Ettema CH, Wardle DA. 2002. Spatial soil ecology. Trends in Ecology & Evolution 17: 177-183.
These authors review the growing use of explicit geospatial analysis techniques in soil biology. As this is a TREE article, there are several helpful boxes that explain fundamentals of geospatial analysis such as the terminology and key case studies. This is also a review article, so there are descriptions of various previous studies that include evidence useful in answering the questions set out in this paper. These questions are 1) What are the scales, patterns and causes of spatial variability in soil organism distributions? 2) What are the implications of spatial variability for the structure and function of soil communities? 3) How do spatial properties of the soil biota influence plant communities?
Regarding question 1, the scales and patterns of spatial variability in soil organisms range from 10s and 100s of metres down to millimetres. Studies of soil microbes including methanogenic Archaea have included soil corers of 1mm diameter (based on a hollow needle) and aggregations of organisms separated by distances of 2 to 4mm.
Soil communities and their influence on plant communities were found to be highly non-uniform, and show predictable though complex spatial patterns. However, while much was made of the role of individual plants (especially trees) to structure the soils around them and create spatial patterns of microbes and invertebrates on the same scale as the trees themselves are distributed, very little was made of the role that small-scale aggregations play in structuring larger patterns. This is surprising, given the highly biased view of soil processes in this paper and more generally in the soil science literature: soil is viewed as something that exists primarily to support plants, rather than a system of its own independent importance. That is the impression I have gotten, at least.
This paper is a very useful overview of geospatial analysis, and the reference list includes a number of similarly useful papers. In particular, further exploration of the statistics of semivariance patterns seems useful.
Tuesday, September 15, 2009
Monday, September 14, 2009
Garten et al. 2007
Garten CT Jr., Kang S, Brice DJ, Schadt CW, Zhou J. 2007. Variability in soil properties at different spatial scales (1m-1km) in a deciduous forest ecosystem. Soil Biology & Biochemistry 39: 2621-2627.
These authors examined some of the fundamental assumptions of geospatial analysis as applied to soil properties, in a pair of transects that I strongly suspect have been used repeatedly for many studies in Tennessee (see, e.g. Zhou et al. 2008). One of the fundamental assumptions is of spatial autocorrelation, that is, samples in close proximity will be more similar to each other than samples separated by greater distances. In this study, this assumption was stated as the null hypothesis “there are no differences in variance at different spatial scales”; a rejection of this null hypothesis can be interpreted as support for the “common sense” (their wording) principle of spatial autocorrelation, at least among the spatial scales discussed here (i.e. metres to kilometres). This and other important assumptions of geospatial analysis come from a series of papers applying these principles to soils, which I should probably read soon.
The 11 soil variables examined in this paper were distinctly non-orthogonal in their relationships. Many of the variables were calculated directly from other variables, and the majority takes the form of either ratios (such as C-to-N) or fractions (such as silt content). The Principle Components Analysis (PCA) these authors conducted on their final, grand-total dataset indicated that the usual suspects of soil properties were important – soil Carbon, soil Nitrogen, and soil Texture are one way to summarize the first three PC variables.
I looked up and read this paper mainly because of the statistical tests used here and the discussion of them. They conducted 5 main statistical tests.
1. Bartlett’s test for equal variances at different distances.
2. Bartlett’s test is sensitive to non-normal data, so they also used the non-parametric Spearman’s Rank Correlation between coefficients of variation (100 x S.D./mean). The other reason a non-parametric test was used was that the functional relationship (linear vs. non-linear) between variance and sampling distance was unknown.
3. Mantel and Partial Mantel tests. These were the central analysis, I think, and provided most of the key results regarding the tests of the main hypotheses. Apparently, Burrough (1993) recommends semivariogram analysis, but the present data set was not amenable to such.
4. PCA, as mentioned above.
5. Power analysis. How many samples would they need to collect to be more certain of being close to the true mean value in their estimates?
Besides the PCA, which produced utterly unsurprising results, I think the statistical tests deployed here will serve as models for my own analysis of 2009 and putative 2010 datasets from the High Arctic. In particular, the Mantel tests and the Power analysis should be very useful in my own examinations.
Overall, the geospatial assumption of spatial autocorrelation was not very well supported by this study. Many soil properties appear to be highly variable at small spatial scales, such that samples collected within a few metres of each other are as variable as samples collected from up to a kilometre away, at least in a temperate forest ecosystem as studied here. This is particularly surprising in light of the consideration of the structure of such a forest, where individual trees presumably have strong impacts on soil properties within perhaps 5 to 10 metres of their trunks.
These authors examined some of the fundamental assumptions of geospatial analysis as applied to soil properties, in a pair of transects that I strongly suspect have been used repeatedly for many studies in Tennessee (see, e.g. Zhou et al. 2008). One of the fundamental assumptions is of spatial autocorrelation, that is, samples in close proximity will be more similar to each other than samples separated by greater distances. In this study, this assumption was stated as the null hypothesis “there are no differences in variance at different spatial scales”; a rejection of this null hypothesis can be interpreted as support for the “common sense” (their wording) principle of spatial autocorrelation, at least among the spatial scales discussed here (i.e. metres to kilometres). This and other important assumptions of geospatial analysis come from a series of papers applying these principles to soils, which I should probably read soon.
The 11 soil variables examined in this paper were distinctly non-orthogonal in their relationships. Many of the variables were calculated directly from other variables, and the majority takes the form of either ratios (such as C-to-N) or fractions (such as silt content). The Principle Components Analysis (PCA) these authors conducted on their final, grand-total dataset indicated that the usual suspects of soil properties were important – soil Carbon, soil Nitrogen, and soil Texture are one way to summarize the first three PC variables.
I looked up and read this paper mainly because of the statistical tests used here and the discussion of them. They conducted 5 main statistical tests.
1. Bartlett’s test for equal variances at different distances.
2. Bartlett’s test is sensitive to non-normal data, so they also used the non-parametric Spearman’s Rank Correlation between coefficients of variation (100 x S.D./mean). The other reason a non-parametric test was used was that the functional relationship (linear vs. non-linear) between variance and sampling distance was unknown.
3. Mantel and Partial Mantel tests. These were the central analysis, I think, and provided most of the key results regarding the tests of the main hypotheses. Apparently, Burrough (1993) recommends semivariogram analysis, but the present data set was not amenable to such.
4. PCA, as mentioned above.
5. Power analysis. How many samples would they need to collect to be more certain of being close to the true mean value in their estimates?
Besides the PCA, which produced utterly unsurprising results, I think the statistical tests deployed here will serve as models for my own analysis of 2009 and putative 2010 datasets from the High Arctic. In particular, the Mantel tests and the Power analysis should be very useful in my own examinations.
Overall, the geospatial assumption of spatial autocorrelation was not very well supported by this study. Many soil properties appear to be highly variable at small spatial scales, such that samples collected within a few metres of each other are as variable as samples collected from up to a kilometre away, at least in a temperate forest ecosystem as studied here. This is particularly surprising in light of the consideration of the structure of such a forest, where individual trees presumably have strong impacts on soil properties within perhaps 5 to 10 metres of their trunks.
Thursday, September 10, 2009
Zhou et al. 2008
Zhou J, Kang S, Schadt CW, Garten CT Jr. 2008. Spatial scaling of functional gene diversity across various microbial taxa. Proceedings of the National Academy of Sciences of the USA 105: 7768-7773.
These authors used a microarray-based technique to estimate biodiversity of soil microbes across a pair of transects in a forest in Tennessee. Their analysis found rates of species turnover through space much lower than rates for macroorganisms such as “higher” plants and animals.
The species-area relationship, generalized to the Taxa-Area-Relationship (TAR), is S=cA^z, where S is the number of species, A is the area, c is the intercept in log-log space, and z is a measure of the rate of species turnover across space. Values of z for macroorganisms have been estimated close the theoretically derived value of 0.25, while previous estimates for microbes have been often much lower, but occasionally much higher. This study found a range of z-values, all a bit less than 0.1.
The key technique used in this study was the GeoChip, a microarray with nearly 25000 50-mer probes for more than 10000 genes in functional groups such as denitrification or heavy-metal resistance. As such, it represents an excellent tool for such investigations, because it reduces or avoids many of the microbe-diversity sampling artifacts such as undersampling that plague other methods.
A large fraction of the observed variation in sequences across the transects was unexplained. These authors speculate that a fraction of this unexplained variation may be driven by unexamined patterns and processes including biotic interactions (competition, trophic interaction), abiotic interactions (O2 concentrations, labile C pool), and microscale effects below 1m scales.
One interesting suggestion by these authors is to use metagenomic approaches to characterize key sequences of interest in a particular system, and then examine biodiversity using a microarray customized for these sequences. This is in line with what I was thinking in regards to using such techniques in polar desert soils – first, characterize what is there; second, look at biodiversity and patterns within diversity relating to groups of interest.
These authors used a microarray-based technique to estimate biodiversity of soil microbes across a pair of transects in a forest in Tennessee. Their analysis found rates of species turnover through space much lower than rates for macroorganisms such as “higher” plants and animals.
The species-area relationship, generalized to the Taxa-Area-Relationship (TAR), is S=cA^z, where S is the number of species, A is the area, c is the intercept in log-log space, and z is a measure of the rate of species turnover across space. Values of z for macroorganisms have been estimated close the theoretically derived value of 0.25, while previous estimates for microbes have been often much lower, but occasionally much higher. This study found a range of z-values, all a bit less than 0.1.
The key technique used in this study was the GeoChip, a microarray with nearly 25000 50-mer probes for more than 10000 genes in functional groups such as denitrification or heavy-metal resistance. As such, it represents an excellent tool for such investigations, because it reduces or avoids many of the microbe-diversity sampling artifacts such as undersampling that plague other methods.
A large fraction of the observed variation in sequences across the transects was unexplained. These authors speculate that a fraction of this unexplained variation may be driven by unexamined patterns and processes including biotic interactions (competition, trophic interaction), abiotic interactions (O2 concentrations, labile C pool), and microscale effects below 1m scales.
One interesting suggestion by these authors is to use metagenomic approaches to characterize key sequences of interest in a particular system, and then examine biodiversity using a microarray customized for these sequences. This is in line with what I was thinking in regards to using such techniques in polar desert soils – first, characterize what is there; second, look at biodiversity and patterns within diversity relating to groups of interest.
Friday, September 4, 2009
Broll et al. 1999
Broll G, Tarnocai C, Mueller G. 1999. Interactions between vegetation, nutrients and moisture in soils in the Pangnirtung Pass area, Baffin island, Canada. Permafrost and Periglacial Processes 10: 265-277.
These authors examined soils from 6 pedons in Pangnirtung Pass, a north-south pass between mountains on Cumberland Peninsula. Three pedons were from moist soils, and three from dry soils. The moisture content drove a major difference in soil structure: dry soils are not cryoturbated, resulting in strong differences in nutrient content and mineralization rates.
The goal of the study was to compare in detail these differences between dry and moist soils. This seems very similar to my PhD goals surrounding examinations of Polar Desert soils. This study thus represents a possible template for some of my own investigations.
These authors examined soils from 6 pedons in Pangnirtung Pass, a north-south pass between mountains on Cumberland Peninsula. Three pedons were from moist soils, and three from dry soils. The moisture content drove a major difference in soil structure: dry soils are not cryoturbated, resulting in strong differences in nutrient content and mineralization rates.
The goal of the study was to compare in detail these differences between dry and moist soils. This seems very similar to my PhD goals surrounding examinations of Polar Desert soils. This study thus represents a possible template for some of my own investigations.
Wednesday, September 2, 2009
Bockheim 1979
Bockheim JG. 1979. Properties and relative age of soils of southwestern Cumberland peninsula, Baffin island, N.W.T., Canada. Arctic and Alpine Research 11: 289-306.
This author sampled soils from more than 60 sites on the Cumberland peninsula of Baffin Island, mostly near the hamlet of Pangnirtung. This covered soils from two tundra vegetations (Dwarf shrub-sedge-moss-lichen on lowlands and coastal, stony sedge-moss-lichen in highlands and northern fjords) and the Polar Desert of Baffin island. The tundra soils ranged from mesic to subxeric, while the desert near Penny icecap was xeric. A similar gradient driven by latitude rather than altitude is referenced in Tedrow (1973).
Descriptions are made of the pH and various exchangeable and free minerals in the soils, along with how those components change with depth in each area. pH increases with depth, for example, especially in the Polar Desert. Phosphorus was found in surprisingly high levels in all soils. The active layer, or at least the layer above the permafrost, is much deeper than found on Ellesmere island, and appears to be deeper than 1m everywhere studied in this paper.
This author sampled soils from more than 60 sites on the Cumberland peninsula of Baffin Island, mostly near the hamlet of Pangnirtung. This covered soils from two tundra vegetations (Dwarf shrub-sedge-moss-lichen on lowlands and coastal, stony sedge-moss-lichen in highlands and northern fjords) and the Polar Desert of Baffin island. The tundra soils ranged from mesic to subxeric, while the desert near Penny icecap was xeric. A similar gradient driven by latitude rather than altitude is referenced in Tedrow (1973).
Descriptions are made of the pH and various exchangeable and free minerals in the soils, along with how those components change with depth in each area. pH increases with depth, for example, especially in the Polar Desert. Phosphorus was found in surprisingly high levels in all soils. The active layer, or at least the layer above the permafrost, is much deeper than found on Ellesmere island, and appears to be deeper than 1m everywhere studied in this paper.
Niederberger et al. 2008
Neiderberger TD, McDonald IR, Hacker AL, Soo RM, Barrett JE, Wall DH, Cary SC. 2008. Microbial community composition in soils of Northern Victoria Land, Antarctica. Environmental Microbiology 10: 1713-1724.
These authors present an analysis of a large collection of data regarding both microbial and metazoan biodiversity at relatively small scales in one part of Taylor Valley, Antarctica, one of the famous Dry Valleys. This contributes to both the Latitudinal Gradient Project, an international effort to characterize Antarctica, and to the biogeographical debate regarding the distribution and community assemblages of microbes and soil microfauna.
Biodiversity was higher than expected based on the physical characteristics of this extreme environment, and was much more variable at small (~200m) spatial scales. While the microbes identified by 16s sequences were not particularly surprising, the changes in community composition between study sites was high. This supports the hypothesis that extreme environments “select for” particular microbial physiologies, and that differences in soil physical features such as moisture and temperature are highly important, in distinct contrast to the “everything is everywhere” hypothesis of microbial biogeography.
NB October 1 2009: the “everything is everwhere” hypothesis (Beijerinck 1913) includes the second clause “the environment selects”, which implies my earlier impressions, above, are incorrect. This paper’s demonstration that extreme environments select for particular soil communities, and that local-scale variables such as moisture and temperature, rather than regional-scale variables such as climate factors, actually supports Beijerinck’s (1913) hypothesis, rather than countering it.
These authors present an analysis of a large collection of data regarding both microbial and metazoan biodiversity at relatively small scales in one part of Taylor Valley, Antarctica, one of the famous Dry Valleys. This contributes to both the Latitudinal Gradient Project, an international effort to characterize Antarctica, and to the biogeographical debate regarding the distribution and community assemblages of microbes and soil microfauna.
Biodiversity was higher than expected based on the physical characteristics of this extreme environment, and was much more variable at small (~200m) spatial scales. While the microbes identified by 16s sequences were not particularly surprising, the changes in community composition between study sites was high. This supports the hypothesis that extreme environments “select for” particular microbial physiologies, and that differences in soil physical features such as moisture and temperature are highly important, in distinct contrast to the “everything is everywhere” hypothesis of microbial biogeography.
NB October 1 2009: the “everything is everwhere” hypothesis (Beijerinck 1913) includes the second clause “the environment selects”, which implies my earlier impressions, above, are incorrect. This paper’s demonstration that extreme environments select for particular soil communities, and that local-scale variables such as moisture and temperature, rather than regional-scale variables such as climate factors, actually supports Beijerinck’s (1913) hypothesis, rather than countering it.
Monday, April 20, 2009
Rolston 1986
Rolston DE. 1986. Gas diffusivity. Pp 1089-1102 in Methods of Soil Analysis Part 1: Physical and Mineralogical Methods 2nd Ed. ed. A Klute. American Society of Agronomy Inc. Soil Science Society of America Inc, Madison, WI.
This author presents a review of the principles and measurement methods of soil gas diffusion, including formulae and relevant calculations. At the heart of all considerations of soil gas diffusion is Fick’s law and the measurement of Dp, the soil gas diffusion constant for a particular gas. There are a wide range of laboratory methods for measuring gas diffusion, but all are based on measuring the passive movement of a target gas through a volume of soil, often by measuring the accumulation of the target gas in a chamber that initially lacks that gas.
Many variables will impact rates of gas diffusion; perhaps of greatest importance is the moisture content of the soil. Wet soils make take hours to measure, while dry soils only a few minutes. Temperature also has a strong effect, and this author urges the reporting of temperatures with all measures of soil diffusion. Additionally, a simple formula is presented that relates diffusion at one temperature to diffusion at a second temperature.
Further calculations surround the determination of Dp, and correcting for errors associated with the details of the measurement chamber. For example, a correction factor can be applied when the ratio of soil air volume to chamber volume is more than about 0.005; this corrects for the gas “stored” in the soil of the measurement apparatus.
This paper is the first of two adjacent chapters in this edited volume by this author, both dealing with soil gas movements. The calculations and formulae here will be very useful in attempts to calibrate the FTIR and the flux chambers.
This author presents a review of the principles and measurement methods of soil gas diffusion, including formulae and relevant calculations. At the heart of all considerations of soil gas diffusion is Fick’s law and the measurement of Dp, the soil gas diffusion constant for a particular gas. There are a wide range of laboratory methods for measuring gas diffusion, but all are based on measuring the passive movement of a target gas through a volume of soil, often by measuring the accumulation of the target gas in a chamber that initially lacks that gas.
Many variables will impact rates of gas diffusion; perhaps of greatest importance is the moisture content of the soil. Wet soils make take hours to measure, while dry soils only a few minutes. Temperature also has a strong effect, and this author urges the reporting of temperatures with all measures of soil diffusion. Additionally, a simple formula is presented that relates diffusion at one temperature to diffusion at a second temperature.
Further calculations surround the determination of Dp, and correcting for errors associated with the details of the measurement chamber. For example, a correction factor can be applied when the ratio of soil air volume to chamber volume is more than about 0.005; this corrects for the gas “stored” in the soil of the measurement apparatus.
This paper is the first of two adjacent chapters in this edited volume by this author, both dealing with soil gas movements. The calculations and formulae here will be very useful in attempts to calibrate the FTIR and the flux chambers.
Subscribe to:
Posts (Atom)
