Geology Reference
In-Depth Information
record of originally abundant organisms may be ex-
pected in environments affected by strong water en-
ergy (e.g. reef environments), lack of sedimentation
(e.g. in deeper hardground areas) and in areas with
strong bioturbation (e.g. deeper-water mud). The fre-
quency of echinoderm fragments, therefore, may not
reflect true abundance, but can be a measure of bio-
diversity (Nebelsick 1996).
• Consider diagenetic alterations (e.g. partial or com-
plete dissolution) of skeletal and other grains that can
lead to misinterpretations of the primary composition
of the sample.
• Note that small silt- and clay-sized terrigenous grains
(e.g. quartz) or small microfossils might not be visible
in thin sections but may be abundant in acid residues.
• If appropriate use adjective terms in describing the
relative frequency of grain categories: Very rare (< 2%
grains), rare (2-5%), sparse (5-10%), common (10-
30%), very common (30-50%), abundant (> 50%). For
semiquantitave methods and estimations use: absent,
rare, present, abundant.
• Note the percentage of thin sections in which the
constituent is present.
Weidlich. O., Bernecker, M., Flügel, E. (1993): Combined
quantitative analysis and microfacies studies of ancient
reefs: an integrated approach to Upper Permian and Up-
per Triassic reef carbonates (Sultanate of Oman). - Fa-
cies,
28
, 115-144
Further reading
: K108
6.3 Multivariate Microfacies Studies
Modal analyses and other frequency measurements sup-
ply the raw data for multivariate microfacies analyses.
These analyses are not intended to replace the evalua-
tion of data derived from qualitative data, but rather
aim at simplification and efficient ordering of datasets.
Statistical methods shed a new, more objective light
on problems related to classifying samples and inter-
preting compositional data, or may point out aspects
which would not come out in a conventional approach
to microfacies studies. Multivariate methods may be
divided two broad groups: Classification refers to the
recognition of discrete groups within a data set and
therefore assumes discontinuity. The outcome of clus-
ter analysis is a two-dimensional diagram, in which no
relationships between the clusters and underlying en-
vironmental gradients can be discerned directly. Ordi-
nation aims to order or arrange samples or variables in
a space in relation to environmental parameters or gra-
dients. Ordination extracts major directions of varia-
tions from the original data, which are assumed to be
in continuity. The outcome of ordination are two- or
three-dimensional scatter plots which need intelligent
and sound interpretations.
Basics: Methods of frequency analysis
Baccelle, L., Bosellini. A. (1965): Diagrammi per la stima
visiva della composizione percentuale nelle rocce sedi-
mentarie. - Annali della Università di Ferrara, Sezione
IX, Science Geologiche e Paleontologiche,
1
, 59-62
Bernecker, M., Weidlich, O., Flügel, E. (1999): Response of
Triassic reef coral communities to sea-level fluctuations,
storms and sedimentation. Evidence from a spectacular
outcrop (Adnet, Austria). - Facies,
40
, 229-280
Chayes, F. (1956): Petrographic modal analysis. - 113 pp.,
New York (Wiley)
Jaanusson, V. (1972): Constituent analysis of an Ordovician
limestone from Sweden. - Lethaia,
5
, 217-237
Matthew, A.E., Woods, A.J., Oliver, C. (1991): Spots before
the eyes: New comparison charts for visual percentage
estimation in archaeological material. - In: Middleton A.,
Freestone, J. (eds.): Recent development in ceramic pe-
trology. - British Museum Occasional Paper,
81
, 211-264
Middleton, A.P., Freestone, I.C., Leese, M.N. (1985): Tex-
tural analysis of ceramic thin sections: evaluation of grain
sampling procedures. - Archaeometry,
27
, 64-74
Patterson, R.T., Fishbein, E. (1989): Re-examination of the
statistical methods used to determine the number of point
counts needed for micropaleontological quantitative re-
search. - J. Paleont.,
63
, 245-248
Russ, J.C. (1999): The image processing handbook. 3rd edi-
tion. - 771 pp., Boca Raton, Florida (CRC Press)
Schäfer, K. (1969): Vergleichs-Schaubilder zur Bestimmung
des Allochemgehalts bioklastischer Karbonatgesteine. -
Neues Jahrbuch für Geologie und Paläontologie, Monats-
hefte,
1969
, 173-184
Soille, P. (ed., 2000): Morphological image analysis. -
314 pp., Berlin (Springer)
Van der Plas, L., Tobi, A.C. (1965): A chart for judging the
reliability of point counting results. - American Journal
of Science,
263
, 87-90
6.3.1 Methods: Variations between
Constituents and between Samples
Multivariate analyses start by creating a numerical
element data matrix, based on frequency measurements.
This matrix serves as basis for calculating correlation
coefficients that describe similarity between variables
and between samples. Various correlation coefficients
(e.g. the product moment correlation) are used to es-
tablish a similarity matrix. This data matrix is used for
cluster analyses in answering the questions: how are
variables (grains) related statistically and how are
samples related.
Cluster and factor analyses applied to microfacies
Variable-per-variable relations are shown by R-mode
cluster analysis, sample-per-sample relations by Q-
mode cluster analysis derived from point-counting data.
Variables are often standardized in order to negate the
strong effects of magnitude. The degree of similarity
