Magnetization of Sediments and the Environment - Paleomagnetism of Sedimentary Rocks: Process and Interpretation - page 105

lake was initially targeted for environmental magnetic
study because its sediment is rhythmically layered.
210 Pb dating indicates that these layers are annual
varves and they allow high-resolution dating of the
sediments, important for correlations with historic
environmental records. The well-preserved lamina-
tions indicate that the lake sediments are not disturbed
by organisms and therefore that the water near the
lake bottom is anoxic throughout the year.
Kodama et al . (1997a, 1998) initially made the
observation that variations in magnetic mineral con-
centration, based on SIRM intensity variations, corre-
late with historic variations in rainfall recorded nearby
at Montrose, Pennsylvania over the past 70 years
(Fig. 8.6). Careful measurement of the organic-rich
dark layers of the varve couplets indicates that they
have higher magnetic mineral concentrations than
the light-colored silt layers in the varves (Kim et al.
2005). Furthermore, the silt layers contain more high-
coercivity antiferromagnetic minerals from the water-
shed soils. While the correlation between magnetic
mineral concentration and historic rainfall could indi-
cate that the magnetic minerals in Lake Ely's sediments
were derived from erosion of the catchment, environ-
mental magnetic parameters suggest instead that the
magnetic minerals were not formed in the watershed
but must have been formed in the lake. In fact, the lake
sediment magnetic minerals turned out to be domi-
nated by magnetosomes created by magnetotactic bac-
teria living in the water column of the lake.
The dominance of magnetosomes in the lake's
environmental magnetic record was shown by a com-
bination of water column measurements. These meas-
urements show that the OAI is situated at a depth of
15 m with a marked decrease in dissolved oxygen at
that level. Furthermore, dissolved sulfi de and Fe(II)
increase below the sharp decrease in dissolved oxygen.
At the OAI, the ARM of material fi ltered from the water
suddenly increases, indicating the presence of stable,
probably single-domain, magnetic mineral particles
(Fig. 8.7). This of course suggests that magnetotactic
bacteria are living at the OAI, confi rmed from the TEM
observation of magnetosome chains in the material
fi ltered from the water column at this depth and lower
in the water column (Fig. 8.7).
Bazylinski (2007, personal communication) has
identifi ed familiar species of magnetotactic bacteria in
the water column as well as a new species of the genus
Magnetospirillum . A sediment trap was installed in the
deepest part of the lake, below the OAI, to see how and

if the magnetosome chains produced by the magneto-
tactic bacteria in the water column were being trans-
ferred to the lake's sediments. TEM observations and
magnetic mineral parameters measured from sediment
collected in the trap, as well as from lake sediment
sampled by a gravity core, show the presence of mag-
netosomes. This indicates that the dominant magnetic
minerals of the lake sediment are indeed formed by
magnetotactic bacteria in the lake. Furthermore, the
initial observation of a correlation between local rain-
fall variations and magnetic mineral concentration
would strongly suggest that the magnetotactic bacteria
are responding to and recording rainfall variations
in the area. If this mechanism is borne out it could
provide a powerful paleoclimate proxy using environ-
mental magnetic measurements of lake sediments,
provided, of course, that the magnetic mineralogy of
the lake sediment is dominated by magnetosomes.
With the confi rmation that magnetotactic bacteria
contribute the bulk of the magnetic minerals in the
lake sediments, several techniques designed to detect
magnetosome chains could be tested. Since the most
defi nitive way of detecting magnetosomes involves
time-consuming and somewhat diffi cult extraction of
magnetic minerals, and then preparation and exami-
nation of the extract under a TEM, quicker and non-
destructive magnetic tests would be highly preferable.
Oldfi eld, a pioneer in the environmental magnetism
fi eld, has proposed a bivariate plot as a test for magne-
tosomes, χ ARM / χ fd versus χ ARM / χ (Oldfi eld 1994 ) where
χ fd is the frequency dependence of susceptibility which,
as mentioned before, can be used to detect superpara-
magnetic grains. An alternating magnetic fi eld is used
to measure susceptibility in most modern instruments
so the samples don't pick up a viscous magnetization
from the application of a DC magnetic fi eld. The Lake
Ely data was used in a modifi cation of Oldfi eld ' s initially
proposed bivariate plot with χ ARM /SIRM plotted versus
χ ARM / χ . The modifi cation, according to Oldfi eld, is
allowed for magnetic grains smaller than 0.07 μ m, as
would be expected for the SD grains that make up mag-
netosome chains. The modifi ed Oldfi eld plot was used
by Snowball et al . (2002) with magnetic data from
Swedish lake sediments inferred to carry magneto-
somes. The Lake Ely data fall perfectly in the same fi eld
as Snowball et al . ' s data (Kim et al. 2005 ).
Moskowitz et al . (1993) proposed a diagnostic test for
magnetosomes that is based on the ratio of a sample's
magnetization when it is cooled to very low tempera-
tures (20 K) in either a strong magnetic fi eld (2.5 T) or