildlife. Yet most work on the reliability of sampling strategies is based on a non-spatial analysis of variance paradigm, and little work has been done assessing the power of alternative spatial methods for creating reliable maps of animal abundance. Such a map forms a critical response variable for multiple scale studies relating landscape structure to biotic function. The power to reconstruct patterns of distribution and abundance is influenced by sample placement strategy and density, the nature of spatial auto-correlation among points, and by the technique used to extrapolate points into an animal abundance map. Faced with uncertainty concerning the influence of these factors, we chose to first synthesize a model reference system of known properties and then evaluate the relative performance of alternative sampling and mapping procedures using it. We used published habitat associations of tree nesting boreal neo-tropical birds, a classified habitat map from the Manitou Lakes area of northwestern Ontario, and point count means and variances determined from field studies in boreal Canada to create 4 simulated models of avian abundance to function as reference maps. Four point sampling strategies were evaluated by 4 spatial mapping methods. We found mixed-cluster sampling to be an effective point sampling strategy, particularly when high habitat fragmentation was avoided by restricting samples to habitat patches >10 ha in size. We also found that of the 4 mapping methods, only stratified ordinary point kriging (OPK) was able to generate maps that reproduced an embedded landscape-scale spatial effect that reduced nesting bird abundance in areas of higher forest age-class fragmentation. Global OPK was effective only for detecting broader, regional-scale differences. This revised version was published online in July 2006 with corrections to the Cover Date.     

Predicting the spatial distribution of ground flora on large domains using a hierarchical Bayesian model

Accomodation of important sources of uncertainty in ecological models is essential to realistically predicting ecological processes. The purpose of this project is to develop a robust methodology for modeling natural processes on a landscape while accounting for the variability in a process by utilizing environmental and spatial random effects. A hierarchical Bayesian framework has allowed the simultaneous integration of these effects. This framework naturally assumes variables to be random and the posterior distribution of the model provides probabilistic information about the process. Two species in the genus Desmodium were used as examples to illustrate the utility of the model in Southeast Missouri, USA. In addition, two validation techniques were applied to evaluate the qualitative and quantitative characteristics of the predictions.This revised version was published online in May 2005 with corrections to the Cover Date.     

A leakiness index for assessing landscape function using remote sensing

The cover, number, size, shape, spatial arrangement and orientation of vegetation patches are attributes that have been used to indicate how well landscapes function to retain, not ‘leak’, vital system resources such as rainwater and soil. We derived and tested a directional leakiness index (DLI) for this resource retention function. We used simulated landscape maps where resource flows over map surfaces were directional and where landscape patch attributes were known. Although DLI was most strongly related to patch cover, it also logically related to patch number, size, shape, arrangement and orientation. If the direction of resource flow is multi-directional, a variant of DLI, the multi-directional leakiness index (MDLI) can be used. The utility of DLI and MDLI was demonstrated by applying these indices to three Australian savanna landscapes differing in their remotely sensed vegetation patch attributes. These leakiness indices clearly positioned these three landscapes along a function-dysfunction continuum, where dysfunctional landscapes are leaky (poorly retain resources). This revised version was published online in July 2006 with corrections to the Cover Date.     

Spatial dependence and the relationship of soil organic carbon and soil moisture in the Luquillo Experimental Forest,Puerto Rico

We used geo-spatial statistical techniques to examine the spatial variation and relationship of soil organic carbon (SOC) and soil moisture (SM) in the Luquillo Experimental Forest (LEF), Puerto Rico, in order to test the hypothesis that mountainous terrain introduces spatial autocorrelation and crosscorrelation in ecosystem and soil properties. Soil samples (n = 100) were collected from the LEF in the summer of 1998 and analyzed for SOC, SM, and bulk density (BD). A global positioning system was used to georeference the location of each sampling site. At each site, elevation, slope and aspect were recorded. We calculated the isotropic and anisotropic semivariograms of soil and topographic properties, as well as the cross-variograms between SOC and SM, and between SOC and elevation. Then we used four models (random, linear, spherical and wave/hole) to test the semi-variances of SOC, SM, BD, elevation, slope and aspect for spatial dependence. Our results indicate that all the studied properties except slope angle exhibit spatial dependence within the scale of sampling (200 – 1000 m sampling interval). The spatially structured variance (the variance due to the location of sampling sites) accounted for a large proportion of the sample variance for elevation (99%), BD (90%), SOC (68%), aspect (56%) and SM (44%). The ranges of spatial dependence (the distances within which parameters are spatially dependent) for aspect, SOC, elevation, SM, and BD were 9810 m, 3070 m, 1120 m, 930 m and 430 m, respectively. Cross correlograms indicate that SOC varies closely with elevation and SM depending on the distances between samples. The correlation can shift from positive to negative as the separation distance increases. Larger ranges of spatial dependence of SOC, aspect and elevation indicate that the distribution of SOC in the LEF is determined by a combination of biotic (e.g., litterfall) and abiotic factors (e.g., microclimate and topographic features) related to elevation and aspect. This demonstrates the importance of both elevation and topographic gradients in controlling climate, vegetation distribution and soil properties as well as the associated biogeochemical processes in the LEF.This revised version was published online in May 2005 with corrections to the Cover Date.     

Ecotonal changes and altered tree spatial patterns in lower mixed-conifer forests, Grand Canyon National Park, Arizona, U.S.A.

This research analyzes patch development and determines tree spatial patterns along the lower mixed-conifer ecotone on the North Rim of Grand Canyon National Park in Arizona (U.S.A.). Patterns of patch development were interpreted from spatial analyses, based on tree age and size, and past records of disturbance and climate. Five plots in the ecotone between mixed conifer forests and monospecific stands of ponderosa pine (Pinus ponderosa) were studied for patterns of patch development. The methods used include: (1) size-structure analyses, to compare species patch development; (2) dendrochronological dating of tree establishment; (3) tree ring master chronology, to determine periods of suppressed growth, compared to a Palmer Drought Severity Index; and (4) spatial analyses by species composition, size and age, with univariate and bivariate analyses of spatial association and spatial autocorrelation. We found an increased density of shade-tolerant and fire-intolerant species namely clusters of pole-sized white fir, and fewer large ponderosa pine. This revised version was published online in May 2005 with corrections to the Cover Date. This revised version was published online in July 2006 with corrections to the Cover Date.     

Modelling land cover transitions: A solution to the problem of spatial dependence in data

Raster-based spatial land cover transition models (LCTMs) are widely used in landscape ecology. However, many LCTMs do not account for spatial dependence of the input data, which may artificially fragment the output spatial configuration. We demonstrate the consequences of ignoring spatial dependence, thus assigning probabilities randomly in space, using a simple LCTM. We ran the model from four different initial conditions with distinct spatial configurations and results indicated that, after 20 simulation steps, all of them converged towards the spatial configuration of the random data set. From an ecological perspective this is a serious problem because ecological data often exhibit distinct spatial configuration related to ecological processes. As a solution, we propose an approach (region approach) that accounts for spatial dependence of LCTM input data. Underlying spatial dependence was used to apply spatial bias to probability assignment within the model. As a case study we applied a region approach to a Vegetation Transition Model (VTM); a semi-Markovian model that simulates forest succession. The VTM was applied to approximately 500,000 ha of boreal forest in Ontario, at 1 ha pixel resolution. When the stochastic transition algorithms were applied without accounting for spatial dependence, spatial configuration of the output data became progressively more fragmented. When the VTM was applied using the region approach to account for spatial dependence output fragmentation was reduced. Accounting for spatial dependence in transition models will create more reliable output for analyzing spatial patterns and relating those patterns to ecological processes.This revised version was published online in May 2005 with corrections to the Cover Date.