Framework Grains
 Point-counting techniques and critiques

The techniques and interpretation of the proportions of the framework grains, have been relatively controversial in the mid-1980s [E.g. Discussion and Reply of Ingersoll et al. (1984)]

The manner in which you count the grains has a significant effect on the interpretation of the point counts - competing "schools" [Indiana vs. Arizona] of thought developed with significant numbers of arguments.

The influence of climatic conditions on framework sandstone composition is addressed in of Suttner and Dutta (1986): Alluvial sandstone composition and paleoclimate, I. Framework Mineralogy, Jour. Sed. Pet., 56, 329-345.

 

General approaches to point counting are primarily related to contrast in emphasis

  • P. Gazzi (Italian) and Bill Dickinson independently came up with a point counting scheme that is meant to minimize the grain size variability - Gazzi-Dickinson approach

  • the "traditional" method that separates grain sizes (considered more in line with climatic influences

TEST OF TWO APPROACHES USING HOLOCENE SANDS

Grain parameters (a bit different than Dickinson and Suczek):

Q

 stable quartzose grains   Framework = Q + F + L + M + D + Misc.

Qm

 monocystalline quartz   QFL%Q = 100Q/(Q+F+L)

Qp

 polycrystalline quartz (including chert)   QFL%F= 100F/(Q+F+L)

F

 monocrystalline feldspar   QFL%L = 100L/(Q+F+L)

P

 plagioclase   Framework%M = 100M/Framework

K

 K-feldspar   Framework%D = 100D/Framework

L

 polycrystalline lithic fragments   LmLvLs%Lm = 100Lm(L-Lp)

Lv

 volcanic-hypabyssal fragments   LmLvLs%Lv = 100Lv(L-Lp)

Ls

 sedimentary lithic fragments   LmLvLs%Ls = 100Ls(L-Lp)

Lm

 metamorphic lithic fragments   Lvm = Lv + xLm [x = proportion of metavolcanics]

Lp

 plutonic lithic fragments (traditional method)   Lsm = Lv + (1-xLm) [x = proportion of metavolcanics]
M phyllosilicates (biotite, muscovite, chlorite)    
D dense (heavy) minerals    
Misc miscellaneous or unidentified  

matrix and cement are not counted

Data collected from 6 unconsolidated Holocene sands of mixed provenance from arroyos in New Mexico

  • prior to sieving, each sample was split with all fragment size fractions

  • other fraction separated into size fractions 4-3Ø [very fine sand=0.0625-0.125 mm], 3-2Ø [fine sand=0.125-0.25mm], 2-1Ø [medium sand=0.25-0.50 mm], 1-0Ø [coarse sand=0.50-1.0 mm], 0-(-1)Ø [very coarse sand=1-2 mm]

  • sand fractions poured in epoxy mold and sectioned

  • stained for plagioclase and K-feldspar

  • each operator counted 300 points [L. van der Plas and A. C. Tobi (1965). A chart for judging the reliability of point counting results. American Journal of Science, 263, 87-90.]

Distinctions between the "traditional" and Gazzi-Dickinson approaches

  • Point counting stage used with a grid spacing for both.

  • Gazzi-Dickinson approach assumes that any sand-sized (>0.0625 mm) mineral, whether as a separate monomineralic grain or as part of lithic grain is counted as a crystal. If the lithic fragments have grains smaller than the sand-sized they are counted as lithic grains. Advantages:

    • modal composition does not change with physical disaggregation

    • counting poorly-sorted sands or sandstones much easier and quicker

  • The "traditional" approach would count the lithic grain with sand-sized crystals as lithic grains.

  • Generally results in higher % of Q, F, M and D values, but lower L

  • There are generally similar results between the two approaches with fine-grained sediments

  • Lithic fragments counted as such if there is a single "other" mineral vs. the Indiana 10% cutoff.

Results of Gazzi-Dickinson vs. Traditional

Point counts of unsieved fractions:

 

 

 

size fractions for GD more clustered in 78, 85, 103 and 110

 

Better clustering, and probably a better source rock indicator.

 

 

 

 

 

 

GD generally shows much less variation with grain size relative to T

 

 

 

 

 

 

 

 

 

Major Critiques of GD method (sed wars)

Suttner and Basu point out that:

  • sacrifice petrographic information at the expense of rapid data collection - unwarranted loss of information

  • similar consistent results can be attained by use of specific grain size populations

  • ultimate goal of sandstone petrology is paleogeology, not general tectonic setting

  • proper counting of inclusions

  • classification of chert fragments as anything but a sedimentary rock seems incorrect and should not be lumped with Qm

  • improper definition of rock fragment, as they define them

Ingersoll et al. response:

  • increased awareness of problems

  • use of GD should be based on goals, and not used for paleoclimatology

  • use chert term a bit loosely

 

Decker and Helmold point out that:

  • does not affect grain size bias

  • loss of genetic information in counting lithic fragment as crystal

  • QFL plots give misleading view of sandstone composition and maturity - under-represents lithic fragments

  • Recommendation: examine medium to coarse sands and record provenance-diagnostic features for all lithic clasts in the sand range.

Ingersoll et al. response:

  • grain size information is transformed

  • they want to incorporate all provenance information

  • dislikes the tone and the discussion of results