The floors of two shallow endorheic lakes, located on volcanic surfaces on James Ross Island, are covered with calcareous organosedimentary structures. Their biological and chemical composition, lake water characteristics, and seasonal variability of the thermal regime are introduced. The lakes are frozen down to the bottom for 8–9 months a year and their water chemistry is characterised by low conductivity and neutral to slightly alkaline pH. The photosynthetic microbial mat is composed of filamentous cyanobacteria and microalgae that are considered to be Antarctic endemic species. The mucilaginous black biofilm is covered by green spots formed by a green microalga and the macroscopic structures are packed together with fine material. Thin sections consist of rock substrate, soft biofilm, calcite spicules and mineral grains originating from different sources. The morphology of the spicules is typical of calcium carbonate monocrystals having a layered structure and specific surface texture, which reflect growth and degradation processes. The spicules' chemical composition and structure correspond to pure calcite. The lakes' age, altitude, morphometry, geomorphological and hydrological stability, including low sedimentation rates, together with thermal regime predispose the existence of this community. We hypothesise that the precipitation of calcite is connected with the photosynthetic activity of the green microalgae that were not recorded in any other lake in the region. This study has shown that the unique community producing biogenic calcite spicules is quite different to any yet described.
The floors of most Antarctic lakes are covered with photosynthetic microbial
mats (Vincent and Laybourn-Parry, 2008). However, the degree of disturbance
plays a key role in the development of microbial mats. When growing in
low-disturbance habitats, interactions between benthic microbial communities
and their environments can produce complex emergent structures. Such
structures are best developed in extreme environments, including benthic
communities of deep, perennially ice-covered Antarctic lakes, where physical
and chemical conditions and/or geographical isolation preclude the
development of larger organisms that could otherwise disrupt organised
microbial structures (Wharton, 1994; Andersen et al., 2011). Many
organosedimentary structures that emerge in these conditions are laminated
and accrete through episodic trapping of sediments or grains and
precipitation of minerals within a growing biogenic matrix (e.g. Arp et al.,
2001; Reid et al., 2003). In perennially ice-covered lakes, the seasonality
of growth imposed by the summer–winter light–dark conditions can induce
annual growth laminations (Hawes et al., 2001), reinforced by calcite
precipitation during growth and sediment diagenesis (Wharton et al., 1982;
Wharton, 1994; Sutherland and Hawes, 2009). Calcite precipitation is not,
however, a prerequisite for laminated, stromatolite-like communities (Walter,
1976; Schieber, 1999; Yamamoto et al., 2009). A diversity of micro- to
nanostructured CaCO
The precipitation of calcite by expulsion (segregation) is also a common process
in nature related to the freezing of common low ionic strength Ca
James Ross Island belongs to a transitory zone between the maritime and
continental Antarctic regions (Øvstedal and Lewis Smith, 2001). Air
temperature records indicate progressive warming trends from 1.5
to 3.0
Location of lakes 1 and 2 and air temperature measurements (AWS) in the Solorina Valley. The upper inset map shows the position of the study site and James Ross Island in the northeastern part of Antarctic Peninsula.
Bathymetric parameters of lakes 1
During two Czech research expeditions (2008 and 2009) to James Ross Island, lake ecosystems of the Ulu Peninsula were studied in respect to their origin, morphometry, physical, chemical and biological characteristics (Nedbalová et al., 2013), together with detailed cyanobacterial and microalgal diversity descriptions (Komárek and Elster, 2008; Komárek et al., 2012, 2015; Kopalová et al., 2013; Škaloud et al., 2013). As part of this study, we encountered 1–5 mm-scale calcareous organosedimentary structures on the floor of two endorheic lakes, 1 and 2, which are quite different to any microbially mediated structures yet described from modern environments. These shallow lakes on higher-lying levelled surfaces originated after the deglaciation of volcanic mesas which became ice-free some 6.5–8 ka ago (Johnson et al., 2011) and are considered among the oldest in the region. However, a later appearance of these lakes is also possible, as we have no exact dates from their sediments (Nedbalová et al., 2013).
The aim of this paper is to describe in detail the chemical and biological composition of the organosedimentary structures together with the limnological characteristics of the two lakes. A hypothesis concerning the formation of calcite spicules is also presented. The results of this study can serve as a baseline for understanding the microbial behaviours in forming these organosedimentary structures, which will provide insight into the interpretation of fossil forms from early Earth.
Endorheic lake 1 (63
Climate conditions of the Ulu Peninsula are characterised by mean annual air
temperatures around
Lake 1 was sampled on 22 February 2008. In 2009, lake 1 was sampled on 5
January, lake 2 on 12 January. Air temperature at 2 m above ground was
measured by an automatic weather station (AWS) located nearby (Fig. 1).
Incident global solar radiation was monitored with a LI-200 pyranometer
(LI-COR, USA) at Mendel Station, located 11 km northwest of the study site
(Fig. 1). The LI-200 spectral response curve covers wavelengths from 400 to
1100 nm with absolute error typically of
Dominant species in the photoautotrophic mats.
Conductivity, pH, temperature and dissolved oxygen were measured in situ with
a portable meter (YSI 600) at the time the lakes were ice free. Water samples
were collected from the surface layer, immediately filtered through a
200
Thin-section analyses were made to observe both rock substrate and inorganic
particles within biofilms. Dry microbial mat were saturated with epoxy
resins in vacuum, subsequently cut perpendicularly and saturated again with
epoxy resin. The sample was cemented to a glass slide after grinding and
polishing, and a thin section was prepared by final sectioning, grinding and
polishing to a desired thickness of 50–55
The morphology of photoautotrophic mats and calcareous spicules was studied using standard methods of scanning electron microscopy (SEM) using back-scattered electrons (BSE) (Jeol JSM-6380, Faculty of Science, Charles University) and optical microscopy (Nikon SMZ-645 using NIS-Elements software). Calcareous spicules were collected directly from the surface of biofilms. Samples studied in SEM were completely dried for 5 months at room temperature, then mounted on stubs with carbon paste and coated with gold prior to photomicrographing.
Photoautotrophic mats in lakes 1 and 2.
The chemical composition of the analysed spicules was measured by using the
Link ISIS 300 system with 10 mm
Pictures and detailed bathymetric parameters of both lakes together with marked lines of water level and the maximum extent of the photosynthetic microbial mat littoral belt in lake 1 are presented in Fig. 2.
The physico-chemical characteristics of the lake water for both lakes are
given in Table 1. The sampling of lake 1 (pH 7.4–7.9, saturation of oxygen
98.9 %) was performed during cloudy days. Oxygen supersaturation
(128 %) together with a relatively high pH (8.6) was observed in lake 2
during a sunny day. Conductivity was below 100
Physico-chemical characteristics and chlorophyll
Figure 3a shows the annual variation of daily mean water temperature in lake
1 and of daily mean air temperature in the Solorina Valley (locations of
temperature sensors are marked in Figs. 1 and 2). Lake 1 was frozen to the
bottom from the end of March to the end of October or beginning of November.
Air temperatures were frequently lower than water temperatures. Minimum daily
mean temperatures on the bottom of the lake were about
The highest night–day air temperature fluctuations (up to 28
The course of global solar radiation (Fig. 3c) was smooth, with the
maximum daily mean of 385 W m
The relative frequency of hourly values of lake 1 water and air temperature
is shown in Fig. S1 in the Supplement. Water temperature fluctuation was narrow, ranging
from
Photoautotrophic mat covering a stone visualised using imaging
fluorometry.
SEM macrographs showing the structure of the dried mat in the lakes:
Perpendicular thin sections of rock substrate covered by dry
biofilms (recorded under cross-polarised light. Note that biofilms are partly detached from the surface of
the rock due to complete drying of the sample.
The temperature at the lake bottom was permanently below
The littoral benthic community in lakes 1 and 2 are dominated by the
heterocytous cyanobacterium
Scanning electron micrographs document the structure of the biofilm (Fig. 7).
Figure 7a shows a lateral view (cross section) of a biofilm with
cyanobacterial filaments (
Thin sections, showing both dry biofilms and rock substrate (Fig. 8), provided information on various inorganic compounds associated with the soft tissue of the cyanobacterial–microalgal community. These inorganic compounds are represented by (1) allochthonous mineral grains that are overgrown and incorporated by biofilms and (2) calcareous spicules of different sizes ranging from 0.5 mm to 1 cm that are precipitated within the cyanobacterial-microalgal community.
SEM macrographs showing the morphology of calcium carbonate
spicules. Spicules were washed away from the living tissue and collected
directly from the surface of biofilms, although residence time on the bottom
cannot be determined.
FSD image of a transversely sectioned, partly recrystallised calcite
spicule acquired in
The rock substrate of biofilms is formed by subangular to subrounded pebbles to boulders of basaltic rock, which is dark grey in colour, compact and usually with a microcrystalline porhyric texture. The rock is not homogeneous, but contains numerous ball-like empty voids, which are often partly filled with feldspathoids (Fig. 8a). Crystals of plagioclase (feldspar group) and augite (pyroxene group) are easily recognisable in thin sections (Fig. 8a–c).
Biofilms are often partly covered with various mineral grains and rock fragments, but all specimens studied also contain these particles incorporated directly within soft cyanobacterial–microalgal biofilm (Fig. 8a–c).
Mineral grains embedded within biofilms close to the basaltic rock surface are mainly angular to subangular crystal fragments of plagioclase and augite (Fig. 8b, c), i.e. the main mineral components of the basaltic rock substrate described above. In the upper part of biofilms, however, partly or fully incorporated grains of quartz occur, being typically rounded or partly rounded (Fig. 8a, b). One of the thin sections shows a calcareous spicule in situ and mineral grains within the biofilm (Fig. 8c, d).
The structure and morphology of calcareous spicules was studied on SEM (Fig. 9). Crystal facets on the surface and cleavage (crystallographic structural planes) in the interior of the spicules (Fig. 9a, b) are typical characteristics of calcium carbonate monocrystals.
The superficial layer of microcrystalline calcite (e.g. Fig. 9b) shows the structure of parallel needle-like calcite microcrystals (Fig. 9d–f). Partial dissolution of spicules show distinct layering of these needle-like microcrystals (Fig. 9d). The layered structure of the spicules is confirmed in the ring-like structures with a possible cyanobacterial filament in the centre (Fig. 10).
The chemical composition of the studied calcareous spicules determined by
FSD corresponds to pure CaCO
The lakes under study are characterised by a low content of major ions due to their volcanic bedrock and lower marine influence. In comparison with other lakes of this area, the two lakes show no specific lake water chemistry characteristics, with moderate SRP and nitrate concentrations frequently below the detection limit (Nedbalová et al., 2013). High pH together with oxygen supersaturation recorded in lake 2 could be associated with high photosynthetic activity of the mats at the time of sampling.
Because water in either liquid or solid form has a large heat storage
capacity, it acts as an important buffer to temperature change. Local
climatic conditions of shallow freshwater lakes is the principal external
factor controlling their ecological functionality. Lake 1 is frozen to the
bottom approximately 8–9 months per year. For most of the year,
however, the temperature of the littoral and lake bottom is only from
In regards to heat balance, the studied shallow lakes are pond (wetlands) environments which freeze solid during the winter. This inevitability is a strong habitat-defining characteristic, which places considerable stress on resident organisms (Hawes et al., 1992; Elster, 2002). In summer, they must withstand drying in large parts of the littoral zone due to a considerable drop in water level. In freezing and desiccation resistance studies of freshwater phytobenthos in shallow Antarctic lakes, several ecological measurements have recorded seasonal, diurnal and year-round temperature fluctuations and changes in water state transitions (e.g. Davey, 1989; Hawes et al., 1992, 1999). In localities with steady moisture and nutrient supplies, the abundance and species diversity of algae is relatively high. However, as the severity and instability of living conditions increases (mainly due to changes in mechanical disturbances, desiccation–rehydration and subsequent changes in salinity), algal abundance and species diversity decreases (Elster and Benson, 2004). The speed at which the water state can change between liquid, ice and complete dryness is one of the most important ecological and physiological factors of these lakes. Studies based on field or laboratory experiments have shown that some cyanobacteria and algae are able to tolerate prolonged periods of desiccation (Pichrtová et al., 2014; Tashyreva and Elster, 2015). It is also obvious that there are strain/species-specific differences in the overwintering strategies, and also between strains/species inhabiting different habitats (Davey, 1989; Hawes et al., 1992; Jacob et al., 1992; Šabacká and Elster, 2006; Elster et al., 2008). The ice and snow which cover the lakes for about 8–9 months per year serve as a natural incubator, which moderates potential mechanical disturbances and stabilises the thermal regimes of the lakes.
Patterns of endemism and alien establishment in Antarctica are very different across taxa and habitat types (terrestrial, freshwater or marine) (Barnes et al., 2006). Environmental conditions, as well as dispersal abilities, are important in limiting alien establishment (Barnes et al., 2006). Antarctic microbial (cyanobacteria, algae) diversity is still poorly known, although recent molecular and ecophysiological evidence support a high level of endemism and speciation/taxon distinctness (Taton et al., 2003; Rybalka et al., 2009; De Wever et al., 2009; Komárek et al., 2012; Strunecký et al., 2012; Škaloud et al., 2013).
The floors of the studied lakes are covered with photosynthetic microbial
mats composed of previously described species of heterocytous cyanobacteria,
mostly
The black leather-like biofilm with mucilaginous marble on its surface is covered by green spots. These macroscopic structures form mats a few millimetres thick consisting of the above-mentioned species packed in mucilage glued together with fine material. The regular leather biofilm structure with distinct cyanobacterial–microalgal composition and incorporated mineral grains is to our knowledge unique. During the limnological survey of the whole Ulu Peninsula (Nedbalová et al., 2013), this specific biofilm structure was observed only in the two endorheic lakes, although lakes with very similar morphometric and chemical characteristics are found in the area. The mat structure is thus apparently tightly linked to the species composition (Andersen et al., 2011).
The low abundance of benthic diatoms in the lakes is unusual, but not unprecedented, as there are other areas in Antarctica where diatoms are scarce or absent (Broady, 1996, Wagner et al., 2004). The reason underlying the absence of diatoms is not immediately obvious, because diatoms are quite a common and frequently dominant component of microbial communities in most freshwater habitats of the Ulu Peninsula, James Ross Island (Kopalová et al., 2013). Local geographical separation of lakes 1 and 2 together with founder effect may have precluded successful colonisation by the subset of diatoms that are common in the surrounding freshwater habitats. Although it has long been held that diatoms are dispersed widely, some recent reports document very small scale microbial distributions and endemism (Kopalová et al., 2012, 2013).
Based on the character of the rock substrate and lake sediments, it is suggested that two of the main prerequisites for existence of this cyanobacterial–microalgal community producing unusual biogenic calcite structures are (1) the flat and stable substrate in both lakes and (2) the low sedimentation rate.
The substrate for biofilms is composed of boulders and pebbles of the stony littoral zone, petrographically corresponding to compact and massive basaltoids (Smellie et al., 2008; Svojtka et al., 2009). Rounded or subrounded quartz grains that are incorporated (“trapped”) within biofilms cannot originate from basaltic volcanic rocks forming the bottom of both lakes and substrate of the studied biofilms. This is evidenced by the petrographic character of the basaltoids, which do not contain any quartz. The presence of abraded quartz grains in lakes 1 and 2 can be easily explained by wind transport (e.g. Shao, 2008).
The specific cyanobacterial–microalgal community described above can prosper in the two shallow endorheic lakes, because of low sedimentation rates resulting from minor water input. Low sedimentary input is the main necessary ecological parameter which facilitates the existence of this special microbial community. The community is, however, well adapted to seasonally elevated sedimentation rates coming from frequent and intense winds. During wind storms, the wind carries a relatively large amount of small mineral grains and rock microfragments (intense eolic erosion; e.g. Shao (2008) and references therein). These grains and particles are usually derived from erosion of the rocks either in the very close vicinity of the locality (weathering of basaltic rocks), but mainly come from remote locations where especially Upper Cretaceous marine sedimentary sequences are outcropping (Smellie et al., 2008; Svojtka et al., 2009). The amount of mineral grains transported into the lake by wind does not stop the growth of cyanobacterial–microalgae biofilms, due to their ability of incorporating and “trapping” mineral grains within the living tissue (Riding, 2011).
This study has shown that inorganic substances precipitated by microbial
lithogenetic processes are exclusively represented by calcite spicules.
Precipitation of carbonate outside of microorganisms during photosynthesis
as a mechanism of carbonate construction was described for many filamentous
cyanobacterial species (Schneider and Le Campion-Alsumard, 1999). However,
the biogenic calcite structures in both lakes are quite different to any
microbially mediated structures yet described from modern environments
(Kremer et al., 2008; Couradeau et al., 2011) and also to structures formed
by abiotic precipitation (e.g. Vogt and Corte, 1996). Although there are
many lakes with thick mats and similar chemical characteristics on the Ulu
Peninsula, the calcite spicules were found exclusively in the two endorheic
lakes. We believe that their formation is linked to the specific
photoautotrophic mats present in the lakes. From Fig. 6g and h it is clearly
visible that the calcareous organosedimentary structures keep the contours of a
viable photosynthetic microbial mat after desiccation or calcite spicules
precipitation. More specifically, the co-dominance of a green microalga is
unique since mats in Antarctic lakes are most frequently formed by
filamentous cyanobacteria (Vincent and Laybourn-Parry, 2008). Therefore, we
hypothesise that the more rapid photosynthesis rate of
Although we interpret the tubular hollow observed in the centre of some spicules as the result of the presence of cyanobacterial filament during the process of crystallisation, such structures may form also as the result of abiotic precipitation of calcite (Vogt and Corte, 1996; Fan and Wang, 2005).
It is striking that some calcite spicules probably exhibit recrystallisation, forming spicules with the structure of calcite monocrystals. However, these spicules could be also interpreted as primary structures: mesostructured carbonate crystals formed through highly oriented growth of micro/nanocrystals and characterised by a specific surface texture (Fig. 9a–b). There is already evidence that some biominerals including calcite are mesocrystals (Cölfen and Antonietti, 2005) and the importance of extracellular polymeric substances for the formation of some types of nanostructured carbonate precipitates was documented (Pedley et al., 2009).
Determining the structure and material of precipitated inorganic substances brought another relevant question: “Do calcite spicules have fossilisation potential”? Microcrystalline calcite forming the recrystallised spicule is a typical material of calcite shells of fossil invertebrates (e.g. Vodrážka, 2009). Although calcite fossils may be partly or completely dissolved during diagenetical processes in the fossil record (e.g. Schneider et al., 2011; Švábenická et al., 2012), their preservation potential is relatively high. Therefore, we expect to find fossil and/or subfossil calcite spicules from the Quaternary lake sediments of the studied area.
This study was conducted during two Czech Antarctic research expeditions of the authors (J. Elster., L. Nedbalová, R. Vodrážka, K. Láska, J. Komárek) to the J. G. Mendel station in 2008 and 2009 (headed by Miloš Barták). We are indebted particularly to the staff and scientific infrastructure of the station. The study was supported by the Ministry of Education, Youth and Sports of the Czech Republic (CzechPolar LM2010009, KONTAKT ME 945 and RVO67985939). R. Vodrážka has been funded through a Research and Development Project of the Ministry of Environment of the Czech Republic No. SPII 1a9/23/07 and project GAČR GP14-31662P. K. Láska was supported by a project of Masaryk University MUNI/A/1370/2014 “Global environmental changes in time and space”. The authors gratefully acknowledge the comments received from Kevin Lepot and two anonymous reviewers. The technical work in laboratories was performed by Jana Šnokhousová and Dana Švehlová. Edited by: S. W. A. Naqvi