Master’s thesis
Physical Geography and Quaternary Geology, 60 Credits
Department of Physical Geography
Diatom distribution in the lower Save River, Mozambique
Taxonomy, salinity gradient and taphonomy
Marie Christiansson
NKA 156
2016
Preface
This Master’s thesis is Marie Christiansson’s degree project in Physical Geography and Quaternary Geology at the Department of Physical Geography, Stockholm University. The Master’s thesis comprises 60 credits (two terms of full-time studies).
Supervisor has been Jan Risberg at the Department of Physical Geography, Stockholm University. Examiner has been Stefan Wastegård at the Department of Physical Geography, Stockholm University.
The author is responsible for the contents of this thesis.
Stockholm, 11 September 2016
Steffen Holzkämper Director of studies
Abstract
In this study diatom distribution within the lower Save River, Mozambique, has been identified from surface sediments, surface water, mangrove cortex and buried sediments.
Sandy units, bracketing a geographically extensive clay layer, have been dated with optical stimulated luminescence (OSL). Diatom analysis has been used to interpret the spatial salinity gradient and to discuss taphonomic processes within the delta. Previously, one study has been performed in the investigated area and it is of great importance to continue to identify diatom distributions since siliceous microfossils are widely used for
paleoenvironmental research. Two diatom taxa, which were not possible to classify to species level have been identified; Cyclotella sp. and Diploneis sp. It is suggested that these represent species not earlier described; however they are assigned a brackish water affinity.
Diatom analysis from surface water, surface sediments and mangrove cortex indicate a transition from ocean water to a dominance of freshwater taxa c. 10 km upstream the delta front. Further, ratios between marine/brackish taxa for samples from surface water and surface sediments do not correspond. It is therefore suggested that diatoms in surface sediments underestimate prevailing salinity conditions in water. In the investigated area extensive taphonomic processes seem to have large impact on diatom frustules in sediments and may bias interpretations. Therefore it is recommended to carefully investigate geology, geomorphology and vegetation before diatom analysis is applied in studies of delta
paleoenvironments.
Table of contents
Abstract ... 1
Table of contents ... 2
1. Introduction ... 3
1.1 Objectives ... 6
1.2 Background ... 6
1.2.1 Salinity stratification ... 6
1.2.2 Diatoms ... 7
1.2.3 SEM/ESEM ... 9
1.2.4 OSL ... 9
2. Description of the investigated area ... 10
3. Methodology... 13
3.1 Siliceous microfossils... 15
3.2 ESEM ... 19
3.3 OSL ... 19
4. Result and interpretation ... 19
4.1 Surface water samples ... 20
4.2 Surface sediment samples ... 21
4.3 Mangrove cortex samples ... 21
4.4 Buried sediments ... 28
4.5 Taxonomic remarks ... 38
5. Discussion... 41
5.1 Taxonomy ... 41
5.2 Salinity gradient ... 42
5.3 Taphonomy ... 44
5.4 OSL and radiocarbon dating ... 49
5.5 Future studies ... 50
6. Conclusions ... 51
7. Acknowledgement ... 52
8. References ... 53
9. Appendix 1 ... 69
10. Appendix 2 ... 71
11. Appendix 3………76
12. Appendix 4………80
13. Appendix 5………81
14. Appendix 6………...82
1. Introduction
Research on past climate is required to forecast future global climate changes (e.g. Palmer &
Abbott, 1986; Shennan et al., 1993; Denys and de Wolf 1999; Zong & Horton, 1999; Church et al., 2013). One aspect implies changes of the global mean sea level, which can involve either a rise or a lowering affecting coastal environments, e.g. deltas (Tamura et al., 2012;
Church et al., 2013). A rising sea-level involves effects on wave-induced erosion,
tides and currents influencing shorelines and mangrove forests holding complex ecosystems (Nicholls & Cazenave, 2010). A lowering could imply an increased erosion by fluvial
and terrestrial processes combined with effects on floras and faunas. As the global sea-level has fluctuated over time, especially during the Quaternary (Lowe & Walker, 1997), changes can be traced in coastal environments (Woodroffe, 1990). To increase knowledge of the effects from approaching eustatic sea-level rise, past fluctuations should be
elaborated. Paleoenvironmental research is often hampered by chronological drawbacks (cf.
Andrews et al., 1999; Wang et al., 2013; Bala et al., 2016). Deltaic sediments have an advantage since they may be dated by optically stimulated luminescence (OSL) as a complement to radiocarbon dating (cf. Bishop et al., 2004; Zhao et al., 2008; Erginal et al., 2009).
One potential approach to study paleo-climate and paleo-environment is analysis of siliceous microfossils, mainly diatoms (e.g. Simonsen, 1969; Palmer & Abbott, 1986; Vos & de Wolf, 1993:1; Espinosa, 1994; Denys, 1999; Denys and de Wolf, 1999). As living diatoms are
sensitive to changes of salinity, pH and nutrition (Cooper, 1999; Jiang, et al., 2001; Hassan, et al., 2006) and fossil frustules (shells) in sediments normally are well preserved (Ferguson Wood, 1967), they may indicate environmental changes such as tidal currents, flooding events and sea level fluctuations (Swan, 1983; Vos & de Wolf, 1988). Previous studies have shown benefits of using diatom analyses to reconstruct these changes in delta environments (cf. e.g. John, 1987; Zalat, 1995; Zong et al., 2009; Ellison, 2008; França et al., 2015). To improve the interpretation of fossil diatoms, further investigations on recent assemblages are beneficial as they are a key to understand fossil communities (John, 1987; Castro et al., 2013; Zong & Horton, 1998; Zong et al., 2006). This study is thus focusing on diatom analysis from surface material from the lower Save River, Mozambique, and factors that may affect
the interpretation of past diatom compositions. Here, the lower Save River include the delta area and c. 70 km upstream.
Most of the global stratigraphic diatom investigations are from the northern hemisphere (Holmgren et al., 2012). Along the east African coast, diatom studies are most frequently conducted in South Africa (e.g. Schuette & Schrader, 1981; Talbot & Bate, 1987; Abrantes, 2000). Norström et al. (2012) conducted a paleoenvironmental study on the Macassa Bay, Mozambique, which is located c. 150 km south of the investigated area. At present, Massuanganhe et al. (2016a) is the only study from the Save River delta.
Parts of the Save River delta are occupied by mangrove forest, which is habitat for complex ecosystems (e.g. Lugo & Snedaker, 1974; Burchett et al., 1984; Smith et al., 1991; Ball &
Pidsley, 1995; Lee, 1999). This vegetation is currently experiencing an increasing degradation (Woodroffe & Grime, 1999; Santos et al., 2014; Chaudhuri et al., 2015), argued to be caused by human activities (Bandeira et al., 2009; Erftemeijer & Hamerlynck, 2015) and/or by changing coastal dynamics as a result of eustatic sea level fluctuations (e.g. Ellison &
Stoddart, 1991; Fujimoto et al., 1996; Behling et al., 2004; Reinhardt et al., 2010; Srivastava
& Farooqui, 2013). Therefore, researchers of previous studies on mangrove wetland dynamics (e.g. Woodroffe et al., 1985; Ellison & Stoddart, 1991; Fujimoto et al., 1996) encourage further studies on coastal areas to increase the understanding of processes within these environments (Nicholls et al., 1999, Nicholls, 2004, Gedan et al., 2011).
Diatom frustules can be affected during sedimentation and after accumulation due to their sensitivity to physical, biological and chemical changes, i.e. taphonomy (Round et al., 1990;
Bennion, 1995; Hillebrand & Sommer, 2000; Riviera & Diaz, 2004; Hassan et al., 2006;
Korhola, 2000). This phenomenon refers to processes affecting diatom frustules after their death. Dynamic environments, such as deltas, imply several of these mechanisms, which can bias the interpretation (Brzezinski et al., 1999; Kato et al., 2003; Ryves et al., 2013).
Sediment traps are often used to determine processes on taphonomy in an area (e.g. Kato et al., 2003; Cameron, 1995). Comparisons between surface and fossil diatom taxa can,
however, also indicate taphonomic mechanisms (cf. Barker et al., 1999; Sawai, 2001). Ryves et al. (2009) emphasize the importance of using living diatom assemblages (biocoenosis) and
fossil communities (thanatocoenosis) to analyze taphonomy, i.e. the connection between present environment and the fossil signal. In tidal environments processes as erosion, resuspension and redepositing of sediments occur on diurnal basis (de Blij et al., 2004).
Taphonomy involves e.g. breakage and/or repositioning of frustules. Breakage can occur during both reworking and compaction of sediments. Biological and chemical processes mainly include presence of roots and their uptake of dissolved silica. Dissolution of silica in soil increases during high temperatures and bacterial activity (Struyf et al., 2005). Diatom frustules can then be affected by chemical corrosion in varying extent (cf. Massuanganhe et al., 2016a) as they have different thicknesses, thus thinly silicified diatoms are less resistant (Castro et al., 2013; Ryves et al., 2013; Brzezinski et al., 1999; Jørgensen, 1955; Lewin, 1961).
Fossil diatom frustules can be of autochthonous or allochthonous origin. Identification of origin of certain species may be useful in environmental studies, as it facilitates
interpretation and understanding of geomorphological processes. Marine planktonic species are known to be allochthonous and represent the tidal water influence. Benthic freshwater taxa are proposed to be autochthonous and characterize the input from the river (Vos & de Wolf, 1993:1). Simonsen (1969) suggests benthic species to be autochthonous thus providing the most reliable information. Taphonomic processes do, however, aggravate analysis of allochthonous and autochthonous species (Andrews, 1972; Beyens & Denys, 1982; Vos & de Wolf, 1993:1). Therefore, further studies of surface and fossil diatom assemblages are required to increase knowledge about taphonomy and effects on diatom frustules.
Furthermore, there is a widespread problem with diatom taxonomy (species identification) and corresponding ecology (Mann, 1999). To ensure accurate interpretations it is
recommended to carefully study the morphology under both light microscope and with SEM Scanning Electron Microscope). In general, the study of environmental conditions in an investigated area can be used to enhance literature information (John, 1987; Castro et al., 2013; Zong & Horton, 1998; Zong et al., 2006; Mann, 1999).
1.1 Objectives
The main aims of present study are to increase knowledge about:
(i) diatom distribution in the lower Save River, (ii) diatom taxonomy and ecology,
(iii) intrusion of marine water into the lower Save River, (iv) taphonomic processes affecting the diatom flora.
1.2 Background
This study is an off-spring of the doctoral thesis by Elidio Massuanganhe (2016). In his thesis, Massuanganhe used siliceous microfossils to evaluate geomorphological and environmental dynamics within the Save River delta. His results, however, showed only scattered
occurrences of diatoms in the investigated sequences and therefore taphonomic processes were discussed. In June 2015, an opportunity was opened to join during the field trip to the area. In connection with this, a number a samples were collected in order to receive more information on diatom distribution and taphonomic processes.
Four subjects are presented here as they are essential for the forthcoming results and discussions.
1.2.1 Salinity stratification
Salinity in seawater is varying with depth as density increases with greater amounts of dissolved salt. Between the surface water of low salinity and the saline deep water there is a layer of rapid change in salinity, i.e. the halocline (Trujillo & Thurman, 2014). Accordingly, different local stratification patterns arise in coastal areas. In dynamic environments such as deltas, a vertical and horizontal salinity gradient is present from the river to the ocean (Cameron & Pritchard, 1963; Pritchard, 1967). The stratification is fluctuating depending on effects from mainly tides, waves, currents, rainfall in the catchment area and amount of evaporation in the delta area. Amounts of suspended sediment do, however, also influence the stratification since it increases the density (Semeniuk, 2016).
In delta environments with perennially flowing rivers, the landward part of the water can consist of almost entirely freshwater. If the river flow is seasonally, a mixture from fresh- to marine water is present. During intense river flows, plumes of freshwater can reach out into the ocean. The stratification in the delta is thus depending on both seaward processes and the features of the river (Haas, 1977; Geyer & Farmer, 1989; MacCready, 1999; Semeniuk, 2016).
During low tide the water in deltas is highly stratified with an almost vertical halocline and a thinning of the freshwater layer towards the delta front. When high tide is present, marine water is pressed further into the river channel and the halocline slightly decline landwards (Largier, 1986; Largier, 1992). Flooding events imply a halocline with a steep gradient landwards as freshwater pressure is high. During both high tide and floods, the freshwater and brackish/marine water is mixed in a greater extent than during low tide (Largier &
Taljaard, 1991; Semeniuk, 2016). Estuaries with several river channels leading to the ocean increase the complexity of mixing and stratification even more (Semeniuk, 2016).
1.2.2 Diatoms
These organisms have been studied since the late 18th century. The initial taxonomy work was done during the early 19th century by Müller, Nitsch and Gray. Further fundamental research was performed by e.g. Ehrenberg, Kützing, W.W. Smith, Gregory, Greville, Ralfs, Donkin, Grunow, P.T. Cleve, A. Cleve-Van Heurck and Hustedt (Ferguson Wood, 1967).
Diatoms are siliceous microfossils and have cell-walls composed of amorphous hydrated silica. They have two shells (also named frustules), which are attached to each other (Lowe &
Walker, 1997). The outer shape of the diatom frustule is divided into two groups; pennate or centric. Pennate diatoms have bilateral symmetric valves and centric are radial symmetric.
Centric diatoms cannot move by themselves; however, some pennate diatoms are capable of moving by the raphe structure (Round et al. 1990; Krammer & Lange-Bertalot, 2000).
These unicellular algae belong to Bacillariophyta (Brasier, 1980; Round et al., 1990), which is the most species-rich group (Mann, 1999). The length of the frustules is generally between c.
5 and 2000 µm, but size can diverge, especially during the early and late stages of the life cycle (Brasier, 1980; Round et al., 1990; Lowe & Walker, 1997).
Siliceous microfossils live in aquatic and semi-aquatic environments, both as planktonic and benthic. Planktonic species are suspended in the pelagic zone and benthic are attached to sediment surfaces and vegetation in the benthic zone. There are, however, benthic diatoms which can be brought up into the plankton habitat through e.g. currents and winds. These are referred to as tychoplanktonic species (Ferguson Wood, 1967; Round et al., 1990).
Diatoms can also be redeposited from their habitat through for example birds (Atkinson, 1980; Figuerola & Green, 2002) and boats (Hallegraeff & Bolch, 1992).
Access to light is an essential factor for diatoms. They are photosynthetic and grow in the photic zone (Round et al., 1990). Diatom blooms, i.e. flourish of species, occur during favorable conditions such as during the spring. Conditions, which imply access to light, optimal temperature and pH, and access to silica, nitrogen and phosphorous are beneficial for diatom reproduction (Furnas, 1990; Round et al., 1990; Martin-Jézéquel et al., 2000;
Litchman et al., 2008).
Diatoms have a significant role in the global carbon, silica and oxygen cycles and therefore contributes to the global ecosystem primary production (e.g. Hsaio, 1988; Cota et al., 1991;
Glud et al., 2002). Furthermore, diatoms are alone accountable for a quarter of the inorganic carbon (e.g. CO2)fixed in the oceans every year, nevertheless researchers have not clarified all aspects of the complex processes which fix the inorganic carbon (Granum et al., 2005).
During phases of cell division in the diatom life cycle, silica is naturally dissolved and partly used by other organisms and vegetation (Mann, 1999; Smetacek, 1999). When the diatom cell divide, cell walls usually become significantly thinner as energy and silica are used to form new frustules (Ehrlich & Newman, 2008). Diatoms reproduce through vegetative division, which implies a cell division inside the parent cell. Therefore, daughter cells are significantly smaller than parent cells. The new cells have one new frustule, which is formed during the division, and one from the parent. Reduction of size of the parent cell thus occurs during every reproduction. The life cycle of a diatom thus implies several stages of frustule size, which is termed “size reduction series” and means young and old diatoms have frustules reduced in size. This is an important factor to consider during diatom analysis (Round et al., 1990; Falkowski & Knoll, 2007; Hense & Beckmann, 2015).
1.2.3 SEM/ESEM
Scanning Electron Microscopes (SEM) are microscopes which allow magnification to atomic level. In SEM the chamber, where the sample for analysis is placed, is a gaseous vacuum environment. Instead of light, which is used in a light-microscope, electrons are used in an SEM to reflect the surface of the specimen. The microscope has a primary concentrated electron ray which is used towards the sample. The secondary rays are collected and create a detailed visualization of the surface of e.g. diatoms (Danilatos, 1993).
Environmental Scanning Electron Microscope (ESEM) has usually the same basics
specifications as SEM, but ESEM tolerates even higher pressure in the chamber. It also allows the device to operate at conditions other than only high vacuum by using different sets of detectors (Danilatos, 1988). SEM and ESEM are frequently used for investigate frustule structures and identification to species level (cf. e.g. Siver et al. 2003; Ponander & Potapova, 2007).
1.2.4 OSL
Optically Stimulated Luminescence (OSL) dating relies on measurements of light
(luminescence) emitted from light sensitive electron traps in crystals, where quartz and feldspar comprises the most commonly used mineral types (Lowe & Walker, 1997). OSL ages ideally represents the timing of burial or last exposure to sunlight of sedimentary deposits After burial the luminescence signal gradually accumulates due to background radiation occurring in the surroundings of the deposit. When the minerals are exposed to sunlight the luminescence signal is zeroed (Yukihara & McKeever, 2011; Preusser et al., 2008). For dating applications, the luminescence signal is released in laboratory conditions through artificial stimulation, during which the signal is measured and further related to a radioactive dose.
Stimulation with green light is typically used for quartz, whereas infrared light are used for feldspar (Lowe & Walker, 1997; Yukihara & McKeever, 2011; Preusser et al., 2008). The luminescence measurements are used to determine the amount of radiation the sample has been exposed to during time of burial. For age calculations, the amount of dose is divided with the rate of radioactive energy acting on the sample per year, also termed dose rate. The dose rate can be measured in field or in laboratory (Preusser et al., 2008).
The luminescence signal can, however, be incompletely zeroed prior to deposition of the sediment, also referred to as incomplete bleaching. This occurs when the minerals have not been sufficiently exposed to sunlight, which results in an overestimation of the age of burial since a signal is preserved. Transport by and/or deposition in water may also increase the risk for incomplete bleaching. During age calculation, it should be considered whether these factors are likely to have affected sediments and certain statistical approaches can detect if samples are incomplete bleached. If so, several measurements of the absorbed dose are made to investigate the dose distribution (Bailey & Arnold, 2006). Furthermore, sedimentary water content in the samples collected for OSL dating can influence the dose rate and could imply uncertainties for the final age.
2. Description of the investigated area
The investigated area includes mainly the Save River delta but also a sampling site c. 70 km up-streams, located in the south-central part of Mozambique (Figure 1). The catchment area for Save River basin is located mainly in the eastern parts of Zimbabwe and covers c. 102,000 km2 (Massuanganhe, 2016). With its classical triangular shape (de Blij et al. 2004) the delta shows similarities, however minor, with the Nile Delta.
The distance of the Save River deltaic plane is c. 20 km from its front to the pre-Holocene bedrock. Caused by south north coastal current it has an elongated shape with a distance of c. 65 km (cf. Massuanganhe et al. 2016b, Figure 1). The investigated area westward from the deltaic plane represents the feeding river channel. The bedrock in the upper part of the drainage system consist mainly of intrusive igneous rocks (Le Maitre, 2002; Schlüter, 2006), which cause acid water conditions. The formation of this sedimentary basin was initiated during the Carboniferous and the Triassic periods (c. 350- 200 Ma ago) (McElhinny & Briden, 1971; Salman & Abdula, 1995). The lower part of the drainage system and the delta area is underlain by calcareous bedrock, which implies alkaline environments. The Save River Delta most likely formed during early Holocene epoch (c. 7-8 cal. yrs BP) when the mean sea level reached approximately the present situation (Fleming et al., 1998; Milne et al., 2005).
The East African rift system is located north of the Save River Delta. It is c. 6000 km long reaching Ethiopia in the north. Several earthquakes have occurred during the last decade along the distal parts of the Save River, indicating tectonic activity in the surrounding area (Chorowicz, 2005). Macassa Bay c. 120 km south of the investigated area has, however, been tectonically stable during the late Quaternary period and no activity has been recognized during the Holocene epoch along the southeastern African coast (Miller et al., 1993; Ramsay, 1995; Ramsay & Cooper, 2002).
Figure 1. An overview map of the investigated area. A – Shows the location of the investigated area in southeastern Africa. B – The subareas within the investigated area and M32, which is located further upstream the Save River. Subarea 1 shows the lower part of the Save River delta, Subarea 2 shows the upper part of the delta and Subarea 3 shows the riverine part. C – SPOT image taken in 2011 with red- green-blue band combination showing landscape structures. The darker red color in the coastal area indicates mangrove forest distribution. Sampling sites are labelled M1-32. M16 and M27 are not shown as these sites are not used in the study. Color and shape of site symbols indicate collected sample type. “M”
represents sites sampled for the present study and “P”
sites used by Massuanganhe (2016). White rectangles display the three subareas.
The coastline of Mozambique is a tidal environment and influenced by southeastern winds and northward ocean currents (Sete et al., 2002) transporting large amounts of sand. When deposited, sand dunes and spits pointing northward are formed.
Mozambique has a humid equatorial climate with dry winters (de Blij et al., 2004) and the rainy season is from October to March (De Boer et al., 2000; Yang et al., 2015). Local
evaporation increases during February to April when the sea surface temperature reaches its maximum (Rouault, 2003; de Blij et al., 2004).
Extreme rainfall within the drainage area cause flooding events in the investigated area and further up-stream the river channel. These, together with the tides, form local differences in salinity stratification in the water (Semeniuk, 2016). The most severe flooding events
commonly co-occur with tropical cyclones. Several major cyclones have affected the area during the last 15 years. During floods, the transport and reworking of sediment in the Save River and its delta increase (Massuanganhe et al., 2015). These events and tides amplitudes of c. 4.5 m (Sete et al., 2002) cause the investigated area to be highly dynamic concerning geomorphological processes.
Parts of the Save River deltaic plane are occupied by mangrove forest (Figure 2), an
ecosystem with rich biodiversity. This type of vegetation is well adapted to rapid changes of e.g. temperature and salinity, which are natural stressors present in coastal environments (Kathiresan & Bingham, 2001). Mangrove thrives in anoxic mud and has characteristic roots, which are partly growing above the water surface to breathe. The roots are efficient
sediment traps and form layers of fines with high organic content (Scoffin, 1970; Kristensen et al., 2008). Intense erosion and reworking of sediment in coastal environments do,
however, occasionally cause mangrove dieback (Alongi, 2002; Massuanganhe et al., 2016a).
During field work for the present study mangrove forest occurred from the delta front to approximately M13 (Figure 1). Mangrove forests are commonly halophytes, resulting in high tolerance to a wide range of salinity and high levels of salt in the water (Thom, 1967;
Fujimoto et al., 1996), however, they are mostly found in estuaries and rarely in freshwater environments (Gilmore & Snedaker, 1993). The pH in mangrove sediments and soils is
usually around 7, but can range between 4 and 8 (Middelburg et al., 1996; Boto &
Wellington, 1984; Joshi & Ghose, 2003).
Figure 2. Mangrove forest in the Save River delta with characteristic roots. Parts of the root systems grow above the water surface to increase uptake of oxygen as mangrove thrive in anoxic mud.
3. Methodology
The material analyzed and interpreted in the present study was collected by the author in the lower Save River (Figure 3). All together 32 sites were visited, however, samples M16 and M27 are not included. M16 was destroyed during transportation to Sweden and M27 was collected as it was most likely aeolian sediment, later considered to be irrelevant for the aim of this study (Table 1). To give a comprehensive distribution of diatoms in the lower Save River, diatom records from sites P1, P2, P3, P6 and P8 (Massuanganhe et al., 2016a) is incorporated and compared with records from the present study. P1, P2, and P3 have the same coordinates as M20, M14 and M24 (Table 1). At site P2/M14 buried sediments have been collected by Elidio Massuanganhe and the author of this study. To compile a further comprehensive study, diatom records from these two sites have been combined. Samples from the present study and Massuanganhe et al. (2016a) are equally labeled according to the site names. Samples/sites beginning with “M” are consequently collected/visited by the author, and samples beginning with “P” are collected/visited by Elidio Massuanganhe.
Table 1. Details of visited sampling sites. Samples at M sites are collected by the author of this study and Elidio
Massuanganhe has collected the sediments from P sites. Coordinate differences for P1/M14, P2/M20 and P3/M24 could be related to erosion between sampling events.
Site Latitude (South)
Longitude (East)
Character of sample*
Physiography Classification of sediments**
M1 20° 57' 36.72" 35° 07' 30.47" SW, SS, MC Shoreline Sand
M2 20° 57' 40.96" 35° 07' 24.31" SW, SS, MC Back-swamp Fines
M3 20° 57' 46.69" 35° 06' 57.02" SW, SS, MC Sandbank middle of river channel Sand M4 20° 57' 43.23" 35°06' 27.54" SW, SS, MC Riverbank. SW; M4a: Riverside, M4b:
Middle of river channel
Fines
M5 20° 57' 43.48" 35° 05' 52.30" SW, MC Sandbank middle of river channel Fines
M6 20° 57' 6.731" 35° 04' 44.44" SW, MC Riverbank Fines
M7 20° 53' 51.82" 35° 05' 41.96" SS Shoreline Fines
M8 20° 54' 37.00" 35° 04' 13.37" SS Riverbank Fines
M9 20° 55' 40.94" 35° 03' 56.05" SW, SS, BS, OSL Riverbank Fines and sand
M10 20° 57' 35.74" 35° 03' 34.52" SW, SS River bay Fines
M11 20° 58' 05.66" 35° 03' 8.89" SW, SS Riverbank Sand
M12 20° 58' 08.01" 35° 03' 11.09" SS SS; M12a: Overflow area, M12b: River channel
M13 20° 58' 37.59" 35°02' 38.54" SW Riverbank Sand
M14 20° 59' 14.81" 35° 00' 39.16" SW,SS, BS, OSL Riverbank Fines and sand
M16 21° 02' 36.45" 34° 52' 26.00" SS Riverbank. Not included in the study, destroyed during transportation.
Fines
M17 21° 03' 22.07" 34° 51' 51.87" SS Building area for irrigation pipes Fines
M18 21° 04' 01.70" 34° 51' 36.03" SS Sand/gravel pit Sand
M19 21° 02' 52.44" 34° 53' 15.43" SS Pond Fines
M20 21° 00' 22.90" 34° 56' 21.58" SW River
M21 20° 59' 07.66" 35° 00' 46.51" SS Riverbank Fines
M22 21° 01' 04.26" 35° 01' 39.83" SS Overflow levee Fines
M23 21° 01' 30.04" 35° 01' 56.28" SW River M24 20° 59' 05.86" 35° 02' 06.07" SW River
M25 20° 58' 19.92" 35° 02' 29.47" BS Riverbank Fines and sand
M26 20° 56' 43.18" 35° 02' 52.58" SW M26a: Overflow area, M26b: River channel
M27 20° 56' 24.97" 35° 02' 9.92" SS. Most likely aeolian sediments
Riverbank. Not included in the study, irrelevant.
Sandy silt
M28 20° 56' 11.47" 35° 02' 46.28" SW, SS Riverbank Fines
M29 20° 56' 47.00" 35° 02' 50.89" SW, SS Riverbank Fines
M30 20° 59' 36.27" 35° 59' 55.42" SS Riverbank Fines
M31 21° 01' 37.88" 34° 54' 23.25" SS Riverbank Sand
M32 21° 07' 58.37" 34° 33' 55.15" SW, SS, Riverbank Sand
P1*** 21° 00' 22.96" 34° 56' 21.44" BS, OSL, 14C Riverbank Fines and sand P2*** 20° 59' 14.82" 35° 00' 39.17" BS, OSL, 14C Riverbank Fines and sand P3*** 20° 59' 06.71" 35° 02' 05.55" BS, OSL, 14C Riverbank Fines and sand P6*** 21° 01' 24.04" 35° 04' 05.87" BS Transition upper-lower deltaic plane Fines and sand P8*** 21° 01' 12.77" 34° 59' 35.22" BS Transition upper-lower deltaic plane Fines and sand
*SW: Surface water. SS: Surface sediments. BS: Buried sediments. MC: Mangrove cortex
**Classified during field work.
***Massuanganhe et al. (2016a)
Figure 3. The tributary river channel in the Save River delta. Mangrove forest grows in parts of the delta and especially along the river channel. The picture is taken close to M5.
The described fieldwork was performed June 9-20, 2015, by the author and concerns M1- M32. Four types of samples were collected; surface sediments, buried sediments, surface water and mangrove cortex. Samples were titled with the same name as the site where they were collected.
3.1 Siliceous microfossils
Samples for siliceous microfossil analysis were collected from surface sediments, buried sediments, surface water and mangrove cortex.
Surface sediment samples were collected with a small spade from ground surfaces. Buried sediments were sampled using the same technique, but collected from riverbanks (Figures 4 and 5). When surface water samples were collected, 1.5 liter plastic bottles were filled at maximum 50 cm water depth. Samples were taken from boat or from the river bank. Bottles were decanted after c. 12 h of sedimentation (Figure 6). Decanting was repeated every second hour until c. 30 ml remained. Samples containing large amounts of sand were stirred and decanted after 5 s of sedimentation. Remaining 30 ml of water were kept in tubes of 45 ml during transportation to Sweden. Two pH measurements were made with litmus in the middle of the river channel at M1 and M10, both indicating neutral pH levels.
The outer most cortexes were scraped with a knife from living mangrove trees growing in water. All scraping were made below the water surface (at the time of sampling).
Figure 4. A section from the river bank of the Save River with vertical layers of alluvial sediment with different grain sizes.
The picture is taken close to M11. Surface and buried sediment samples were taken from similar sections along the river bank.
Figure 5. Parts of a section showing a fine grained layer at M9, which was sampled for diatom analysis. Sand with iron precipitation is present below and above.
Subsamples of 0.5-1.0 cm3 were extracted from surface and buried sediments and put into 100 ml beakers. Regarding samples from mangrove cortex, the outermost from the cortex were scraped off and also put into 100 ml beakers. Surface water samples were decanted after 2 hours of sedimentation and then poured into 100 ml beakers. 10 % HCl were added to remove carbonates and organics were removed by boiling samples in 17-35 % H2O2 until reaction terminated (Battarbee, 1986). Samples were then repeatedly decanted to remove sand and clay particles using settling time times in water based on Stoke’s law. To dissolve clay flocculates NH3 was added and the decanting procedure repeated until the liquid was clear. Remaining water and fractions of silt-size were mounted in Naphrax® on microscope slides to increase the refraction index.
Figure 6. Plastic bags containing surface and buried sediment samples. The 1.5 liter plastic bottles are samples collected from surface water. They were repeatedly decanted until c. 30 ml of water and sediments remained. Residuals were poured into plastic tubes of 45 ml and transported to Sweden.
Siliceous microfossils were analyzed under a Zeiss Axiophot light microscope using
immersion oil and X1008 magnification. Frustules were identified following Foged (1975), Gasse (1986), Snoeijs (1993), Snoeijs & Vilbaste (1994), Snoeijs & Potapova (1995), Snoeijs &
Kasperovičienė (1996), Krammer & Lange-Bertalot (1986, 1988, 1991a, b, 2000), Metzeltin (1998) and Witkowski et al. (2000). For practical reasons and available literature, new names of certain species have not been applied. For example Synedra ulna and Biddulphia aurita
are at present named Ulnaria ulna and Odontella aurita, respectively (Lange-Bertalot &
Compére, 2001; Hoppenrath, 2004).
Eight groups based on salinity tolerance of diatom species were established; marine,
brackish, halophilic, indifferent, freshwater, aerophilic, unknown and extinct. Halophilic taxa prefer increased salinities, which freshwater species do not. Indifferent taxa tolerate low salinity levels, they can thus live in both fresh and brackish water. The group unknown includes diatom frustules, which were not possible to identify because of breakage, chemical dissolution or large amounts of mineral fractions covering the frustule. Characteristic/
commonly occurring species for a sample were based on frustule occurrences in relation to other taxa.
Diatom assemblages within samples from surface sediments, surface water and mangrove cortex are presented in pie charts on SPOT images of the investigated area. The groups brackish and halophilic, indifferent and aerophilic were combined to explicate results. Based on the same argument, the investigated area was divided into subareas 1, 2 and 3. Data for the pie charts is presented in a geographical order perpendicular to the river, from east to west (Appendix 1). M22 and M23 are not incorporated into the pie charts as they are not located close to the present river channel. Samples where no frustules were found are not visualized. Samples with low basic sums (<50 diatom frustules) were interpreted accordingly.
Diatom assemblages from buried sediment samples are presented as diagrams made in Tilia 1.7.16. P2/M14 is presented as percentage and P1, P3, P6, P8, M9 and M25 are presented as counts since basic sums were too low. Diatom assemblage zones within P1, P3, P6 and P8 are modified from Massuanganhe et al. (2016a) to emphasize data important for the aim of this study. In Massuanganhe et al. (2016a) zones are based on the lithological units while in the present study they are defined according to diatom occurrences and ecologies (cf.
Appendices 2 and 3). The lithology is based on Troels-Smith (1955) classification of
sediments. Cluster analysis based on CONISS and eye matching was added into P2/M14 to display zonation. Cluster analysis was not applied to diagrams with counts.
3.2 ESEM
Residuals from surface sediment samples M1-M4, M6, M10, from mangrove cortex samples M1, M2, M5 and from surface water sample M1. Subsamples were paved with gold and studied using ESEM (FEI, Quanta FEG 650) high vacuum. Diatom taxa particularly examined to identify to species level were e.g. Diploneis interrupta, Opephora minuta, Luticola mutica, Diploneis sp. and Cyclotella sp.
3.3 OSL
Two OSL-samples were collected at site M9. Opaque plastic tubes were driven into the sandy layers. During the sampling procedure, black tarpaulin was used for light protection and tubes were then covered with several layers of black tape (Figure 7). OSL-sample M9V was collected vertically because of rising tidal water, which implied limited time for digging.
Approximately 1 kg of sediments was collected for dose rate measurement. Preparation technique of luminescence samples is presented in Massuanganhe et al. (2016a). Sediment for dose rate measurements for M9 and M9V was sent to VKTA, Laboratory for
Environmental and Radionuclide Analyses, Dresden, Germany.
Figure 7. The vertical OSL samples from M9. Black tarpaulin was used for light protection during collection. The plastic sample tube was covered with black tape to ensure tight coverage.
4. Result and interpretation
Four categories of material have been collected; surface water (20), surface sediments (21), mangrove cortex (5) and buried sediments (16). The latter category has partly been collected by Elidio Massuanganhe and analyzed by Annika Berntsson and the author of this study
(Massuanganhe et al., 2016a). During analysis 63 diatom genera and 258 species were identified (Appendix 3). Phyotoliths, sponge spiculae and chrysophyte cysts were identified but not counted as they were not relevant for the objectives of this investigation. Most common species are Nitzschia granulata, N. littoralis, Hantzschia distinctepunctata,
Cyclotella sp., Thalassiosira eccentrica, Diploneis sp. and D. interrupta. (Appendices 3 and 6).
Numbers of brackish and marine diatom species are overall higher in samples from surface water, surface sediment and cortex than in buried sediments, exemplified by Amphora ventricosa, Opephora pacifica and Navicula alpha.
Diatom assemblages in surface sediments, surface water and mangrove cortex are presented as pie charts overlying geographical maps to show the salinity gradient in the lower Save River (Figures 8, 9, 10 and Appendix 1). Identified diatoms are grouped according to associated ecology and presented as percentages. The basic sums of identified frustules in each sample vary between 0 and 353.5 (Figures 8C, 9D and 10B). Differences in diatom assemblages indicate a spatial variation in salinity. Marine and brackish species are generally decreasing up-streams while indifferent and freshwater taxa are increasing. Diatoms can thus indicate a latitudinal salinity gradient in the lower Save River. Brackish taxa are present in nearly all samples, which indicate influences of tidal water up-stream.
4.1 Surface water samples
20 surface water samples were collected (Figure 8 A and B). No surface water samples were collected in Subarea 3. In Subarea 1, brackish and halophilic taxa are dominant (Figure 8 A).
The number of indifferent and aerophilic taxa in the diagram at site M1 is relatively low as the basic sum is only eight and should be interpreted accordingly (Figure 8 B). The sample at site M11 shows a high occurrence of marine species in relation to samples closer to the ocean.
Samples from Subarea 2 show a significantly higher occurrence of brackish and halophilic taxa than in Subarea 1 (Figure 8 B). Low numbers of marine diatoms are identified in the sample at site M13. The occurrence of indifferent and aerophilic taxa starts to increase at site M24 and continuous in samples at site M20 and M32. The occurrence of freshwater diatoms is, however, limited in water samples throughout the delta and the river channel.
4.2 Surface sediment samples
21 surface sediment samples were collected (Figure 9 A, B and C). Brackish taxa are
dominating most of the samples from Subarea 1 (Figure 9 A). M11 contain a relatively high number of freshwater species. In sample M25 marine taxa occur in the same proportions as in Subarea 1 (Figure 9 B). In the other two samples visualized in Subarea 2 (M21 and M30) the indifferent and aerophilic taxa increase and marine species decrease. The brackish taxa are still dominating. Amount of freshwater species are significantly higher in M31 and M32, Subarea 3, (Figure 9 C), however, they do not occur in M15, M17 and M18. Brackish taxa are highly represented in M15 and M17. Note the low basic sums in samples M15, M17 and M18 (Figure 9C).
4.3 Mangrove cortex samples
During fieldwork five mangrove cortex samples were collected. Samples from mangrove cortex were collected within Subarea 1 at sites M1-M5 (Figure 10), i.e. in the delta front. The diatom communities indicate mostly brackish water conditions. In M1 and M3 there is a marine signal, however, the basic sum in M1 is too low to be valid (Figure 10). The diatom assemblage in M5 indicates some influence of freshwater since frustules of Cymbella spp are present.
Figure 8 A. Identified diatom frustules from surface water samples. Counts are grouped according to optimal salinity and presented as percentages in pie charts. Colors in the pie charts represent different salinity levels and are explained in the legend. The overview map in the lower right corner display the geographical position of subarea 1.
Figure 8 B. Counted diatoms from surface water samples within Subarea 2. The table below the map display basic sums for each sample. See Figure 8 A for additional details.
M32 267
M29 120
M28 194.5
M26b 353.5
M26a 177
M24 94
M20 39.5
M14 53
M13 91
M12 229
M11 136
M10 49
M 9 33
M6 97.5
M5 72.5
M4b 37
M4a 32.5
M3 44
M2 43.5
M1 8
Sample Basic sum
Figure 9 A. Diatom counts from surface sediment samples displayed as percentages. Counts are grouped according to optimal salinity. The colors within the pie charts represent different salinities and are explained in the legend in Figure 8A.
Overview map in the lower right corner show the geographical position of Subarea
1
.
Figure 9 B. Diatom counts from surface sediment samples within Subarea 2. See Figure 9 A for further details.
Figure 9 C. Diatom frustules from surface sediment samples within Subarea 3. See Figure 9 A for further details. The table below the figure show basic sums for the samples.
M32 7.5
M18 11
M17 1
M15 14.5
M31 36
M30 205
M21 138
M25 134
M12b 202
M12a 208.5
M11 33
M29 110
M28 63.5
M10 162.5
M6 369
M8 14
M4 42
M3 70
M7 91
M2 204.5
M1 32
Sample Basic sum
Figure 10. Identified diatoms from mangrove cortex. Pie charts contain identified frustules counted as percentages and represented with different colors depending on their optimal salinity. Legend of the colors are available in Figure 8 A. The table below the map show basic sums for the samples. Geographical position of Subarea 1 is showed in the overview map in the lower right corner.
M5 359
M4 75
M3 54
M2 45
M1 4.5
Sample Basic Sum
4.4 Buried sediments
Buried sediment samples were collected at P1, P2/M14, P3, P6, P8, M9, and M25 and presented in seven Tilia diagrams (Figures 11-17). Detailed lithologic information of P1, P3, P6 and P8 is modified from Massuanganhe et al. (2016a). 16 samples were collected at M- sites and 126 samples at P-sites. Samples P2 and M14 are merged as their geographical location coincide. The lithology at all sites shows a clay layer bracketed by units consisting of sand and/or silt.
The diagram from P1 is divided into two zones, however diatom frustules only occur in zone 1) (Figure 11). Basic sums of diatoms are high in the lowest part of the core, c. 620-560 cm depth. In zone 1 variations between the ecological groups are minor. The frustules counted represent mainly brackish (e.g. Diploneis interrupta and D. pseudovalis) and indifferent conditions (e.g. Amphora copulata), however, marine (Nitzschia granulata) and freshwater (e.g. Fragilaria ulna) diatoms co-occur. Two samples were collected for radiocarbon dating suggesting ages of 3216-2980 cal. yrs BP (Poz-67397) at c. 570 cm depth and 1072-956 cal.
yrs BP (Poz-60019) at c. 320 cm depth (Figure 11 and Table 2). At c. 350 cm depth a sample for OSL dating was collected, indicating an age of 1300±160 years (Figure 11 and Table 3).
In P3, zone 1 identified diatom frustules are few, but those occurring represent marine and brackish conditions (Figure 12). Zone 2 contain a higher number of diatoms and the marine and brackish signal is more clear and represented by e.g. Paralia sulcata and Diploneis interrupta. In zone 3 the brackish and marine signal remains strong, additionally
characterized by Nitzschia granulata and Hyalodiscus sp. Aerophilic species (e.g. Hantzschia amphioxys) are highly occurring in zone 4, however, brackish, indifferent and freshwater taxa co-occur, mostly represented by Diploneis interrupta, Fragilaria brevistrata and Eunotia spp. The OSL date at c. 370 cm depth indicates an age of 210±20 years (Figure 12 and Table 3).
Figure 11. Diatom diagram from P1. Sums of identified diatom frustules are displayed as counts. DAZ (diatom assemblage zones) are constructed based on frustule occurrences. The chronology comprises both OSL and radiocarbon dates. OSL is marked with ± and radiocarbon as a time interval.
240 290 340 390 440 490 540 590 640
ep D (cm th )
1300±160
1072-956 3216-2980
hr C ol on y og
Nitzsc hia gran ulat a
2040
Diplo neis inte rrupt a
20406080
Diplo neis psu edo valis
20
Hantzschia distinctepunctata
20
Hyalo discu s sp
. 20
Nitzsc hia cocc one
iformis 2040
Amp hora cop ulat
a Epit
hem ia a dna ta
Rho palo dia gibb a
Rho palo dia ope rcula ta
Diplo neis ellip tica
20
Frag ilaria
ulna Gyro
sigma ob tusa tum
Hantzschia amphioxys
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 1 0 0 0 4 0 3 36 0 28 55 32 97 152 304
Bas ic su m
20
Marin e ta xa
50100150
Brac kish taxa
Halo philic tax
a 20406080100
Ind iffere nt t axa
2040
Fresh wate r tax a
Aero philic tax a
20
Unknown taxa AZ D
2 1
Marine taxa Brackish taxaIndifferent taxa Freshwater taxaAerophilic taxa Modified from Berntsson 2015-2016
thLi og ol y
SiltSandClaySilty sandSilty clay
Figure 12. Diatom diagram from P3. Sums of identified diatom frustules are displayed as counts. DAZ (diatom assemblage zones) is constructed based on frustule occurrences. One OSL date is visible in the chronology.
The occurrence of diatoms is relatively low in samples from P6 and merely 11 of 29 samples contain frustules (Figure 13). In zone 1 there is a brackish-marine signal dominated by mainly Diploneis interrupta and Terpsinoë americana. Zone 2 indicates the same signal as zone 1, however, the occurrence of frustules is higher. Aerophilic species is most common in zone 3.
Significant taxa are Navicula mutica and Pinnularia borealis.
70 120 170 220 270 320 370
ep D (cm th )
210±20
hr C on og ol y
ip D ne lo c is fra af
20
itz N hi sc gr a ul an a at
20
ar P ia al ul s
ta ca 20
ip D ne lo in is rr te ta up
ip D ne lo p is ud se al ov is
an H sc tz a hi st di ct in un ep at ct a
20
ya H di lo us sc
p.vi sre b ria ila ag Fr
ria st ta
un E ia ot
pp s 2040
an H sc tz a hi am io ph s xy
av N ul ic m a ic ut a
av N ul ic pa a m ra ic ut a
18 33 2 17 37 65 4 6 3 52 48 86 60 11
as B s ic um
2040
ar M e in xa ta
2040
ra B is ck ta h xa
al H hi op ta lic xa
di In re ffe ta nt xa
20
es Fr hw er at
xa ta 2040
er A hi op ta lic xa
nk U w no ta n xa
AZ D
4 3 2 1
Marine taxaBrackish taxa Indifferent taxaFreshwater ta
xa Aerophilic taxa Modified from Berntsson 2015-2016
th Li og ol y
SandSilty claySilt
