Boehrer - Stratification of Lakes, 2008

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/248781047
Stratiﬁcation of lakes
Article in Reviews of Geophysics · May 2008

DOI: 10.1029/2006RG000210
CITATIONS READS
346 4,753
2 authors:
Bertram Boehrer Martin Schultze

Helmholtz-Zentrum für Umweltforschung Helmholtz-Zentrum für Umweltforschung
108 PUBLICATIONS 1,507 CITATIONS 91 PUBLICATIONS 1,577 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Ecology of Meromictic Lakes View project
Building resilience of Lake Bosumtwi to climate change - RELAB Project (Funded by Ministry of Foreign Affairs of Denmark) View project
All content following this page was uploaded by Bertram Boehrer on 10 August 2015.
The user has requested enhancement of the downloaded file.

Click
Here
for
Full
Article
STRATIFICATION OF LAKES
Bertram Boehrer1 and Martin Schultze1
Received 10 July 2006; revised 26 September 2007; accepted 4 January 2008; published 30 May 2008.
[1] Many lakes show vertical stratification of their water do not experience a complete overturn because of pressure
masses, at least for some extended time periods. Density dependence of temperature of maximum density. Such lakes
differences in water bodies facilitate an evolution of must be sufficiently deep and lie in the appropriate climate
chemical differences with many consequences for living zone. Although these lakes are permanently stratified, deep
organisms in lakes. Temperature and dissolved substances waters are well ventilated, and chemical differences are
contribute to density differences in water. The atmosphere small. Turbulent mixing and convective deep water renewal
imposes a temperature signal on the lake surface. As a must be very effective. As a consequence, these lakes
result, thermal stratification can be established during the usually are not termed meromictic. Permanent stratification
warm season if a lake is sufficiently deep. On the contrary, may also be created by episodic partial recharging of the
during the cold period, surface cooling forces vertical deep water layer. This mechanism resembles the cycling of
circulation of water masses and removal of gradients of the ocean: horizontal gradients result from gradients at the
water properties. However, gradients of dissolved substances surface, such as differential cooling or enhanced evaporation
may be sustained for periods much longer than one annual in adjacent shallow side bays. Dense water parcels can be
cycle. Such lakes do not experience full overturns. Gradients formed which intrude the deep water layer. In the final
may be a consequence of external inflows or groundwater section, stratification relevant physical properties, such as
seepage. In addition, photosynthesis at the lake surface and sound speed, hydrostatic pressure, electrical conductivity,
subsequent decomposition of organic material in the deeper and density, are discussed. The assumptions behind salinity,
layers of a lake can sustain a gradient of dissolved electrical conductance, potential density, and potential
substances. Three more geochemical cycles, namely, temperature are introduced. Finally, empirical and
calcite precipitation, iron cycle, and manganese cycle, are theoretical approaches for quantitative evaluation from easy
known for sustaining meromixis. A limited number of lakes to measure properties conclude this contribution.
Citation: Boehrer, B., and M. Schultze (2008), Stratification of lakes, Rev. Geophys., 46, RG2005, doi:10.1029/2006RG000210.
1. INTRODUCTION human impact, e.g., Dead Sea [Gat, 1995] and Mono Lake
[Jellison and Melack, 1993] or by climatic variability, e.g.,
[2] Humanity needs to be prepared for the changes it Caspian Sea [Peeters et al., 2000].
imposes upon the Earth, especially in times of global [3] Understanding environmental processes is the first step
change and of direct human impact on the hydrological to forecasting such evolutions. Furthermore, a good under-
cycle. The anthropogenic impact of the last decades on our standing of stratification is required not only to prognosticate
aquatic environment has shown that a responsible use of our the future of our lakes but also to better interpret the past from
natural resources is mandatory to guarantee sustainable lacustrine sediment cores [e.g., Kjensmo, 1988; Brauer,
conditions. Over the last decades, not only have new large 2004]. Those interpretation’s intent is not only to document
water bodies, e.g., reservoirs, been created on the Earth’s the lake history but also to draw conclusions about the climate
surface, but also entirely novel aquatic systems developed in in the past by relating variations of the lake sediments to
the aftermath of mining in abandoned opencasts [e.g., Miller circulation patterns in lakes and finally climatic conditions
et al., 1996; Krüger et al., 2002]. In addition, the hydro- triggering these changes.
logical regime of many lakes has been modified to the [4] A considerable portion of the lakes on Earth are
extent that lakes have fundamentally changed their appear- permanently stratified. Deep lakes especially show this
ance, e.g., Aral Sea [Létolle and Mainguet, 1993; Crétaux et feature. As a consequence, many of the largest lakes (e.g.,
al., 2005], or their stratification pattern has been altered by Caspian Sea, Baikal, Tanganyika, Malawi-Nyasa) and many
middle size and small lakes do not circulate completely in
the vertical and do not show a homogenized, overturning
1
Helmholtz Centre for Environmental Research – UFZ, Magdeburg, water body at any time during the annual cycle. The
Germany.
permanent stratification has decisive impact on the redistri-
Copyright 2008 by the American Geophysical Union. Reviews of Geophysics, 46, RG2005 / 2008
1 of 27
8755-1209/08/2006RG000210$15.00 Paper number 2006RG000210
RG2005
RG2005 Boehrer and Schultze: STRATIFICATION OF LAKES RG2005
bution of dissolved substances, such as nutrients or oxygen, physical limnologists how biogeochemical processes deter-
and hence determines the biocenosis that can form in the mine the physical properties of stratification. In order not to
lake [e.g., Gaedke et al., 1998; George and Hewitt, 2006; interfere with the progress of reading, we tried to remove all
Thackeray et al., 2006]. quantifications of processes from sections 1 – 4. Such nu-
[5] Observations of stratified lakes probably date back to merical approximations are included in section 5, which is
long before the start of scientific literature on the environment. dedicated to the evaluation of stratification relevant physical
The Romans likely already knew about stratification. More quantities.
recently, solar ponds, i.e., saltwater ponds capped with a less [8] Referring to the classification of ectogenic, creno-
saline surface layer, have been utilized for breeding oysters at genic, and biogenic meromixis by Hutchinson [1957], we
least since the 19th century in Norway [see Klaveness, insert more recent observations into these classes and place
1990]. Findenegg [1933] brought chemical differences in lakes that stay permanently stratified because of thermobaric
connection with incomplete mixing in winter. Since then, effect with the meromictic lakes. In addition, we include
considerable work has been published on ecological con- lakes with episodic partial deep water renewal, where a
sequences of meromixis. Also, the reasons for permanent process can be identified that recharges and replaces part of
stratification have been investigated, and the understanding the deep water on a regular basis. Section 2.1, introducing
of involved processes has progressed. Textbooks on hydro- the annual temperature stratification cycle, briefly repeats
biology and limnology present basic knowledge about the formation of an epilimnion and the deep recirculation,
stratification and circulation of lakes [e.g., Forel, 1901; while clarifying the naming convention of mixis types and
Ruttner, 1940; Cole, 1983; Lampert and Sommer, 1993; layers in a lake. Section 5 introduces and explains physical
Schwoerbel, 1999; Wetzel, 2001; Schönborn, 2003]. Some of quantities which are relevant for the stratification in lakes
these books refer to permanent stratification and meromixis and lists the most common quantitative expressions for their
and mention some ecological consequences. However, the approximation for given conditions in lakes.
space dedicated to the physical and chemical preconditions
for permanent stratification in any of these texts is limited. 2. STRATIFICATION FEATURES OF HEAT AND
Hutchinson [1957] has attempted to summarize the knowl- DISSOLVED SUBSTANCES
edge on meromixis of his time, and his work is still a valid
reference on meromixis at the beginning of the 21st century. 2.1. Annual Temperature Cycle in Lakes
In addition, there are some studies on special aspects of [9] The lake surface shows a pronounced temperature
limnology [Tilzer and Serruya, 1990; Margalef, 1994; Ler- cycle over the year in most climate zones on Earth (Figure 1).
man et al., 1995; Pourriot and Meybeck, 1995; Johnson and Besides the thermal contact with the atmosphere, the com-
Odada, 1996; Taub, 1984] which touch upon interesting mon exposition of lakes and atmosphere to seasons, namely,
aspects of physical limnology, especially Lerman et al. the variation of incoming solar radiation, is the main reason
[1995], who deal with chemical peculiarities which may be for the parallel evolution. The depicted temperature cycle is
linked to permanent stratification. On the contrary, the studies typical for lakes of the temperate climate zone. Only lakes in
on physical limnology, i.e., stratification and mixing in lakes close proximity of the equator (or permanently ice-covered
[e.g., Fischer et al., 1979; Imboden and Wüest, 1995; lakes) may show little variation in the surface temperatures. A
Imberger and Patterson, 1990], and modeling currents in closer look reveals several processes contributing to the heat
lakes [e.g., Hutter, 1987] have concentrated on thermally transfer through the lake surface: solar radiation, long-wave
stratified lakes and have widely neglected the topic of radiation of atmosphere and surface waters, sensible heat
meromixis. For considerations about internal waves and exchange, and heat flux connected with evaporation and
currents in lakes, the oceanographic literature [e.g., Pedlosky, precipitation [see, e.g., Imboden and Wüest, 1995]. In addi-
1988; Gill, 1982; Pickard and Emery, 1982; Pond and tion, contributions of inflow and outflow of surface and
Pickard, 1983] has delivered the necessary theoretical groundwater, as well as thermal contact with the lakebed,
knowledge for physical limnology but to date has not must be considered in some cases.
contributed a comprehensive work with focus on permanent [10] Weather conditions set up stratification and currents
stratification in adjacent seas, such as the Black Sea and the in lakes. They directly and indirectly control the biological
Baltic Sea. processes in the lake. Hence for the detailed investigation of
[6] Other than in the open ocean, chemical transforma- physical processes in lakes, observation of the meteorolog-
tions can play an important role in stratifying a lake. Lakes ical conditions above the water is mandatory (Figure 2).
greatly vary in salinity and chemical composition of dis- Waves and currents are driven by wind. For heat transfer
solved substances. As a consequence, freshwater assump- measurements, air temperature and net radiation sensors are
tions and ocean water approximations may be acceptable for also required, which refer to the part of incoming solar
some considerations, while for many lakes, or more detailed radiation that is not reflected (albedo) but deposits its
investigations, they fail. Such cases require separate energy as heat in the water. In addition, air humidity
approaches to satisfactorily reflect given conditions. measurements are required to estimate the evaporation
[7] This review on stratification of lakes is also intended from the water surface, and finally, many meteorological
to be useful for limnologists of educational backgrounds stations also include air pressure as a central meteorological
different from physics. Furthermore, it shall also show magnitude.
2 of 27
[14] Epilimnion and atmosphere exchange heat and vol-

atile substances such as gases. In addition, the epilimnion is
recirculated episodically by wind events or periods of lower
temperatures during the stratification period. On the con-
trary, the hypolimnion is insulated from exchange with the
atmosphere during the stratification period. Transport of
dissolved matter across the vertical density gradient of the
thermocline is usually small. In most lakes, only chemical
precipitation and settling of particles are able to carry larger
amounts of matter across the thermocline.
[15] In nearly all lakes, wind is the decisive factor for the
thickness of the epilimnion. There are exceptions, like lakes
with extremely weak winds, where other factors can evolve
to be the most important. One example is a very clear lake,
where light penetrates beyond the mixing depth of wind
Figure 1. Temperatures at several depths in Lake Goitsche
[e.g., Simpson and Dickey, 1981a, 1981b; Kling, 1988;
near Bitterfeld, about 100 km southwest of Berlin in
Germany during 2005 (hourly mean). Depths of tempera- Mazumder and Taylor, 1994; Fee et al., 1996; Pérez-
ture tracks are identified from higher temperatures to lower Fuentetaja et al., 1999]. In addition, forming an ice cover
temperatures for increasing depth. Lake Goitsche formed in may be regarded as a special case where molecular diffusion
a decommissioned lignite mine pit over the years 1999 to of heat is responsible for the thickness of the surface ice
2002. It has a surface area of 13.5 km2 and a maximum layer.
depth of about 50 m. [16] The epilimnion thickness is not constant over the
stratification period (Figure 4). In spring, a thin layer is
[11] A look at temperature time series from the surface formed, which throughout the summer gradually gains
waters of lakes within the same climate (Figure 3) reveals thickness because of wind action. It is not until autumn
that lake surface temperatures reflect a climate signal. The that colder water from the lake surface can erode the
daily average temperature is very similar between the stratification to the hypolimnion at a larger rate. During
displayed lakes. Only slight differences can be seen in this later period of thermal stratification, substances dis-
spring, when the smaller lakes tend to gain heat faster, solved in hypolimnetic waters, such as nutrients, become
and in hot periods during summer, when the highest peaks available in the epilimnion again. If surface temperature
of smaller lakes are 1! or 2! higher [see Boehrer et al., falls enough, epilimnion and hypolimnion can be mixed,
2000b]. The situation is different for temperature tracks and the entire lake is homogenized (circulation period) to
acquired at a water depth of 7 m in the same lakes. one layer, the so-called mixolimnion. As an example, the
Obviously, the mixing characteristics of each lake determine acidic residual lakes in Goitsche Mine Pit, which were
how effectively heat is being transferred vertically. neutralized by introduction of river water, are presented
[12] While the surface water is exposed to solar radiation, (see Figure 5). Chemical gradients (such as pH) could only
heat loss by long-wave radiation, and thermal contact with be sustained for extended time periods if they were stabi-
the atmosphere, the deeper layers are shielded from the lized by density stratification (Figure 5, bottom).
major sources of heat. Diffusive heat transport on a molec- [17] Epilimnion thickness is important for many organ-
ular level is very slow and requires a time period of the isms. Hence regressions relating epilimnion thickness with a
order of a month for the transport of heat over a vertical priori known or easy to measure quantities have been
distance of 1 m. A much more efficient heat transport can be proposed. Wind speed is considered through lake surface
facilitated by the kinetic energy input through wind stress at area A and a fitted exponent. Jöhnk [1999, 2000] collected
the lake surface. the results of several authors (Figure 6). All data sets rely on
[13] If surface temperatures exceed the 4!C threshold, measurements from moderate latitudes of the Northern
warmer and hence less dense water can only be mixed to a Hemisphere, i.e., central Europe, North America, and Japan.
limited depth, which corresponds to the limited energy The most central regression originates from Patalas [1984]
budget supplied by the wind. A density difference has been zepi = 4.6A0.205 which is close to previously used formulas
established, and the thermal stratification period of summer from Ventz [1972] and Fachbereichsstandard [1983] (as
starts. In sufficiently deep lakes, thermal stratification holds cited by Klapper [1992]). The differences between fitted
until cooler autumn and winter temperatures force a circu- curves (factor 1.5) give a good impression of the accuracy at
lation into deep waters. The warm surface water layer is which epilimnion thickness can be parameterized with
called epilimnion, while the colder water layer beneath, surface area only. Also, Gorham and Boyce [1989] confirm
which has not been mixed into the epilimnion over the surface area and fetch as the most important factors for the
stratification period, is called hypolimnion. A sharp tem- epilimnion thickness, as does Davies-Colley [1988] for
perature gradient (thermocline) forms in the contact zone lakes in New Zealand.
between both layers (Figure 4). [18] Because of its high gradients, the thermocline can
form a special habitat. Organisms that can regulate their
3 of 27
Figure 2. Meteorological data (hourly mean) of air pressure, net radiation, global radiation, wind speed,
wind direction (0! = 360! = north, 90! = east), relative humidity, and air temperature acquired at 2.6 m
above the lake surface of Lake Goitsche during 2005.
density can position themselves in the strong density advantages of both layers), epilimnion, and hypolimnion.
gradient. Also, inanimate particles can accumulate on a As a consequence, a separate layer may be identified
level representing their density. In addition, motile organ- because of its own characteristic. Such a layer is called
isms can dwell in the thermocline (to profit from the the metalimnion. Especially in nutrient-rich lakes, the
4 of 27
Figure 3. Daily average temperatures in several lakes of the northern German plain (MA2 Wallendorfer
See, MB1 und MB3 Rassnitzer See, R12 Mining Lake 111, VO1 Vollert-Süd, GOJ subbasin Niemgk of
Lake Goitsche) acquired at 0.5, 3, and 7 m below the surface during 1998 (modified from Boehrer et al.
[2000b], with permission from Elsevier).
decomposition of organic material can cause a depletion of

oxygen, resulting in a so-called metalimnetic oxygen min-
imum (Figure 7). On the contrary, if light can penetrate to
the thermocline and photosythesis can overcome the oxygen
uptake in this zone, a metalimnetic oxygen maximum can
be encountered. Occasionally, this phenomenon can be
observed in mesotrophic lakes.
[19] Traditionally, the imagination of processes and nam-
ing convention are dominated by experiences from lakes in
moderate climates. In the vicinity of the equator, some
features are different, e.g., the annual temperature variation
is relatively small. As a consequence, the temperature
difference between epilimnion and hypolimnion is also
small. Under favorable conditions, night temperatures may
Figure 4. Temperature profiles of Lake Mondsee, Austria,
suffice to form cold waters in shallow areas dense enough to on several dates during the stratification period of the year
intrude into the hypolimnion (similar to deep water renewal, 1999 (data in part from Dokulil and Teubner [2003]).
5 of 27
Figure 5. Contour plot of temperature, electrical conductance (k25), and pH versus time since flooding
started on 7 May 1999 and depth in mining lake Goitsche (station XN3) during flooding with river water
[see also Boehrer et al., 2003] (with permission from ecomed). The increasing shaded area reflects rising
water surface.
see section 4) or even trigger a complete overturn (‘‘poly- [23] 3. Episodic partial deep water renewal lakes can
mictic’’ [e.g., Zauke et al., 1995]). Hence some features of have the appearance of meromictic lakes despite the pres-
the stratification cycle may look different in a tropical or ence of recycling mechanisms. In the end, it turns out that
polar climate. We refer to Lewis [1987] for more detail on these recycling mechanisms keep the lakes stratified.
the special features of tropical lakes; for stratification [24] 4. Amictic lakes do not experience a deep recircula-
features in a polar climate, see, e.g., Gibson [1999]. tion. Usually, permanently ice-covered lakes are included in
[20] The classification of lakes according to their circu- this class. Lakes, however, can circulate underneath an ice
lation pattern has proven very useful within the limnology sheet. Transport of both heat and matter happens because of
community [e.g., Hutchinson, 1957]: diffusion, or double diffusion [e.g., Brandt and Fernando,
[21] 1. Holomictic lakes overturn and homogenize at least 1996]. On top of this, they can also be circulated by external
once a year. forcing, such as solar radiation that penetrates to the lake
[22] 2. Meromictic lakes are lakes in which the deep bed and geothermal heat flux, or by salinity gradients
recirculation does not include the entire water body. Beyond created when ice is forming on a salt lake.
this traditional definition, most limnologists pragmatically [25] Holomictic lakes are subdivided into classes indicat-
call a lake meromictic if a chemically different (e.g., anoxic) ing the frequency and time of overturn. We refer to the most
bottom layer, called monimolimnion, has continuously been common classes, while Lewis [1983] subdivided the classes
present for at least one annual cycle. even further. ‘‘Polymictic’’ refers to lakes which are not
6 of 27
[27] In the absence of any clear boundary, lakes are

referred to as salt lakes when their salt content lies above
3 g in a kilogram of lake water; i.e., 3 g kg!1 = 3%. From
this concentration, humans can clearly taste the salt, and
ecological consequences become obvious [e.g., Williams,
1996, 1999]. The salt content in lakes, however, can be as
high as 300 g kg!1 [e.g., Williams, 1998]. For limnetic
waters, salinity has been defined as the sum of all ion
concentrations [Williams, 1994]. It is measured in g kg!1 or
part per thousand (ppt). Direct salinity measurements hence
require a full chemical analysis. By definition, freshwater
lakes have a salt content below 3 g kg – 1. However,
salinities normally lie below 0.5 g kg!1. Even salinity
gradients in this range can determine the circulation pattern
of lakes.
[28] The composition of dissolved substances in lakes
Figure 6. Graphical representation of several approxima- varies greatly. The surrounding geological formations
tions of epilimnion thickness zepi versus surface area of release substances that are carried into the lake [e.g.,
respective lakes (slightly modified from Jöhnk [2000] with Rodhe, 1949; Jellison et al., 1999]. The contributions of
author’s permission). the chemical species to physical properties, such as elec-
trical conductivity or density, are specific, and as a
deep enough to form a hypolimnion. The entire lake consequence, lake-specific approaches are required for
behaves like an epilimnion, which is mixed by sporadic numerical approximations of electrical conductance and
strong wind events over the year or even on a daily basis in density from conductivity-temperature-depth (CTD) probe
response to strong daily temperature cycle. ‘‘Dimictic’’ data. Highly accurate approximations are only available for
lakes are handled as prototypes of lakes in moderate to a limited number of lakes. One particular composition of
cold climates. A closer look at the lakes, however, reveals salt water that is familiar and accessible to many limnol-
that in most cases an ice cover or a great maximum depth is ogists is brackish water, i.e., water mixed from ocean
required to guarantee a stratification period during the cold water and freshwater. Lakes close to the coast that have
season. Intermittent between ice cover and summer stratifi- received inflows from the open sea may well represent
cation the lake can be circulated in the vertical, the easiest such conditions.
when in freshwater lakes, the temperature traverses 4!C (see [29] As a consequence, effects of salinity on the proper-
Figure 1). ‘‘Monomictic’’ lakes possess one circulation ties of lake water will be introduced for brackish water, and
period in addition to the stratification period. Many lakes
in the temperate climate belong into this class if they do not
develop an ice cover during winter. Sometimes such lakes
are also referred to as warm monomictic to distinguish them
from cold monomictic lakes, which show an ice cover for
most of the year and circulate during the short period
without ice. ‘‘Oligomictic’’ lakes circulate less frequently
than once a year, normally at irregular intervals, triggered
by extreme weather conditions such as unusually cold
winters for the respective location.
2.2. Salt Stratification
[26] A considerable portion of the Earth’s inland water is
salty [Williams, 1996]. Many large salt lakes, e.g., Caspian
Sea, Issyk-Kyl, Aral Sea, Lake Van, Great Salt Lake, and
the Dead Sea, are located in endorheic basins, i.e., areas on
the Earth’s surface without surficial hydraulic connection to
the world ocean [e.g., Meybeck, 1995]. Williams [1996]
presents a world map of the largest of these basins but also
points out that salt lakes occur outside these areas, such as Figure 7. Profiles of temperature (T), (in situ) conductiv-
ity (C), and concentration of dissolved oxygen (O2) from 6
solar ponds [e.g., Kirkland et al., 1983] or basins filled with
September 2000 in Arendsee, Germany (adapted from
seawater that lost the connection to the sea. In addition, the Boehrer and Schultze [2005] with permission from
high salinity of few inland salt lakes derives from under- ecomed). The boundaries between layers were drawn along
ground or local salt sources, often associated with mining gradients in oxygen profiles. (Oxygen concentration
[e.g., Kirkland et al., 1983; Böhrer et al., 1998]. numerically corrected for response time of 7.5 s of sensor.)
7 of 27
the thermal contribution. Some lakes show extreme salinity

stratification between low salinity and water beyond ocean
concentration. Also, in deep waters at a temperature close to
the maximum density, supposedly small salinity gradients
can dominate the stratification.
[31] The salinity contribution to density shows a temper-
ature dependence. The same amount of salt increases the
density of colder water slightly more than of warmer water.
As a consequence, temperature of maximum density
decreases with increasing salinity. At about 24 psu, temper-
ature of maximum density intersects with the freezing point.
Water of higher salinity does not show the anomaly of fresh
water. This affects the hypolimnion temperature in cases
close to the temperature of maximum density. In Solbad
Stassfurt, the salt composition is similar to the ocean.
Figure 8. Density contours versus temperature and During summer, hypolimnion temperatures stay below
salinity of mixtures of seawater and pure water in kg m!3 4!C because of high salinity of the mixolimnion (Figure 9)
at normal atmospheric pressure. The dashed line shows [Hausmann and Boehrer, 2006]. For the same reason, deep
freezing point respective to salinity, and the thinner solid ocean waters are considerably colder than 4!C.
line represents temperature of maximum density at given [32] Increasing salinity also lowers the freezing point of
salinity. water. Seawater freezes at !2!C (Figure 8); lake water of
even higher salinity freezes at correspondingly lower tem-
we selected lakes where brackish conditions are a good peratures. Kerry et al. [1977] (as cited by Hobbie [1996])
approximation. As such conditions can commonly be en- report that the highly saline Deep Lake in Antarctica
countered, appropriate approaches and highly accurate nu- circulates in winter at temperatures of !15!C, while in
merical approximations have been developed and are the southern summer, an epilimnion forms with temper-
widely accepted. Salinity in the ocean environment is atures up to 10!C.
calculated with the UNESCO formula from measurements 2.3. Suspended Matter
of electrical conductivity and temperature (see section 5.3).
[33] Rivers can transport high loads of suspended mate-
The practical salinity unit (psu) follows the idea of above
rial, especially at high flows. Concentrations of several
mentioned salt content in g kg!1 = % and represents a
hundred milligrams per liter have regularly been recorded.
reasonably good value for the ocean water (salinity of about
For settling particles, their mass adds to bulk density of the
35 psu) and brackish water, though concerns about accuracy
surrounding water, as long as the average speed of precip-
with the contained salt have resulted in debates about a new
itation is constant. In some cases, suspended material is the
ocean standard for salinity (see also section 5.3.2). The
crucial contribution to facilitate deep water renewal [e.g.,
composition of salts in lakes can, however, greatly deviate
Hürzeler et al., 1996]. Alterations in the catchment of
from ocean conditions. In such cases, salinity in psu can (if
inflows into Lake Brienz strongly reduced the amount of
at all) only be used with reservation.
suspended matter, and as a consequence, incoming water
[30] Salt contributes to water density [e.g., Pickard and
could not replace the deepest water layers anymore [Finger
Emery, 1982] (Figure 8 and section 5.4). This is important,
et al., 2006]. An extreme example was the Mulde flood of
as density gradients are implied by salinity differences due
2002, which washed out a new river bed from fine sandy
to inflowing rivers or evaporation and precipitation. In
material. Concentration of suspended material was so high
many cases, density gradients due to salinity overcome
Figure 9. Profiles of temperature, salinity, and oxygen concentration in Solbad Stassfurt on 3 May 2005
(see also Hausmann and Boehrer [2006] with permission from Weißensee).
8 of 27
sufficiently deep, gravel pit lakes in the upper Rhine Valley

tend to be meromictic [e.g., Landesanstalt für Umweltschutz
Baden-Württemberg, 2000; Ilmberger and von Rohden,
2001].
[37] Monimolimnia are excluded from gas exchange with
the atmosphere for years. In most cases, anoxia is estab-
lished after sufficient time. Nitrate and sulphate can function
as electron acceptors for microbial oxidation of organic
material. In this zone, substances can be produced that
would not be chemically stable in the mixolimnion and
would soon be oxidized under oxic conditions. The mon-
imolimnion is permanently exposed to the hydrostatic
pressure of the water column above. Hence gases (CO2,
H2S, and others) can accumulate in concentrations far
beyond concentrations in mixolimnia (e.g., Lake Monoun
Figure 10. Profiles of temperature, conductance, density, or Lake Nyos in Cameroon, Africa [Halbwachs et al.,
and dissolved oxygen from Rassnitzer See in former mining 2004]).
area Merseburg-Ost on 7 October 2003 (adapted from [38] In many meromictic lakes, deep recirculation erodes
Boehrer and Schultze [2005] with permission from the monimolimnion, leaving a sharp gradient at the end of
ecomed). (Oxygen concentration numerically corrected for the circulation period. The transition of all water properties
response time of 7.5 s of the sensor). happens within a few decimeters from mixolimnetic values
to monimolimnetic values (see Figure 10). This sharp
(1 " 106 m3 of suspended material in 100 " 106 m3 of flood gradient is called halocline, chemocline, or pycnocline,
water) that it could replace an entire hypolimnion despite depending on whether salinity gradient, chemical gradient,
the seasonal temperature stratification [Boehrer et al., or density gradient is referred to. From observations of
2005]. Frey [1955] even claims that suspended matter in intensive colonization with only few different species, it is
deep waters initiated meromixis (see section 3) in Längsee, known that some plankton species can take advantage of
Austria. such gradients (e.g., Lago Cadagno in the Swiss Alps
[34] Casamitjana and Roget [1993] report fluidized lake [Camacho et al., 2001; Tonolla et al., 2003] and Lake
beds where thermal springs keep sediment suspended. The Bolvod, Gek Gel, and Maral Gel [see Sorokin, 1970]).
resulting bulk density suffices to overcome the adverse [ 39 ] Monimolimnia are thermally locked between
temperature gradient. Two layers clearly separated by tem- groundwater (#10!C in the depicted case of Figure 10)
perature and suspended matter are the consequence. Finally, and the hypolimnion. While groundwater roughly repre-
particles that are precipitated in the contact zone of chem- sents annual average temperature, the hypolimnion con-
ically different layers, e.g., mixolimnion and monimolimn- serves surface temperature of the lake during the cold
ion, contribute their weight to the density structure (see season or lies close to temperature of maximum density.
section 5.4). Particles can be identified through higher As a consequence, monimolimnia show temperatures within
turbidity in profiles (see also section 5.4). these boundaries. Usually, a continuous temperature profile
with depth can be observed from the hypolimnion toward
3. MEROMIXIS: PERMANENT STRATIFICATION groundwater temperature.
[40] Traditionally, meromictic lakes are classified after
[35] Lakes are called meromictic if a chemically different main reasons for the creation of permanent stratification.
bottom layer, called monimolimnion, has continuously been Early studies (see Yoshimura [1937] and Findenegg
present for at least one annual cycle. Higher concentrations [1937], as cited by Hakala [2004]) identified both exter-
of dissolved substances have increased density sufficiently nal forcing and internal processes responsible for sustain-
to resist deep recirculation. Exchange rates with the mixing meriomixis. Finally, Hutchinson [1957] proposed
olimnion are small enough that chemically different con- three classes that found broad acceptance: ectogenic,
ditions are sustained continuously. crenogenic, and biogenic meromixis. Walker and Likens
[36] Famous examples are the lakes of Carinthia, Austria [1975] assigned about 200 meromictic lakes into various
[Findenegg, 1933, 1935], but many mining lakes are also classes. Other authors [e.g., Lemmin, 1995] also claimed
meromictic [e.g., Davis and Ashenberg, 1989; Böhrer et al., a separate class for morphogenic meromixis, which re-
1998; Stevens and Lawrence, 1998; Rücker et al., 1999; ferred to the importance of basin size and shape for
Fisher, 2002; Denimal et al., 2005]. Some small and deep sustaining meromixis. Often, monimolimnia are found in
maar lakes [Scharf and Oehms, 1992; Scharf and Menn, well-defined depressions in a lake bed, which are only
1992], as well as natural lakes in southern Norway or marginally impacted by lake-wide currents of water
Finland [Strøm, 1945; Merilainen, 1970; Hongve, 1997, above. Other lakes have great relative depths, which
2002], are permanently stratified by small concentration Berger [1955] already realized favor meromixis. A recent
differences between mixolimnion and monimolimnion. If
9 of 27
overview of various classification schemes is given by the United States [Ludlam and Duval, 2001] or lakes in
Hakala [2004]. Tasmania [King and Tyler, 1981]). The episodic recharge of
[41] In this contribution, we do not try to promote a new both freshwater and salt water sustains the stratification.
classification scheme for meromictic lakes. It is obvious that Moreover, fjords were separated from the sea by falling sea
several processes can create and sustain meromixis. We level or rising terrain but still contain salt water in the deep
discuss them in sections 3.1 –3.3, well aware of the fact that layers [e.g., Strøm, 1963]. Powell Lake in British Columbia,
a close enough investigation would reveal more processes Canada, is a beautiful example. Sanderson et al. [1986]
involved in forming and sustaining stratification. In conclu- showed that the salt in its deep layer has been confined there
sion, some lakes are difficult to classify. We looked for for at least 10,000 years.
examples that represented particular processes as clearly as [46] Gibson [1999] listed 34 permanently stratified water
possible. bodies in Vestfold Hills, Antarctica. When freezing, brine
[42] Lakes do not necessarily show the same circulation exclusion raised the salinity of the water body below the ice.
pattern every year. A dimictic lake, which does not freeze in Salinity increased continuously and eroded progressively
a particularly warm winter, may experience a monomictic deeper into the stratified water body below. Maximum
year, and a usually monomictic lake may be dimictic in a recirculation depth was reached at a time when the lakes
particularly cold winter. An extreme storm event may were covered with ice. After ice melt, a layer of relatively
destroy a monimolimnion. Depending on conditions, a lake fresh water protected the stratification below and inhibited
may turn meromictic again. In some cases, the circulation deep mixing. This mechanism was also present in saline
patterns of lakes have changed permanently. After dropping lakes in Saskatchewan, Canada. Rawson and Moore [1944]
the water level, the Dead Sea (Israel, Jordan) has been had already reported that in some of these highly saline
holomictic since 1979 (see section 4), Mono Lake (Cali- lakes, sodium sulphate was precipitated at low water tem-
fornia, United States) turned meromictic for several years by peratures and by brine exclusion while freezing. The crys-
reconnecting the freshwater inflow (see section 3.1), and tals settled to the lake bed and redissolved when
Lake Ikeda (Kyushu, Japan) turned meromictic in 1985 after temperatures rose during summer. Hammer [1994] claimed
a series of colder winters (H. Kikukawa, personal commu- that the precipitation of sodium sulphate contributed to the
nication, 2005), to name just a few well-known examples. stability of the meromixis in several saline lakes (e.g.,
Waldsea and Deadmoose in Saskatchewan).
3.1. Ectogenic and Crenogenic Meromixis [47] Crenogenically meromictic lakes owe their perma-
[43] Ectogenically meromictic lakes have become mero- nent stratification to groundwater inflows. The above men-
mictic through inflowing saline surface water. Hutchinson tioned lakes Rassnitzer See and Wallendorfer See, northern
[1957] refers to Hemmelsdorfer See, near Kiel in Germany, Germany (Figure 10) [Böhrer et al., 1998; Heidenreich et
which was flooded by a storm from the Baltic in 1872 and al., 1999], receive fresh water from a shallow aquifer, while
remained meromictic until the 1930s. Other deep lakes have a deep aquifer carries highly saline water from salt deposits
become meromictic by inflowing salt water from deicing in the deeper underground into the lake. Also, in Lago
roads (Schalkenmehrener Maar [see Scharf and Oehms, Cadagno in the Swiss Alps [Del Don et al., 2001], perma-
1992]) or lost meromictic stability by inflow of road salt nent stratification is a consequence of incoming groundwa-
to the epilimnion [Kjensmo, 1997]. ter, while in Kongressvatn, Spitsbergen, Norway, the
[44] The class of ectogenically meromictic lakes also monimolimnion is fed by mineral spring water [Bøyum
includes saline lakes with a freshwater inflow [Walker and and Kjensmo, 1970]. Moncur et al. [2006] present the case
Likens, 1975]. After having been deprived of its tributaries of a small Canadian shield lake in Manitoba (Camp Lake)
for decades, Mono Lake, California, turned meromictic by which became meromictic by inflowing groundwater that
reestablishing the natural hydrologic connection with fresh- had been contaminated with acid rock drainage. Lake
water runoff entering the lake [Jellison et al., 1998]. The Monoun, Lake Nyos, and Lake Kivu, central Africa [Lorke
concentration difference sufficed to resist deep recirculation et al., 2004], can be assigned to this class, as they are
for 6 years. The most spectacular example of an ectogeni- located in volcanically active regions where vents carry
cally meromictic lake is probably Island Copper Mine Lake substances into the deep waters of the lake where they
(Vancouver Island, British Columbia, Canada). After dissolve and keep density stratification stable. Lac Pavin, in
decommissioning, a 330 m deep ore mine was filled with France, has a groundwater component to its meromictic
ocean water and capped with a 7 m thick freshwater layer [Aeschbach-Hertig et al., 1999, 2002] behavior, but more
[Fisher and Lawrence, 2006] and hence designed to be recent work about the distribution indicates that endogenic
meromictic to confine undesirable substances to the deep (see section 3.2) processes may also contribute [Bonhomme,
waters for further treatment (other examples are given by 2008].
Hamblin et al. [1999], Stevens and Lawrence [1997], [48] The extent to which gases can be dissolved in a lake
Stottmeister et al. [1998], and Stottmeister and Weißbrodt is displayed in Figure 11 for Lake Monoun. Depending on
[2000]). the chemical specification, dissolved gases contribute pos-
[45] We also include lakes close to the coast which itively or negatively to the density of the water (see
receive seawater at spring tide or seasonally but otherwise section 5.4). In Lake Nyos, the contribution of CO2 is
have freshwater throughflow (e.g., Lower Mystic Lake in decisive for the stability of the density stratification [Schmid
10 of 27
Respiration
CH2 O þ O2 ! CO2 þ H2 O; ð2aÞ
Denitrification
5CH2 O þ 4NO! þ
3 þ 4H ! 5CO2 þ 2N2 þ 7H2 O; ð2bÞ
Manganese reduction
CH2 O þ 2MnO2 þ 4Hþ ! CO2 þ 2Mn2þ þ H2 O; ð2cÞ
Iron reduction
CH2 O þ 4FeðOHÞ3 þ 8Hþ ! CO2 þ 4Fe2þ þ 11H2 O; ð2dÞ
Figure 11. Profiles of partial pressures of dissolved gases Sulphate reduction

in deep water of Lake Monoun, Cameroon, in direct
comparison with hydrostatic pressure (solid line) (adapted 2CH2 O þ SO2! þ
4 þ 2H ! 2CO2 þ H2 S þ 2H2 O; ð2eÞ
from Halbwachs et al. [2004]).
Methanogenesis
et al., 2004]. If more gas were advected so that the sum of
the partial pressures grew higher than the hydrostatic 4CH3 OH ! 3CH4 þ CO2 þ 2H2 O; ð2f Þ
pressure, bubbles could be created. If these bubbles form
a chimney, the lake releases the dissolved gas in a limnic CH3 COOH þ H2 O ! CH4 þ CO2 þ H2 O: ð2gÞ
eruption, as happened in Lake Nyos in 1986.
3.2. Endogenic or Biogenic Meromixis Equations (2a) – (2g) are gross summaries for initial and end
reagents of several microbially mediated intermittent steps.
[49] Biogenic meromixis originates from decomposition
More details are available in the appropriate studies [e.g.,
of organic material in deep water of a lake and dissolution
Schlesinger, 2005; Wetzel, 2001]. End products dissolve in
of its end products. Walker and Likens [1975] used the term
deep lake water and contribute their part to density. The
endogenic meromixis, which also allows for the inclusion of
more productive a lake is, the more important this process
geochemical processes that are not controlled by biological
can be for stabilizing stratification. Some of the maar lakes
activity and carbon cycle. In the epilimnion, photosynthet-
in Germany are typical examples for biogenic meromixis
ically active plankton uses the incoming solar radiation for
[Scharf and Oehms, 1992; Stewart and Hollan, 1975].
primary production. In addition, organic material is carried
Findenegg [1935] argues that the meromictic lakes of
into the system through the surface or by inflowing streams
Carinthia are permanently stratified because of dissolution
(allochthonous material). A portion of this organic material
of end products from decomposing organic material.
settles on the lake bed. Its decomposition is facilitated by
Presented CO2 profiles and pH profiles with high CO2
the presence of oxygen or other oxidizing agents in the deep
concentrations and low pH in the monimolimnion support
layers of the lake (Figure 12). Equations (1a), (1b), and
his arguments.
(2a) – (2g) show the corresponding reactions where CH2O is
used for organic substance for simplicity.
Photosynthesis
light
CO2 þ H2 O !! CH2 O þ O2 : ð1aÞ
as performed by green algae and cyanobacteria. A similar

equation is valid for the formation of organic matter by
photosythetic sulphur bacteria [van Niel, 1936; Schwoerbel,
1999; Wetzel, 2001]. Besides oxygen and sulphur, other
substances can also act as electron/hydrogen carriers:
light
CO2 þ 2H2 S !! CH2 O þ H2 O þ 2S: ð1bÞ
Figure 12. Sketch of a biogenically meromictic lake.
Decomposition of organic material is accomplished by the Organic material is assembled from photosythesis in the
following processes: epilimnion; it can precipitate and decompose and dissolve in
the monimolimnion.
11 of 27
Figure 13. Calcite precipitation by photosythetically Figure 14. Oxidation of ferrous iron to ferric iron and
increasing pH in the epilimnion and redissolution in the subsequent precipitation in oxic layers of the lake. In the
monimolimnion by carbon dioxide from decomposition of monimolimnion, reduction to ferrous iron is facilitated by
organic material. using organic material as reduction agent. Diffusion across
the chemocline and mixing of mixolimnetic waters with
monimolimnetic waters is inefficient in terms of transport of
[50] Since the study by Hutchinson [1957], some more iron, as iron is oxidized and precipitated out of the oxic
geochemical cycles have been shown to create or sustain layer.
meromixis [Walker and Likens, 1975; Hongve, 2004].
Rodrigo et al. [2001] refer to high Ca concentrations in
the monimolimnion to prove that the monimolimnion in [1980] saw the same feature in lakes in Ontario, Canada.
Lake La Cruz, Spain, is stabilized by calcite precipitation in More recent work by Hongve [1997, 2002] confirms this
the epilimnion and partial dissolution in the monimolimn- stratification in lakes of southern Norway. He also states
ion. In this geochemical cycle, calcite is precipitated out of that the contribution of dissolved CO2 plays an important
the epilimnion by pH shift due to photosynthetic activity, role for stability. Most importantly, this iron cycling can be
while in the monimolimnion carbon dioxide is formed by observed in many mining lakes, e.g., in eastern Germany
the decomposition of organic material which facilitates the (Waldsee near Döbern [see Schimmele and Herzsprung,
solution of calcite in monimolimnetic waters (see Figure 13 2000], Moritzteich [see Stellmacher, 2004] (Figure 15),
and equations (2a) –(2d), (3), and (4)). Photosynthesis and Mining Lake 111 [see Karakas et al., 2003], and Goitsche-
calcite precipitation see [see Boehrer et al., 2003]). In Berkeley Pit Lake (Butte,
Montana), even pyrite is oxidized by ferric iron in the
light monimolimnion, liberating additional ferrous iron, sulphate,
2HCO!
3 þ Ca
2þ
!! CH2 O þ O2 þ CaCO3 ð3Þ
and acidity [Pellicori et al., 2005]. Reactions (5a) and (5b)
release protons in sequence. A formation of local pH
Calcite dissolution through carbon dioxide
minima can be the consequence in the contact zones
between oxic and anoxic layers, as has been observed in a
CaCO3 þ CO2 þ H2 O ! Ca2þ þ 2HCO!
3: ð4Þ
number of lakes (e.g., Lake Goitsche [see Boehrer et al.,
A considerable number of meromictic lakes are stabilized 2003]).
by iron cycling. Inflowing groundwater advects dissolved
ferrous iron. This is oxidized to ferric iron and precipitated
4Fe2þ þ O2 þ 4Hþ ! 4Fe3þ þ 2H2 O ð5aÞ
Fe3þ þ 3H2 O ! FeðOHÞ3 þ 3Hþ : ð5bÞ
Under favorable conditions a portion of this iron can be

reduced to ferrous iron again and become dissolved in
monimolimnetic waters (Figure 14), where it contributes to
density and stabilizes the stratification. Both oxidation and
reduction are facilitated by the presence of microbial
organisms. Reduction of iron was already shown in
equation (2d).
[51] Kjensmo [1967, 1968] reported the case of an anoxic
monimolimnion with high iron concentrations beneath an Figure 15. Temperature profile and concentration of iron
oxic mixolimnion. Merilainen [1970] has found the same in Moritzteich, Germany, on 28 October 2003. Oxic
layering in lakes of Finland, and Campbell and Torgersen mixolimnion is 0– 10 m, and anoxic monimolimnion is
below 10 m (data from Stellmacher [2004]).
12 of 27
responsible for forming a monimolimnion. In addition, if

sulphate reduction occurs at the same time and at a similar
rate as iron reduction, the formation of a microstratification
of sulphur bacteria in the chemocline, as known from
Sorokin [1970] and Tonolla et al. [2003], will be sup-
pressed. Similarly, precipitation of iron sulphide will re-
move H2S, which is required by autotrophic sulphur
bacteria for photosynthesis.
3.3. Thermobaric Stratification
[55] We encounter one additional barrier for the deep
water recirculation in very deep lakes that show low
concentrations of dissolved substances and that traverse
the temperature of maximum density (4!C) during the
annual cycle in their surface waters. Deep water of these
Figure 16. Temperature profiles of thermobarically stra- lakes is close to the temperature of maximum density Tmd.
tified lakes: (left) Tinnsjø, Norway, during (early) summer
The barrier for deep water recirculation is imposed by the
stratification above temperature of maximum density Tmd;
(middle) Lake Baikal, Siberia, Russia, during (late) winter (small) compressibility of water and its temperature depen-
stratification with vertical transition through Tmd; (right) dence: cold water is more compressible than warmer water.
Lake Shikotsu, Hokkaido, Japan, with nearly isothermal As a consequence, temperature of maximum density Tmd
deep water body below the Tmd transition. Tinnsjø data from decreases as pressure, i.e., depth, increases (by about 0.21 K
B. Boehrer (unpublished data, 2008); Lake Baikal data from over 100 m water depth [Eklund, 1963], see also equation (26)
Wüest et al. [2005], (with permission of the American in section 5).
Society of Limnology and Oceanography, Inc.); and Lake [56] Tmd can be displayed as profile versus depth. To the
Shikotsu data from B. Boehrer et al. (Stratification in very right of the Tmd line, usual temperature profiles with rising
deep, purely thermally stratified lakes, manuscript in temperatures toward the surface can be found, while to its
preparation, 2008). left, stable temperature stratification is inverse. Particularly
interesting are cases where surface waters cross the temper-
[52] The same cycle between oxic and anoxic zones can ature of maximum density during spring and autumn. Both
be accomplished by manganese. In a few lakes, manganese summer and winter profiles then show peculiarities.
can take the leading role, e.g., Lake Nordbytjernet in [57] Starting at a surface temperature above 4!C in sum-
southern Norway [Hongve, 1997]. In most cases of ob- mer, falling temperatures with increasing depth imply a
served enrichment with manganese in the monimolimnion stable stratification, as long as temperatures lie above Tmd
(e.g., Berkeley Pit Lake [Pellicori et al., 2005] and Lake of corresponding pressure. At greater depth, temperatures
Goitsche, Rassnitzer See, and Wallendorfer See (M. Schul- can well be lower than 4!C (Figure 16, left). The first profile
tze, unpublished data, 2003; B. Boehrer, unpublished data, in a purely temperature stratified freshwater lake with a
2003)) and in deep depressions of the Baltic Sea [Pohl et transition through 4!C was measured in Lake Mjøsa,
al., 2004], the manganese only contributes a small portion Norway, on 23 June 1883 cited by Strøm [1945]. Similar
to the density difference between mixolimnion and mon- temperature profiles were also found in Lake Ladoga,
imolimnion, while either salinity or iron and bicarbonate Russia, in 1900 by Pettersson [1902] (as cited by Strøm
concentration is responsible for sustaining meromixis. [1945]), in Crater Lake, United States, in 1913 by Kemmerer
[53] Beyond the contribution to density stratification, et al. [1924] (as cited by Strøm [1945]), and in Lake
cycling of iron and manganese plays an important role for Shikotsu, Japan, by Yoshimura [1936a, 1936b], proving
the enrichment of trace elements in the monimolimnion. the presence of thermobaric stratification in freshwater lakes
These substances often form coprecipitates with iron or if climatic and morphometric conditions allowed for it.
manganese or are adsorbed by solid oxidized species of iron [58] In winter, when surface temperatures are forced
and manganese and hence are transferred by sedimentation below 4!C, only inverse temperature stratification is stable
into the monimolimnion. Reductive dissolution of solid just below the surface. At greater depths, however, water
oxidized species of iron and manganese results in the may show temperatures between 4!C and Tmd at the
liberation of trace substances into monimolimnetic water. respective pressure. Under these conditions, profiles must
An example is given by Hongve [1997] for phosphorus. intersect with the Tmd profile. Conditions are stable only for
[54] Some of the processes discussed in connection with a vertical transition, i.e., no temperature gradient [Eklund,
biogenic meromixis cannot be present at the same time. If, 1963, 1965]. In fact, Lake Baikal shows a temperature
at the base of a lake, sulphate is reduced in considerable maximum at the intersection (see Figure 16, middle). In
quantities and thus high concentrations of H2S are present Lake Shikotsu (Hokkaido, Japan, Figure 16, right) the
(see equation (2e)), ferrous iron will be precipitated as iron temperature gradient vanishes from the intersection down-
sulphide and iron will continuously be removed from the ward (see also Crater Lake [Crawford and Collier, 1997,
system. Under such conditions, the iron cycle cannot be 2007]).
13 of 27
[ 62 ] From tracer measurements (chlorofluorocarbon

(CFC), helium-tritium, sulphur hexafluoride (SF6)), an
age, i.e., time since the last gas exchange with the atmo-
sphere took place, is inferred which often lies in the range of
one to a few decades. Monimolimnetic waters representing
a mixture of waters of different age are assigned the
apparent bulk age. In addition, tracer gases are also trans-
ported by smaller-scale turbulent transport processes
through the stratified water column. This contribution also
needs separate consideration when estimating deep water
renewal from tracer gas data [e.g., Ravens et al., 2000;
Weiss et al., 1991].
[63] A famous and easy to follow example for deep water
Figure 17. Sketch of deep water renewal in Lake Malawi. renewal is 703 m deep Lake Malawi (or Nyasa) in East
At the southern end ‘‘S,’’ surface water becomes cold Africa. Lake Malawi is usually referred to as a meromictic
enough during cold periods in southern winter (shaded lake, as it is permanently stratified and a clear chemocline
dark) that its density increases sufficiently to sink and delimits the oxic mixolimnion from the anoxic monimo-
intrude into the anoxic monimolimnion.
limnion at a depth of 200 m. The age of monimolimnetic
water has been measured with CFC as 20 to 25 years
[59] In winter, while waters warmer than Tmd can be [Vollmer et al., 2002a]. Lake Malawi extends from latitudes
conserved at great depth, all waters colder than Tmd will be 9!S to 15!S. Thus, at least in the southern winter, it is
removed when the surface water becomes warmer. As a exposed to a climatic gradient. It is suspected that the lowest
consequence, some of the deep crater lakes reflect the Tmd water temperatures are created at the southern end during
profile over a wide range of depths. Many of the sufficiently southern winter (Figure 17). Water can become cold enough
deep lakes in Norway, however, show a temperature profile to exceed the density of the surrounding mixolimnion and to
between isothermal and the Tmd profile. Eklund [1965] flow down side slopes to the abyss. In each episode, only
reasons that the lake stratification remains close to maxi- part of the monimolimnion is replaced [e.g., Halfman,
mum stability, which he has found at a temperature profile 1993]. As part of the tracer gases were carried by turbulent
of half the Tmd gradient. transport through the stratified water column, the water age
[60] Observations show that the water bodies of these of 20 to 25 years gives an upper bound for the amount of
lakes (Crater Lake, Lake Shikotsu, Lake Baikal, Lake annually renewed deep water.
Tinnsjø) are well supplied with oxygen. This can in part [64] The deep water circulation in moderately saline
be attributed to low temperatures, as productivity and (#6 g kg!1) Issyk-Kul (Kyrgyzstan, central Asia [Peeters
depletion of oxygen happen at a slow rate. However, it is et al., 2003]) is similar. From the deepest point (668 m) up
also an indication that a considerable portion of bottom to 100 m depth, deep water shows a temperature between
water is replaced each year and turbulent diffusive trans- 4.3 and 4.8!C. In winter, cold winds descending from the
ports are highly effective because of the very weak density Tien Shan (Mountains) traverse the lake from east to west.
gradient. As a consequence, chemical gradients do not In shallow areas of the lake, water temperature falls enough
appear in these lakes, and (most) scientists have refrained to form dense water, which enters submerged river valleys
from calling these lakes meromictic, despite their permanent to proceed into the abyss of the lake and to recharge deep
stratification. waters. Tracer measurements indicate a deep water age of
about 10 years [Hofer et al., 2002; Vollmer et al., 2002b;
4. EPISODIC PARTIAL DEEP WATER RENEWAL Peeters et al., 2003]. Though the water age is only a factor
of 2 or 3 smaller than in Lake Malawi, Issyk-Kyl is not
[61] A number of lakes do not fit well into the classifi- called meromictic, as a pronounced chemocline to a chem-
cation of holomixis and meromixis. These lakes do not ically different monimolimnion is missing.
experience a complete overturn; thus they are not holomic- [65] Deep water renewal also takes place in lakes that are
tic, but neither does their circulation fit into the picture of commonly called holomictic/oligomictic. Hollan [1998,
meromictic lakes. While in meromictic lakes the circulation 1999], for example, argues that in years of no complete
is confined to the mixolimnion and the water body is often overturn, shallow areas of Lake Constance, Germany, could
homogenized from the surface to the chemocline over the become cold enough to form intrusions that proceed into the
entire horizontal dimension, lakes with episodic deep water deepest parts of the lake (250 m). In exceptionally cold
renewal form water parcels of monimolimnetic density winters, Lake Constance may stratify inversely [1987] or
within the mixolimnion. These water parcels manage to even freeze [1963] [Zenger et al., 1990; Bäuerle et al.,
proceed through the surrounding mixolimnetic waters down 1998]. However, a more recent study by Rossknecht [2003]
into the monimolimnion. In general, this process is similar found that the renewal of the deepest waters was not
to the ocean circulation, and hence it is not surprising that particularly efficient during those seemingly dimictic years.
some of the largest lakes show a similar behavior. This is probably connected to remaining thermobaric strat-
14 of 27
while it mixes with surrounding water (thermobaric insta-

bility [Carmack and Weiss, 1991]). The exact interplay of all
processes, however, has not yet been deciphered. CFC and
helium-tritium measurements indicate a monimolimnion age
of less than 20 years [Weiss et al., 1991; Hohmann et al.,
1998; Kodenev, 2001], where turbulent diffusive transport
vertically through the stratification also contributes its part.
5. QUANTITATIVE APPROXIMATIONS FOR

STRATIFICATION RELEVANT QUANTITIES
[69] For evaluation of stratification related magnitudes

Figure 18. Sketch of deep water renewal in a deep and from field measurements, a number of approaches have
long lake (e.g., Lake Baikal) by wind stress and consecutive been documented. At low salinities, the best reference is
convective propagation of a dense intrusion (1, epilimnion Chen and Millero [1986], while for ocean conditions the
without strong wind stress; 2, downwelling forced by wind valid reference is Fofonoff and Millard [1983]. From both
stress on the epilimnion; 3, convective propagation after studies, we cite most often used approximations. However,
crossing temperature of maximum density). Longshore both studies include many more physical quantities related
wind and onshore wind represent contribution of Coriolis to density, which cannot be shown within this review.
deflection and direct surface draft for accumulation of Instead, we present ways to approach stratification relevant
surface water at the shore.
magnitudes if both ocean assumptions and freshwater
assumptions fail.
ification during inverse temperature stratification in very
5.1. Depth and Sound Speed
deep lakes (see section 3.3).
[66] Intrusions recharging monimolimnetic waters can 5.1.1. Depth
also be driven by a higher salt concentration. Before 1979, [70] Depth of a probe could be measured from length of a
evaporation created water of higher salt concentration in the rope. However, many ropes change their length consider-
shallow southern basin of the highly saline Dead Sea. ably under tension. Other ropes shrink when absorbing
Because of its very high density, it could flow underneath water. Steel wires or metal pipes might solve some of the
the mixolimnetic waters of the main basin [Nissenbaum, problems [e.g., von Rohden and Ilmberger, 2001], but
1969; Gat, 1995]. This process kept the Dead Sea stratified currents or drift of a vessel will curve the wire and shift a
for decades, perhaps centuries. However, since the southern probe sideways.
basin has basically dried up and does not connect to the main [71] For this reason, depth is most commonly measured
basin as before, the lake has been holomictic since 1979. In as pressure. In water, hydrostatic pressure p increases with z
addition, capping with fresh water from the Jordan River (taken as positive downward) as
(ectogenic meromixis) has weakened over the same period.
dp
[67] High-salinity waters can also be formed by freezing. ¼ grin situ : ð6Þ
dz
The liquid phase remains with increased salinity and thus
can gain enough density to form intrusions for recharging a
[72] Earth’s acceleration g ( 9.8 m s!2can be assumed to
monimolimnion [Goldman et al., 1972]. Walker and Likens
be constant in nearly all lakes. In situ density rin situ in
[1975] have listed five lakes in Antarctica which stay
freshwater lakes differs from 1000 kg m!3 as a consequence
permanently stratified by this process [also Gibson, 1999]
of higher temperature, dissolved substances, and pressure
and proposed a separate meromixis class, as they felt that
but only in the range of a few permille. Many probes
these lakes do not fit into any of the three meromixis classes
therefore indicate pressure in dbar, which is close to m. In
of Hutchinson [1957].
the ocean and in lakes of high salinity, depth needs to be
[68] Lake Baikal fulfils the conditions for a classic ther-
evaluated properly following equation (6), which normally
mobaric stratification (see section 3.3). Late during the
is facilitated by the measurements in the profile above. In all
winter, surface waters are below 4!C and ‘‘inversely’’ strat-
cases, atmospheric pressure, which varies with altitude and
ified to a depth of 250 m, where the temperature profile
time, needs to be eliminated (‘‘calibration’’).
intersects vertically with the Tmd line. However, below the
5.1.2. Sound Speed
intersection, temperatures fall below 3.3!C toward the bot-
[73] Depth can also be measured from sound speed. An
tom of the southern basin of the lake (Figure 16, middle)
echo sounder measures the travel time t for a sound ping to
[Wüest et al., 2005]. Hence a process recharging the deep
complete the path of length s
water from the surface can be suspected. Most probably
during winter periods when the lake is not covered with ice, receiver
1
Z
wind pushes cold surface water deep enough that its density is
t¼ ds; ð7Þ
higher than surrounding water (Figure 18, ‘‘2’’). From there, c
sender
it can proceed convectively to greater depth (Figure 18, ‘‘3’’)
15 of 27
[77] On the contrary, positioned senders and receivers can

record the variations in the travel time for a fixed distance,
and variations in the temperature field can be measured. A
symmetrical setup may facilitate measurements of current
speeds. A field of sensors can provide data for tomographic
procedures, which can evaluate three-dimensional tempera-
ture fields and currents by noninvasive methods [e.g.,
Dushaw et al., 2001].
5.2. Temperature
[78] Temperatures recorded in lakes by field equipment
are so-called in situ temperatures. Without any further
annotation, temperature data will be understood as in situ
temperatures. Nearly all calculations refer to this value, as it
is the physically, chemically, and ecologically relevant
magnitude.
Figure 19. Profile of measured sound speed in Lake [79] However, if detailed considerations of stability and
Constance on 13 June 2002 at a 200 m deep location.
vertical temperature gradients are intended, potential tem-
Profiles of temperature, pressure, and salinity (barely visible
on the left) in roughly the magnitude of their contribution to perature may be useful. This includes the effect of energy
sound speed. required for the expansion. Potential temperature is defined
as temperature that a water parcel would acquire if trans-
ferred adiabatically to a reference pressure, often taken as
where sound speed c can be calculated from density and atmospheric pressure [e.g., Gill, 1982]. Derivation is along
adiabatic compressibility k: lines of the Maxwell identities (see Joule-Thomson effect
[e.g., Reif, 1987]) of the statistical mechanics. Adiabatic
c2 ¼ 1=ðrin situ kÞ: ð8Þ lapse rate is given by
[74] Hence the sound propagation is dependent on tem- gaðT þ 273:15Þ

! "
dT
perature, salinity, and pressure (see Figure 19). In addition, ¼ ; ð10Þ
dz ad cp
it is also influenced by currents, but in most lakes this effect
is a subordinate contribution and will not be discussed here. where a is thermal expansion coefficient and cp ( 4200
The temperature dependence in pure water was evaluated by J(kg K)!1 is specific heat at constant pressure. Hence
Del Grosso and Mader [1972], and the approximation was potential temperature is evaluated as
extended for pressure and salinity by Chen and Millero
[1986]: Z0 ! Z0
gaðT þ 273:15Þ 0
"
dT
Tpot ¼ T ð zÞ þ dz0 ¼ T ð zÞ þ dz ;
5 2 2 dz ad cp
X X X z0 ¼z z0 ¼z
c¼ i
ai Tpot þS i
bi Tpot þp i
ci Tpot ! 5:58 " 10!5 Sp
i¼0 i¼0 i¼0
ð11Þ
!5 2
þ 1:593 " 10 p ; ð9Þ
where z is positive downward. For most limnological
!2
with ai = [1402.388; 5.0371; ! 5.8085 " 10 ; 3.342 " 10 ; !4 purposes, a, g, and cp can be considered constant over the
!1.478 " 10!6; 3.146 " 10!9], bi = [1.322; !7.01 " 10!3; entire depth range. Using an estimate for a of
4.9 " 10!5], and ci = [0.15564; 4.046 " 10!4; !8.15 " 10!7].
1
[75] Coefficients have been listed without units for better 1 . . . 1:5 " 10!5 ðT ! Tmd Þ
legibility. The unit for pressure p is bar, and for salinity it is K2
psu. Chen and Millero [1986] use a slightly different equation (10) yields an adiabatic lapse rate (dT/dz)ad of the
definition for salinity in the ocean (see section 5.3.2) and order of (T ! Tmd) " 10!5 m!1. Hence in most cases the
salinity in this formula, S = 1.00488Socean (for more about temperature stratification is much larger than the adiabatic
salinity, see section 5.3.2). lapse rate. The difference between in situ and potential
[76] Sound speed can vary by several percent within one temperature can be neglected in most cases. However, to be
lake mainly because of the temperature variation. Hence this exact, we will add a subscript Tpot whenever it is appropriate
must be considered for accurate measurements. Sound in this contribution. A considerable difference between
beams propagating in horizontal directions within vertical potential temperature and in situ temperature can be found in
temperature and pressure gradients are deflected from deep tropical lakes. While, for example, the in situ
straight lines and thus can take unexpected paths from temperature profile in Lake Malawi is not monotonous,
sender to receiver. Depending on conditions, several differ- Wüest et al. [1996] showed that potential temperature rises
ent paths may be possible [e.g., Dushaw et al., 2001]. monotonously; that is, the lake is stably density stratified in
respect to temperature only. In Figure 20, we present a profile
16 of 27
[82] For the hypolimnion water of highly acidic Mining

Lake 111 (5 m sample), Karakas et al. [2003] evaluated
b=a ¼ 35:2
and a very small value aref = 0.0166 (see also section 5.3.1)
at a reference temperature Tref = 25!C:
C ðT Þ
k25 ¼ : ð13Þ
0:0166ðT ! 25+ CÞ þ 1
[83] For better clarity, we dropped the unit K!1. Equation

(13) was applied to CTD profiles of years 1996– 2001 (see
Figure 22). The result can be interpreted as an indication for
concentration of dissolved substances. The depiction clearly
Figure 20. Profiles of (in situ) temperature T and potential shows the high concentration in the monimolimnion and its
temperature Tpot near the deepest location of Lake Malawi intermittent absence. In Mining Lake 111, the precipitation-
on 13 September 1997 (data from Vollmer et al. [2002a]). evaporation deficit during summer months increases the
conductance in the epilimnion [Karakas et al., 2003]. Heinz
measured in Lake Malawi by Vollmer et al. [2002a]. Potential et al. [1990], for example, used changes in conductance as
temperature was calculated following equation (11), where an indication of vertical transport in Lake Constance and
we assumed constant values for g and cp. The thermal calculated transport coefficients from it.
expansion coefficient [84] The advantage of restricting oneself to a linear
regression for conductance bears the advantage of only
1 @V 1 @rin situ one parameter
a¼ ¼!
V @T rin situ @T
ðTref þ ðb=aÞÞ!1 ¼ aref
was evaluated from equation (17) (see below). The depiction
is instructive in so far as it indicates what temperature is defining the calculation entirely. This facilitates the
required at the surface for deep water renewal. comparison of data of different reference temperatures.
Measurements of various years would directly indicate to
5.3. Electrical Conductivity, Electrical Conductance, what extent changes between various layers and years must
and Salinity be considered. As in Mining Lake 111, such differences can
5.3.1. Electrical Conductivity and be found, especially within other meromictic lakes. Karakas
Electrical Conductance et al. [2003] decided to use the hypolimnion regression for
[80] Many substances dissolve in lake water, dissociate, the entire lake, as the epilimnetic fit turned out to be very
and form electroactive ionic species. Hence they contribute to
the electrical conductivity. Electrical conductivity of a water
sample depends on the dissolved constituents and tempera-
ture of the sample [e.g., Rodhe, 1949]. To compensate for
temperature effects, electric conductivity is recorded while
scanning the relevant temperature interval (Figure 21).
[81] In most cases, a linear regression C(T) = aT + b is
satisfactory to define conductance, i.e., electrical conduc-
tivity kref = C(Tref) = aTref + b at a certain reference
temperature Tref. Most commonly, a reference temperature
of 25!C is used. Dividing C(T) by k ref and solving for
the latter, we find the user’s formula for calculating
conductance
C ðT Þ
kref ¼ ; ð12Þ
aref ðT ! Tref Þ þ 1
where
Figure 21. Electrical conductivity of three water samples
aref ¼ ½Tref þ ðb=aÞ*!1 : of Mining Lake 111 versus temperature [see Karakas et al.,
2003]. Symbols represent measured conductivity. Solid
lines show linear regression (with permission from
Birkhäuser).
17 of 27
Figure 22. Electrical conductance k25 in Mining Lake 111 versus depth and time (adapted from
Karakas et al. [2003] with permission from Birkhäuser).
similar, while in the monimolimnion, gradients were very Most scientists still use a temperature compensation
high anyway; hence no particularly high accuracy for following ISO [1985] or variations thereof.
conductance was required. [87] If the lake water properties cannot be evaluated in a
[85] In highly acidic Mining Lake 111, a very small value conductivity measurement versus temperature, coefficients
a25 = 0.0166 was found. In most surface water, a value close for the most common ions can be taken from Sorensen and
to a25 = 0.02 is appropriate [Wissenschaftlich-technische Glass [1987]. The coefficients of most of the common ions
Werkstätten, 1993]. For the neutral mixolimnion of lake are close to each other; only H+ shows a different behavior.
Goitsche, a25 = 0.0197 was evaluated, while epilimnion, Consequently, a good approximation is gained by replacing
hypolimnion, and monimolimnion of salt lake Rassnitzer See K in equation (14) by K = 0.941 ! 0.00639[H+], where H+is
showed values 0.0191 < a25 < 0.0199. In acidic Roter See given in mmol L!l. Their data relied on Canadian lakes with
near Burgkemnitz, Germany, a25 = 0.0185, while several low concentration of other dissolved salts. How well this
samples of the Elbe river near Magdeburg, Germany, showed regression would suit highly mineralized waters has not
0.0196 < a25 < 0.0208 (G. Götz, personal communication, been checked.
2000). [88] From oceanography it is known that electrical con-
[86] This approach follows the International Organization ductivity is also influenced by pressure. Wüest et al. [1996]
for Standardization (ISO) [1985], in that only the factor a25 is assume that the electrical conductivity is increased by a
adjusted for varying composition of constituents contributing factor of 1.0066 at 700 m depth in Lake Malawi in
to electrical conductivity. A close look at Figure 21 reveals comparison to surface pressure by extrapolating ocean
that data points show a curvature against temperature, assumption to zero salinity. While in Lake Malawi this
which cannot be reproduced by a linear regression. Hence effect makes up for nearly the entire conductivity gradient in
a better parameterization has been looked for. Sorensen and the monimolimnion, this effect is negligible in most other
Glass [1987] consider various approaches for fitting tem- lakes.
perature dependence of electrical conductivity. They con- 5.3.2. Total Dissolved Substances and Salinity
clude that the ISO [1985] delivers good results, compared [89] For various purposes, depiction and numerical eval-
to other methods. However, they also prove that an ap- uation of the quantity of total dissolved substances (TDS)
proximation of the form may be useful. In some applications, this value is used as an
intermediate step to calculating density in limnic waters
k25 ¼ C ðT Þð½ðhðT Þ=h25 *ÞK ; ð14Þ [Chen and Millero, 1986; McManus et al., 1992]. Evaluat-
ing TDS involves the full chemical analysis of all dissolved
would be better. The term h(T)/h25 represents the ratio of substances including gases and organic material, both of
viscosity of water at the temperature of measurement and which can pose sampling and analytical difficulty.
the reference temperature at 25!C. Equation (14) requires [90] The easiest approach connects TDS with conduc-
only one constant K, which ISO supplies in a table for the tance by a regression factor [e.g., Snoeyink and Jenkins,
most common ions. One shortcoming of equation (14) lies 1980] (as cited by McManus et al. [1992]). But there are
in the fact that h(T)/h25 is not necessarily available, at least more elaborate approaches as given by higher-order approx-
not for the considered lake water [e.g., Jellison et al., 1999].
18 of 27
imations of McManus et al. [1992]. If the nonconductive even pronounced vertical gradients within the same lake.
dissolved species do not show a parallel profile to conduc- Also, there may be substances contributing considerably to
tive species, their contribution must be added in separate density but not to electrical conductivity (e.g., suspended
(e.g., Wüest et al. [1996] in their approach for Lake particles, dissolved organic compounds, silicate). As a
Malawi). consequence, salinity can only be used with reservation
[91] Oceanography uses electrical conductivity and tem- and when chemical composition of lake water is close
perature to calculate salinity in psu, which gives a good enough to ocean conditions. Annotation of the psu or
indication for dissolved salt in g kg!1 for ocean water and referring to salinity according to UNESCO indicates refer-
brackish water: ence to a numerical value and not salt content.
[94] Oceanography has long been aware of limitations of
S ¼ S ðC; T ; pÞ: ð15aÞ current definitions of salinity and psu scale. Measurements
during the last decades showed that an approximation of
Salinity is evaluated over several steps: dissolved substances could be achieved at a much better
accuracy. The idea is now to compile a Gibb’s function,
R ¼ C=42:914 mS cm!1 ; ð15bÞ from which thermodynamic properties of seawater can be
derived, including deviations from standard compositions of
4
salts [e.g., Feistel and Hagen, 1995]. The novel approach is
based on recent measurements and is planned to be agreed
X
i
rt ¼ ci Tpot ; ð15cÞ
i¼0 upon and issued in 2008 [see Feistel and Marion, 2007].
5.4. Density
3
X h%
2
$ &i [95] Density r is the crucial magnitude for the stability of
ei pi = d0 þ d1 Tpot þ d2 Tpot
#
Rp ¼ 1 þ þ d3 þ d4 Tpot R ; stratification. Density differences drive currents. Whether
i¼0
and how stable a stratification is can be concluded from
ð15dÞ
comparing density of water parcels under the same pressure
conditions. Usually, the reference of normal conditions
# $
Rt ¼ R= Rp rt ; ð15eÞ (atmospheric pressure) is used, and the term potential
density may be more correct. Hence density is a magnitude
and of central importance in limnophysics, but direct measure-
ment in the field is not feasible at the required accuracy [see
5 5
Gräfe et al., 2002]. As a consequence, the most practical
X i=2 Tpot ! 15 X i=2 way is to evaluate density from measurements of electrical
Socean ¼ ai Rt þ # $ b i Rt : ð15f Þ
i¼0 1 þ k Tpot ! 15 i¼0 conductivity and temperature.
[96] In lakes of a composition of dissolved substances
similar to the ocean, the so-called UNESCO formula by
Coefficients have been evaluated at high accuracy:
Fofonoff and Millard [1983] may be applied (e.g., Rassnitzer
ai ¼ ½0:0080; !0:1692; 25:3851; 14:0941; !7:0261; 2:7081*;
See in Figure 10, lakes close to the coast [see Ludlam, 1996]),
which according to Fofonoff and Millard, is only applicable
bi ¼ ½0:0005; !0:0056; !0:0066; !0:0375; 0:0636; !0:0144*;
for salinities between 2 and 42 psu. Hence it does not apply
ci ¼ 0:6766097; 2:00564 " 10!2 ; 1:104259 " 10!4 ; !6:9698
'
for hyperhaline lagoon waters, though the salt is imported
" 10!7 ; 1:0031 " 10!9 *; from the sea [e.g., Por, 1972]. In limnology, depiction of
di ¼ 1; 3:426 " 10!2 ; 4:464 " 10!4 ; 0:4215; !3:107 " 10!3 ; density as ‘‘density minus 1000 kg m!3’’ may be used and
' (
ei ¼ 0; 2:070 " 10!5 ; !6:370 " 10!10 ; 3:989 " 10!15 ;

' ( referred to as sigma (s).
k ¼ 0:0162:
r ¼ r Socean ; Tpot
# $
5
X 4
X 2
X
[92] At the time of writing this review, UNESCO offered i i 3=2 i 2
¼ ai Tpot þ Socean bi Tpot þ Socean ci Tpot þ Socean d0 ;
an online calculator ( http://ioc.unesco.org/) for properties i¼0 i¼0 i¼0
ð16Þ
of ocean water using coefficients from Fofonoff [1985]. For
low-salinity limnetic water (<0.6 psu), Chen and Millero
[1986] introduced a correction factor to evaluate salinity for
ai ¼ 999:842594; 6:793952 " 10!2 ; !9:095290 " 10!3 ;
'
weakly conductive limnetic waters:
, 1:001685 " 10!4 ; !1:120083 " 10!6 ; 6:536332 " 10!9 *;
bi ¼ 0:824493; !4:0899 " 10!3 ; 7:6438 " 10!5 ; !8:2467
'
S ¼ 1:00488Socean : ð15gÞ
" 10!7 ; 5:3875 " 10!9 *;
ci ¼ !5:72466 " 10!3 ; 1:0227 " 10!4 ; !1:6546 " 10!6
' (
[93] In limnetic systems, the composition of dissolved
d0 ¼ 4:8314 " 10!4
substances differs from the ocean. In some cases, there are
19 of 27
[97] For low salinities (<0.6 psu), the following formula TABLE 1. Contribution of Dissolved Substances to the
by Chen and Millero [1986] may be used for density Density of Watera
calculation:
$ X6 2
X Substance bn, g kg!1
r ¼ r S; Tpot ¼ i i
#
ai Tpot þS bi Tpot ; ð17Þ
i¼0 i¼0 Ca(HCO3)2 0.807 " 10!3
Mg(HCO3)2 0.861 " 10!3
using Na(HCO3) 0.727 " 10!3
K(HCO3) 0.669 " 10!3
Fe(HCO3)2 0.838 " 10!3
ai ¼ 999:8395; 6:7914 " 10!2 ; !9:0894 " 10!3 ; 0.462 " 10!3
'
NH4(HCO3)
CO2 0.273 " 10!3
, 1:0171 " 10!4 ; !1:2846 " 10!6 1:1592 " 10!8 ; !5:0125 CH4 !1.250 " 10!3
" 10!11 * Air !0.090 " 10!3
a
Data from Imboden and Wüest [1995].
bi ¼ 0:8181; !3:85 " 10!3 ; 4:96 " 10!5
' (
[98] In cases where salinity cannot be used, calculation of

density may directly be based on measurements of temper- The resulting 10 coefficients aik are specific for the case of
ature and conductivity. Bührer and Ambühl [1975], for the flooding of Lake Goitsche in 1999 and can be retrieved
example, propose a formula for Lake Constance where they from Gräfe and Boehrer [2001].
separate temperature dependence rT from contribution of [101] An alternative method is applicable if the composi-
dissolved substances G. Heinz et al. [1990] and Bäuerle et tion of dissolved substances is known. Density can be
al. [1998] used the following variation for Upper Lake calculated by using a temperature curve and adding sub-
Constance (Obersee): stance-specific contributions to the density temperature
curve.
r ¼ rT þ G ¼ 999:8429 þ 10!3 !
% & X
3 2
þ 65:4891Tpot þ G: r ¼ rT Tpot bn cn
# $
, 0:059385Tpot ! 8:56272Tpot ð18Þ 1þ ð22Þ
n
They used 20!C as reference for the conductance and

evaluated the coefficient connecting k20 with density 1 @r
empirically: bn ¼ ð23Þ
r @cn
$!1
G ¼ gk20 andg ¼ 0:67 " 10!3 kgm!3 mScm!1 :
#
ð19Þ represent coefficients correlating density contribution with
concentration of a dissolved substance; cn is the concentra-
[99] An approach that separates temperature and conduc- tion of substance n. Values of b n are given in Table 1 for
tivity contributions cannot reflect the effect of lowering common salts and gases. As the assumption is made that the
temperature of maximum density by dissolved substances ions do not interfere with each other, given values are only
(see section 2.2). Hence Karakas et al. [2003] allow for a valid for diluted solutions [Imboden and Wüest, 1995; see
temperature-dependent g. They use 25!C as reference also Wüest et al., 1996].
temperature in the highly acidic Mining Lake 111: [102] In addition, suspended material also contributes to
density. If no volume effect is considered, the bulk density
G ¼ g ðT Þk25 ; ð20Þ rb can be calculated by numerically filling the space of
suspended material with its density rs in the surrounding
where liquid density rl at given properties
g Tpot ¼ ! 7:522 " 10!6 Tpot

3
þ 2:613 " 10!4 Tpot
2
# $
! 5:146 rb ¼ rl þ c " ðrs ! rl Þ=rs ; ð24Þ
" 10!3 Tpot þ 0:7918:
where c is the mass of suspended material per unit water

[100] In general, many samples can be used for measuring volume (kg m!3).
temperature dependence of density. In a second step, a 5.4.1. In Situ Density
regression curve is created over the two-dimensional field [103] Besides stability considerations, density was al-
of data. This was done for Lake Goitsche in 1999, for ready mentioned in section 5.1.2. There, however, the term
example, when river water was deviated into an abandoned used was in situ density rin situ, which is the value for
mine pit, where Gräfe and Boehrer [2001] included all density under given environmental conditions, e.g., higher
terms to third order: pressure. Density (or potential density) and in situ density
differ by the contribution of (little) compressibility from
$ iþk-3
each other. For most considerations, except for stability, in
X
r ¼ r k25 ; Tpot ¼ aik ki25 Tpot
k
#
: ð21Þ
i;k.0 situ density is the proper reference. However, also in cases
20 of 27
of complex stratification, e.g., quantitative considerations maximum frequency (w) for internal waves that can prop-
about the thermobaric stratification, the concept of potential agate in respective stratification. N2 indicates how much
density cannot reflect conditions appropriately. energy is required to exchange water parcels vertically [e.g.,
[104] Adiabatic compressibility of water lies around 5 " Boehrer et al., 2000a]. Especially in cases where the
10!10 Pa!1. This means at 200 m depth (potential) density variability of the thermal expansion coefficient becomes
and in situ density differ by about 10!3. This is the same obvious (see sections 3.3 and 5.4.1), derivation of stability
magnitude as weak thermal or weak salinity stratification. as potential temperature gradient finds its limitations [see
Hence the compressibility is the leading term for the also Gill, 1982].
increase of in situ density in weakly stratified deep waters [110] As a consequence, chemical gradients can only
of very deep lakes. persist for longer time periods where density gradients limit
[105] Fofonoff and Millard [1983] present a system of vertical transport of dissolved substances (see Figure 23).
equations to evaluate rin situ for waters under ocean con- Lake basin Niemegk of Lake Goitsche (Germany) has been
ditions. This involves a number of constants which can be neutralized by introducing buffering river water to the
extracted from the respective paper. As under limnological epilimnion (Figure 5). During summer 2000, vertical trans-
conditions (smaller pressure range than the ocean, limited port through the temperature stratification was limited, and
validity of the salinity concept), the in situ density is of a chemical gradient in pH could be sustained. However, in
limited importance for lake waters of high salinity. winter, temperature stratification vanished, vertical transport
[106] On the contrary, the evaluation of rin situ is useful in was enhanced, and consequently chemical gradients were
connection with the thermobaric stratification. Starting with removed.
(potential) density evaluated from equation (16), rin situ-
under given pressure p bar is calculated as 6. CONCLUDING REMARKS AND PERSPECTIVES
rinsitu ¼ rð1 ! p=K Þ; ð25Þ [111] This review presents an overview of lacustrine
stratification features. It is designed to enable a limnologist
to correctly interpret a stratification that he/she may en-
where for freshwater applications [Chen and Millero, 1977, counter in field measurements. We hope that after reading
1986] this contribution, limnologists and geoscientists are able to
understand what processes create stratification and which
4 2 1
X
i
X
i
X
i approaches can be taken to evaluate fundamental quantities
K ½bar* ¼ ci Tpot þp di Tpot þS ei Tpot þ fpS;
i¼0 i¼0 i¼0
such as conductance or density. References are given to
ci ¼ 19652:17; 148:113; !2:293; 1:256 " 10 ; !4:18 " 10!5 ;
' !2
( provide the reader with access to seminal publications in the
field.
di ¼ 3:2726; !2:147 " 10!4 ; 1:128 " 10!4 ;
' (
[112] Though actual stratification in a lake results from a
ei ¼ ½53:238; !0:313*; f ¼ 5:728 " 10!3 : balance between stratifying processes and mixing processes,
the former have been investigated less quantitatively than
the latter. This may have been a consequence of the
[107] From this, Chen and Millero [1977, 1986] present similarities of mixing processes in atmosphere, ocean, and
an equation to calculate the temperature of maximum lakes, while many stratifying processes do not occur in
density for low-salinity water; that is, the temperature atmosphere or ocean, or at least do not play an important
dependence of density at constant pressure and constant role, e.g., geochemical processes or thermobaric stratifica-
salinity vanishes:(@rin situ/@Tpot)p,S = 0: tion close to the temperature of maximum density. As a
consequence, this paper has focused on stratifying processes
Tmd ¼ 3:9839 ! 1:9911 " 10!2 p ! 5:822 " 10!6 p2 in lakes. Contributions of heat, dissolved or suspended
! 0:2219 þ 1:106 " 10!4 p S:
# $
ð26Þ substances, and pressure on the density have been dis-
cussed. Also, most widely accepted approaches for quanti-
tatively evaluating stability of stratification have been
5.4.2. Stability introduced.
[108] Stability of a water column derives from density [113] Reliable numerical simulations of annual tempera-
increase along the vertical coordinate. The usual quantity ture cycles of lakes deceive us into believing that problems
for stability is concerning stratification prognostications have all been
solved. However, not much quantitative material is avail-
g dr able on transport in strongly stratified deep waters. Prog-
N2 ¼ : ð27Þ nosticating the evolution of meromixis even in the (simple)
r dz
case of conservative salts remains a challenge. Thermobaric
stratification can be observed in lakes which are deep
[109] The quantity N is also called stability frequency or enough and located in the corresponding climate zone.
Brunt-Väisälä-frequency (unit s!1), which also indicates the The stratification features encountered in these lakes can
21 of 27
Figure 23. Turbulent diffusive transport of an artificial tracer (SF6) in the strongly stratified
monimolimnion of Rassnitzer See versus density gradient, N2 = !g/r(dr/dz) (adapted from von Rohden
and Ilmberger [2001] with permission from Birkhäuser).
be explained with the current knowledge. Numerical con- responsibly compromise between use of the environment
siderations show that deep water renewal can be facilitated and the conservation of its sound ecological status.
by wind action in Lake Baikal, for example. However, a
demonstration that a stratification model is capable of [116] ACKNOWLEDGMENTS. Many thanks to three anon-
reproducing temperature profiles as found in these lakes ymous reviewers for carefully reading the manuscript and making
and possibly illuminating the complex conditions when helpful suggestions; Walter Geller, Craig Stevens, and Helmut
Klapper for their comments on earlier drafts; and Olaf Büttner,
replacing deep waters is still outstanding.
Karsten Rahn, Uwe Kiwel, and Jörg Hausmann for measurements
[114] A number of geophysical processes have been
and figure preparation. Martin Vollmer, Martin Dokulil, and Alfred
identified which play a role in sustaining stratification. Johny Wüest generously let us use their original data on Lake
Regardless of whether more geochemical cycles need con- Malawi, Mondsee, and Lake Baikal, respectively.
sidering, quantifying precipitation and coprecipitation as [117] The Editor responsible for this paper was Ian Fairchild. He
well as dissolution and liberation of substances in the thanks two technical reviewers and one anonymous cross disci-
monimolimnetic environment has just started. This is espe- plinary reviewer.
cially true for the effect on lake stratification at the accuracy
needed for prognostications. In some cases, such processes REFERENCES
can be decisive factors for whether a lake turns meromictic
or not. Aware of many limitations, better prognostication Aeschbach-Hertig, W., M. Hofer, R. Kipfer, D. M. Imboden, and
R. Wieler (1999), Accumulation of mantle gases in a permanently
instruments are nevertheless required. Even in complex stratified volcanic lake (Lac Pavin, France), Geochim. Cosmo-
situations, uncertainties can be limited to a smaller number chim. Acta, 63(19 – 20), 3357 – 3372, doi:10.1016/S0016-7037
of factors. A large number of lakes would profit from (99)00257-4.
improving numerical tools. Aeschbach-Hertig, W., M. Hofer, M. Schmid, R. Kipfer, and D. M.
[115] A healthy aquatic environment is a key component Imboden (2002), The physical structure and dynamics of a deep,
meromictic crater lake (Lac Pavin, France), Hydrobiologia,
for a good quality of life. In the perspective of global 487(1), 111 – 136, doi:10.1023/A:1022942226198.
change, predictive tools for the evolution of ecological Arai, T. (1991), Climatic and geomorphological influences on lake
function of lacustrine ecosystems are indispensable. Climate temperature, Verh. Int. Ver. Theor. Angew. Limnol., 21, 130 – 134.
change, changing land use, decommissioning disrupted Bäuerle, E., D. Ollinger, and J. Ilmberger (1998), Some meteoro-
mining or industrial landscapes, or the creation of artificial logical, hydrological and hydrodynamical aspects of Upper Lake
Constance, Ergeb. Limnol., 53, 31 – 83.
water bodies poses new questions to limnology. If timely Berger, F. (1955), Die Dichte natürlicher Wässer und die Konzen-
prognosticated, some undesirable evolutions may be trations-Stabilität in Seen, Arch. Hydrobiol. Suppl., 22(3/4),
avoided, while others may have at least been forecast in 286 – 294.
time to initiate legal and social action to prevent the worst. Boehrer, B., and M. Schultze (2005), Schichtung von Seen, in
Understanding and communicating the causal connection is Handbuch Angewandte Limnologie, edited by C. E. W. Steinberg
et al., Chapter IV-2.2, pp. 1 – 21, ecomed, Landsberg, Germany.
fundamental for public acceptance of legislative steering to
22 of 27
Boehrer, B., J. Ilmberger, and K. O. Münnich (2000a), Vertical Davis, A., and D. Ashenberg (1989), The aqueous geochemistry of
structure of currents in western Lake Constance, J. Geophys. the Berkeley Pit, Butte, Montana, U.S.A., Appl. Geochem., 4(1),
Res., 105(C12), 28823 – 28,835, doi:10.1029/2000JC900139. 23 – 36, doi:10.1016/0883-2927(89)90056-5.
Boehrer, B., A. Matzinger, and M. Schimmele (2000b), Similarities Davies-Colley, R. D.-C. (1988), Mixing depths in New Zealand
and differences in the annual temperature cycles of east German lakes, N. Z. J. Mar. Freshwater Res., 22, 517 – 527.
mining lakes, Limnologica, 30(3), 271 – 279, doi:10.1016/S0075- Del Don, C., K. W. Hanselmann, R. Peduzzi, and R. Bachofen
9511 (00)80058-4. (2001), The meromictic alpine Lake Cadagno: Orographical
Boehrer, B., M. Schultze, S. Liefold, G. Behlau, K. Rahn, S. Frimel, and biogeochemical description, Aquat. Sci., 63(1), 70 – 90,
U. Kiwel, B. Kuehn, I. Brookland, and O. Büttner (2003), Strati- doi:10.1007/PL00001345.
fication of mining Lake Goitsche during flooding with river water, Del Grosso, V. A., and C. W. Mader (1972), Speed of sound in pure
in Tailings and Mine Waste ’03, pp. 223 – 231, A. A. Balkema, water, J. Acoust. Soc. Am., 52(5B), 1442 – 1446, doi:10.1121/
Rotterdam, Netherlands. 1.1913258.
Boehrer, B., M. Schultze, K. Ockenfeld, and W. Geller (2005), Denimal, S., C. Bertrand, J. Mudry, Y. Paquette, M. Hochart, and
Path of the 2002 Mulde flood through Lake Goitsche, Germany, M. Steinmann (2005), Evolution of the aqueous geochemistry of
Verh. Int. Ver. Theor. Angew. Limnol., 29, 369 – 372. mine pit lakes—Blanzy-Montceau-les-Mines coal basin (Massif
Böhrer, B., H. Heidenreich, M. Schimmele, and M. Schultze Central, France): Origin of sulfate contents; effects of stratifica-
(1998), Numerical prognosis for salinity profiles of future lakes tion on water quality, Appl. Geochem., 20(5), 825 – 839,
in the open cast mine Merseburg-Ost, Int. J. Salt Lake Res., 7(3), doi:10.1016/j.apgeochem.2004.11.015.
235 – 260, doi:10.1007/BF02441877. Dokulil, M. T., and K. Teubner (2003), Steady state phytoplankton
Bonhomme, C. (2008), Turbulence et ondes en milieu naturel stra- assemblages during thermal stratification in deep alpine lakes,
tifié, Ph.D. thesis, Ecole Nationale des Ponts et Chaussées, Uni- Do they occur?, Hydrobiologia, 502(1 – 3), 65 – 72, doi:10.1023/
versité Paris-Est, Champs sur Marne, France. B:HYDR.0000004270.70364.f3.
Bøyum, A., and J. Kjensmo (1970), Kongressvatn. A crenogenic Dushaw, B., et al. (2001), Observing the ocean in the 2000s: A
meromictic lake at western Spitsbergen, Arch. Hydrobiol., 67(4), strategy for the role of acoustic tomography in ocean climate
542 – 552. observation, in Observing the Oceans in the 21st Century, edited
Brandt, A., and H. J. S. Fernando (Eds.) (1996), Double Diffusive by C. J. Koblinsky and N. R. Smith, pp. 391 – 418, Bur. of
Convection, Geophys. Monogr. Ser., vol. 94, 334 pp., AGU, Meteorol., Melbourne, Victoria, Australia.
Washington, D. C. Eklund, H. (1963), Fresh water: Temperature of maximum density
Brauer, A. (2004), Annually laminated lake sediments and their calculated from compressibility, Science, 142, 1457 – 1458,
paleoclimatic relevance, in The Climate in Historical Times: To- doi:10.1126/science.142.3598.1457.
wards a Synthesis of Holocene Proxi Data and Climate Models, Eklund, H. (1965), Stability of lakes near the temperature of max-
edited by H. Fischer et al., pp. 109 – 128, Springer, Heidelberg, imum density, Science, 149, 632 – 633, doi:10.1126/science.149.
Germany. 3684.632.
Bührer, H., and H. Ambühl (1975), Die Einleitung von gereinigtem Fachbereichsstandard (1983), Nutzung und Schutz der Gewässer-
Abwasser in Seen, Schweiz. Z. Hydrol., 37(2), 347 – 368, Stehende Binnengewässer-Wassergütebewirtschaftung Seen,
doi:10.1007/BF02503411. DDR-Fachbereichsstandard TGL 27 885/03, Verlag für Standar-
Camacho, A., J. Erez, A. Chicote, M. Florin, M. M. Squires, disierung, Leipzig, Germany.
C. Lehmann, and R. Bachofen (2001), Microbial microstratifica- Fee, E. J., R. E. Hecky, S. E. M. Kaisan, and D. R. Cruikshank
tion, inorganic carbon photoassimilation and dark carbon fixation (1996), Effects of lake size, water clarity and climate variability
at the chemocline of the meromictic Lake Cadagno (Switzerland) on the mixing depths in Canadian shield lakes, Limnol. Ocea-
and its relevance to the food web, Aquat. Sci., 63(1), 91 – 106, nogr., 41(5), 912 – 920.
doi:10.1007/PL00001346. Feistel, R., and E. Hagen (1995), On the Gibbs thermodynamic
Campbell, P., and T. Torgersen (1980), Maintenance of iron mer- potential of seawater, Prog. Oceanogr., 36, 249 – 327,
omixis by iron redeposition in a rapidly flushed monimolimnion, doi:10.1016/0079-6611 (96)00001-8.
Can. J. Fish. Aquat. Sci., 37, 1303 – 1313. Feistel, R., and G. M. Marion (2007), A Gibbs-Pitzer function for
Carmack, E. C., and R. F. Weiss (1991), Convection in Lake Baikal: high-salinity seawater thermodynamics, Prog. Oceanogr., 74,
An example of thermobaric instability, in Deep Convection and 515 – 539, doi:10.1016/j.pocean.2007.04.020.
Deep Water Formation in the Oceans, edited by P. C. Chu and Findenegg, I. (1933), Alpenseen ohne Vollzirkulation, Int, Rev. Ge-
J. C. Gascard, pp. 215 – 228, Elsevier, Amsterdam. samten Hydrobiol., 28, 295 – 311, doi:10.1002/iroh.19330280309.
Casamitjana, X., and E. Roget (1993), Resuspension of sediment Findenegg, I. (1935), Limnologische Untersuchungen in Kärntner
by focused groundwater in Lake Banyoles, Limnol. Oceanogr., Seengebiete, Int. Rev. Gesamten Hydrobiol., 32, 369 – 423.
38(3), 643 – 656. Findenegg, I. (1937), Holomiktische und meromiktische Seen, Int.
Chen, C.-T. A., and F. J. Millero (1977), The use and misuse of Rev. Gesamten Hydrobiol., 35, 586 – 610, doi:10.1002/
pure water PVT properties for lake waters, Nature, 266, 707 – iroh.19370350130.
708, doi:10.1038/266707a0. Finger, D., M. Schmid, and A. Wüest (2006), Effects of upstream
Chen, C.-T. A., and F. J. Millero (1986), Precise thermodynamic hydropower operation on riverine particle transport and turbidity
properties for natural waters covering only the limnological in downstream lakes, Water Resour. Res., 42, W08429,
range, Limnol. Oceanogr., 31(3), 657 – 662. doi:10.1029/2005WR004751.
Cole, G. A. (1983), Textbook of Limnology, 3rd ed., Waveland, Fischer, H. B., E. J. List, R. C. Y. Koh, J. Imberger, and N. H.
Prospects Heights, Ill. Brooks (1979), Mixing in Inland and Coastal Waters, Academic,
Crawford, G. B., and R. W. Collier (1997), Observations of deep San Diego, Calif.
mixing in Crater Lake, Oregon, Limnol. Oceanogr., 42(2), 299 – Fisher, T. S. R. (2002), Limnology of the meromictic Island Copper
306. Mine pit lake, Ph.D. thesis, Univ. of B. C., Vancouver, Canada.
Crawford, G. B., and R. W. Collier (2007), Longterm observations Fisher, T. S. R., and G. A. Lawrence (2006), Treatment of acid rock
of hypolimnetic mixing in Crater Lake, Oregon, Hydrobiologia, drainage in a meromictic pit lake, J. Environ. Eng., 132(4), 515 –
574(1), 47 – 68, doi:10.1007/s10750-006-0345-3. 526, doi:10.1061/(ASCE)0733-9372(2006)132:4(515).
Crétaux, J.-F., A. V. Kouraev, F. Papa, B. Bergé-Nguyen, A. Cazenave, Fofonoff, N. P. (1985), Physical properties of seawater: A new
N. Aladin, and I. S. Plotnikov (2005), Evolution of sea level in the salinity scale and equation of state for seawater, J. Geophys.
big Aral Sea from satellite altimetry and its implications for water Res., 90(C2), 3332 – 3342, doi:10.1029/JC090iC02p03332.
balance, J. Great Lakes Res., 31(4), 520–534.
23 of 27
Fofonoff, N. P., and R. C. Millard Jr. (1983), Algorithms for com- (2002), Rapid deep-water renewal in Lake Issyk-Kul (Kyrgyzstan)
mutation of fundamental properties of seawater, UNESCO Tech. indicated by transient tracers, Limnol. Oceanogr., 47(4), 1210–
Pap. Mar. Sci., 44, 58 pp. 1216.
Forel, F. A. (1901), Handbuch der Seenkunde. Allgemeine Limno- Hohmann, R., M. Hofer, R. Kipfer, F. Peeters, D. M. Imboden,
logie, J. von Engelhorn, Stuttgart, Germany. H. Baur, and M. N. Shimarev (1998), Distribution of helium
Frey, D. G. (1955), Längsee: A history of meromixis, Mem. Ist. and tritium in Lake Baikal, J. Geophys. Res., 103(C6), 12823 –
Ital. Idrobiol., Suppl., 8, 141 – 164. 12,838, doi:10.1029/97JC02218.
Gaedke, U., D. Ollinger, P. Kirner, and E. Bäuerle (1998), The influ- Hollan, E. (1998), Idealized calculation of the horizontal circula-
ence of weather conditions on the seasonal plankton development tion in Lake Constance driven by deep density currents, in Third
in a large and deep lake (L. Constance). III. The impact of water Workshop on Physical Processes in Natural Waters, UFZ-Ber.
column stability on spring algal development, in Management of 23/1998, edited by B. Boehrer and M. Schimmele, pp. 71 – 75,
Lakes and Reservoirs During Global Climate Change, edited by UFZ-Umweltforschungszentrum Leipzig-Halle GmbH, Leipzig,
D. G. George et al., pp. 71 – 84, Springer, Berlin, Germany. Germany.
Gat, J. R. (1995), Stable isotopes of fresh and saline lakes, in Hollan, E. (1999), Erneuerung des Tiefenwassers im Bodensee
Physics and Chemistry of Lakes, edited by A.|Lerman et al., durch weitreichende see-eigene Dichteströmungen, in Tagungs-
pp. 139 – 165, Springer, Berlin, Germany. bericht 1998 (Klagenfurt), pp. 82 – 96, Dtsch. Ges. für Limnol.,
George, D. G., and D. P. Hewitt (2006), The impact of year-to-year Tutzing, Germany.
changes in the weather on the dynamics of Daphnia in a ther- Hongve, D. (1997), Cycling of iron, manganese, and phosphate in
mally stratified lake, Aquat. Ecol., 40(1), 33 – 47, doi:10.1007/ a meromictic lake, Limnol. Oceanogr., 42(4), 635 – 647.
s10452-005-9013-3. Hongve, D. (2002), Seasonal mixing and genesis of endogenic
Gibson, J. A. E. (1999), The meromictic lakes and stratified marine meromixis in small lakes in southeast Norway, Nord. Hydrol.,
basins of the Vestfold Hills, east Antarctica, Antarct. Sci., 11(2), 33(2 – 3), 189 – 206.
175 – 192, doi:10.1017/S0954102099000243. Hongve, D. (2004), Endogenic meromixis—Studies of Nordbytjer-
Gill, A. E. (1982), Atmosphere-Ocean Dynamics, Academic, New net and other meromictic lakes in the upper Romereike area,
York. Ph.D. thesis, Univ. of Oslo, Oslo, Norway.
Goldman, C. R., D. T. Mason, and B. J. B. Wood (1972), Com- Hürzeler, B. E., J. Imberger, and G. N. Ivey (1996), Dynamics of
parative study of the limnology of two small lakes on Ross turbidity current with reversing buoyancy, J. Hydrol. Eng., 122(5),
Island, Antarctica, in Antarctic Terrestrial Biology, Antarct. Res. 230 – 236, doi:10.1061/(ASCE)0733-9429(1996)122:5(230).
Ser., vol. 20, edited by G. A. Llano, pp. 1 – 50, AGU, Washington, Hutchinson, G. E. (1957), A Treatise on Limnology, vol. 1. Geo-
D. C. graphy, Physics and Chemistry, John Wiley, New York.
Gorham, E., and F. M. Boyce (1989), Influence of lake surface area Hutter, K. (1987), Hydrodynamic modelling of lakes, in Encyclo-
and depth upon thermal stratification and the depth of the sum- pedia of Fluid Mechanics, edited by N. P. Cheremesinoff, pp.
mer thermocline, J. Great Lakes Res., 15(2), 233 – 245. 897 – 998, Gulf, Houston, Tex.
Gräfe, H., and B. Boehrer (2001), Dichtebestimmung im Tagebau- Ilmberger, J., and von C. Rohden (2001), SF6 als Tracer für Trans-
see Niemegk (Goitsche), in Tagungsbericht 2000 (Magdeburg), port- und Mischungsprozesse in Tagebaurestseen, in Tagungs-
pp. 483 – 487, Dtsch. Ges. für Limnol., Tutzing, Germany. bericht 2000 (Magdeburg), pp. 503 – 507, Dtsch. Ges. für
Gräfe, H., B. Boehrer, N. Hoppe, S. C. Müller, and P. Hauptmann Limnol., Tutzing, Germany.
(2002), Ultrasonic measurements of density, adiabatic compressibil- Imberger, J., and J. C. Patterson (1990), Physical limnology, Adv.
ity and stability frequency, Limnol. Oceanogr., 47(4), 1255–1260. Appl, Mech., 27, 303 – 475.
Hakala, A. (2004), Meromixis as a part of lake evolution - obser- Imboden, D. M., and A. Wüest (1995), Mixing mechanisms in
vations and a revised classification of true meromictic lakes in lakes, in Physics and Chemistry of Lakes, edited by A. Lerman
Finland, Boreal Environ. Res., 9(1), 37 – 53. et al., pp. 83 – 138, Springer, Berlin, Germany.
Halbwachs, M., et al. (2004), Degassing the ‘‘Killer Lakes’’ Nyos International Organization for Standardization (ISO) (1985), Water
and Monoun, Cameroon, Eos Trans. AGU, 85(30), 281, quality: Determination of electrical conductivity, Standard 7888,
doi:10.1029/2004EO300001. Geneva, Switzerland.
Halfman, J. D. (1993), Water column characteristics from modern Jellison, R., and J. M. Melack (1993), Meromixis in hypersaline
CTD data, Lake Malawi, East Africa, J. Great Lakes Res., 19(3), Mono Lake, California. 1. Vertical mixing and density stratifica-
512 – 520. tion during the onset, persistence, and breakdown of meromixis,
Hamblin, P. F., C. L. Stevens, and G. A. Lawrence (1999), Simulation Limnol. Oceanogr., 38, 1008 – 1019.
of vertical transport in mining pit lake, J. Hydrol. Eng., 125(10), Jellison, R., J. Romero, and J. M. Melack (1998), The onset of
1029 – 1038, doi:10.1061/(ASCE)0733-9429(1999)125:10 (1029). meromixis during restoration of Mono Lake, California: Unin-
Hammer, U. T. (1994), Life and times of five Saskatchewan saline tended consequences of reducing water diversions, Limnol.
meromictic lakes, Int. Rev. Gesamten Hydrobiol., 79, 235 – 248, Oceanogr., 43(4), 706 – 711.
doi:10.1002/iroh.19940790209. Jellison, R., S. MacIntyre, and F. J. Millero (1999), Density and
Hausmann, J., and B. Boehrer (2006), Hypolimniontemperaturen conductivity properties of NA-CO3-Cl-SO4 brine from Mono
unter 4!C in einem Salzsee-das Solbad Leopoldshall bei Stassfurt Lake, California, USA, Int. J. Salt Lake Res., 8(1), 41 – 53,
in Sachsen-Anhalt, in Tagungsbericht 2005 (Karlsruhe), pp. 87 – doi:10.1007/BF02442136.
90, Dtsch. Ges. für Limnol., Berlin, Germany. Jöhnk, K. D. (1999), Meromixis aus physikalischer Perspektive, in
Heidenreich, H., B. Boehrer, R. Kater, and G. Hennig (1999), Tagungsbericht 1998 (Klagenfurt), pp. 82 – 96, Dtsch. Ges. für
Gekoppelte Modellierung geohydraulischer und limnophysika- Limnol., Tutzing, Germany.
lischer Vorgänge in Tagebaurestseen und ihrer Umgebung, Jöhnk, K. D. (2000), 1D hydrodynamische Modelle in der Limno-
Grundwasser, 4(2), 49 – 54, doi:10.1007/s767-1999-8604-4. physik—Turbulenz, Meromixis, Sauerstoff, Habilitationsschrift,
Heinz, G., J. Ilmberger, and M. Schimmele (1990), Vertical mixing Tech.Univ. Darmstadt, Darmstadt, Germany.
in Überlinger See, western part of Lake Constance, Aquat. Sci., Johnson, T. C., , and E. O. Odada (Eds.) (1996), The Limnology,
52(3), 256 – 268, doi:10.1007/BF00877283. Climatology and Paleoclimatology of the East African Lakes,
Hobbie, J. E. (1996), Polar limnology, in Lakes and Reservoirs, Gordon and Breach, New York.
edited by F. B. Taub, pp. 63 – 105, Elsevier, Amsterdam, Karakas, G., I. Brookland, and B. Boehrer (2003), Physical char-
Netherlands. acteristics of acidic Mining Lake 111, Aquat. Sci., 65(3), 297 –
Hofer, M., F. Peeters, W. Aeschbach-Hertig, M. Brennwald, 307, doi:10.1007/s00027-003-0651-z.
J. Holocher, D. M. Livingstone, V. Romanovski, and R. Kipfer
24 of 27
Kemmerer, G., J. F. Bovard, and W. R. Boorman (1924), North- determination of a conductivity- and temperature-dependent
western lakes of the United States. Biological and chemical stu- expression for salinity, Limnol. Oceanogr., 37, 4l – 53.
dies with reference to possibilities in production of fish, Bull. Merilainen, J. (1970), On the limnology of the meromictic
U. S. Bur. Fish., 39, 51 – 140. Lake Valkiajarvi, in the Finnish lake district, Ann. Bot. Fenn.,
Kerry, K. R., D. R. Grace, R. Williams, and H. R. Burton (1977), 7, 29 – 51.
Studies on some saline lakes of the Vestfold Hills, Antarctica, in Meybeck, M. (1995), Global dirstribution of lakes, in Physics and
Adaptations Within Antarctic Ecosystems, edited by G. A. S. Chemistry of Lakes, edited by A. Lerman, D. Imboden, and J. Gat,
Llano, pp. 839 – 858, Gulf, Houston, Tex. pp. 1 – 35, Springer, Berlin, Germany.
King, R. D., and P. A. Tyler (1981), Meromictic lakes of south- Miller, G. C., W. B. Lyons, and A. Davis (1996), Understanding pit
west Tasmania, Aust. J. Mar. Freshwater Res., 32(5), 741 – 756, water quality, Environ. Sci. Technol., 30(3), 118A – 123A.
doi:10.1071/MF9810741. Moncur, M. C., C. J. Ptacek, D. W. Blowes, and J. L. Jambor
Kirkland, D. W., J. Platt Bradbury, and W. E. Dean (1983), The (2006), Spatial variations in water composition at a northern
heliothermic lake - a direct method of collecting and storing solar Canadian lake impacted by mine drainage, Appl. Geochem.,
energy, Arch. Hydrobiol. Suppl., 65, 1 – 60. 21, 1799 – 1817, doi:10.1016/j.apgeochem.2006.06.016.
Kjensmo, J. (1967), The development and some main features of ‘‘iron- Nissenbaum, A. (1969), Studies in the geochemistry of the Jordan
meromictic’’ soft water lakes, Arch. Hydrobiol., 32, 137–312. River-Dead Sea system, Ph.D. thesis, Univ. of Calif., Los Angeles.
Kjensmo, J. (1968), The primary production and its influence on Osgood, R. A. (1988), Lake mixis and internal phosphorusdy-
the meromictic stability in Lake Svinsjøen, Schweiz. Z. Hydrol., namics, Arch. Hydrobiol., 113, 629 – 638.
30(2), 297 – 317, doi:10.1007/BF02505840. Patalas, K. (1961), Wind- und morphologiebedingte Wasserbewe-
Kjensmo, J. (1988), Post-glacial sediments and the stagnation his- gungstypen als bestimmender Faktor für die Intensität des Stoffk-
tory of the iron-meromictic Lake Skjennungen, eastern Norway, reislaufs in nordpolnischen Seen, Verh. Int. Ver. Theor. Angew.
Arch. Hydrobiol., 113, 481 – 499. Limnol., 14, 59 – 64.
Kjensmo, J. (1997), The influence of road salts on the salinity and the Patalas, K. (1984), Mid-summer mixing depths of lakes of different
meromictic stability of Lake Svinsjøen, southeastern Norway, Hy- latitudes, Verh. Int. Ver. Theor. Angew. Limnol., 22, 97 – 102.
drobiologia, 347(1 – 3), 151 – 158, doi:10.1023/A:1003035705729. Pedlosky, J. (1988), Geophysical Fluid Dynamics, Springer, Berlin,
Klapper, H. (1992), Eutrophierung und Gewässerschutz, G. Fischer, Germany.
Jena, Stuttgart, Germany. Peeters, F., D. Finger, M. Hofer, M. Brennwald, D. M. Livingstone,
Klaveness, D. (1990), Size structure and potential food value of the and R. Kipfer (2003), Deep-water renewal in Lake Issyk-Kul
plankton community to Ostrea edulis L. in a traditional Norwe- driven by differential cooling, Limnol. Oceanogr., 48(4),
gian ‘‘østerspoll,’’ Aquaculture, 86(2 – 3), 231 – 247, doi:10.1016/ 1419 – 1431.
0044-8486(90)90116-5. Peeters, F., R. Kipfer, D. Achermann, M. Hofer, W. Aeschbach-
Kling, G. W. (1988), Comparative transparency, depth of mixing Hertig, U. Beyerle, D. M. Imboden, K. Rozanski, and K. Fröh-
and stability of stratification in lakes of Cameroon, West Africa, lich (2000), Analysis of deep-water exchange in the Caspian Sea
Limnol. Oceanogr., 33(1), 27 – 40. based on environmental tracers, Deep Sea Res., Part I, 47(4),
Kodenev, G. G. (2001), Deep-water renewal in Lake Baikal, Geol. 621 – 654, doi:10.1016/S0967-0637(99)00066-7.
Geofiz., 42(7), 1127 – 1136. Pellicori, D. A., C. H. Gammons, and S. R. Poulson (2005), Geo-
Krüger, B., A. Kadler, and M. Fischer (2002), The creation of post- chemistry and stable isotope composition of the Berkeley pit lake
mining landscapes of lignite mining in the new federal states, and surrounding mine waters, Butte, Montana, Appl. Geochem.,
Surf. Min. Braunkohle Other Miner., 54(2), 161 – 169. 20(11), 2116 – 2137, doi:10.1016/j.apgeochem.2005.07.010.
Lampert, W., and U. Sommer (1993), Limnoökologie, Thieme, Pérez-Fuentetaja, A., P. J. Dillon, N. D. Yan, and D. J. McQueen
Stuttgart, Germany. (1999), Significance of dissolved organic carbon in the predic-
Landesanstalt für Umweltschutz Baden-Württemberg (2000), tion of thermocline depth in small Canadian shield lakes, Aquat.
Seenphysikalische Prozesse in Baggerseen, FfU Ber. 62/2000, Ecol., 33(2), 127 – 133, doi:10.1023/A:1009998118504.
Karlsruhe, Germany. Pettersson, O. (1902), Resultaten af den internationella undersökin-
Lemmin, U. (1995), Limnologie physique, in Limnologie Générale, gen af norra Europas djupa sjöar och innanhaf ar 1900, K. Sven.
edited by R. Pourriot and M. Meybeck, pp. 61 – 106, Masson, Vetenskapsakad. Handl., 28(2), 2.
Paris. Pickard, G., and W. J. Emery (1982), Descriptive Oceanography,
Lerman, A., D. Imboden, and J. Gat (Eds.) (1995), Physics and Pergamon, Oxford, UK.
Chemistry of Lakes, 2nd ed., 334 pp., Springer, Berlin, Germany. Pohl, C., A. Löffler, and U. Hennings (2000), A sediment trap
Létolle, R., and M. Mainguet (1993), Aral, Springer, Paris. fluxstudy for trace metals under seasonal aspects in the stratified
Lewis, W. M., Jr. (1983), A revised classification of lakes based on Baltic Sea (Gotland Basin, 57819,20VN, 20803,00VE), Mar.
mixing, Can. J. Fish. Aquat. Sci., 40, 1779 – 1787. Chem., 84(3 – 4), 143 – 160, doi:10.1016/j.marchem.2003.07.002.
Lewis, W. M., Jr. (1987), Tropical limnology, Annu. Rev, Ecol. Pond, S., and G. L. Pickard (1983), Introductory Dynamical Ocea-
Syst., 18, 159 – 184, doi:10.1146/annurev.es.18.110187.001111. nography, Pergamon, Oxford, UK.
Lorke, A., K. Tietze, M. Halbwachs, and A. Wüest (2004), Por, F. D. (1972), Hydrobiological notes on the high-salinity waters
Response of Lake Kivu stratification to lava inflow and cli- of the Sinai Peninsula, Mar. Biol., 14(2), 111 – 119, doi:10.1007/
mate warming, Limnol. Oceanogr., 49(3), 778 – 783. BF00373210.
Ludlam, S. D. (1996), The comparative limnology of high arctic, Pourriot, R., and M. Meybeck (Eds.) (1995), Limnologie Générale,
coastal, meromictic lakes, J. Paleolimnol., 16(2), 111 – 131, Masson0, Paris, France.
doi:10.1007/BF00176931. Ravens, T. M., O. Kocsis, A. Wüest, and N. Granin (2000), Small-
Ludlam, S. D., and B. Duval (2001), Natural and management- scale turbulence and vertical mixing in Lake Baikal, Limnol.
induced reduction in monimolimnetic volume and stability in a Oceanogr., 45(1), 159 – 173.
coastal, meromictic lake, Lake Reserv. Manage., 17(2), 71 – 81. Rawson, D. S., and J. E. Moore (1944), The saline lakes of Sas-
Margalef, R. (1994), Limnology Now: A Paradigm of Planetary katchewan, Can. J. Res., Sect. D, 22, 141 – 201.
Problems, Elsevier, Amsterdam, Netherlands. Reif, F. (1987), Statistische Physik und Theorie der Wärme, De
Mazumder, A., and W. D. Taylor (1994), Thermal structure of Gruyter, Berlin, Germany.
lakes varying in size and water clarity, Limnol. Oceanogr., Rodhe, W. (1949), The ionic composition of lake waters, Verh. Int.
39(4), 968 – 976. Ver. Theor. Angew. Limnol., 10, 377 – 386.
McManus, J., R. Collier, T. A. C.-Chen, and J. Dymond (1992),
Physical properties of Crater Lake, Oregon: A method for the
25 of 27
Rodrigo, M. A., M. R. Miracle, and E. Vicente (2001), The mer- Strøm, K. M. (1945), Lakes with stagnant deeps. The meromictic
omictic Lake La Cruz (central Spain), Patterns of stratification, lake type in Norway and the question of measuring productivity
Aquat. Sci., 63(4), 406 – 416, doi:10.1007/s00027-001-8041-x. by hypolimnetic oxygen deficit, Skr. Nor, Vidensk. Akad. Kl. 1
Rossknecht, H. (2003), Der Alpenrhein im Bodensee: Chemische Mat. Naturvidensk. Kl., 7, 1 – 16.
Indikatoren zur Aufklärung physikalischer Prozesse im Boden- Strøm, K. M. (1963), Trapped sea water, New Sci, 274, 384 – 386.
see-Obersee, Landesanstalt für den Umweltschutz Baden-Würt- Taub, F. B. (Ed.) (1984), Lakes and Reservoirs, Elsevier, Amster-
temberg, Karlsruhe, Germany. dam, Netherlands.
Rücker, J., A. Fyson, R. Deneke, and G. Packroff (1999), Meromik- Thackeray, S. J., D. G. George, R. I. Jones, and I. J. Winfield
tische Seen mit saurem Mixolimnion - eine Besonderheit der Lausit- (2006), Statistical quantification of the effect of thermal stratifi-
zer Bergbaufolgelandschaft, in Tagungsbericht 1998 (Klagenfurt), cation on patterns of dispersion in a freshwater zooplankton
pp. 239–242, Dtsch. Ges. für Limnol., Tutzing, Germany. community, Aquat. Ecol., 40(1), 23 – 32, doi:10.1007/s10452-
Ruttner, F. (1940), Grundriss der Limnologie (Hydrobiologie des 005-9021-3.
Süsswassers), Walter de Gruyter, Berlin, Germany. Tilzer, M. M., and C. Serruya (Eds.) (1990), Large Lakes: Ecolo-
Sanderson, B., K. Perry, and T. Pedersen (1986), Vertical diffusion gical Structure and Function, Springer, Berlin, Germany.
in meromictic Powell Lake, British Columbia, J. Geophys. Res., Tonolla, M., S. Peduzzi, D. Hahn, and R. Peduzzi (2003), Spatio-
91(C6), 7647 – 7655, doi:10.1029/JC091iC06p07647. temporal distribution patterns of phototrophic sulfur bacteria in
Scharf, B. W., and U. Menn (1992), Hydrology and morphometry, the chemocline of meromictic Lake Cadagno (Switzerland),
Arch, Hydrobiol., 38, 43 – 62. FEMS Microbiol. Ecol., 43(1), 89 – 98, doi:10.1111/j.1574-
Scharf, B. W., and M. Oehms (1992), Physical and chemical char- 6941.2003.tb01048.x.
acteristics, Arch. Hydrobiol., 38, 63 – 83. van Niel, C. B. (1936), Les Photosynthèses Bactériennes, Bull.
Schimmele, M., and P. Herzsprung (2000), Limnology of sulfur- Assoc. Diplomes Microbiol. Fac. Pharm. Nancy, 13, 3 – 18.
acidic lignite mining lakes. I. Physical properties: Influence of Ventz, D. (1972), Die Einzugsgebietsgröße, ein Geofaktoer für den
dissolved substances on electrical conductivity and density, Verh. Trophiezustand stehender Gewässer, Fortschr. Wasserchem.
Int. Ver. Theor. Angew. Limnol., 27, 251 – 255. Ihrer Grenzgeb., 14, 105 – 118.
Schlesinger, W. H. (Ed.) (2005), Treatise on Geochemistry, vol. 8, Vollmer, M. K., R. F. Weiss, and H. A. Bootsma (2002a), Ventilation
Biogeochemistry, Elsevier, Oxford, UK. of Lake Malawi/Nyasa, in The East African Great Lakes: Limnol-
Schmid, M., K. Tietze, M. Halbwachs, A. Lorke, D. McGinnis, and ogy, Paleolimnology and Biodiversity, edited by E. O. Odada and
A. Wüest (2004), How hazardous is the gas accumulation in D. O. Olago, pp. 209– 233, Kluwer, Dordrecht, Netherlands.
Lake Kivu? Arguments for a risk assessment in light of the Vollmer, M. K., R. F. Weiss, R. T. Williams, K. K. Falkner, X. Qiu, E. A.
Nyiragongo Volcano eruption of 2002, Acta Vulcanol., 14/ Ralph, and V. V. Romanovsky (2002b), Physical and chemical prop-
15(2002 – 3), 115 – 122, doi:10.1400/19084. erties of the waters of saline lakes and their importance for deep-water
Schönborn, W. (2003), Lehrbuch der Limnologie, Schweizerbart, renewal: Lake Issyk-Kul, Kyrgyzstan, Geochim. Cosmochim. Acta,
Stuttgart, Germany. 66(24), 4235–4246, doi:10.1016/S0016-7037(02)01052-9.
Schwoerbel, J. (1999), Einführung in die Limnologie, 6th ed., von Rohden, C., and J. Ilmberger (2001), Tracer experiment with
Gustav Fischer, Stuttgart, Germany. sulfurhexafluoride to quantify the vertical transport in a mero-
Simpson, J. J., and T. D. Dickey (1981a), The relationship between mictic pit lake, Aquat. Sci., 63(4), 417 – 431, doi:10.1007/
downward irradiance and upper ocean structure, J. Phys. Oceanogr., s00027-001-8042-9.
11(3), 309 – 323, doi:10.1175/1520-0485(1981)011<0309: Walker, K. F., and G. E. Likens (1975), Meromixis and a recon-
TRBDIA>2.0.CO;2. sidered typology of lake circulation patterns, Verh. Int. Ver. The-
Simpson, J. J., and T. D. Dickey (1981b), Alternative parameteriza- or. Angew. Limnol., 19, 442 – 458.
tions of downward irradiance and their dynamical significance, Weiss, R. F., E. C. Carmack, and V. M. Koropalov (1991), Deep-
J. Phys. Oceanogr., 11(6), 876 – 882, doi:10.1175/1520- water renewal and biological production in Lake Baikal, Nature,
0485(1981)011<0876:APODIA>2.0.CO;2. 349, 665 – 669, doi:10.1038/349665a0.
Snoeyink, V. L., and D. Jenkins (1980), Water Chemistry, John Wetzel, R. G. (2001), Limnology: Lake and River Ecosystems, 3rd
Wiley, New York. ed., Academic, San Diego, Calif.
Sorensen, J. A., and G. E. Glass (1987), Ion and temperature Williams, W. D. (1994), Definition and measurement of salinity in salt
dependence of electrical conductance for natural waters, Anal. lakes, Int. J. Salt Lake Res., 3(1), 53 – 63, doi:10.1007/BF01990642.
Chem., 59(13), 1594 – 1597, doi:10.1021/ac00140a003. Williams, W. D. (1996), The largest, highest and lowest lakes of
Sorokin, Y. I. (1970), Interrelations between sulfur and carbon the world: Saline lakes, Verh. Int. Ver. Theor. Angew. Limnol., 26,
turnover in a meromictic lake, Arch. Hydrobiol., 66, 391 – 446. 61 – 79.
Stellmacher, R. (2004), Der Moritzteich bei Döbern—Analogon Williams, W. D. (1998), Management of Inland Saline Waters,
für die Sanierung saurer Bergbaugewässer? Hydro- und Sedi- Guidelines Lake Manage. Ser., vol. 6, Int. Lake Environ. Comm.
mentgeochemie eines meromiktischen Restsees, Master thesis, Found., Shiga, Japan.
Tech. Univ. Bergakad. Freiberg, Freiberg, Germany. Williams, W. D. (1999), Salinisation: A major threat to water re-
Stevens, C. L., and G. A. Lawrence (1997), The effect of subaqu- sources in the arid and semi-arid regions of the world, Lakes
eous disposal of tailings in standing waters, J. Hydraul. Res., Reservoirs Res. Manage., 4(3 – 4), 85 – 91, doi:10.1046/j.1440-
35(2), 147 – 159. 1770.1999.00089.x.
Stevens, C. L., and G. A. Lawrence (1998), Stability and meromixis Wissenschaftlich-technische Werkstätten (1993), Leitfähigkeits-
in a water-filled mine pit, Limnol. Oceanogr., 43(5), 946– 954. Fibel, Weilheim, Germany.
Stewart, K. M., and E. Hollan (1975), Meromixis in Ulmener Maar Wüest, A., G. Piepke, and J. D. Halfmann (1996), Combined effects of
(Germany), Verh. Int. Ver. Theor. Angew. Limnol., 19, 1211 – 1219. dissolved solids and temperature on the density stratification of
Stottmeister, U., and E. Weißbrodt (2000), ‘‘Enhanced Bioattenua- Lake Malawi (East Africa), in The Limnology, Climatology and
tion’’: Anwendung natürlicher Prozesse zur Sanierung carbo- Paleoclimatology of the East African Lakes, edited by T. C. Johnson
chemischer Altlasten, TerraTech, 45 – 48. and E. O. Odada, pp. 183 – 202, Gordon and Breach, New York.
Stottmeister, U., E. Weißbrodt, P. M. Becker, J. Pörschmann, F. D. Wüest, A., T. M. Ravens, N. G. Granin, O. Kocsis, M. Schurter,
Kopinke, G. M. Martius, A. Wießner, and C. Kennedy (1998), and M. Stur (2005), Cold intrusions in Lake Baikal—Direct ob-
Analysis, behaviour and fate of a lignite pyrolysis wastewater servational evidence for deep water renewal, Limnol. Oceanogr.,
deposit, in Contaminated Soil 98, pp. 113 – 121, Thomas Telford, 50(1), 184 – 196.
London.
26 of 27
Yoshimura, S. (1936a), A contribution to the knowledge of deep Zenger, A., J. Ilmberger, G. Heinz, M. Schimmele, P. Schlosser,
water temperatures of Japanese lakes. part 1. Summer tempera- D. Imboden, and K. O. Münnich (1990), The behavior of a medium-
ture,, Jpn. J. Astron. Geophys., 13, 2. sized basin connected to a large lake. The case of Überlingersee and
Yoshimura, S. (1936b), Deep water temperatures of lakes of Japan Obersee, in Large Lakes: Ecological Structure and Function, edited
in winter, Sea Air, 15, 195 – 208. by M. M. Tilzer and C. Serruya, pp. 133– 155, Springer, Berlin,
Yoshimura, S. (1937), Abnormal thermal stratifications of inland Germany.
lakes, Proc. Imp. Acad., 13, 316 – 319.
Zauke, G.-P., R. G. Niemeyer, and K.-P. Gilles (1995), Tropen, in !!!!!!!!!!!!!!!!!!!!!!!!
!
Handbuch Angewandte Limnologie, edited by C. E. W. Steinberg B. Boehrer and M. Schultze, Helmholtz Centre for Environmental
et al., Chapter VIII-3.2, pp. 1 – 86, ecomed, Landsberg, Germany. Research – UFZ, Brueckstrasse 3a, D-39114, Magdeburg, Germany.
(bertram.boehrer@ufz.de; martin.schultze@ufz.de)
27 of 27
View publication stats

Boehrer - Stratification of Lakes, 2008

Uploaded by

Copyright:

Available Formats

Boehrer - Stratification of Lakes, 2008

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Boehrer - Stratification of Lakes, 2008

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Article in Reviews of Geophysics · May 2008

Bertram Boehrer Martin Schultze

SEE PROFILE SEE PROFILE

Ecology of Meromictic Lakes View project

The user has requested enhancement of the downloaded file.

[14] Epilimnion and atmosphere exchange heat and vol-

decomposition of organic material can cause a depletion of

[27] In the absence of any clear boundary, lakes are

the thermal contribution. Some lakes show extreme salinity

sufficiently deep, gravel pit lakes in the upper Rhine Valley

CH2 O þ O2 ! CO2 þ H2 O; ð2aÞ

CH2 O þ 2MnO2 þ 4Hþ ! CO2 þ 2Mn2þ þ H2 O; ð2cÞ

CH2 O þ 4FeðOHÞ3 þ 8Hþ ! CO2 þ 4Fe2þ þ 11H2 O; ð2dÞ

Figure 11. Profiles of partial pressures of dissolved gases Sulphate reduction

as performed by green algae and cyanobacteria. A similar

4Fe2þ þ O2 þ 4Hþ ! 4Fe3þ þ 2H2 O ð5aÞ

Fe3þ þ 3H2 O ! FeðOHÞ3 þ 3Hþ : ð5bÞ

Under favorable conditions a portion of this iron can be

responsible for forming a monimolimnion. In addition, if

[ 62 ] From tracer measurements (chlorofluorocarbon

while it mixes with surrounding water (thermobaric insta-

5. QUANTITATIVE APPROXIMATIONS FOR

[69] For evaluation of stratification related magnitudes

[77] On the contrary, positioned senders and receivers can

[74] Hence the sound propagation is dependent on tem- gaðT þ 273:15Þ

[82] For the hypolimnion water of highly acidic Mining

[83] For better clarity, we dropped the unit K!1. Equation

ei ¼ 0; 2:070 " 10!5 ; !6:370 " 10!10 ; 3:989 " 10!15 ;

[98] In cases where salinity cannot be used, calculation of

They used 20!C as reference for the conductance and

g Tpot ¼ ! 7:522 " 10!6 Tpot

where c is the mass of suspended material per unit water

You might also like