Bathymetric control of Holocene spit migration in a lacustrine environment

Abstract

Wave processes are well known for developing spit systems in large or elongated lakes by inducing longshore sediment transport for spit migration, while bathymetric interaction is less studied. In this study, we investigate the combined effects of wave processes, antecedent topography and lake level changes on the development of Holocene spit systems in the Danish lake Mossø. Wave climate prediction and a digital elevation model were used to provide a conceptual model for the development of the spits system, while optically stimulated luminescence (OSL) dating of spit sand and antecedent topography analyzed via boreholes and seismic survey data were used to evaluate effects on spit migration other than wave setting. We found that spit migration stagnated during the Holocene when reaching areas of deeper waters, but continued following shallowing after deep basin infilling with lacustrine sediments. During periods of bathymetrically induced stagnation of prominent spits, less stable or more slowly migrating spits became prominent in the development of the spit system. No clear effects on spit migration caused by lake level fluctuations could be demonstrated. However, such fluctuations may have been important for the stabilization of spits and subsequent development of a major barrier shoreline.

Keywords

bathymetry chronology lake simple wave mode spit water level

Introduction

Spits are common and distinctive coastal landforms along exposed marine shorelines, but they are also found in lakes and fjords with sufficient wave energies to produce prominent longshore sediment transport (LST) (Krist and Schaetzl, 2001; Kroon et al., 2017; Prieto, 1995). The growth rate and orientation of spits depend on the longshore drift induced by waves oriented obliquely to the shoreline. Theoretically, the strongest drift is produced by an angle of c. 45° between the shoreline and the wave crests, whereas LST is not induced by wave crests which are perpendicular or parallel to the shoreline (CERC, 1984; Petersen et al., 2008). Accumulation of sediment occurs when the LST decreases down-drift (i.e. in the direction of transport), and may be located at undulations in the shoreline direction. A qualitative description of spit accumulation for various shoreline and wave settings is given by Zenkovich et al. (1967), while models by Ashton et al. (2001) show that high angles (>45°) between the shoreline and wave crests induce spit-like growth from protuberances of the shoreline and self-organization of the coastline morphology. Furthermore, numerous field studies and numerically studied examples of spit growth at an embayment show that continuous spit accumulation results in a connection to the opposite shore or spit and eventually enclosure of the bay by a barrier (Hoan et al., 2011; Zenkovich et al., 1967).

The directional fetch limitation in elongated lakes or lagoons stimulates dominant longshore currents in the longitudinal direction and, consequently, formation of spits (Zenkovich, 1959) which, depending on the intensity of the conditions, are only slightly affected by changes in the dominant wind direction (Ashton et al., 2009). Generally, spits in small freshwater lake systems are relatively small compared to those found in large lakes and at sea or on ocean coastlines, and although depending on sediment availability, deposits are expected to be composed of sand only. Likewise, build-up of a thick spit platform is not common in small lakes because of a generally lower sediment transport (LST) capacity. The tidal influence is often insignificant and the spit migration depends entirely on the wind-induced waves, and indirectly on the lake geometry, as water level fluctuations may be caused by the wind conditions. Finally, lake basins of glacial or tectonic origin typically have very irregular bathymetry in contrast to the regular bathymetry of lagoons and marine shorefaces in older landscapes (Ashton et al., 2001; Petersen et al., 2008).

Unlike wave influences, the influence of bathymetry and lake level variations on spit migration in lake systems has so far not been widely studied. In this paper, our aim is to investigate the Holocene development of a complex spit system in the downwind end of an elongated glacial lake basin, and to propose a site-specific conceptual model of the conditions, processes and past morphology which established the present spit system. We hypothesize that bathymetry and lake levels have strongly limited spit migration across relatively deep waters, and that these mechanisms therefore had major implications on the basin development throughout the Holocene. To achieve this we used a highly interdisciplinary research approach with seismic survey and age–depth modeling of cores in the present-day lake while we on-ground used a combination of remote sensing data, soil pits, cores and geochronology dating of sediments.

Study site

Lake Mossø is situated in a wide east-west oriented glacial valley carved through the highest elevated moraine plateaus (up to 173 m a.s.l.) on the Jutland peninsula (Figure 1). The average lake level is 22.6 m a.s.l. (DVR90) (Tjellden et al., 2016).

Figure 1.

Digital elevation map of the glacial landscape around Lake Mossø. Lake shorelines are exaggerated to provide a better illustration of the abundance of small lakes in the area. Lake Mossø’s average level is 22.6 m a.s.l. (m above sea level), with depth contours (2 m intervals starting at 1 m of depth).

The elongated Lake Mossø has an aspect ratio of 4:1 (Figure 1). The long east-west fetch of Lake Mossø implies that the prevailing westerly winds are likely to be capable of generating waves with sufficient energy to potentially change the morphology of the eastern shorelines. Here, lacustrine spit deposits of five individual spits are found (outlined in Figure 2). Søe et al. (2017a) described the spits and found that due to subsidence of deposits (caused by modern agricultural drainage) and basin infilling, the morphology of the oldest spits is only partly visible in present-day topography. Herein, figures with profiles and images of cores taken from the lake and the survey pits were shown. It was also shown that the barrier coast (S1) representing the present eastern shore of Lake Mossø consists of core deposits from two opposite spits migrating from the north (S1n) and the south (S1s), respectively. Parts of spit S1s have been studied in some detail by ground-conductivity meter analysis (Christiansen et al., 2016). Two older relict spits in the same area (S2 and S3) are located in an eastern embayment enclosed by S1 where a fenland (Alken Enge) is found today. On the northern shore of Lake Mossø, an eastwardly directed spit (N1) is seen to have propagated into a north-eastern bay. At present, N1 is densely vegetated and hence all spits are inactive.

Figure 2.

Outline of spits and location of seismic lines, optically stimulated luminescence (OSL) samples, wave roses, cores and jetty. The insert is zoomed in on the hill-shading of spit S2. The present-day erosional buffs are shown by black triangles while palaeo-erosion buffs are shown by white triangles.

The southern shoreline of Lake Mossø follows the steep slope of the valley margin, and an erosional bluff is prominent at the eastern end of the lake (Figure 2). Numerous remnant cobbles and small boulders are found here on the nearshore lake bed. The northern shoreline of the eastern part of the lake borders a moraine hill from where N1 progrades. The sediment of the moraines is sandy clay till overlain by meltwater sand. The spit deposits are predominantly medium to coarse sands and typically contain thin layers of allochthonous gyttja or peat, while the surrounding lacustrine sediments are highly organic gyttjas which become increasingly calcareous in the downward direction (Søe et al., 2017a).

Archaeological excavations have shown that the tip of spit S2 was once connected to S1n by a short wooden bridge, and numerous human bones were sacrificed during the Iron Age around this location (Holst et al., in press), in a landscape which according to sedimentary pollen evidence had reduced land-use during the period AD 0–1000 (Søe et al., 2017a, 2017b). The bones and the bridge date to the 1st and 5th centuries AD, respectively, and mark the period after the enclosure of the embayment.

Methods

Geodata

The topography and morphology of the area were evaluated using the Danish LiDAR scanned digital elevation model (DEM) with a raster grid of 1.6 m x 1.6 m (Rosenkranz and Frederiksen, 2011). Historical maps from 1780, 1899 and 1925 as well as aerial photos taken since 1944 were studied to evaluate possible historical changes in the morphology and vegetation of the shoreline. Existing bathymetric mapping revealed a highly irregular lake bottom topography with current lake depths up to a maximum of 22 m (available online Supplementary Figure 1).

An on-land coring campaign in the infilled basins to the east and northeast of Lake Mossø was conducted in 2012–2014 using a closed-chamber Russian corer for soft fine-grained sediments, and a gouge auger, an Edelman auger, or a helical screw Ø90 mm mounted on a drilling rig for thick sandy layers (60 cores in total; from 0.6 to 12 m deep, 8 were sampled). The sediments were described in the field, and the locations of the boreholes were measured using a differential GPS device. In May 2013, four sediment cores (MOS1-4) were collected from Lake Mossø to the north of the N1 spit terminus (Figure 2), using a Russian corer with lengths from 1.8 to 5.0 m, and at water depths of 2.1 to 3.8 m. See Søe et al. (2017a) for the exact locations of the 60 on-land augering sites and geological information from all 64 boreholes.

Radiocarbon and OSL dating

From well-defined, stratified layers of a few centimeters in thickness, seven macrofossil assemblages in the MOS3 core were radiocarbon dated together with a piece of charcoal from a gyttja layer in a core (K32) located in a cove east of S3 in Alken Enge (locations shown in Figure 2). The terrestrial plant macrofossils were ¹⁴C dated using the standard radiocarbon protocols of the accelerator mass spectrometry (AMS) Center at Aarhus University and age-calibrated with OxCal v4.1.3 using the IntCal13 curve (Reimar et al., 2013). Optically stimulated luminescence (OSL) samples were obtained from excavation pits (locations shown in Figure 2) using opaque 20 cm long plastic tubes (Ø5 cm). The time since burial was measured at the Nordic Luminescence Laboratory at Risø using the single-aliquot regenerative-dose measurement procedure (Murray and Wintle, 2000).

Geophysical surveys

Seismic reflection data were acquired in the easternmost part of Lake Mossø in May 2013. The acquisition system consisted of an EG&G Uniboom 240 source and a DesignProjects single channel streamer equipped with eight hydrophones. The data were recorded on a Cheasapeake Technologies SonarWiz 5 recording system. GPS navigation data were recorded every 10 s by an Eiva Navipac system. The cruising speed was approximately 3 knots, and consecutive shots were fired 3 times per second, corresponding to a step distance of approximately 0.5 m. The sampling frequency of the seismic data was 8 kHz and the recording length 50 ms. The seismic data processing included the translation of recorded reflection times to vertical travel times needed for the conversion to vertical depths. The P-wave velocity of 1445 m/s used for this conversion was calculated on the basis of information on the salinity and temperature of the water. The field data were bandpass filtered in the interval 900–2000 Hz.

Wave climate analysis and LST

Wave conditions in the eastern part of Lake Mossø before shoaling were predicted using modern data on wind climate and with measured stretches of water surface as estimated effective fetch lengths. Wave roses of significant wave heights were estimated using the fetch limited deep water forecast equation based on the simple Sverndrup, Munk-Bretchneider (SMB) method (CERC, 1984): H₀ = 3.63*10⁻⁴·U_w^1.23·F^1/2, where H₀ is the significant wave height (m), U_w is the wind speed (m/s) and F is the fetch length (m). The wind speeds were estimated based on a 10 m elevation wind speed rose (Figure 5) provided by the Danish Meteorological Institute and obtained at Billund Airport (c. 50 km southwest of Lake Mossø) during a 10-year period (1989–1999). The rose separates wind directions into 30° sections and the speeds at 0.2, 5 and 11 m/s. For the calculation of wave heights, we used uniformly distributed wind speeds in the intervals (0.2–5 m/s and 5–11 m/s) and 11 m/s distributed wave heights induced by the frequent moderate wind climate and the minimum height of waves induced by the infrequent strong wind climate.

To compute the potential LST rate for sand at the breaker line, we used the so-called Costal Engineering Research Center (CERC) formula: Q_mass = 128 (H_s)^2.5 sin(2α) with Q_mass in kg s^-1 dry mass (here converted into tons/year). In the formula, H_s is the significant wave height, while α is the angle, and the longshore transport rate per wave height class and direction was multiplied by the frequency of occurrence. The rates were summarized for all wave height classes, resulting in the potential LST rate for each direction. Effective fetch lengths in each wind direction class were estimated by the cosine-weighted length of fetch rays in an angle section: F_eff = ∑(F_i*cosα_i)/∑cosα_i, where F_eff is the effective fetch length (m) of a wind direction class, α_i is the relative angle to the wind direction class, and F_i is the fetch ray length (m). We used angle sections of 30° (the same as for the wind direction classes) and a spacing between the fetch rays of 3°. Wind set-up was disregarded due to the effectively fetch limiting system of Lake Mossø.

Results

Spit morphology

The morphology of the spits can be partly evaluated from the DEM shown in Figure 2. First, spit S3 is short, curves toward an NNW direction and has a maximum elevation of c. 23 m a.s.l. (0.4 m above the average winter lake level). A sediment core from the distal part of S3 showed approximately 2 m of Holocene lake marl separating the c. 2 m thick spit sand deposit from the underlying glacial sand. Spit S2 has a total length of c. 400 m, but only the proximal southernmost 250 m are visible on the DEM. This proximal part of S2 has a northern direction that gradually curves toward the east and several generations of overtaken distal spit tips pointing in a NNE direction can be observed on the DEM (the insert in Figure 2). The distal 150 m of S2 are observed in excavation pits, and characterized as being narrow (20–30 m) with an estimated maximum thickness of 3 m and having a NNW direction. Furthermore, boreholes revealed that the distal part of the spit is deposited on c. 10 m thick layers of Holocene gyttja and lake marl. The maximum elevation of S2 is 23.9 m a.s.l. (1.3 m above present-day average lake level). The exact morphology of the S1 spits cannot be evaluated from the DEM because of modern-day disturbances. However, the boreholes show that the intermediate part of S1n is c. 2.5 m thick and deposited upon c. 3 m of Holocene lake sediment; the distal part of S1s is c. 1.5 m thick and deposited upon c. 1.5 m of Holocene lake sediment. In the area where the S1n and S1s connect, c. 7 m of Holocene lake sediment is found below approximately 2 m spit deposits. The fact that there is deeper water between the bluffs from which S1, S2 and S3 start suggests that initially there were several sets of overtaking spits instead of just one spit with a series of progradational beach ridges building lakewards from the initial shoreline. The series of historical maps and aerial photos revealed no progradation of N1 and little change of the eastern shoreline during the past c. 230 years. Judging from available aerial and satellite images, the shore vegetation at the eastern end of Mossø has generally become more lush since 1944 with build-up of reed vegetation in more sheltered places and generally more shore shrub vegetation.

Spit deposit chronology

A chronology of evolution of the spit system can be established via the two OSL samples in Table 1 and OSL samples previously reported by Søe et al. (2017a). The locations of the OSL samples are shown in Figure 2. Spit S3 is the oldest spit which was active only during the early Holocene until approximately 8200 ± 500 years BC (see OSL13, Table 1). The time when S2 started prograding is unknown, but two samples (OSL4, OSL9) of the more distal part show a progradation of 60 m during the period from 630 ± 200 to 80 ± 100 BC. The intermediate parts of S1n and S1s show ages of 2210 ± 200 and 1710 ± 200 BC (OSL3, OSL5), respectively, and indicate contemporary migration. In the area where the northern and southern spits connect, the deposit of S1 shows ages between 500 ± 150 BC and 110 ± AD 120 (OSL8, OSL6) consistent with the ages of the archeological finds (Holst et al., in press). Subsequently, S1 was still the target of depositional and erosional processes, indicated by an overwash deposit on the intermediate part of S1s dated to AD 630 ± 70 (OSL14) and by the Iron Age shoreface deposits of artifacts and radiocarbon dated human bones described in Søe et al. (2017a). Later barrier deposits dated to the 17th-century AD indicate a stabilization of the barrier S1.

Table 1.

Details of (optically stimulated luminescence OSL) ages of spit sand at Mossø, Denmark. OSL samples marked with * have previously been reported by Søe et al. (2017a) with indication of profiles and sampling sites.

Sample	Lab. ID	Depth	Dose	N	Dose rate	Water content	Age
Sample	No.	(cm)	(Gy)		(Gy/ka)	(%)	(ka)^a
OSL13	134402	50	16.1 ± 0.5	42	1.58 ± 0.06	42	10.2 ± 0.5
OSL14	114904	50	1.95 ± 0.05	30	1.43 ± 0.06	28	1.37 ± 0.07
OSL2*	114902	60	0.32 ± 0.01	22	1.29 ± 0.06	18	0.25 ± 0.02
OSL3*	134401	155	4.75 ± 0.14	38	1.13 ± 0.05	19	4.20 ± 0.2
OSL4*	144901	140	4.26 ± 0.27	26	1.62 ± 0.06	37	2.63 ± 0.20
OSL5*	144903	120	4.79 ± 0.14	26	1.29 ± 0.05	41	3.71±0.20
OSL6*	124901	250	2.11 ± 0.05	24	0.84 ± 0.04	34	2.51 ± 0.15
OSL8*	124903	240	3.16 ± 0.14	21	1.66 ± 0.07	19	1.90 ± 0.12
OSL9*	134403	176	4.71 ± 0.11	18	2.27 ± 0.09	22	2.08 ± 0.10

The OSL ages are given as times since burial. Years of analysis are 2011–2014, as indicated by the first two digits in the Lab. ID No.

Depositional limits

The four cores obtained in Lake Mossø showed an erosional level observed as a sharp, irregular upper limit of lake marl and a boundary containing coarse sand, gravel and coarse woody fragments before gradually turning into highly organic gyttja upward (Figure 3). The erosion level was found down to 16.3 m a.s.l. in core MOS1, that is, 6.3 m below the present-day average lake level. Four of the radiocarbon dates in MOS3 (MOSC, MOSD, MOSF, and MOSG in Table 2) constitute a uniform deposition rate model, the extrapolation of which shows that deposition most likely resumed at this site around or shortly before AD 1000, as the terrestrial macro-plant remains were dated to AD 1050–1150 (see Table 2 and available online Supplementary Figure 2).

Table 2.

Details of radiocarbon dated material from cores K32 and MOS3 in Lake Mossø. The sampling depths are shown in Figure 3.

Sample	AMS lab. number (AAR-)	Sample type	δ¹³ C (‰ VPDB)	Conventional ¹⁴C age (BP)^a	Calibrated age (cal. yr BP)^b
K32B	18781	Charcoal	–26.54 ± 0.05	8116 ± 33	9088–9046 (34.5%) 9036–9006 (33.1%)
MOSA	21301	Phragmites stem	–26.8 ± 0.05	266 ± 25	422–407 (14.1%) 316–287 (54.1%)
MOSB	21302	Alnus seeds, bud scales, small twigs	–28.6 ± 0.05	245 ± 25	306–285 (54.1%) 165–157 (14.1%)
MOSC	21303	Bud scales, leaf fragments	–27.34 ± 0.53	895 ± 27	902–865 (30.8%) 826–813 (8.6%) 800–763 (28.8%)
MOSD	21304	Moss shoots (Neckera pumila, Hypnum cupressiforme), bud scales	–27.25 ± 0.05	929 ± 26	908–892 (12.7%) 875–844 (25.5%) 833–797 (30.0%)
MOSE	21305	Betula seeds, Alnus seed fragments, Cirsium seeds	–24.26 ± 0.6	1360 ± 27	1302–1279 (68.2%)
MOSF	21306	Alnus seeds, Betula seeds, bud scales, leaf fragments, moss shoots (Hypnum cupressiforme)	–27.42 ± 0.05	941 ± 25	915–900 (13.5%) 868–823 (40.1%) 816–799 (14.6%)
MOSG	21307	Betula seeds and female catkin scales, Alnus seeds, bud scales	–28.18 ± 0.55	1042 ± 26	963–931 (68.2%)

The ¹⁴C ages are given in conventional radiocarbon years BP (before present = 1950) with a measuring uncertainty of one standard deviation.

Calibrated ages in years BP (before present = 1950) from the 68.2% probability distribution.

Figure 3.

Optical picture of the MOS-3 core (see location in Figure 2) and depth of radiocarbon dating (details of ¹⁴C dating are shown in Table 2). The erosion contact at approximately 298 cm below lake level with the sharp transition from lake marl to calcareous gyttja is clearly seen.

Basin morphology of Lake Mossø

The seismic survey in the north-eastern part of the lake showed an irregular bathymetry composed of alternating depositional basins and ridges (Figure 4a). The erosional level seen in the MOS1 core could be correlated with a characteristic reflector denoted Basis2 in the seismic profile crossing the core location. The other core positions were shoreward of the seismic lines.

Figure 4.

Results from seismic survey at Lake Mossø. (a) Map of the lake bottom with 0.5 m contour lines. (b) Map of Basis2 (erosion level in cores) with 0.5 m contour lines. The white area shows where Basis2 is absent from the seismic profiles. (c) Seismic profile of line 10 marked in (b) with an example of two depositional basins separated by a ridge and the interpretation of Basis2.

The gyttja found in K32 is located at up to 24 m a.s.l. and indicates a palaeo-lake level above 26 m a.s.l. assuming a water depth limit of 2 m for the deposition of gyttja. This high lake stand is supported by continuous slopes at approximately 26 m a.s.l. on the northern and western shores of Lake Mossø that possibly reflect high-stand palaeo-shorelines. The charcoal in K32 was found at an elevation of 23.9 m a.s.l. and had a calibrated radiocarbon age of c. 9050 cal. yr BP, that is, 7050–7150 BC (Table 2).

The reflector could be followed through the network of seismic lines (Figure 2) and recognized on the other sides of the ridges. Hence, the erosional level was mapped in three separate depositional basins (Figure 4b). The reflector either thinned out or had an angular discordance with the lake bottom when approaching the ridges that separate the depositional basin (Figure 4c). In the deeper part of the basins it had disappeared.

Wave settings

The wave rise for three locations (W1, W2 and W3 in Figure 2) shows the distribution of wave directions and wave heights (Figure 5). Each location is situated off the shoreline and estimates the wave climate at the southern erosional bluff (W1), S1s, S2 and S3 (W2) and S1n (W3).

Figure 5.

Wind rose for Billund Airport (1989–1999) and wave roses for the three locations (W1, W2, W3 seen in Figure 2) in Lake Mossø.

At W1 substantial waves are produced by westerly winds and c. 1% of the time, the wave height is predicted to be above 0.5 m. Indeed, wave heights of more than 1 m are predicted by the forecast equation at W1 during extreme wind speeds of >20 m/s. Similarly, the wave climate at W2 is predicted to come from westerly directions but is of less energy. The most substantial waves come from the W-SW directions. At W3 the highest waves come from a SW direction, while a substantial part of the smaller waves come from a W-NW direction. The net LST rates of the three locations are calculated for W1, W2 and W3 using the CERC formula. For W1, this reveals a shore normal component of 305° and an annual net transport rate of 13,700 tons/year to the east. Location W2 has a shore normal component of 248° and an annual net transport rate of 4530 tons/year to the south. This net transport rate is far lower here, and there is also significant transport in the other direction. Location W3 has a shore normal component of 203° and its annual net transport rate is 7560 tons/year to the east. The net transport rate is almost similar to the total transport rate in this case.

Discussion

Wave climates

The W1 wave climate in Figure 5 has the potential to induce shore erosion and a high LST toward the spits, produced by high-energy waves from a westerly direction. Thus, the erosional bluff on the southern shoreline behind wave location W1 is a large sediment source of the migration of spits S3, S2 and S1s. However, this erosional bluff is also shadowing the most proximal part of S1s toward waves from WSW, which means that the initial migration of S1s would be reduced by a low sediment supply and wave energy.

The W3 wave rose in front of spit S1n shows a strong wave climate only in the SW direction (Figure 5). This would not induce the LST expected for S1n to migrate south-eastwards as the sediment transport rate is toward the east only. However, diffraction of waves around the bluff of N1 and the subsequent refraction caused by ridges on the lake bottom (Figure 4) may induce a low-energy eastern longshore current in the bay. Also, the more low-energy waves from WNW may induce LST for migration of S1n. Indeed, an eastward LST at the northern shore updrift from S1n is evidenced by historical sedimentation on the eastern side of an > 120 years old jetty (Figure 2). The sediment source is presumed to have originated partly from the moraine hill onto which spit N1 is attached, and partly from the eroding ridges on the lake bottom. The low-energy wave climate at W3 suggests a slow but steady spit migration without major erosion due to the westward shielded location behind the moraine hill.

Bathymetric effect on the evolution of spits

The preferential spit migration across elevated antecedent topographies has been described for several relatively high-energy locations (Noe-Nygaard and Surlyk, 1988; Raynal et al., 2009). The lower depth at the elevated antecedent topography ensures littoral processes for spit formation (e.g. LST), and hence accumulation of a thick spit platform deposit is not required for spit migration to occur. Bathymetry is expected to have a vital effect on the development of spit systems in Lake Mossø, due, in particular, to the low-energy environment with relatively low sediment transport rates, which prevents deposition of a thick spit platform. A shadowing effect is also of importance for the northward migrating spits in Lake Mossø. The proximity of the S3, S2 and S1s spits implies that a spit located to the west shadows wave energy and sediment supply toward a more easterly located spit. Therefore, the continuous spit accumulation of S2 and S3 was depending on a relatively higher migration rate than for the westward spits.

S2 is evaluated to be positioned at the most advantageous position in the conceptual model for spit migration in Lake Mossø (Figure 6). Spit growth was favored by its proximity to the large sediment source of LST and the connection to a highly exposed bluff. A ground conductivity meter survey indicates that the proximal part of S2 migrated fast due to the advantageous position and a relative high glacial topography in front of the bluff (Christiansen et al., 2016). Hence, the apparent short lifetime of spit S3 was probably a result of shadowing caused by the faster migrating S2 during the early Holocene. The late-Holocene ages of the distal S2 deposits indicate that the fast migration ceased at some point. The migration of S2 most likely stagnated when the spit approached increasingly deep waters. The thick Holocene lacustrine deposit beneath the more distal part of S2 shows decreasing elevation of glacial topography in the direction of progradation, and the slope is believed to be situated where the spit disappears in the DEM (see the insert in Figure 2). Unlike the distal part of spit S2, the more elevated and wider proximal part suggests a high lake level for substantial spit deposit build-up and limited subsequent subsidence by compaction of relatively shallow lacustrine organic sediment. The continued migration of S2 is expected to have progressed slowly after the deeper basin in front of the spit was sufficiently filled with lacustrine sediment for the littoral processes to be effective. However, the continuation of S2 also seems temporally correlated with a rise of the water level in Lake Mossø around 800–300 BC (Søe et al., 2017a), that is, after an increase in the water depth, as hypothesized by Møller (1984). Spit S2 was probably shadowed by S1s at that time, but the rise in the lake level implied exposure of S2 to wave energies for further migration by erosion of S1s. Furthermore, the placement of a late Iron Age wooden bridge between distal parts of S2 and S1n (Figure 2, Holst et al., in press; Søe et al., 2017a) may suggest that S1s was found too unstable for anchoring a connection. During the late Holocene, a shadowing of S2 is indicated by the mediate part of S1s being OSL dated to c. 1700 BC.

Figure 6.

Conceptual model of the spit development and shore positions in lake Mossø’s eastern part during the Holocene. Palaeo-lake levels are interpreted from radiocarbon and OSL dating, while current digital elevation models, cadastral maps, seismic and ground conductivity meter data, and archaeological evidence are used as background to suggest the positions of shorelines and spits. Partly based on the data from Søe et al. (2017a), Christiansen et al. (2016), and Rasmussen (in press). The present-day shoreline (based on an average winter level of 22.6 m above sea level) is indicated by a white dotted line.

Water level fluctuations and evolution of spits

The lake level fluctuations of several meters during the Holocene are indicated by the depositional limits in the cores K32 and MOS-1, but the effect of these changes on the spit migration in Lake Mossø is not clear. All spit deposits are situated approximately at the same elevation, and thus, a significantly increased lake level seems not to have been crucial for the progradation of any spits. On the other hand, the lake level controls the potential for sediment erosion of both the spit deposits and the sediment source because of the higher energy of the waves reaching the shoreline when the lake level is increased. Accordingly, an increase of the lake level would both potentially increase spit accumulation rates and/or erosion of the exposed spits, for example, spit S1s. In contrast, a lowering of the lake level would shift the littoral processes and any associated sediment accumulation lakewards. In that case, a subsequent rise in the lake level could result in increased landward cross-shore sediment transport (Masselink et al., 2006; Stockdon et al., 2006). Accordingly, the process of stabilization of spit S1s and development into a barrier shoreline could perhaps have been enhanced by a temporary drop in the lake level correlating with the erosion level in the sediment cores (Figure 3) from Lake Mossø dated to prior to or around AD 1000. Furthermore, it should be kept in mind that when a sandy spit propagates on top of unconsolidated organic lake sediments, substantial subsidence – even subaqueously – is to be expected (Rosati et al., 2010). Therefore, the present elevations of the spits in Lake Mossø which progressed over organic sediments do not necessarily reflect the lake level during active deposition, as exemplified by the subsided distal part of spit S2.

The coastline of spit N1 is currently entirely vegetated, lacking the more obvious white sandy beaches easily seen to front the coastal section of spit S1. Consistent with the inactive vegetated shore, the shape of spit N1 has not changed since the first cadastral maps were made 230 years ago.

Conceptual model for spit evolution

Based on the wave roses, shoreline directions, and palaeo-lake levels, a conceptual model for the Holocene spit system development in the eastern end of Lake Mossø is sketched in Figure 6. The bluffs from which S1s, S2 and S3 prograde are presumed to have been protuberances on a general NE directional southern shoreline that started before the erosional bluff, although some shore-face longshore transport from a more distal source cannot be excluded (Figure 2). Thus, according to the model of Ashton et al. (2001), accumulating spits will grow from the bluffs during winds from WSW that induce high-angle waves toward the general shoreline. Such a wind climate is indeed prominent at W2 in front of these bluffs and further up-drift at W1 (Figure 5). However, the similarly prominent waves from the west at W2 would approach the general shoreline approximately at the angle of maximum longshore transport. Whether these waves would continue the spit accumulation or erode the spits would depend highly on the wave energy, nearshore wave refraction and, especially, the orientation of the spit in relation to the waves. Here, it is assumed that the environmental settings stayed roughly the same throughout the Holocene. Wind patterns and wind strengths may have varied, although westerly winds have predominated throughout the entire Holocene in Denmark. The sediment input from the slopes and the surrounding landscape have also increased considerably during the past approx. 1000 years due to anthropogenic disturbance of the vegetation cover (Søe et al., 2017b). However, for the spit dynamics in a mainly groundwater-fed lake like Mossø, the input of silt and clay from surface erosion and rivers is of relatively less importance compared to sources of coarser sediment such as shore and lake-floor erosion and is hence governed mainly by the lake levels and the east-west direction of the lake.

Nevertheless, the palaeo-bluffs at the southern shores provide clear evidence of strong erosion. The largest spits S1s and N1 are also located downstream from where the most prominent bluffs are located, but today no bluff erosion seems to take place. As the sediment sources at the modern lake shore are limited because of the lush reed beds, the modern shoreline is presumably rather resilient to perturbation, although it is difficult to extrapolate this vegetation back in the past. More vegetation along lake shores and in general is, however, universal in Denmark and probably reflects the overall decline in domestic grazing during the past decades, so the modern-day shores are not representative of the conditions of the past. Changes in spit development, if you assume that the wind did all the work, are normally triggered by a major event, but following a change in vegetation (human induced) or climate that have caused a less resilient coast. However, our calculated values for LST are theoretical and entirely based on fetch and wind force, and our temporal resolution are likely too coarse to resolve such causal-relationships.

Having investigated the sediments at the western rims of lake Mossø and river depth-profiles down-stream, Møller (1984) suggested that Lake Mossø had a lower level before the construction of a medieval dam. Our data confirm that an increase in the water level occurred around AD 1000, which is presumably related to the activities at a nearby monastery, including construction of the dam. A man-made straight channel (Munkekanalen) between Mossø and river Gudenå is attributed to this monastery, however, it is not precisely dated and its exact influence on the observed lake level changes is hence difficult to discuss. Møller (1984) also proposed that natural deposition of river sediments along the Gudenå River could have caused the river to be cut off from flowing in and out of Lake Mossø in the later Holocene. He further suggested that this might have caused a lake level increase. However, no evidence was given to support these scenarios, and a rising lake level due to sediment build-up at the outlet of Lake Mossø needs to be confirmed and dated, as our data are non-conclusive as to the local causes of the observed rise in the lake level up to around year AD 1.

The OSL age of c. 2000 BC of spit S1n indicates a middle to late-Holocene migration, although when looking at the conceptual model for spit migration (Figure 6), a spit is not incited to migrate where we find S1n. The sediment supply for S1n was lower than for the northward migrating spits, and low water depths were thus probably essential for the spit migration here. The spit was presumably migrating on top of a ridge of the glacial topography such as those seen on the seismic map (Figure 4) and indicated by the elevated glacial topography in the boreholes. This may explain why less sediment was needed for spit migration compared to other nearby sites at the southern coast. In contrast, spit N1 is situated at a much more advantageous position for spit migration with higher sediment transport and wave energy. However, the progradation of spit N1 stagnated at a deep north-south basin (Figure 4). Spit N1 would be presumed to have isolated the north-eastern bay if the bathymetry was regular and shallow.

The spit development model (Figure 6) is based on palaeo-lake levels, interpreted from both radiocarbon and OSL dating (Rasmussen, in press; Søe et al., 2017a). The early Holocene (7000 BC) maximum lake level was around 25–26 m a.s.l. (compared to the present-day average lake level of 22.6 m a.s.l.). Later, around 3000 BC, it declined to a minimum level of 20–21 m a.s.l., while around 500 BC to AD 100 it reached 22–23 m a.s.l. again. The causes of these lake level changes must also be related to threshold changes of the major inlet and outlet in the western part of Mossø. These changes are probably a result of first erosion of the threshold downstream in the outlet river Gudenå that lowered the lake level, and second slow build-up of levees and sediments from the Gudenå river that, according to Møller (1984), could have caused damming of Lake Mossø in late Holocene as discussed above. One interesting aspect of the observed lake level of 26 m a.s.l. in early Holocene is that the lake basins of Mossø, Gudensø, Skanderborg Sø and Salten Langsø are likely to have merged to form one lake. Such a ‘grand’ palaeo-lake Mossø would then have been at least 25 km long from the east to the west, and consequently candidates as the largest lake in Denmark. However, more investigations are required in order to confirm this hypothesis.

In the model in Figure 6a, the migration of S3 ceased as it became shadowed by S2. Further S2 migration was limited as it approached a deep basin. The migration of S1s was partly shadowed by the erosional bluff and perhaps limited by the deep basin, while Sn1 migrated slowly. Figure 6b, around c. 2000 BC shows that S3 was completely shadowed by S2, which, in turn, was limited by the deep basin and presumably temporarily shadowed by S1s. Spit S1s still prograded but was unstable, as in a westerly wave climate and high wave energies it was exposed to erosion and cross-shore sediment transport. Figure 6c shows that around c. 1000 BC, the basin infilling had resulted in shallow water, and S2 started to migrate again as the water level of the lake increased. Presumably S1 was prograding but was unstable. Spit Sn1 migrated slowly or was already limited by deep water at the southernmost position of the spit deposit. Figure 6d, around AD 1, shows a spit system development that had terminated. Spits S2 and S3 were now shadowed, while S1s and S1n turned into the present-day system with barrier islands and only an inlet channel, with the Illerup stream connecting the eastern Alken Enge basin to Lake Mossø. The cut-off lake in Alken Enge was later infilled with peat and gyttja. Finally, it has to be stressed that the exact spatio-temporal development of spit N1 is unclear due to a lack of profile observations and dating.

Conclusion

A conceptual model for spit migration throughout the Holocene based on wave setting and shoreline morphology in Lake Mossø was compared with the ages of spit deposits and suggested bathymetric control of the spit migration. Subaerial spits first develop on ridges in the antecedent glacial topography and then across deeper basins when infilling of these basins has caused the water to become sufficiently shallow for the spit processes to prevail. As the infilling sediment is lacustrine gyttja, the gradual build-up in deeper basins slowly creates shallower water, upon which sandy spit sediment and morphological structures are able to develop over time. Since they rest on unconsolidated sediments, the process is even slower due to the partial compression of the underlying gyttja as the spit grows. Consequently, the stagnation of an accumulating spit at a basin margin may allow other spits to develop at locations with less advantageous wave settings but with shallower water depths. This suggests that the Holocene development of spits in Lake Mossø was most likely determined by low sedimentation rates which were insufficient to deposit thick spit platforms in a complex interplay with the initially glacially determined, variable lake bathymetry. Such a possible early Holocene ‘grand’ palaeo-lake Mossø with a 25 km east-west length is also a hypothesis that requires more investigation, as later erosions at the thresholds which connect Mossø to the adjacent lakes cannot be excluded, and as the exact position and early Holocene elevation of river Gudenå’s threshold remain uncertain.

Supplemental Material

Supplmentary_Material – Supplemental material for Bathymetric control of Holocene spit migration in a lacustrine environment

Supplemental material, Supplmentary_Material for Bathymetric control of Holocene spit migration in a lacustrine environment by Niels Emil Søe, Aart Kroon, Bent Vad Odgaard, Holger Lykke-Andersen and Søren Munch Kristiansen in The Holocene

Footnotes

Acknowledgements

The authors wish to acknowledge useful discussions with Ian A. Simpson, Sterling University, Scotland; Catherine Jessen, The National Museum of Denmark; Lucia Petersen and Thomas Ljungberg, both Aarhus University; and Casper S Andersen, Mads K Holst, Uffe Rasmussen, and Anna E.K. Tjellden, all at Moesgaard Museum, and Ejvind Hertz, Skanderborg Museum. The authors would also like to thank an anonymous reviewer for valuable inputs to this manuscript.

Funding

This research was financially supported by The Carlsberg Foundation (2012_01_0495) and carried out during PhD studies co-funded by the Graduate School of Science and Technology, Aarhus University.

References

Ashton

Murray

Arnoult

(2001) Formation of coastline features by large-scale instabilities induced by high-angle waves. Nature 414: 296–300.

Ashton

Murray

Littlewood

et al . (2009) Fetch-limited self-organization of elongate water bodies. Geology 37: 187–190.

CERC (1984) Shore Protection Manual, vol. 1. Vicksburg, MI: Waterways Experiment Station, Corps of Engineers, Coastal Engineering Research Center (CERC), 337 pp.

Christiansen

Pedersen

Søe

et al . (2016) Improved geoarchaeological mapping of paleo-landscapes with electromagnetic induction instruments from dedicated processing and inversion. Remote Sensing 8(12): 1022.

Hoan

Hanson

Larson

et al . (2011) A mathematical model of spit growth and barrier elongation: Application to fire Island Inlet (USA) and Badreveln Spit (Sweden). Estuaries, Coastal and Shelf Science 93: 468–477.

Holst

Heinemeier

Hertz

et al . (in press) Direct evidence of a large North European roman time martial event and post battle corpse manipulation. PNAS.

Krist

Schaetzl

(2001) Paleowind (11,000 BP) directions derived from lake spits in northern Michigan. Geomorphology 38: 1–18.

Kroon

Abermann

Bendixen

et al . (2017) Deltas, fresh water discharge and waves along the young sound, NE Greenland. Ambio 46 (Suppl. 1): S132–S145.

Masselink

Kroon

Davidson-Arnott

RGD

(2006) Morphodynamics of intertidal bars in wave dominated settings–A review. Geomorpholpgy 73: 33–49.

10.

Møller

(1984) Klostermølle og Gudenåen. En beskrivelse af vandløbene ved Voer og Vissing Klostre. Miljøministeriets Klostermølleudvalg. Copenhagen: Miljøministeriet, 38 pp.

11.

Murray

Wintle

(2000) Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol. Radiation Measurements 32: 57–73.

12.

Noe-Nygaard

Surlyk

(1988) Washover fan and brackish bay sedimentation in the Berriasian-Valanginian of Bornholm, Denmark. Sedimentology 35: 197–217.

13.

Petersen

Deigaard

Fredsoe

(2008) Modelling the morphology of sandy spits. Coastal Engineering 55: 671–684.

14.

Prieto

(1995) Shoreline forms and deposits in Gallocanta Lake (NE Spain). Geomorphology 11: 323–335.

15.

Rasmussen

(in press) Fund og landskab bag odderne i Alken Enge. In: Løvschal

Iversen

Holst

(eds) Hær og efterkrigsritual ved Alken Enge: En introduktion (in Danish). Aarhus: Moesgaard Museum.

16.

Raynal

Bouchette

Certain

et al . (2009) Control of alongshore-oriented sand spits on the dynamics of a wave-dominated coastal system (Holocene deposits, northern Gulf of Lions, France). Marine Geology 264: 242–257.

17.

Reimar

Bard

Bayliss

et al . (2013) INTCAL13 and MARINE13 radiocarbon calibration curves 0-50,000 cal BP. Radiocarbon 55: 1869–1887.

18.

Rosati

Dean

Stone

(2010) A cross-shore model of barrier island migration over a compressible substrate. Marine Geology 271: 1–16.

19.

Rosenkranz

Frederiksen

(2011) Quality Assessment of the Danish Elevation Model (DK-DEM). Technical Report no. 12. Aalborg: Danish Ministry of the Environment, National Survey and Cadastre, May 2011.

20.

Søe

Odgaard

Hertz

et al . (2017a) Geomorphological setting of a sacred landscape: iron age post battle deposits of human remains at Alken Enge, Denmark. Geoarchaeology. Epub ahead of print 22 May. DOI: 10.1002/gea.21622.

21.

Søe

Odgaard

Nielsen

et al . (2017b) Late-Holocene landscape development around a roman iron age mass grave, Alken Enge, Denmark. Vegetation History and Archaeobotany 26: 277–292.

22.

Stockdon

Holman

Howd

et al . (2006) Empirical parameterization of setup, swash, and runup. Coastal Engineering 53: 573–588.

23.

Tjellden

AKE

Mathiessen

Petersen

LMM

et al . (2016) In situ preservation and deterioration of sacrified iron age human bones at Alken Enge, Denmark. Conservation and Management of Archaeological Sites 18: 126–138.

24.

Zenkovich

(1959) On the genesis of cuspate spits along lagoon shores. Journal Geololgy 67: 269–277.

25.

Zenkovich

Steers

King

CAM

(1967) Processes of coastal development. New York: Wiley Interscience, 738 pp.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.45 MB