Clark-Evans Index Research Articles

What Are We Missing? Occlusion in Laser Scanning Point Clouds and Its Impact on the Detection of Single-Tree Morphologies and Stand Structural Variables

Laser scanning has revolutionized the ability to quantify single-tree morphologies and stand structural variables. In this study, we address the issue of occlusion when scanning a spruce (Picea abies (L.) H.Karst.) and beech (Fagus sylvatica L.) forest with a mobile laser scanner by making use of a unique study site setup. We scanned forest stands (1) from the ground only and (2) from the ground and from above by using a crane. We also examined the occlusion effect by scanning in the summer (leaf-on) and in the winter (leaf-off). Especially at the canopy level of the forest stands, occlusion was very pronounced, and we were able to quantify its impact in more detail. Occlusion was not as noticeable as expected for crown-related variables but, on average, resulted in smaller values for tree height in particular. Between the species, the total tree height underestimation for spruce was more pronounced than that for beech. At the stand level, significant information was lost in the canopy area when scanning from the ground alone. This information shortage is reflected in the relative point counts, the Clark–Evans index and the box dimension. Increasing the voxel size can compensate for this loss of information but comes with the trade-off of losing details in the point clouds. From our analysis, we conclude that the voxelization of point clouds prior to the extraction of stand or tree measurements with a voxel size of at least 20 cm is appropriate to reduce occlusion effects while still providing a high level of detail.

Impact of long and short-term conservation periods on structure of English yew (Taxus baccata L.) in Arasbaran forests, Iran

Yew (Taxus baccata L.) is one of the most important and threatened tree species in the Arasbaran region of northwestern Iran. To understand the natural stand structure of yew forests to inform forest management, we assessed the structural characteristics and composition of yew communities using the nearest neighbour and full callipering method at three sites with different conservation histories. Within a one-hectare sampling area, tree species identity, diameter, height, and crown diameter were measured. In each of these sampling areas, 56 sample points were surveyed in a 25 m × 25 m grid for tree species identity, diameter, height, and distance from reference to neighbour trees. To quantify the structural characteristics in areas of different conservation status, some indices were calculated including mingling, distance between reference tree and its nearest neighbouring trees, diameter and height differentiation, uniform angle, and Clark-Evans index. Results revealed that four species – hornbeam (68%), maple (8%), yew (7%), and oak (5.2%) – composed 88% of the tree species. The majority of trees had a short distance (2–3 m) between neighbours. The mean diameter differentiation index for long-term and short-term conservation areas was 0.59 and 0.06, respectively. The uniform angle index showed that there was no class value = “1” at all three sites. In the long-term enclosed area, Clark-Evans index was 1.18. In short-term enclosed areas, it was less than 1 (0.82). At all sites, yew trees were in the least vital class. We conclude that enclosing affects the yew stand structure, specifically in long-term periods of enclosure.

A 3D Individual-Based Model to Study Effects of Chemotaxis, Competition and Diffusion on the Motile-Phytoplankton Aggregation.

In this paper, we develop a 3D-individual-based model (IBM) to understand effect of various small-scale mechanisms in phytoplankton cells, on the cellular aggregation process. These mechanisms are: spatial interactions between cells due to their chemosensory abilities (chemotaxis), a molecular diffusion and a demographical process. The latter is considered as a branching process with a density-dependent death rate to take into account the local competition on resources. We implement the IBM and simulate various scenarios under real parameter values for phytoplankton cells. To quantify the effects of the different processes quoted above on the spatial and temporal distribution of phytoplankton, we used two spatial statistics: the Clark-Evans index and the group belonging percentage. Our simulation study highlights the role of the branching process with a weak-to-medium competition in reinforcing the aggregating structure that forms from attraction mechanisms (under suitable conditions for diffusion and attraction forces), and shows by contrast that aggregations cannot form when competition is high.

White spruce recruitment distances relative to lodgepole pine in southern Yukon, Canada

ABSTRACTEleven seral, postfire forest stands in southern Yukon (Canada) were sampled to determine where western white spruce (Picea albertiana ssp. albertiana) seedlings occurred with respect to distance to the nearest lodgepole pine (Pinus contorta var. latifolia) tree. Seedling-to-nearest tree distances were assessed at 10-cm increments up to 220 cm. On average, seedlings occurred 54 cm from the nearest pine (n = 490), but peak frequencies were 20–50 cm away, compared to a potential separation distance of 103 cm. Greatest average seedling density occurred 10–20 cm from pine (0.81 m−2), with values between 10 and 120 cm decreasing logarithmatically with increasing distance from pine (r = 0.994, p < .001, n = 11). Spruce seedling densities were <0.02 m−2 beyond 120 cm. The differences in frequency, which represented moderately strong aggregation (Clark-Evans Index 0.34–0.52), and density suggest greater spruce recruitment near lodgepole pine was facilitated by more favorable ecological conditions than further away, for example, greater nutrient availability. The bias in seedling-to-nearest tree distances occurred regardless of stand age (57–165 years), pine density (1599–5935 stems ha−1), or understory vegetation type, although the bias may be weakened by the abundant presence of feathermosses (Hylocomium splendens) on the forest floor.

Zastosowanie wskaźników koncentracji przestrzennej w badaniu procesów urban sprawl

This article reviews selected indicative methods allowing for analysis of the concentration and dispersion of settlement. A further aim is to evaluate the utility of these measures in studying the spontaneous process of suburbanisation known as “urban sprawl”. Following the model of the “dispersed city”, as opposed to the “compact city”, it is assumed that urban sprawl is associated with scattering of development. It is therefore reasonable to assume that spatial concentration indicators will allow for the at least partial description of its physiognomy. Urban sprawl is described as a multi-dimensional spatial phenomenon related to metropolitan deconcentration. Three fundamental spatial forms are observed: lowdensity sprawl, ribbon sprawl and leapfrog sprawl. Thereafter, issues are described in relation to the nature of the spatial dispersion and diffusion, the influence of centripetal and centrifugal forces, and the occurrence of the modified areal unit problem (MAUP), in the analysis of urbanisation. The four different measures chosen for actual review were the Gini coefficient, the C index of B. Kostrubiec, the average nearest neighbour method (Clark-Evans index) and Shannon entropy. Each of the indicators is analysed, with account taken of its theoretical and mathematical underpinnings, the adopted understanding of the spatial concentration concept, the impact of the delimitation of basic units on the results of spatial analyses, and available methods by which results may be presented. The Gini coefficient, based on the Lorenz curve, and initially used in econometrics, determines the cumulated concentration of features within a smaller or larger number of spatial units. It measures the unevenness of spatial distribution, but does not consider the mutual location of the basic units. A further limitation of this indicator in studying urban sprawl is that it takes no account of the precise locations of the objects. The spatial concentration index C, as proposed by B. Kostrubiec, is a measure of the concentration or dispersion of a set of elements – on a scale between concentration at one point and a spread across the maximum distance (range) it is possible to achieve within the boundaries of a certain area. The indicator is rarely used, but is of clear applicability, given the way it allows additional statistical parameters based on marginal distributions to be calculated. The average nearest neighbour method (Clark-Evans index), as derived from ecology, is widely known and applied in urbanisation studies. It allows for observation of the attractive forces associated with the locating of buildings and other new developments. This indicator resembles the previous one in combining recognition of the level of dispersion and the randomness of a set of features. Shannon entropy is a probabilistic measure of “disorder” and – in geography – a measure of segregation, the spatial organisation of an area, or, most simply, the proportion of the share of a phenomenon in territorial units. Entropy defines fragmentation or the filling of terrain with settlement. It is often used in researching land use and land cover change. This article concludes with a table describing the main features of the four indicators. Methods of multidimensional analysis of urban sprawl are also highlighted. These are important because the morphology of sprawl cannot be defined solely by reference to the degree of spatial concentration, which is understood and defined in various ways. Other important spatial dimensions include density, continuity, clustering, centralisation or the mixed use of land. In the author’s view, the most comprehensive assessment of the phenomenon of sprawl will be made possible if several methods are selected, and parallel analyses carried out using them. In these circumstances, complementary information will be obtained as regards the concentration and dispersion of development in an area.

Efficiency of sample-based indices for spatial pattern recognition of wild pistachio (Pistacia atlantica) trees in semi-arid woodlands

The efficiency of sample-based indices proposed to quantify the spatial distribution of trees is influenced by the structure of tree stands, environmental heterogeneity and degree of aggregation. We evaluated 10 commonly used distance-based and 10 density-based indices using two structurally different stands of wild pistachio trees in the Zagros woodlands, Iran, to assess the reliability of each in revealing stand structure in woodlands. All trees were completely stem-mapped in a nearly pure (40 ha) and a mixed (45 ha) stand. First, the inhomogeneous pair correlation function [g(r)] and the Clark–Evans index (CEI) were used as references to reveal the true spatial arrangement of all trees in these stands. The sampled data were then evaluated using the 20 indices. Sampling was undertaken in a grid based on a square lattice using square plots (30 m × 30 m) and nearest neighbor distances at the sample points. The g(r) and CEI statistics showed that the wild pistachio trees were aggregated in both stands, although the degree of aggregation was markedly higher in the pure stand. Three distance- and six density-based indices statistically verified that the wild pistachio trees were aggregated in both stands. The distance-based Hines and Hines statistic (h t ) and the density-based standardised Morisita (I p ), patchiness (IP) and Cassie (C A ) indices revealed aggregation of the trees in the two structurally different stands in the Zagros woodlands and the higher clumping in the pure stand, whereas the other indices were not sensitive enough.

Spatial Distribution Patterns in the Very Rare and Species-Rich Picea chihuahuana Tree Community (Mexico).

The very rare Mexican Picea chihuahuana tree community covers an area of no more than 300 ha in the Sierra Madre Occidental. This special tree community has been the subject of several studies aimed at learning more about the genetic structure and ecology of the species and the potential effects of climate change. The spatial distribution of trees is a result of many ecological processes and can affect the degree of competition between neighbouring trees, tree density, variability in size and distribution, regeneration, survival, growth, mortality, crown formation and the biological diversity within forest communities. Numerous scale-dependent measures have been established in order to describe spatial forest structure. The overall aim of most of these studies has been to obtain data to help design preservation and conservation strategies. In this study, we examined the spatial distribution pattern of trees in the P. chihuahuana tree community in 12 localities, in relation to i) tree stand density, ii) diameter distribution (vertical structure), iii) tree species diversity, iv) geographical latitude and v) tree dominance at a fine scale (in 0.25 ha plots), with the aim of obtaining a better understanding of the complex ecosystem processes and biological diversity. Because of the strongly mixed nature of this tree community, which often produces low population densities of each tree species and random tree fall gaps caused by tree death, we expect aggregated patterns in individual Picea chihuahuana trees and in the P. chihuahuana tree community, repulsive Picea patterns to other tree species and repulsive patterns of young to adult trees. Each location was represented by one plot of 50 x 50 m (0.25 ha) established in the centre of the tree community. The findings demonstrate that the hypothesis of aggregated tree pattern is not applicable to the mean pattern measured by Clark-Evans index, Uniform Angle index and Mean Directional index of the uneven-aged P. chihuahuana trees and P. chihuahuana tree community and but to specific spatial scales measured by the univariate L-function. The spatial distribution pattern of P. chihuahuana trees was found to be independent of patches of other tree species measured by the bivariate L-function. The spatial distribution was not significantly related to tree density, diameter distribution or tree species diversity. The index of Clark and Evans decreased significantly from the southern to northern plots containing all tree species. Self-thinning due to intra and inter-specific competition-induced mortality is probably the main cause of the decrease in aggregation intensity during the course of population development in this tree community. We recommend the use of larger sampling plots (> 0.25 ha) in uneven-aged and species-rich forest ecosystems to detect less obvious, but important, relationships between spatial tree pattern and functioning and diversity in these forests.

Dead wood and stand structure - relationships for forest plots across Europe

Dead wood and stand structural parameters were sampled in eleven countries using standardized methods at about 90 intensive forest monitoring sites across large parts of Europe. Besides descriptions and correlation analyses of dead wood and stand structure parameters, a joint evaluation of both fields was performed by principal component analysis (PCA). The extracted principal components were subsequently regressed against important numerical and categorical site-related parameters like soil pH, altitude, or forest type. Dead wood volumes varied largely across plots, however, 77 percent of them had volumes below 25 cubic meter per hectare. While all fractions of dead wood - except cut stumps - reveal high intercorrelation, different aspects of stand structure varied more independently. Clark-Evans index, number of tree species and standard deviation of tree trunk diameters revealed as most self-contained. The 1st PCA axis covered 46 percent of the total variance and was mostly loaded by total dead wood volume denoting it as the feature differentiating forests most. The 2nd axis was primarily loaded by tree species diversity together with stem density and the Clark-Evans index. On the 3rd axis diameter differentiation of trees together with the volume of cut stumps prevailed, while the 4th was mainly related to the decay class of woody debris. Bivariate ex post analyses revealed country as a significant predictor of all PCA axes, underlining national forest legislations and management rules as crucial for all investigated structural features of forests. Forest type was related only to the 3rd and 2nd axis. Only the 3rd axis revealed significant relationships with some ecological site factors (age, number of tree layers, latitude, altitude). The outcome underlines the significance of nationally enacted forest legislations for both important structural and biodiversity-relevant features of forest ecosystems and encourages similar approaches with data from national forest inventories or monitoring systems.

Posible influencia de los rametes en la estructura poblacional y distribución del árbol gigante Warszewiczia uxpanapensis, Cordillera de Tilarán, Costa Rica

The distribution and population structure of Warszewiczia (=Elaeagia) uxpanapensis, a tree that reaches a height of 45 m, a DBH of 1.40 m and 8 m tall buttresses, was studied on two 1-ha plots of premontane pluvial forest. It had a frequency of 49 % and a mean dominance of 6.5 %. The adult distribution was random (mean distance to the nearest neighboring tree in plot 1, it was 14.26 m, n = 15; Clark-Evans Index (IC-E) = 1.07; in plot 2, it was 15.33 m, n = 18, IC-E = 1.34,). The seedlings represented 60.3 % of the sampled population (plot 1, n = 94; plot 2, n = 85). This species has a very high capability of propagating itself vegetatively through clones ("ramets"), which influence its distribution and population structure. With a ramet density in plot 1 of 0.8 and in plot 2 of 0.3/100 m2, its propagation could be from fallen trees or branches (which also create clearings in the forest) or from standing trees and even live branches that produce adventitious roots.

Structural diversity and regeneration patterns of the Norway spruce (Picea abies (L.) Karst) dominated virgin forest in NNR Zadná Poľana

The paper analyzes the structural diversity of the spruce dominated old-growth forest in National Nature Reserve (NNR) Zadná Poľana on the series of three permanent research plots (PRP) with the area of 0.5 ha each. The height and diameter diversity as well as the spatial distribution were expressed using standard structural indices (Füldner’s TH and TD index, Gini index, Clark-Evans index). Obtained results confirmed the low structural diversity regardless of the developmental stage of investigated forest. The extensive ungulate browsing retards the height growth of the subsequent tree generation and leads toward to the structural homogenization of the future forest stand.

Spatial diversity of planted and untended silver birch (Betula pendula L.) stands

The aim of this paper was to describe changes in the spatial distribution of living trees in planted and untended birch stands, as well as to determine the spatial differentiation of the diameters and height of their constituent trees. Analysis was performed in 9- and 39-year old stands. The structural diversity of birch stands was described using the following spatial indices: Clark-Evans index (CE), spatial differentiation index for DBH (TD) and height (TH). The CE index describes the horizontal distribution of trees within a stand and the TD/TH indices indicate the size differentiation between neighbouring trees within a stand (DBH and H, respectively). The indices mentioned above concern small spatial scales. Our results indicate a regular spatial distribution of living trees on the ground at age nine, shown by high CE index. After 30 further years of stand development CE values were distinctly lower and no longer significantly different from the value for a random distribution of trees. At age 39, the 2 stands had a random spatial distribution and in the planted stand living trees still retained a regular spatial arrangement. Analysing the spatial differentiation of diameter and height of trees revealed that there was little differentiation between neighbours in forest stands in neither their diameter and height. The highest average values for the TD index decreased for stands during the 30 years since their initial measurement at age nine. A similar trend was observed for height differentiation (TH),which decreased from a maximum at age nine.

Spatio-temporal structure of natural forest: A structural index approach

The submitted paper deals with the analysis of stand structure of mixed species natural forest located in the Carpathians, Middle Europe. In order to evaluation of structural diversity we chose the combination of distance independent and nearest-neighbor indices. Following structural indices were calculated: Shanon-Weaver index (S-W), diameter differentiation index (Td), Gigi coefficient (G), mingling index (DMi) and Clark Evans index (CEd). The permanent research plot with area of 2.5ha was established in National Naure Reserve Hrončecký grúň and subdivided into 25×25m subplots. The combination of mentioned indices was calculated for each subplot separately. Regarding the diameter heterogeneity, the most part of analyzed structures showed the moderate level of tree size differentiation (S-W value 0.393±0.074). The presence of high number of juvenile individuals which already reached the diameter category “above 2cm” did not significantly affect the diameter variation of investigated stand. On the other hand, a massive establishment of subsequent generation led to creation of spatial structure characteristic by high level of vertical differentiation (Gini coefficient 0.47 in the structure characterized as the breakdown with successive generation). High climax species richness was confirmed by calculated values of DMi index (maximum relative presence in the category from 0.3 to 0.4). Each tree was surrounded by two and more different species individuals. Analysis discovers the negative developmental shift in the DMi values towards to less differentiated structures. Regarding the horizontal distribution the analysis of CEd values indicated that investigated natural forest can reveal a random distribution of individuals regardless of developmental stage.

Predicting forest structural parameters using the image texture derived from WorldView-2 multispectral imagery in a dryland forest, Israel

Estimation of forest structural parameters by field-based data collection methods is both expensive and time consuming. Satellite remote sensing is a low-cost alternative in modeling and mapping structural parameters in large forest areas. The current study investigates the potential of using WordView-2 multispectral satellite imagery for predicting forest structural parameters in a dryland plantation forest in Israel. The relationships between image texture features and the several structural parameters such as Number of Trees (NT), Basal Area (BA), Stem Volume (SV), Clark-Evans Index (CEI), Diameter Differentiation Index (DDI), Contagion Index (CI), Gini Coefficient (GC), and Standard Deviation of Diameters at Breast Heights (SDDBH) were examined using correlation analyses. These variables were obtained from 30m×30m square-shaped plots. The Standard Deviation of Gray Levels (SDGL) as a first order texture feature and the second order texture variables based on Gray Level Co-occurrence Matrix (GLCM) were calculated for the pixels that corresponds to field plots. The results of the correlation analysis indicate that the forest structural parameters are significantly correlated with the image texture features. The highest correlation coefficients were calculated for the relationships between the SDDBH and the contrast of red band (r=0.75, p<0.01), the BA and the entropy of blue band (r=0.73, p<0.01), and the GC and the contrast of blue band (r=0.71, p<0.01). Each forest structural parameter was modeled as a function of texture measures derived from the satellite image using stepwise multi linear regression analyses. The determination coefficient (R2) and root mean square error (RMSE) values of the best fitting models, respectively, are 0.38 and 109.56ha−1 for the NT; 0.54 and 1.79m2ha−1 for the BA; 0.42 and 27.18m3ha−1 for the SV; 0.23 and 0.16 for the CEI; 0.32 and 0.05 for the DDI; 0.25 and 0.06 for the CI; 0.50 and 0.05 for the GC; and 0.67 and 0.70 for the SDDBH. The leave-one-out cross-validation technique was applied for validation of the best-fitted models (R2>0.50). In conclusion, cross-validated statistics confirmed that the structural parameters including the BA, SDDBH, and GC can be predicted and mapped with a reasonable accuracy using the texture features extracted from the spectral bands of WorldView-2 image.

Horizontal structure of forest stands on permanent research plots in the Krkonoše Mts. and its development

Horizontal structure on 38 permanent research plots is described for juvenile growth and developmental phases (natural seeding, advance growth, plantations) and tree layer of a parent stand. Hopkins-Skellam index, Pielou-Mountford index, Clark-Evans index and Ripley's K-function were computed. The results are presented separately for beech stands, mixed stands, spruce stands, stands in the timberline ecotone and relict pinewood. The numbers and distribution of natural and combined regeneration recruits are mostly sufficient from the aspect of ecological, environmental and production functions of forest. The horizontal structure of juvenile growth and developmental phases of natural and combined regeneration shows mostly clustering; it is random or moderately regular at places with a single dominant proportion of artificial regeneration. In the tree layer the horizontal structure of forest stands is mostly random to moderately regular. In the future silvicultural measures should be aimed to support the structure of homogeneous stands of younger growth phases that have originated on a large scale after the air-pollution disaster.

Posible influencia de los rametes en la estructura poblacional y distribución del árbol gigante Warszewiczia uxpanapensis, Cordillera de Tilarán, Costa Rica

Se estudio los rametes, la distribucion y estructura de la poblacion de Warszewiczia uxpanapensis, en 2 ha de bosque pluvial premontano. Este arbol alcanza hasta 45 m de altura, DAPs hasta de 1.40 m y gambas de 8 m de altura. Su frecuencia fue de 49 % y su cobertura promedio de 6.5 %. La distribucion de los adultos fue aleatoria (distancia promedio al vecino mas cercano en la parcela 1 de 14.26 m, IC-E= 1.07, n= 15 en la parcela 2 de 15.33 m, n= 18, IC-E= 1.34). Sus plantulas representaron el 60.3 % de la poblacion estudiada (parcela 1, n= 94; parcela 2, n= 85). Se descubrio que esta especie tiene una alta capacidad de propagarse vegetativamente por rametes, los cuales influyen en la estructura de poblacion y distribucion de la especie (con una densidad de rametes parcela 1, 0.8 y parcela 2, 0.3 rametes/100 m2). Su propagacion vegetativa puede ser a partir de troncos caidos o ramas desprendidas (las cuales abren claros en el bosque), rebrotes de arboles, o ramas todavia adheridas a los arboles que son capaces de producir raices adventicias aereas.

CHARACTERIZATION OF DISPERSION STATE OF CARBON FILLER IN CARBON BLACK-SBR COMPOSITE MATERIALS BY STATISTICAL PATTERN PROCESSING

The sections of carbon black filled SBR composite materials, prepared by breaking in liquid nitrogen, was observed by SEM. The state of dispersion of carbon filler were characterized by two statistical processings: the distance method using Clark-Evans index CEI and the quadrat method using Morisita's Iδ value. Both were able to characterize the state of dispersion. The Morisita's Iδ value is more effective to characterize the state than CEI. The electrical conductivity of the composite materials were measured. The results were compared with these of the statistical analysis. The sudden increase in electrical conductivity with increasing carbon content was more adequately described by the Morisita's Iδ value method than the CEI method.