Geochemical Variability of Soils in Geochemical Variability of Soils - - PowerPoint PPT Presentation

Geochemical Variability of Soils in Geochemical Variability of Soils in the Maritime Provinces of Canada and the New England States of the and the New England States of the United States

E.C. Grunsky1, L.G. Woodruff2, L.J. Drew2, P.W.B. Friske1, D.M. Sutphin2 National Environmental Monitoring Conference August 2011

1 Natural Resources Canada 2U.S. Department of the Interior

Geological Survey of Canada U.S. Geological Survey

The North American Soil Geochemistry Landscape Project (NASGLP) Landscape Project (NASGLP)

• The NASGLP was developed as a joint Mexico, USA, and

Canada project for characterizing the soil cover over North Canada project for characterizing the soil cover over North America.

• Goals of the NASGLP are:

– develop a continental-scale framework for generating and managing develop a continental scale framework for generating and managing geochemical and mineralogical data. – produce a continental array of soil data using consistent sampling and analytical protocols. t bli h hi f il l f f t h – establish an archive of soil samples for future researchers

• In 2004 two transects (W-E and N-S) that crossed the US and

Canada were conducted as a pilot study – results reported in a special issue of Applied Geochemistry, Vol. 24 (8), 2009. p pp y, ( ),

• Spatial sampling in the US and Canada began in 2007 – Maritime

provinces sampled by GSC and provincial surveys; New England states + New York sampled by USGS.

Soil Horizons

PH{0 } PH{0-5} Three samples i thi in this analysis

Analytical Protocols

• Geochemical analysis of the 3 soil horizons (PH (0-5 cm depth), A &

C hori ons) C horizons)

• Samples sieved to < 2 mm & milled to < 150 mm
• 4 acid digestion (HNO3-HF-HCl-HClO4)

ac d d gest o ( O3 C C O4)

• ICP-MS/ICP-AES instrumentation
• Hg – Cold Vapour AA (US samples)
• Hg – Aqua-Regia ICP-MS (CD samples)
• As – Hydride generation (US samples)
• As – 4 acid ICP-MS (CD samples)

Sample Sites

~1 sample/1600 km2 in US; double density in Canada ~1 sample/1600 km in US; double density in Canada

(ACPB) (EGLHL) (EDP) (LEL) (MPH) (ML) (NCA) (NECZ) (NAPU) (NAAMH) (RV)

Basic Questions

• What is the chemical and mineralogical

What is the chemical and mineralogical variability in soil profile throughout different geological units and ecoregions?

• How can this be understood and visualized?
• What can a statistical analysis of the data tell

at ca a stat st ca a a ys s o t e data te us?

Compositional Aspects of the Geochemistry Geochemistry

• Soil geochemistry (% mg/kg) is

Soil geochemistry (%, mg/kg) is compositional and subject to closure.

• Centered logratio (clr) and isometric
• Centered logratio (clr) and isometric

logratio (ilr) transformations were used. R l ti hi i th d t l

• Relationships in the data reveal processes
• f deposition, erosion, weathering and

lt ti ( d t ff t ) alteration (e.g., groundwater effects).

Quantile-Quantile Plots – Soil Geochemistry Soil Geochemistry

Al dominant in C horizon As nearly equal in all 3 horizons Al dominant in C horizon As nearly equal in all 3 horizons

Quantile-Quantile Plots – Soil Geochemistry Soil Geochemistry

Organic carbon dominant in the A horizon Cd dominant in the PH & A horizons Organic carbon dominant in the A horizon Cd dominant in the PH & A horizons Carbon not determined in PH horizon

Quantile-Quantile Plots – Soil Geochemistry Soil Geochemistry

Hg dominant in PH & A horizons K dominant in the C horizon Hg dominant in PH & A horizons K dominant in the C horizon

Cd – PH Horizon

Cd associated with organic material in the PH horizon

Principal Component Analysis – Soil Geochemistry Soil Geochemistry

Physical Processes F l i

0.3

Nb

New England States & Maritime Provinces Soils [clr transform]

Soil Horizon PH A

Weathering &

Felsic

0.1 0.2

Ba Ce Ga K La Mo Na Pb Rb Sn Sr Th Ti U W

A C

Weathering & Mixing

Random Processes

Parent Material

.1 0.0 C 2

Al Bi Ca Cd Ce Hg La P Pb S Sb Se Y

Organic

3

• 0.2
• 0.

As Fe Li Mn Sb Sc V Zn

Mafic

• 0.4
• 0.3

Co Cr Cu Mg Ni Zn

• 0.4
• 0.3
• 0.2
• 0.1

0.0 0.1 0.2 0.3 0.4 C 1

PC1 – PH Horizon

Increased Cd-S-Pb-Se-Sn-Mo-Bi-Sb in NB and NS

PC2 – C Horizon

Mafic Mafic Felsic Carbonates Carbonates/Mafic rocks PC2< 0; Felsic rocks PC2 > 0

PC2-PC3 Biplots – Chemistry

Carbonates

F l i Sili t Mafic Silicates Felsic Silicates

Biplot shows bedrock affinities without the influence of the soil profile

PC3 – C Horizon

Linear Discriminant Analysis (LDA) – ILR Transform Soil Geochemistry ILR Transform – Soil Geochemistry

A Horizon with high C content

LDA Accuracy Matrix – Soil Geochemistry Soil Geochemistry

LDA Based on EcoRegions

(ACPB) (ACPB) (EGLHL) (EDP) (LEL) (LEL) (MPH) (ML) (NCA) (NCA) (NECZ) (NAPU) (NAAMH) (NAAMH) (RV)

LDA – EcoRegions – Accuracy Matrix Accuracy Matrix

LDA – Rock Type Prediction – Soil Geochemistry Soil Geochemistry

LDA – Rock Type – Accuracy Matrix Accuracy Matrix

Conclusions

• Logratios and a multivariate approach yields

g pp y patterns that infer geological processes.

• Results of the survey show transitional

distinctions between the upper (A PH) and distinctions between the upper (A, PH) and lower (C) horizons.

• C horizon correlates well with underlying

l geology.

• These results provide insight into the process of

soil formation and the corresponding so

• at o

a d t e co espo d g geochemical response over a substantial area of eastern Canada and the north-eastern United States States.

Organic Carbon - A Horizon

Sampling Protocol Problem High C in A horizon soils in New Brunswick

For more information

• Contact Eric Grunsky

– Eric.Grunsky@NRCan-RNCan.gc.ca

• Read:

– Drew, L.J., Grunsky, E.C., Sutphin, D.M., and Woodruff, L.G., 2010, Multivariate analysis of the geochemistry and mineralogy y g y gy

• f soils along two continental-scale transects in North America,

Science of the Total Environment, v. 409, p. 218-227.