Parameter
Proxy
Basic assumption(sl Empirical relationship
Selected reference(s)'
temperature
3a
[Mg/Ca]
Mg incorporation into T _ a
* I O(b~ lLVl~' aJ)
Chave (1954a, 1954b)
biogenic calcite is
controlled by a TMicroprobe:
dependent
Nurnberg el al. (1996a;
physiological process
I 996b)
GC-MS:
Mashiotta et al. (subm)
3b
[Sr/Ca]
see 3a
T
= a
* [Sr/Ca]
Beck et al. (1992);
(Smith et al. 1979)
Guilderson et al. (1994)
4
U"7 (alkenone
unknown temperature
T - (U' 37
+ a)1b
Brassell et al.
(1986);
unsat. ratio)
dependent bioPrahl and Wakeham
chemical process
where U K
'37
=
[C37:2]/
(1987);
[C37:2
+ C37:3]
Muller et al. (1998)
5
artificial neural
depending
on input
fuzzy logic
Malmgren und Nordlund
networks
in learning process
(1997)
salinity
6
transfer function
see 1
statistical treatment of
Pflaumann (1985)
faunal or floral
assemblage composition
7
artificial neural
see 5
see 5
Wolff et al. (this
networks
volume)
- -
Precision (lab.
Accurac/
reproducib.)
± 30 ppm Mg
± 1-1.6 °C
(=>±
1.2°C, depending on the
calibration curve)
±? °C
±
1°C
=>±
0.1 °C
[Sr/Ca] in
(Guilderson et al.
sfe. water can
1994)
vary by up to
2 %, which
translates to
2_3°C
(Pisias et al.
1995)
±0.017U'37
± 1-1.5 °C
(=>±
0.5 0c)
depending on
(Muller et al. 1998) calibration
curve (global
or regional)
not relevant
± 0.7-0.8°C
(Malmgren
and
Nordlund
1997)
not relevant
?
not relevant
0.5 - 0.8 %,
Wolffetal.
(this volume)
Pitfalls
I) vital effects (mode of
calcification)
2) dissolution
3) residence time of Mg is 13 my
(Broecker and Peng 1982)
additionally for Sr:
I) growth rate artifact (higher
[Sr/Ca] are associated with slower
growth rates (Pisias et al. 1995)
2) residence time of Sr is 2.5 my
(Palmer and Edmond 1989); of Ca
is 0.6-1.1 my (Broecker and Peng
1982); [Sr/Ca] constant, at least, for
timescales less than 0.6 my.
1 ) mechanism unknown
2) uncertainty about species
composition
3) seasonality and depth-habitat of
species
4) selective degradation ofC37:2 ,
C37:3,
(and C37:4) does not seem to
be a problem (Teece et al. 1995,
1998)
same problems as with transfer
functions
salinity correlates with temperature;
see 1
see 5
- - - - -
n
CD;
S
o
(')
'"
§
::r:
!'
'" to
::l.
'" ,...,
o
-<
~
~.
o
,...,
~
~.
V>
--l