Báo cáo khoa học: " Spatial variability of humus forms in some coastal forest ecosystems of British Columbia"

Chia sẻ: Nguyễn Minh Thắng | Ngày: | Loại File: PDF | Số trang:14

Thêm vào BST

Báo xấu

55
lượt xem 4
download

Download Vui lòng tải xuống để xem tài liệu đầy đủ

Tuyển tập các báo cáo nghiên cứu về lâm nghiệp được đăng trên tạp chí lâm nghiệp quốc tế đề tài: " Spatial variability of humus forms in some coastal forest ecosystems of British Columbia...

Chủ đề:

Bình luận(0) Đăng nhập để gửi bình luận!

Lưu

Nội dung Text: Báo cáo khoa học: " Spatial variability of humus forms in some coastal forest ecosystems of British Columbia"

Original article Spatial variability of humus forms in some coastal forest ecosystems of British Columbia H Qian, K Klinka Forest Sciences Department, University of British Columbia, Vancouver, BC, Canada V6T 1Z4 1 (Received 19 June February 1994; accepted 1995) Summary — The spatial variability of 5 humus form properties (thickness, acidity, total C, total N and mineralizable-N) was examined in 3 coastal forest sites of different tree species composition (western hemlock, Douglas-fir and western redcedar), humus forms, and ecological site quality using variogram and kriging. Humus form properties were found spatially dependent and the kriging interpolation between sample locations unbiased for all 5 properties and in all 3 sites. The overall range of spatial dependence ranged from 46 to 1 251 cm, but varied with property and site. The average range for the humus form properties increased from 109 cm (total N) to 704 cm (mineralizable-N), and that for the sites increased from 275 cm (western hemlock) to 581 cm (Douglas-fir). It appears that humus forms in each site occur in polygons with the lateral dimension ranging from 100 to 700 cm. The spatial pat- tern of each property in each site was portrayed in contour maps. humus form / spatial variability / variogram / kriging Résumé — Variabilité spatiale des types d’humus dans quelques écosystèmes forestiers côtiers de Colombie britannique. La variabilité spatiale de 5 caractéristiques de l’humus (épaisseur, aci- dité, carbone total, azote total et minéralisable) a été étudiée dans 3 sites forestiers côtiers, différant par l’espèce dominante (pruche de l’Ouest, douglas et thuya géant), le type d’humus et le type de station. Elle est analysée par variogramme et krigeage. Ces propriétés des types d’humus sont dépen- dantes spatialement, et l’interpolation par krigeage entre les points d’échantillonnage est non biaisée pour les 5 propriétés et les 3 sites. La portée globale de dépendance spatiale varie de 46 à 1 251 cm, mais dépend de la propriété considérée et du site. La portée moyenne pour les propriétés de l’humus varie entre 109 cm (pour l’azote total) à 704 cm (pour l’azote minéralisable), et cella des sites varie entre 275 cm (sous pruche de l’Ouest) à 581 cm (sous douglas). Il apparaît que les types d’hu- mus dans chaque site sont groupés en polygones dont la dimension varie entre 100 et 700 cm. La varia- bilité spatiale de chaque propriété dans chaque site est illustrée par des cartes obtenues par kri- geage. type d’humus/ variabilité spatiale/ variogramme / krigeage
INTRODUCTION jbregts, 1978; Yost et al, 1982a; Robertson, 1987; Rossi et al, 1992). Geostatistics can be used to quantify the spatial dependence Humus form is a group of soil horizons between sampling locations and to provide located at or near the surface of a pedon, optimal estimates for unsampled locations which have formed from organic residues, (Matheron, 1963, 1971; Burgess and Web- either separate from, or intermixed with, ster, 1980a; Vieira et al, 1981; Yost et al, mineral materials (Green et al, 1993). In 1982b). Central to geostatistics is the vari- consequence, humus forms may be com- ogram, which models the average degree prised of entirely organic or both organic of similarity between the values as a function and mineral (melanized A) horizons. Due of their separation distance, and kriging, to the difficulties in combining organic and which estimates values for unsampled loca- mineral horizons in chemical and data anal- tions without bias and with minimum vari- yses (Lowe and Klinka, 1981),this study ance. examined only the organic or the forest floor portion of humus forms. Geostatistics has been extensively used in mining (eg Matheron, 1963, 1971; Krige, product of biologically mediated As the 1966; David, 1977; Clark, 1979; Journel and decomposition processes, the humus form Huijbregts, 1978) and, more recently applied that has developed on a particular site in soil science (eg Nielsen et al, 1973; Big- depends on the biota and environment of gar and Nielsen,1976; Campbell, 1978; that site. Both biota and environment may Burgess and Webster, 1980a, b; Vieira et change over a short distance, yielding a al, 1981; Yost et al, 1982a, b; Xu and Web- variety of microsites which support the ster, 1984), hydrology (eg McCullagh, 1975; development of different humus forms. The Delhomme, 1976, 1978, 1979; Hajrasuliha nature of spatial variability in humus forms is et al, 1980; Kitandis, 1983), ecology (eg itself scale-dependent because the factors Robertson, 1987; Kemp et al, 1989), veg- and processes of humus formation interact etation science (eg Palmer, 1988; Fortin et over many different spatial scales. It seems al, 1989), but no systematic effort has yet reasonable to assume that, on average, the been made to apply it to humus form stud- closer humus forms are to each other, ies. whether in space or time, the more likely it is their properties will be similar. This assump- The objective of this study was to exam- tion calls for an inquiry into the nature and ine the spatial variation of 5 selected humus degree of spatial dependence between the form properties - thickness, acidity, total C, humus forms, particularly in the sample plots total N and mineralizable-N - in disturbed chosen to represent individual ecosystems, and undisturbed coastal forest ecosystems. ie segments of landscape relatively uniform This objective was accomplished by employ- in climate, soil and vegetation (Pojar et al, ing variogram and kriging for the analysis 1987). of spatial variability of these properties. The thickness was thought the most variable Classical statistical techniques are unable morphological property, reflecting difference adequately the spatial aspect of data to treat in the deposition and decomposition of in which neighboring samples may not be organic residues in both space and time. independent of each other; furthermore, The significance of the 4 selected chemical they do not consistently provide unbiased properties has been long recognized in estimates for unsampled points, or estimate optimal variances for the interpolated val- humus form classification (Green et al, ues (Matheron, 1963; Journel and Hui- 1993).
developed shrub and herb layers dominated by MATERIALS AND METHODS Athyrium filix-femina (L) Roth, Rubus spectabilis Pursh and Tiarella trifoliata L, and Leptomoders All study sites were located near Vancouver, and Mullmoders were the prevailing humus forms British Columbia, and were within the Coastal (table III). Using the methods described by Klinka Western Hemlock (CWH) zone, which delineates et al (1984, 1989), the western hemlock site was the sphere of influence a cool mesothermal cli- considered slightly dry and nitrogen-poor; the mate (Klinka et al, 1991). The soils in the area Douglas-fir site, fresh and nitrogen-rich and the are typically coarse-textured humo-ferric podzols western redcedar site, moist and nitrogen-very (Canada Soil Survey Committee, 1978) derived rich. from granitic morainal deposits. At each a 20 x 20 m (0.04 ha) sam- study site, The study sites were deliberately chosen to located to represent an individual ple plot was represent forest ecosystems with different veg- ecosystem. Within each plot, a 10 x 10 grid, 1 x 1 m, and a 7 x 7 grid, 15 x 15 cm, were laid out for etation, humus forms, ecological site quality and sampling humus forms. One-hundred discontin- history of disturbance (table I). The first site (Hw) uous samples were collected from the large, 10 x was dominated by western hemlock (Tsuga het- 10 grid at the center of each 1 x 1 m quadrant, erophylla [Raf] Sarg), the second (Fd) by Dou- and 49 contiguous samples were taken from the glas-fir(Pseudotsuga menziesii [Mirbel] Franco), small, 7 x7 grid - a total of 149 humus form sam- and the third (Cw) by western redcedar (Thuja ples per site. The small grid provided data for the plicata Donn ex D Don). The western hemlock analysis of a small-scale pattern (the sampling site had a well-developed moss layer dominated interval of 15 cm), while the large grid provided by Plagiothecium undulatum (Hedw) BSG, and data for the analysis of a large-scale pattern (the Mors (Hemimors and Lignomors) (Green et al, sampling interval of 1 m). 1993) were the prevailing humus forms; the Dou- glas-fir site had a well-developed herb layer with Each humus form sample was a composite of abundant Polystichum munitum (Kaulf) Presl and all of its organic horizons (except recently shed lit- Dryopteris expansa (K Presl) Fraser-Jenkins & ter), and represented a uniform, 15 x 15 cm col- umn cut by knife from the ground surface to the Jermy, and Mormoders were the prevailing humus boundary with mineral soil. Each sample was western redcedar site had well- forms; and the
described and identified according to Green et (1983), Webster (1985), Trangmar et al (1985) al (1993), its grid location recorded and its thick- and lsaaks and Srivastava (1989). Consider that ness determined by taking 4 measurements at a humus form property is a regionalized variable each cardinal direction with a steel ruler. Z(x) and that its measurements at places x i= 1, , i n, constitute n discrete points in space, 2, 3, All samples were air-dried to constant mass ..., where x a set of spatial coordinates in 2 idenotes and ground in a Wiley mill to pass through a 2-mm dimensions. The measurements give a set of val- sieve. The chemical analysis was done by Pacific ues z(x and the semivariance that summarizes ), i Soil Analysis Inc (Vancouver, BC) and the results the spatial variation for all possible pairing of data were expressed per unit of mass (tables II and is calculated by: III). Humus form pH was measured with a pH meter and glass electrode in water using a 1:5 suspension. Total C (tC) was determined using a Leco Induction Furnace (Bremner and Tabatabai, 1971).Total N (tN) was determined by semimicro- kjeldahl digestion followed by determination of where the value &jadnr;(h) is the estimated half- or NH using a Technicon Autoanalyzer (Anony- -N 4 semivariance for h, which is a vector known as the mous, 1976). Mineralizable-N (min-N) was deter- lag, with both distance and direction, and N(h) is mined by an anaerobic incubation procedure of the number of pairs of points separated by h. A Powers (1980) with released NH determined 4 plot of the estimated &jadnr;((h) values against h is called colorimetrically using a Technicon Analyzer. semivariogram or variogram. a For the geostatistical analyses, we used the By definition, the variogram value at zero lag GS geostatistical package (Gamma Design Soft- + should be zero, but in practice it usually inter- ware, 1992) following the theory and principles cepts the ordinate at a positive value known as the given by Matheron (1963, 1971), Journel and Huijbregts (1978), David (1977), Delhomme nugget variance (c The nugget represents mea- ). 0 (1978), Vieira et al (1981, 1983), Vauclin et al unexplained or random spa- surement error and
tion criterion (AIC), the spherical (eq [2]) and tial variability at distances smaller than the small- exponential (eq [3]) isotropic models were found sampling interval. The variogram value at est best fitting the data: which the plotted points level off is known as the sill, which is the sum of nugget variance (c and ) 0 structural variance (c), and the lag distance (a) at which the variogram levels off is known as the range (or the zone of influence) beyond which there is no longer spatial correlation and, hence, no longer spatial dependence. Local estimation by kriging required fitting a continuous function to the computed experimen- where c c, a and a are nugget variance, struc- tal semivariance values. The most commonly , 0 0 tural variance, range and range parameter, used models are: linear, linear with sill, spheri- respectively. Because the semivariance from an cal, exponential and gaussian (Journel and Hui- exponential isotropic model approaches the sill jbregts, 1978; Tabor et al, 1984; McBratney and asymptotically, there is no absolute range. A work- Webster, 1986; Oliver and Webster, 1986). Exper- ing range of a 3 a a lag at which the semi- imental variogram values for each humus form , 0 = variance is 95% of the sill values, was estimated property were fitted to each model by least square for practical purposes (Oliver and Webster, 1986). approximation. Using Akaike’s (1973) informa-
With appropriate variogram models defined, properties but dissimilar distributions, except kriging was used to interpolate between sample for mineralizable-N (table II). The values of points and to estimate the values for unsampled coefficient of variation and variance implied locations. Kriging is a weighted moving average trends of a low variability around mean acid- with estimator: an ity and total C (except in the western red- cedar site), a moderate variability around mean total N and a high variability around mean thickness and mineralizable-N. Skew- ness values indicated an asymmetric dis- where n is the number of values z(x for the sam- ) i tribution for each property in 1 or 2 study pled locations involved in the estimation of the sites (table II). When considering the skew- unsampled location x and &iadbmal ; are the weights , 0 ness values (table II) and the univariate associated with each sampled location value. summary of data stratified according to both Kriging is considered an optimal estimation humus form taxa and study sites (table III), method as it estimates values for unsampled locations without bias and with minimum vari- the acidity data for the Douglas-fir site were ance. No estimation method is without estima- strongly skewed to the right, reflecting the tion errors, thus there is an error associated with presence of relatively less-acid Leptomod- kriging. The magnitude of this error will be a mea- ers occupying mineral mounds. The acidity sure of the validity of estimation. The goodness of and carbon data for the western redcedar estimation can be determined by comparing the site were skewed to the right and left, difference between the measured value at a given location with its kriged value at the same loca- respectively, attesting to the presence of tion, using neighborhood values but not the mea- more-acid and carbon-richer Lignomoders sured value itself. Thus, if for each location with a relative to dominant Leptomoders. The total measured value z(x where i = 1, 2, 3, ..., n, the ), i N data for both Douglas-fir and western estimated value is &jadnr;(x where i= 1, 2, 3, ), i n, ..., hemlock sites were strongly skewed to the then the calculated set of estimated errors is &ivispe ; ) - jadnr;(x i ), &jadnr;(x &i where i 1, 2, 3, n. The good- left, indicating the presence of nitrogen- = = ..., ness of estimation is expressed by 2 conditions on richer Mormoders relative to the other the estimated error: 1) a mean error, m close, e humus forms on these sites. In the Dou- to zero - this property of the estimator is known as glas-fir site, the distribution of mineralizable- unbiasedness, and 2) dispersion of the errors N was skewed to the right, manifesting the was to be concentrated around m this being ϵ - presence of Lignomors - the humus form expressed by a small value of the estimated vari- with the lowest concentration of available 2 &ϵ (table VI). sigma; ance N. The distribution of thickness data in both For statistical analyses, we used the SYSTAT (Wilkinson, 1990a, b). Prior to geostatistical anal- Douglas-fir and western hemlock sites was ysis, humus form variables for each study stand highly asymmetric and strongly skewed to were examined for normality, using probability the right, indicating the presence of dis- distribution diagrams (Wilkinson, 1990a). The turbed microsites (mineral mounds) with thin thickness values in the western hemlock and forest floors. Douglas-fir sites and the acidity and min-N values in the Douglas-fir site were log-transformed as Although univariate measures provided they were found log-normally distributed. useful summaries, they did not describe spatial continuity of the data, ie the rela- tionship between the value for a property in RESULTS AND DISCUSSION one location and the values for the same property at another The spatial ’location. A univariate summary of humus form data continuity of each humus form property and according to study sites suggested the pres- study site was examined by the variograms ence of comparable mean values for the 5 computed as an average overall direction
properties and sites. It appears that in all using equation [1] and assuming isotropy - study sites humus forms have developed similar spatial continuity with direction. The in polygons with the lateral dimension rang- data collected from the small, 7 x7 grids ing from about 100 to 700 cm, and that their were used for the lag distance (h) ≤ 100 cm, spatial continuity increases somewhat from and those collected from the large 10 x 10 disturbed to undisturbed sites. grid were used for the lag distance > 100 cm. Although the maximum lag dis- The property with the absolutely short- tance could have been 1 000 cm, the max- est range (46 cm) was total N in the dis- imumh of 800 cm was used in order to have turbed western hemlock site (table IV, fig each lag class adequately represented by a 1).This feature manifests a nearly random sufficient number of data. spatial pattern of Hemimors and Mormoders versus Lignomors and Lignomoders, each The parameters of the models fitted to pair with strongly contrasting N concentra- experimental variograms are given in table tions (table III). The property with the abso- IV, and the fitted regression lines are shown lutely longest range (1 251 cm) was miner- in figure 1. The models used for fitting pro- alizable-N in the Douglas-fir site (table IV). duced transitive variograms, which are forms This feature indicates a low spatial variabil- of second-order stationarity with finite vari- ity, which might be related to a uniform for- ances represented by the sill; the spherical est floor cover resulting from disturbance. models represent the variograms with fixed range, the exponential models the vari- To compare the nugget effect within- and ograms without fixed range. between-site, relative nugget variances, ie (real) nugget variances out of sills in per- The computed and plotted variograms centage, were calculated (table IV). These showed that the distribution of each of the 5 variances also varied with property and site humus properties is not random but spa- (fig 1).The relative nuggets for easily mea- tially-dependent as their estimated vari- sured thickness and acidity were clearly ogram values increase with increasing lags smaller than those for total C, total N and to their sills, at a finite lag or approaching mineralizable-N (table IV), ie the properties the sill asymptotically (table IV, fig 1).Over- with a greater likelihood of analytical error. all, the variograms were generically similar, The low relative nuggets for thickness and reflecting relatively small differences in spa- acidity, ranging from 0.2 to 14.0%, indicated tial continuity of their properties, and imply- that their structural variances account for ing a small-scale spatial pattern of humus more than 85% of their sill variances and form variability. Despite the overall similar- approach their overall sample variances. ity, the variograms varied with property and The high relative nuggets for total C, total site.This suggested that each property has N and mineralizable-N, ranging from 32 to a somewhat different spatial pattern 70%, indicated that their nuggets represent imposed by the property itself, the factors a large proportion of their total variance that controlling humus form development in each can be modelled as spatial dependence site, and the history of site disturbance. from the available sampling scheme. The average range values for the humus Using the variogram models (table IV) form properties increased from 109 cm for with kriging algorithm (eq [4]), the values total N to 708 cm for mineralizable-N, and for each of the 5 humus form properties those for the study sites increased from 275 were estimated for a total of 1 581 unsam- cm in the western hemlock site to 581 cm in pled locations in each large (10 x 10 m) grid. the Douglas-fir site. Thus, the ranges Since the configuration of sampling loca- beyond which humus forms are no longer tions had the regular, 100 cm sampling inter- spatially dependant were short for both the
val and the interval for kriging was 25 cm, within m ± 2σ except a few ϵ ϵ , errors were each of the 1 681 measured-plus-kriged where the errors were slightly smaller cases points was located at the nodes of the 25 x than 95%. 25 cm grid. Each kriged point was estimated As a supplement to the spatial analysis, using 16 measured points around it. The the contour maps based on the measured- means and standard deviations for the mea- plus-kriged values were produced for each sured values (n= 100) and the measured- of the 5 humus form properties in each of plus-kriged values (n 1 681) are given in = the 3 10 x 10 m study sites (fig 2). We con- table V. sider these maps more precise (with the The mean estimated errors were sub- precision definable in terms of the kriging mitted to t-test (Zar, 1984; table VI). Com- variance) than those which would be pro- pared to the value of 1.984 for t (2), 99 duced from the original samples, as 16.81 0.05 , all the mean estimated errors were signifi- times more values were used to construe cantly equal to zero, except for mineraliz- a picture of spatial continuity. The maps able-N in the Douglas-fir site with mean esti- illustrate the interpretations made earlier mated error close to 1.984. The verification from variograms, ie the distribution of all 5 of the low variance also showed that the humus form properties is spatially-depen- percentages of the observed estimation dent and generically similar, and that the 5 humus form properties measured in the 3
CONCLUSION sites are spatially continuous over a study short distance. Furthermore, the maps illus- trate an aspect which was not examined in The spatial analysis of 5 humus form prop- this study - a joint spatial dependence erties in 3 sites showed the presence of a between humus form properties. For exam- distinct pattern that reflected spatial depen- ple, the right center and lower right regions dence. The structural spatial dependence of the 10 x 10 m grid for the Douglas-fir site ranged from 46 to 1 251 cm, and varied (fig 2, center) shows relatively thicker humus somewhat with property and site. The most forms. Relative to other regions of the grid, spatially continuous property was mineral- the same area is also shown to have a izable-N, and the most spatially discontinu- higher acidity, higher total C concentration ous property was total N. The results sug- and lower total N and mineralizable-N con- gest a relatively low spatial continuity and centrations, ie the characteristics of Lig- small-scale pattern of humus form devel- nomors and Lignomoders, which, in fact, opment which appears to occur in polygons were the prevailing humus forms in these with the lateral dimension ranging from about 100 to 700 cm. 2 regions.
Delhomme JP (1979) Spatial variability and uncertainty ACKNOWLEDGMENTS in groundwater flow parameters: a geostatistical approach. Water Resour Res 15, 269-280 We thank Dr H Schreier, Department of Soil Sci- Fortin MJ, Drapeau P, Legendre P (1989) Spatial auto- University of British Columbia, Dr A Franc, correlation and sampling design in plant ecology. ence, Département de Mathématiques Appliquées et Vegetatio 83, 209-222 Informatique, ENGREF, and one anonymous Gamma Design Software (1992) GS professional geo- : + reviewer for helpful comments on the manuscript. statistics for the PC. Version 2. Plainwell, MI, USA Financial support for this study was provided, in Green RN, Trowbridge RL, Klinka K (1993) Toward a part, by the Natural Science and Engineering taxonomic classification of humus forms. For Sci Council of Canada. Monogr 29, 1-48 Hajrasuliha S, Baniabbassi N, Metthey J, Nielsen DR (1980) Spatial variability of soil sampling for salinity studies in south-west Iran. Irrigation Science 1, 197- REFERENCES 208 Isaaks EH, Srivastava RM (1989) Spatial continuity mea- Akaike H (1973) Information theory and an extension sures for probabilistic and deterministic geostatis- of maximum likelihood principle. In: Second Inter- tics. Math Geol 20, 313-341 national Symposium on Information Theory (BN Journel AG, Huijbregts CJ (1978) Mining geostatistics. Petrov, F Csáki, eds), Akadémia Kiadó, Budapest, Academic Press, London, UK 267-281 Kemp WP, Kalaris TM, Quimby WF (1989) Rangeland Anonymous (1976) Technicon autoanalyzer. II. Method- grasshopper (Orthoptera: Acrididae) spatial vari- ology: industrial individual/simultaneous determina- ability: macroscale population assessment. J Econ tion of nitrogen and/or phosphorus in BD acid digest. Entom 82, 1270-1276 Industrial Method 329/4W/A, Technicon Corpora- (1983) Statistical estimation of polynomial Kitandis DK tion. Tarrytown, New York, USA generalized covariance functions and hydrological Biggar JW, Nielsen DR (1976) Spatial variability of the applications. Water Resour Res 19, 909-921 leaching characteristics of a field soil. Water Resour Klinka K, Green RN, Courtin PJ, Nuzsdorfer FC (1984) Res 12, 78-84 Site diagnosis tree species selection, and slashburn- Bremner JM, Tabatabai MA (1971) Use of automated ing guidelines for the Vancouver Forest Region. BC combustion techniques for total carbon, total nitrogen, Min For, Land Manage Rep no 8, Victoria, BC, Canada and total sulphur analysis of soils. In: Instrumental Klinka K, Krajina VJ, Ceska A, Scagel AM (1989) Indi- methods for analysis of soils and plant tissue (LM catorplants of coastal British Columbia. University of Walsh, ed), Soil Science Society of America, Madi- British Columbia Press, Vancouver, BC, Canada son, WI, USA, 1-16 Pojar J, Meidinger DV (1991) Revision of bio- Klinka K, Burgess TM, Webster R (1980a) Optimal interpolation geoclimatic units of coastal British Columbia. North- and isarithmic mapping of soil properties. I. The semi- west Sci 65, 32-47 variogram and punctual kriging. J Soil Sci 31, 315- 331 Krige DG (1966) Two-dimensional weighted moving average trend surfaces for evaluation. In: Proc ore Burgess TM, Webster R (1980b) Optimal interpolation Symp on Mathematics, Statistics, and Computer and isarithmic mapping of soil properties. II. Block Applications in Ore Evaluation. South African Institute kriging. J Soil Sci 31, 333-341 of Mining and Metallurgy, Johannesburg Campbell JB (1978) Spatial variation of sand content Lowe LE, Klinka K (1981) Forest humus in the Coastal and pH within single contiguous delineations of 2 Western Hemlock biogeoclimatic zone of British soil mapping units. Soil Sci Soc Amer J 42, 460-464 Columbia in relation to forest productivity and pedo- Canada Soil Survey Committee (1978) The Canadian genesis. BC Min For, Res Note no 89, Victoria, BC, system of soil classification. Publ 1646, Can Dept Canada Agric, Ottawa, Ontario Matheron G (1963) Principles of geostatistics. Econ Geol Clark I (1979) Practical geostatistics. Applied Science 58, 1246-1266 Publishers, London, UK Matheron G (1971) The theory of regionalized variables David M (1977) Geostatistical estimation. ore reserve and its applications. Les Cahiers du Centre de Mor- Elsevier, Amsterdam phologie Mathématique de Fontainebleau, no 5, ENSMP, Paris, France Delhomme JP (1976) Kriging in hydrosciences. Centre d’informatique géologique, Fontainebleau, France McBratney AB, Webster R (1986) Choosing functions for semi-variograms of soil properties and fitting them Delhomme JP (1978) Kriging in the hydrosciences. Adv to sampling estimates. J Soil Sci 37, 617-639 Water Resour 1, 251-266
McCullagh MJ (1975) Estimation by kriging of the relia- Advances in agronomy, vol 38 (NC Brady, ed), Aca- bility of the proposed trend telemetry network. Com- demic Press, New York, NY, USA, 45-94 puter Applications 2, 357-374 Vauclin M, Vieira SR, Vachaud G, Nielsen DR (1983) Nielsen DR, Biggar JW, Erh KT (1973) Spatial variabil- The use of cokriging with limited field soil observa- tions. Soil Sci Soc Amer J 47, 175-184 ity of field-measured soil water properties. Hilgardia 42, 215-260 Vieira SR, Nielsen DR, Biggar JW (1981) Spatial vari- ability of field-measured infiltration rate. Soil Sci Soc Oliver MA, Webster R (1986) Semi-variograms for mod- Amer J 45, 1040-1048 elling spatial pattern of landform and soil properties. Earth Surface Processes and Landforms 11, 491-504 Vieira SR, Hatfield JL, Nielsen DR, Biggar JW (1983) Geostatistical theory and application to variability of Palmer MW (1988) Fractal geometry: a tool for describ- some agronomical properties. Hilgardia 51, 1-75 ing spatial patterns of plant communities. Vegetatio 75, 91-102 Webster R (1985) Quantiative spatial analysis of soil in the field. Adv Soil Sci 3, 1-70 DV (1987) Biogeoclimatic Pojar J, Klinka K, Meidinger ecosystem classification in British Columbia. For Wilkinson L (1990a) SYSTAT. the system for statistics. Ecol Manage 22, 119-154 SYSTAT Inc, Evanston, IL, USA Powers RF (1980) Mineralizable soil nitrogen as an Wilkinson L (1990b) SYGRAPH: the system for graphics. index of nitrogen availability to forest trees. Soil Sci SYSTAT Inc, Evanston, IL, USA Soc Amer J 44, 1314-1320 Xu JY, Webster R (1984) A geostatistical study of topsoil Robertson GP (1987) Geostatistics in ecology: interpo- properties in Zhangwu County, China. Catena 11, lating with known variance. Ecology 68, 744-748 13-26 Rossi RE, Mulla DJ, Journel AG, Franz EH (1992) Geo- Yost RS, Uehara G, Fox RL (1982a) Geostatistical anal- statistical tools for spatial modelling and interpret- ysis of soil chemical properties of large land areas. I. ing ecological spatial dependence. Ecol Monogr 62, Semi-variograms. Soil Sci Soc Amer J 46, 1028- 277-314 1032 Tabor JA, Warrick AW, Penningto DA, Myers DE (1984) Yost RS, Uehara G, Fox RL (1982b) Geostatistical anal- Spatial variability of nitrate in irrigated cotton. I. Peti- ysis of soil chemical properties of large land areas. II. oles. Soil Sci Soc Amer J 48, 602-607 Kriging. Soil Sci Soc Amer J 46, 1033-1037 Trangmar BB, Yost RS, Uehara G (1985) Application Zar JH (1984) Biostatistical analysis. Prentice-Hall, Engle- of geostatistics to spatial studies of soil properties. In: wood Cliffs, NJ, USA