Unmix模型污染源解析研究进展及应用前景

基于国内外近十年的期刊文献,归纳总结了Unmix模型在污染源解析研究中的应用情况,对比分析了在不同介质和研究对象上的差异,对该模型的应用前景进行了展望.现阶段Unmix模型在国外主要用于大气颗粒物的源解析,而国内则多应用于多环芳烃(PAHs)和土壤重金属的源解析.Unmix模型与主成分分析法(PCA)、正定矩阵因子分析...     

基于多元统计学和地统计学的土壤重金属源解析

The main objectives of this study were to introduce an integrated method for effectively identifying soil heavy metal pollution sources and apportioning their contributions, and apply it to a case study. The method combines the principal component analysis/absolute principal component scores (PCA/APCS) receptor model and geostatistics. The case study was conducted in an area of 31 km² in the urban-rural transition zone of Wuhan, a metropolis of central China. 124 topsoil samples were collected for measuring the concentrations of eight heavy metal elements (Mn, Cu, Zn, Pb, Cd, Cr, Ni and Co). PCA results revealed that three major factors were responsible for soil heavy metal pollution, which were initially identified as “steel production”, “agronomic input” and “coal consumption”. The APCS technique, combined with multiple linear regression analysis, was then applied for source apportionment. Steel production appeared to be the main source for Ni, Co, Cd, Zn and Mn, agronomic input for Cu, and coal consumption for Pb and Cr. Geostatistical interpolation using ordinary kriging was finally used to map the spatial distributions of the contributions of pollution sources and further confirm the result interpretations. The introduced method appears to be an effective tool in soil pollution source apportionment and identification, and might provide valuable reference information for pollution control and environmental management.     

An Assessment of the Mobile Source Contribution to PM10 and PM2.5 in the United States

Mobile sources are significant contributors to ambient particulate matter (PM) in the United States. As the emphasis shifts from PM₁₀ to PM_2.5, it becomes particularly important to account for the mobile source contribution to observed particulate levels since these sources may be the major contributor to the fine particle fraction. This is due to the fact that most mobile source mass emissions have an aerodynamic diameter less than 2.5 µm, while the particles of geological origin that tend to dominate the PM₁₀ fraction generally have an aerodynamic diameter greater than 2.5 µm. A common approach to assess the relative contributions of sources to observed particulate mass concentrations is the application of source apportionment methods. These methods include material balance, chemical mass balance (CMB), and multivariate receptor models. This paper describes a number recent source attribution studies performed in the United States in order to evaluate the range of the mobile source contribution to observed PM. In addition, a review of the methods used to apportion source contributions to ambient particulate loadings is presented.     

土壤污染物源解析方法及其应用研究进展

本文介绍了污染物源解析的方法体系,重点对化学质量平衡法、多元统计模型、混合方法、分子标志物与同位素技术等受体模型研究方法的原理与应用进展加以总结,论述了化学质量平衡法和多元统计模型的优缺点、适用范围及应用现状;指出以这两种方法为基础形成的混合方法可避免前者选择排放源种类时的主观性,同时又可检验后者在鉴别过程中是否遗漏其他未知源,从而能够发挥最大的优势;最后对源解析方法的发展趋势及土壤学应用加以展望。     

Size and Geographical Variation in PM1, PM2.5 and PM10: Source Profiles from Soils in the Western United States

Source profiles (fractional mass abundances and uncertainties) for use in chemical mass balance (CMB) source apportionment are reported for soil dust from unpaved roads and desert lands. The objectives were to add to the library of source composition profiles available for CMB studies, and to provide data on the variation in wind-blown dust composition between particle sizes and geographical locations. Samples were resuspended and sampled through PM₁, PM_2.5 and PM₁₀ inlets onto filters, then chemically analyzed for 40 elements, seven ions and eight carbon fractions. Air quality management would benefit from technology that can distinguish soil dust from sites in different geographical locations and sites with different land uses. Five geographical clusters representing Ft Bliss, TX, Dona Ana County, NM, Ft Irwin, CA, the west desert and the Canyonlands, UT, were examined. Distinctive differences were found between sites within a cluster and these differences were comparable to the differences between cluster averages. Some sites showed small differences in chemical composition between particle sizes, but these differences were less than the differences observed between nearby geographical locations. Dust emissions from sites with uniform geology have distinctive source profiles that may be useful for CMB receptor modeling. Heterogeneous regions, defined by land management boundaries, will require extensive field sampling to develop representative composite source profiles.     

Evaluating tracer selection for catchment sediment fingerprinting

Purpose

Recent sediment fingerprinting research has shown the sensitivity of source apportionment results to data treatments, tracer number, and mixing model type. In light of these developments, there is a need to revisit procedures associated with tracer selection in sediment fingerprinting studies. Here, we evaluate the accuracy and precision of different procedures to select tracers for un-mixing sediment sources.

Materials and methods

We present a new approach to tracer selection based on identifying and removing tracers that exhibit non-conservative behaviour during sediment transport. This removes tracers on the basis of non-conservative behaviour identified using (1) tracer-particle size relationships and (2) source mixing polygons. We test source apportionment results using six sets of tracers with three different synthetic mixtures comprising one, five, and ten mixture samples. Source tracer data was obtained from an agricultural catchment in northwest England where time-integrated suspended sediment samples were also collected over a 12-month period. Source un-mixing used MixSIAR, a Bayesian mixing model developed for ecological food web studies, which is increasingly being applied in catchment sediment fingerprinting research.

Results and discussion

We found that the most accurate source apportionment results were achieved by the selection procedure that only removed tracers on the basis of non-conservative behaviour. Furthermore, accuracy and precision were improved with five or ten mixture samples compared to the use of a single mixture sample. Combining this approach with a further step to exclude additional tracers based on source group non-normality reduced accuracy, which supports relaxation of the assumption of source normality in MixSIAR. Source apportionment based on the widely used Kruskal-Wallis H test and discriminant function analysis approach was less accurate and had larger uncertainty that the procedure focused on excluding non-conservative tracers.

Conclusions

Source apportionment results are sensitive to tracer selection. Our findings show that prioritising tracer exclusion due to non-conservative behaviour produces more accurate results than selection based on the minimum number of tracers that maximise source discrimination. Future sediment fingerprinting studies should aim to maximise the number of tracers used in source un-mixing constrained only by the need to ensure conservative behaviour. Our procedure provides a quantitative approach for identifying and excluding those non-conservative tracers.

     

GIS-Based Emission Inventory, Dispersion Modeling, and Assessment for Source Contributions of Particulate Matter in an Urban Environment

The Industrial Source Complex Short Term (ISCST3) model was used to discern the sources responsible for high PM₁₀ levels in Kanpur City, a typical urban area in the Ganga basin, India. A systematic geographic information system-based emission inventory was developed for PM₁₀ in each of 85 grids of 2?×?2 km. The total emission of PM₁₀ was estimated at 11 t day^?1 with an overall breakup as follows: (a) industrial point sources, 2.9 t day^?1 (26%); (b) vehicles, 2.3 t day^?1 (21%); (c) domestic fuel burning, 2.1 t day^?1 (19%); (d) paved and unpaved road dust, 1.6 t day^?1 (15%); and the rest as other sources. To validate the ISCST3 model and to assess air-quality status, sampling was done in summer and winter at seven sampling sites for over 85 days; PM₁₀ levels were very high (89?C632 ??g m^?3). The results show that the model-predicted concentrations are in good agreement with observed values, and the model performance was found satisfactory. The validated model was run for each source on each day of sampling. The overall source contribution to ambient air pollution was as follows: vehicular traffic (16%), domestic fuel uses (16%), paved and unpaved road dust (14%), and industries (7%). Interestingly, the largest point source (coal-based power plant) did not contribute significantly to ambient air pollution. The reason might be due to release of pollutant at high stack height. The ISCST3 model was shown to produce source apportionment results like receptor modeling that could generate source apportionment results at any desired time and space resolution.     

Source identification and apportionment of soil cadmium in cropland of Eastern China: a combined approach of models and geographic information system

Purpose

Cadmium (Cd) is regarded as one of the most toxic heavy metals in the environment and can undermine the ecosystem function and human health at trace level due to its high toxicity. In order to reduce the anthropogenic Cd input into agricultural soil, it is of utmost importance to pinpoint the sources of Cd in soils and apportion their respective contributions.

Materials and methods

One hundred twenty-seven topsoil samples and 21 subsoil samples were collected from croplands of Meishan Basin, Changxing County, Zhejiang Province, Eastern China, and analyzed for concentrations of Cd and other heavy metals. Finite mixture distribution model (FMDM) was employed to fit the data to obtain the local soil Cd threshold value, a critical indicator to assess soil heavy metal contamination. Then, principal component analysis (PCA) and geographic information system (GIS) were used to identify the potential sources of Cd. Finally, positive matrix factorization (PMF) was applied to apportion the source contributions.

Results and discussion

Among the 127 topsoil samples, 71 were subject to Cd contamination with a mean concentration of 0.66 mg kg^?1 while the others were considered as background with a lower mean concentration of 0.145 mg kg^?1, close to the local background concentration of 0.142 mg kg^?1. Further, three components were extracted by PCA and interpreted as natural background, lead-acid battery manufacturing plants, and construction material associated activities, respectively. Additionally, most of the topsoil samples around the lead-acid battery manufacturing plants, construction material plants, and limestone/marble quarries were classified as Cd contaminated. However, PMF failed to get a successful portioning.

Conclusions

Lead-acid battery manufacturing plants and construction material associated activities were the main anthropogenic sources of soil Cd contamination. With the help of FMDM, it is possible to distinguish the contaminated soil and estimate the contribution of anthropogenic sources to soil Cd. The apportionment by PMF was not successful in this paper due to the high skewness or outliers of Cd concentration in sampling sites and violation of the assumption that all samples have the same sources.

     

The use of magnetic susceptibility to measure long-term soil redistribution

Several studies have documented the severity of recent soil erosion on the Canadian prairies where cultivation started about a century ago. Little quantitative information is available on erosion before 1960. This study attempts to quantify post- and pre-1960 soil erosion in a small cultivated basin near Saskatoon, Saskatchewan, Canada, by measuring ¹³⁷Cs and magnetic susceptibility (χ) distribution with depth. Soil cores were collected along six transects (three across closed depressions and three across the drainage channel) in the cultivated field, and one transect across an uncultivated depression. The cores were sliced into 3-cm layers and the soil analyzed for ¹³⁷Cs, χ, and organic and inorganic C. High variability in χ with depth in eroding areas (as indicated by ¹³⁷Cs) made it impossible to use χ to quantify past soil losses in these locations. However, these eroding upper and middle slope positions have a much higher χ than lower slope areas where soil deposition occurs and where the variability in χ with depth could be used to estimate soil deposition. Estimating soil deposition from the χ vs. depth profiles was more successful in the closed depressions than in the drainage channel, where the χ profiles may reflect the variable source areas of the materials rather than the pedological conditions. The data indicated that soil deposition since 1960 has been about 30 to 50% of that prior to 1960. This suggests that soil deposition rates, and by implication, soil erosion rates, have been relatively steady since cultivation of these soils started, although there are clear indications that the spatial pattern of deposition has varied.     

Comparison of sampling designs for sediment source fingerprinting in an agricultural watershed in Atlantic Canada

Purpose

Sediment fingerprinting is increasingly being used to improve the understanding of sediment dynamics within the critical zone and provide information that can help guide management decisions at the watershed scale. The objectives of this study were to investigate both the implications of different sediment fingerprinting sampling designs and spatial scales on the characterization of sediment dynamics in a predominantly agricultural watershed in northwestern New Brunswick, Canada.

Materials and methods

Color and radionuclide fingerprints were used to discriminate between three potential sediment sources: agricultural topsoil, agricultural streambanks, and forested areas (topsoil and streambanks). Suspended sediment was collected seasonally, between 2008 and 2014, at five sites with drainage areas ranging from 3.0 to 13.4 km². Using the same source and sediment data set, multiple-, nested-, and local-location fingerprinting sampling designs were employed to investigate the influence of scale of observation, geomorphic connectivity, land use, and the heterogeneity of source fingerprints on apportionment results.

Results and discussion

Sediment collected in the headwaters was primarily derived from forested areas while the sediment collected at the outlet of the watershed was primarily from agricultural topsoil. When comparing the multiple- and nested-location designs, it was found that accounting for the spatial variability in the fingerprint properties of each source had a small difference in the sediment apportionment results. Furthermore, the local-location design demonstrated that the sediment collected at each location was composed of predominately local sources as opposed to upstream sediment entering the local catchment.

Conclusions

Assessment of the sources of sediment at a range of spatial scales better accounts for both geomorphic connectivity and differences in land use throughout the watershed. Overall, each of the three fingerprinting sampling designs provided different information that can be used to guide soil and water conservation management.

     

Source Identification of Rural Precipitation Chemistry in Japan

Precipitation chemistry was discussed from the viewpoint of potential sources for four rural sites where wet-only daily-basis measurement data sets were available during the period from April 1996 to March 1997 in Japan. Annual volume-weighted mean concentrations of nss-SO₄ ^2? and NO₃ ^? ranged from 18.0 to 34.6 µeq L^?1, and from 9.3 to 23.1 µeq L^?1, respectively. The degree of neutralization of input acidity in terms of the concentration ratio, [H⁺] / ([nss-SO₄ ^2?] + [NO₃ ^?]), ranged from 0.46 to 0.63. This suggests that about half of the input acidity due to H₂SO₄ and HNO₃ was neutralized by NH₄ ⁺ and nss-Ca²⁺ to produce the pH values of 4.46 to 4.82 for these sites. Maximum likelihood factor analysis was then performed on the logarithmically transformed daily wet deposition of major ions. Two factors successfully explained a total of about 80% of the variance in the data for each site. Interpreting varimax rotated factor loadings, we could identify two source types: (1) acid source with large loadings on ln(H⁺), ln(nss-SO₄ ^2?), ln(NO₃ ^?) and ln(NH₄ ⁺), (2) sea-salt source with large loadings on ln(Na⁺), ln(Cl^?), ln(Mg²⁺) and ln(K⁺). The rural wet deposition over Japan appears to have a similar structure in terms of the kinds of sources and their relative location.     

Sediment–pore water partition of PAH source contributions to the Yellow River using two receptor models

Purpose

Understanding the fate and behavior of polycyclic aromatic hydrocarbon (PAH) sources in aquatic systems is important for the efficiency of control policies. In this work, a new approach??organic carbon-normalized sediment?Cpore water partition coefficients of PAH source contributions (logK??_Osource)??was developed to study the sediment?Cpore water partition of PAH source contributions. The focus of this study was the Yellow River, which is the second largest river in China and one of the largest rivers in the world.

Materials and methods

Sixteen priority US Environmental Protection Agency PAHs were analyzed in 14 surface sediments and 11 pore water samples. Principal component analysis?Cmultiple linear regression (PCA-MLR) and Unmix models were employed to estimate the source contributions of PAHs in sediments and pore water samples. Finally, logK??_Osource values were calculated according to the modeled source contributions of PAHs.

Results and discussion

??PAHs (sum of the 16 PAH concentrations) in 14 sediment samples and 11 pore water samples from the Yellow River were 1,415?±?726?ng?g^?1 dry weight (dw) and 123?±?57.4???g?l^?1, respectively. The source apportionment results indicate the following: (1) for sediment samples, the contributions to ??PAHs from vehicular emissions, coal combustion, and petrogenic sources were 41.07?C61.05, 38.83?C45.56, and 11.18?C14.92?%, respectively, and (2) for pore water samples, vehicular emissions were the most significant contributor (45.51?C69.39?%), followed by petrogenic sources (29.80?C34.22?%) and coal combustion (7.35?C21.59?%). Coal combustion had the highest logK??_Osource values (4.15?C4.26) among the three categories, followed by vehicular emissions (3.51?C3.57) and petrogenic sources (3.30?C3.43).

Conclusions

The possible categories of PAH sources identified by hierarchical cluster analysis, PCA-MLR, and Unmix models were consistent, indicating that vehicular emissions, coal combustion, and petrogenic sources were three important categories. The logK??_Osource values indicate that contributions from coal combustion had a higher partition for the sediment phase compared with the other two source categories.     

Source apportionment of lake bed sediments to watersheds in an Upper Mississippi basin using a chemical mass balance method

Sediment samples were collected from three known sources contributing to a depositional basin of the Upper Mississippi River, USA. The heavy silt fraction (2–63 μm, density >2.95) isolated from the sediments was analyzed for 42 elements by inductively coupled plasma-mass spectrophotometry (ICP-MS) to give signature profiles characteristic of each source catchment. Known proportions of heavy silt fractions from each source were mixed and the composite samples analyzed in the same way. The proportions of components in the composite samples were compared to their apportionments calculated statistically using the US Environmental Protection Agency (USEPA) chemical mass balance (CMB) air quality receptor model. This showed that the CMB technique can apportion sediments to their catchment sources reliably. The technique was then applied to dated sediments of Lake Pepin, in the Upper Mississippi basin, to apportion them to their contributing catchments. The results show that almost all the recent sediments deposited in the basin originated in the Minnesota River catchment, and that the sediment load increased over time as anthropogenic impacts increased. This technique could be used elsewhere to determine sediment provenance and identify changes in sediment sources over time.     

Simulating badland erosion with KINEROS2 in a small Mediterranean mountain basin (Vallcebre,Eastern Pyrenees)

The Ca l?Isard catchment (1.32 km²), a sub-basin of the Vallcebre experimental catchments, yields large amounts of sediments (about 580 Mg km^− 2 year^− 1) that are produced in relatively small but very active eroded areas (badlands). Several lines of evidence suggest that there is a delay between sediment production, caused by intense summer rainstorms, and sediment transport, occasioned by the main floods produced by large precipitation events following wet antecedent conditions. First, a calibration–validation exercise was carried out with sediment yield data obtained using containers provided with slot divisors in a badlands micro-catchment (1240 m²). Then, the model was applied to the main badlands areas in the Ca l?Isard sub-catchment for a 4-year period and the simulated sediment yields were compared with the records at the gauging station. The test was performed with the Generalized Likelihood Uncertainty Estimation (GLUE) approach for assessing the uncertainty associated with model predictions, which assumes that many parameter sets can give acceptable simulations. The results demonstrated the capacity of KINEROS2 to simulate badland erosion, although it showed limited robustness. A clear temporal mismatch between erosion and sediment transport and the relevance of sediment stores in the catchment were confirmed, while the total weights of sediment were generally under-predicted. The limited suitability of the area used for calibration or the role of sediment sources not simulated in the approach may account for this shortcoming.     

Trace inorganic species in precipitation and their potential use in source apportionment studies

Trace species in atmospheric particles have frequently been used to apportion pollutants in acid rain to their sources by comparing the relative concentrations of trace species at the source and receptor sites. Another approach is to use the trace species in precipitation directly in a source apportionment. This has rarely been done due to the paucity of data on trace species in precipitation and the unknown fractionation of the species during the scavenging process. A study was undertaken to analyze precipitation for a large number of trace species and to compare the concentrations with those in atmospheric particles. Precipitation was collected in Lenox, MA during the summer of 1984. Extreme measures were taken to avoid contamination during collection and analysis. Using five analytical methods, including the novel method of freeze-drying the precipitation followed by neutron-activation analysis, 31 species were analyzed in the rain. The trace concentrations measured in this study were lower than those found in earlier studies, where contamination was believed to be a problem. Concentrations of trace species in the precipitation were compared with those in atmospheric particles. A few species were enriched in precipitation relative to particles, including some large-particle species (Ca, Ti, Fe) and species with gaseous precursors (SO₄, NO₃, Br). Due to this enrichment, these species should be carefully considered before using them in source apportionment calculations. However, most species showed similar enrichments in precipitation and in atmospheric particles, suggesting that negligible fractionation had taken place during the scavenging process. These species could be used effectively in the future to apportion the pollutants in precipitation to their various sources.     

指纹技术识别泥沙来源:进展与展望

泥沙来源信息有助于研究流域土壤侵蚀、泥沙输移与沉积特征,对于制定流域泥沙及非点源污染控制战略具有重要意义。泥沙指纹技术是识别泥沙来源的一个可靠方法被广泛应用于世界各地不同流域,最近20多年在指纹因子选择和混合模型优化等技术方法上以及不同时空尺度和源地分类等应用类型上都得到了快速发展,正在从一种研究工具逐步发展为一种管理工具。综述了国内外指纹技术识别泥沙来源在研究尺度、指纹因子筛选、源地贡献分配、不确定性分析等方面开展的研究,并对泥沙指纹技术研究当前存在的问题及今后研究趋势作了分析展望。     

sulphur isotope composition of stream water,moss and humus from eight arctic catchments in the Kola Peninsula region (NW Russia,N Finland,NE Norway)

In summer 1994, stream water, moss and humus samples were collected for sulphur isotopic analysis from eight catchments located in the western Kola Peninsula region, where several industrial centres emit high loads of SO₂ and other elements to the atmosphere. Three potential sources of sulphur and their isotopic signatures were identified: (1) marine (δ ³⁴S+20 to +21‰ CDT), (2) anthropogenic emissions (<+10‰), and (3) geogenic (variableδ ³⁴S, mostly <+10‰). Averaged per catchment, the sulphur isotopic composition varies between +6.0 and +16.3‰ for stream water sulphate, +6.0 and +8.4‰ for moss sulphur, and +5.2 and +12.2‰ for humus sulphur. Theδ ³⁴S composition of stream water from the more remote catchments is quite variable, reflecting several natural (geogenic) sources, but it becomes restricted to the range +8 to +10‰ near the pollution sources. A plot ofδ ³⁴S vs. 1:SO₄ in stream water suggests that sulphate originating from the smelters has aδ ³⁴S value ≈+9.5‰, and is a dominant source. Sulphur isotope values for moss and humus are consistent with the deduced composition for the emitted sulphur, though for humus a component of geogenic sulphur incorporated via vegetation uptake may play a role. Further isotopic characterisation of atmospheric emissions, together with environmental samples, is needed to better understand sulphur sources and sinks in the area.     

Polycyclic aromatic hydrocarbons (PAHs) in rural soils of Dongjiang River Basin: occurrence,source apportionment,and potential human health risk

Purpose

A comprehensive study was conducted to investigate the presence of polycyclic aromatic hydrocarbons (PAHs) in Dongjiang River Basin (DRB) soils and to evaluate their sources and ecological and health risk. In addition, factors affecting the distribution and fate of PAHs in the soils such as emission density, soil organic matter, degradation, etc. were studied.

Materials and methods

Surface soil (0–20 cm) samples from 30 sampling sites in the rural areas of DRB were collected and analyzed for 17 polycyclic aromatic hydrocarbons (16 EPA priority PAHs and perylene). Positive matrix factorization model was used to investigate the source apportionment of these PAHs, and an incremental lifetime cancer risk (ILCR) was used to estimate the integrated lifetime risks of exposure to soil-borne PAHs through direct ingestion, dermal contact, and inhalation collectively.

Results and discussion

The total PAH concentrations in the rural soils in DRB range from 23.5 to 231 μg/kg with a mean concentration of 116 μg/kg. The predominant PAHs in the rural soils were naphthalene, fluoranthene, phenanthrene, and benzo(b)fluoranthene. Cluster analysis was performed to classify the soil PAHs into three clusters, which could be indicative of the soil PAHs with different origins and different properties. Source apportionment results showed that coal, biomass, oil, commercial creosotes, and vehicle contributed 24 %, 24 %, 17 %, 17 %, and 18 % of the total soil PAH burden, respectively. The ILCR results indicated that exposure to these soil-borne PAHs through direct ingestion, dermal contact, and inhalation collectively produces some risk.

Conclusions

PAHs in the soils of the DRB will produce long-term influences on rivers and oceans via soil erosion and river transport. Therefore, PAHs in rural soils of DRB have potential impacts on the water supply and human health risk.     

Direct comparison of individual substrate utilization from a CLPP study: A new analysis for metabolic diversity data

For over a decade, community level physiological profile (CLPP) assays, which assess a microbial community's capacity to metabolize specific sole carbon sources under defined laboratory conditions, have been popular for study of environmental soil samples. One such assay, BiOLOG™ allows for the colorimetric measurement of metabolism through the reduction of a tetrazolium dye, which yields optical density (OD) data for each substrate. Bacterial communities are extracted from soil and 150 μL of this extract is inoculated directly into each well of the microtitre plate. The combined metabolic data obtained are most often analyzed with multivariate statistical analyses, such as principal component analysis (PCA). The objectives of this study were (1) to develop a simple, visual, statistically valid method of determining community capabilities to utilize specific substrates in CLPP studies and (2) to test the number of samples needed for such discrimination to be reliable. This was done by direct comparison of the OD values obtained for two closely related microbial communities (surface and subsurface soil), plotted against a one-to-one (y=x) line. Due to variability in the portion of the soil microbial community inoculated into the individual test wells, the accuracy of the method was dependent on the number of replicates analyzed. A variety of data set sizes were tested, from n=3 samples/soil depth to n=40 samples/depth. The method was statistically valid for all data sets tested. Those substrates that deviated from the one-to-one line consistently had F values greater than 1. Additionally, data sets of n=30,35 and 40 samples/depth consistently allowed identification of the eight substrates whose metabolism varied significantly between the two test soil communities. In conclusion, this one-to-one comparison has been shown to be a statistically valid analytical method to compare individual substrate usage between soils.