Potentials and Limitations of Testate Amoebae from Tidal Marshes As Bio-Indicators of Environmental Change in the Scheldt Estuary

Faculteit Wetenschappen Departement Biologie Onderzoeksgroep Ecosysteembeheer

Potentials and limitations of testate amoebae from tidal marshes as bio-indicators of environmental change in the Scheldt estuary

Potentieel en grenzen van het gebruik van thecamoeben uit schorren als bio-indicatoren voor omgevingsveranderingen in het Schelde estuarium

Proefschrift voorgelegd tot het behalen van de graad doctor in de Wetenschappen aan de Universiteit Antwerpen te verdedigen door

Marijke OOMS

Promotor: Prof. Dr. Stijn Temmerman Co-promotor: Prof. em. Dr. Louis Beyens

Antwerpen, 2013

Contents

Samenvatting 5 Summary 9

Chapter 1 13 Introduction

Chapter 2 37 Testate amoebae as estuarine water-level indicators

Chapter 3 65 Testaceae of a brackish tidal marsh

Chapter 4 95 Comparison of testate amoebae assemblages along a salinity gradient

Chapter 5 115 Sea level reconstruction and selective preservation of testaceae

Chapter 6 149 Role of testate amoebae in Si cycle

Chapter 7 173 General Discussion

References 187 Appendices 203 Abbreviations 233 Dankwoord 237

Samenvatting

De huidige stijging van het zeeniveau wordt in estuaria versterkt door een combinatie van de getijwerking en antropogene modificaties aan het estuariene stroombekken. Al deze aanpassingen hebben een diepere landinwaartse doordringing en verdere opstuwing van de getijgolf tot gevolg. Dit resulteert in een verhoogd overstromingsrisico van haven- en kuststeden. Om het overstromingsrisico te doen dalen wordt in vele estuaria extra waterbergingscapaciteit (door ontpoldering) voorzien voor tijden van verhoogd overstromingsrisico (vb. tijdens stormen).

Reconstructie van vroegere waterpeilveranderingen kan helpen om de toekomstige waterpeilveranderingen binnen het estuarium te voorspellen. Waterpeilveranderingen kunnen gereconstrueerd worden aan de hand van protisten (o.a. foraminiferen, diatomeeën en thecamoeben) waarvan de omhulsels (schaaltjes), die de soortspecifieke determinatiekenmerken dragen, kunnen fossiliseren. Er zijn reeds meerdere studies verschenen die aantonen dat deze protisten goede proxies zijn voor het reconstrueren van zeespiegelveranderingen. Deze reconstructies zijn gebaseerd op het gebruik van een transferfunctie, welke de relatie tussen moderne protistengemeenschappen en waterpeilveranderingen gebruikt om vroegere waterpeilveranderingen af te leiden uit fossiele protistengemeenschappen.

In deze studie wordt gefocust op het gebruik van thecamoeben als proxy voor de reconstructie van waterpeilveranderingen binnen het Schelde-estuarium. Thecamoeben behoren tot de protozoa en zijn geschaalde amoeben (gemiddelde grootte tussen 20 µm - 200 µm) die voorkomen in de schorbodem. Ze hebben een korte generatietijd ( ± 60 generaties jaar-1), wat maakt dat ze snel reageren op omgevingsveranderingen en hun schaaltjes, met soortspecifieke eigenschappen, kunnen fossiliseren.

Samenvatting

Moderne thecamoebensoortensamenstellingen van een zoetwaterschor (Notelaarschor), brakwaterschor (Groot buitenschoor) en zoutwaterschor (Paulina) werden onderzocht. Op het zoutwaterschor werden onvoldoende thecamoeben teruggevonden om te spreken van thecamoebengemeenschappen (450 schaaltjes g-1). De thecamoebensoortensamenstellingen van het zoet- en brakwaterschor varieerde in de eerste plaats met schorhoogte t.o.v. gemiddelde hoogwaterpeil. Deze relatie werd gebruikt om een “zoete” en “brakke” transferfunctie voor waterpeilveranderingen te maken. De resulterende transferfuncties hebben een nauwkeurigheid die vergelijkbaar is met andere gepubliceerde transferfuncties voor zeespiegelreconstructies.

Sub-fossiele thecamoebengemeenschappen werden onderzocht in boorkernen van het zoetwaterschor (Notelaar). De zoetwater transferfunctie, gebaseerd op de moderne data van de Notelaar, werd gebruikt om vroegere waterpeilveranderingen te reconstrueren. Uit vergelijking van de gereconstrueerde waterpeilveranderingen met gemeten waterpeilveranderingen, is duidelijk geworden dat de reconstructiemethode op basis van fossiele schaaltjes werkte tot een diepte van 50 cm (~ 1965). De limitatie van de transferfunctie werd veroorzaakt door een snelle daling in de concentratie van fossiele thecamoeben in de schorbodem, waardoor nog maar weinig schaaltjes overblijven dieper dan 50 cm.

Onze studie van de afname van fossiele thecamoebenconcentraties met de diepte hebben aan het licht gebracht dat er mogelijk selectieve oplossing is, waarbij voornamelijk thecamoeben met biogene silica schaaltjes (idiosome thecamoeben) snel verdwijnen. Aangezien de biogene idiosome thecamoebenschaaltjes snel oplossen, is de relatie tussen de silicium cyclus en thecamoeben verder onderzocht.

Samenvatting

Schorren spelen een belangrijke rol in de Si cyclus, omdat zij de hoeveelheid opgelost silicium in het estuarium aanvullen in tijden van tekorten om zo toxische algenbloeien tegen te gaan. Tot nu toe is de afkomst van dit opgelost silicium toegeschreven aan oplossing van diatomeeën en fytolieten (biogene siliciumstructuren van planten). Ons onderzoek van proportionele afname van idiosome thecamoebenschaaltjes, diatomeeën en organisch materiaal gehalte (als proxy voor fytolieten) in de bodem, heeft aangetoond dat thecamoeben een belangrijke rol spelen in de Si cyclering. Hun relatief snelle oplossing zorgt ervoor dat ze gemakkelijk bijdragen bij het export van opgeloste silicium naar het estuarium in tijden van opgeloste silicium limitatie.

We kunnen besluiten dat er een goede relatie is tussen moderne thecamoebensoortensamenstellingen en de hoogteligging van de schorbodem ten opzichte van het gemiddelde hoogwaterpeil. De thecamoebenschaaltjes fossilizeren echter slecht en er komt selectieve bewaring van thecamoeben in de schorbodem voor. Hierdoor zijn thecamoeben toch geen geschikte bio-indicatoren voor het reconstrueren van waterpeilveranderingen in estuaria. Ons verkennend onderzoek over de bijdrage van thecamoeben in schorsedimenten aan de Si cyclus zijn veelbelovend. Verdere studie naar de rol van thecamoeben in de Si cyclus is nodig om een beter beeld te krijgen van hun algehele invloed op de Si cyclus.

Summary

The present sea-level rise is enhanced in estuaries due to a combination of tidal waves and estuarine anthropogenic modifications, like embankments, channeling and dredging. All these modifications result in further inland penetration and amplification of tidal waves. The increase in tidal amplitude leads to a higher flooding risk of harbor and coastal cities. In order to reduce flood risks, extra water bearing capacity (by restoring floodplain) is foreseen for times of enhanced flooding risk (e.g. storm surges).

Reconstruction of past water level changes can help to predict future water level changes within the estuary. Water level changes can be reconstructed using protists (e.g. foraminifera, diatoms, testate amoebae) that have shells which preserve in the soil and carry the species specific traits. Multiple studies have been published that acknowledge protists as good proxies for the reconstruction of sea level changes. The reconstructions are based on the use of a transfer function. A transfer function uses the relationship between modern protist assemblages and water level changes (environmental variable) to infer past water level changes from fossil protist assemblages.

This PhD study focuses on the use of testate amoebae as proxy for the reconstruction of water level changes of the Scheldt estuary. Testate amoebae, belonging to the protozoa, are shelled amoebae (average size between 20 µm - 200 µm) that habituate the marsh sediments. Their short generation time ( ± 60 generations year-1) makes that they react quick to environmental changes and their shells, with species specific traits, can fossilize.

Modern testate amoebae assemblages of a freshwater tidal marsh (Notelaar), brackish marsh (Groot buitenschoor) and salt marsh (Paulina) were investigated. The found testate amoebae concentrations

Summary of the salt marsh Paulina were too low to speak of testate amoebae assemblages ( ± 450 shells g-1). Modern testate amoebae assemblages of the freshwater tidal marsh and brackish marsh varied mainly with marsh elevation relative to mean high water level. This relationship is exploited to make a “freshwater” and “brackish” transfer function. The resulting transfer functions have a comparable accuracy for water level reconstructions to sea level reconstruction transfer functions.

Sub-fossil testate amoebae assemblages were investigated in sediment cores of the freshwater tidal marsh (Notelaar). The freshwater transfer function, based on the modern data of the Notelaar, was used to infer past water level changes of this location at the estuary. Comparison of the reconstructed water level changes with measured water level changes showed that the transfer function gives reliable water level reconstruction, based on fossil testate amoebae shells, until a depth of 50 cm (~ 1965). The limitation of the use of the transfer function was related to the quick disappearance of fossil testate amoebae in the soil.

Investigation of the decrease in fossil testate amoebae with depth have demonstrated that there might be selective dissolution. Mainly testate amoebae that have biogenic shells (idiosomic testate amoebae) disappear fast out of the fossil record. The finding of quick dissolution of idiosomic testate amoebae shells, made us investigate the relationship between silica cycle and testate amoebae.

It is known that marshes play an important role in the Si cycle by providing dissolved Si to the estuary in times of depletion, in order to prevent toxic algal blooms (Struyf et al., 2006). The origin of this dissolved silica is thought to come from dissolution of diatoms and phytoliths (biogenic structures of plants).

Summary

Our investigation on proportional loss of idiosomic testate amoebae shells, diatoms and organic matter (as proxy for phytoliths) in the sediment profiles, has demonstrated that testate amoebae play an important role in the Si cycle. Their relatively quick dissolution makes them easy contributors to the export of dissolved silica to estuarine waters in times of dissolved silica limitation.

We can conclude that there is a good relationship between modern testate amoebae assemblages and the elevation of the marshsurface relative to mean high water level. However, testate amoebae shells have insufficient preservation in the marsh sediments and selective preservation of testate amoebae occurs. Therefore, testate amoebae are not good bio-indicators for the reconstruction of water level changes of an estuary. Our exploratory investigation on the role of testate amoebae of marsh sediments in the Si cycle are promising. Further study on the role of testate amoebae in the Si cycle is necessary to get a general and good understanding.

Chapter 1

Introduction

Present-day sea level rise, as a consequence of climate change, forms a severe threat for millions of people living along coasts estuaries by enhanced flood risks. These enhanced flood risks are not solely due to climate change, but are also strongly influenced by the anthropogenic modifications to estuaries during the past centuries. In order to reduce flood risks, there are recent plans to give more room to the tidal river by restoring tidal marshes. The gain in water storage capacity will reduce inland flood propagation during times of high flood risk. The present-day effects of sea level rise on estuarine water levels can be better understood if there is knowledge about the relation between historical embankments and the effect on water levels. Reconstruction of past estuarine water levels can be done by the use of bio-indicator species, that can fossilize in estuarine marsh soils, and a transfer function, which translates the modern relationship between bio-indicator assemblages and tidal water levels into past water levels from fossil bio-indicator assemblages. In this study, we will investigate the use of testate amoebae as bio- indicators for the reconstruction of past water level changes in the Scheldt estuary (Belgium).

Introduction

1.1. Changes in sea level along coasts and estuaries

Sea-level rise There is strong evidence that sea level is rising (Solomon et al., 2007). This sea level rise threatens millions of people living along coasts, estuaries and on islands. It is estimated that by 2070, around 150 million people will be exposed to flooding events together with US $35,000 billion worth of assets (e.g. buildings, public transport) (Nicholls et al., 2007). Sea level rise may also have a big impact on low lying coastal and tidal ecosystems, such as tidal marshes and mangroves (Nicholls and Cavenave, 2010). Vegetation die-off and/or subsidence of these ecosystems will facilitate the occurrence of flood events. Their dense vegetation acts like a natural flood barrier, as friction between the vegetation and the flood reduces flood waves and inland flood water levels (Constanza et al., 2008, Krauss et al., 2009, Wamsley et al., 2010).

Global warming is responsible for the two major causes of eustatic or absolute sea level rise. Firstly, global warming has an effect on the thermal expansion of water. The warming of the ocean water decreases the density of sea water, which results in a volume increase (Johnson and Wijffels, 2011). Secondly, ice caps and glaciers melt quicker as a result of global warming. The mass of melting water contributes to the rise of sea level. A secondary effect of the addition of fresh melt water to the saline sea water is that the resulting salinity reduction further contributes to the reduced water density, which further enlarges the sea level rise (Johnson and Wijffels, 2011). The effect of thermal expansion is estimated to be twice as big as the effect of ice caps and glacier melt on the total eustatic sea level rise (Raper and Braithwaite, 2006).

Chapter 1

Sea level does not rise uniformly over the entire planet, but is determined by a combination of local factors. One of these factors is the unevenly distributed heating of surface water. Surface water at the tropics is more heated compared to surface water of higher latitudes (Griffies and Greatbatch, 2012). This results in high evaporation rates and higher salinity in the tropics. Further, the melting of large ice masses and glaciers resulting in freshwater addition to the saline sea water is mostly located at the higher latitudes. Following IPCC estimates (Bindoff et al., 2007), the above mentioned factors, and others, might lead to lower sea level rise rates in the tropics and higher sea level rise rates at the high latitudes.

Relative sea level is the sea level measured by a local datum. It is determined by the eustatic sea level rise in combination with land movement. If we assume that the eustatic sea level does not change, land movement can induce a sea level fall by tectonic uplift and a sea level rise by subsidence of land. Especially in deltas and lowland estuaries, subsidence is an important process that contributes significantly to the relative sea level rise.

Tide Gauge data indicate that there was a rise in eustatic sea level of -1 th 1.5 - 2.0 mm year for the 20 century (Miller and Douglas, 2004). Though, since the 1990s it already accelerated to ~ 3.0 mm year-1 (Shepard et al., 2012). Future estimations predict a further acceleration of the rate of global sea level rise between 20 cm and 80 cm by 2100 (Bindoff et al., 2007), with recent estimates indicating a rise at rates close to the upper boundary of that range (Church et al., 2011).

Water level changes in estuaries Estuaries are the transition zone between land and sea, where mixing occurs of fresh water, from rivers and streams, with saline water from the sea. The estuarine part of a river is influenced by tides. Tides are

Introduction generated on the oceans and propagate from the oceans as so-called tidal waves into shallow coastal seas, such as the North Sea, and subsequently propagate into estuarine river mouths.

The tidal wave is often amplified further inland into estuaries as a result of convergence, i.e. the decreasing width and depth of estuaries in the landward direction. As a consequence of this convergence, the tidal wave is pushed up in the landward direction and the height difference between the high and low water level (i.e., the tidal range) increases. This convergence effect is called tidal amplification (Friedrichs and Aubrey, 1994, Lanzoni and Seminara, 1998, Savenije, 2001). Apart from convergence, the decreasing estuarine width and depth also causes friction to the tidal wave propagation. At a certain point along the estuary, the friction effect can balance out the convergence effect and upstream from this point the tidal range decreases again. This friction effect is called tidal dampening (Friedrichs and Aubrey, 1994, Lanzoni and Seminara, 1998, Savenije, 2001). The exact amount of tidal amplification and dampening, and the position along the estuary where amplification transitions into dampening, depends on the specific morphological characteristics of the estuary.

The tidal propagation through an estuary can be seriously impacted by anthropogenic modifications to the estuarine morphology. Humans have altered estuarine morphology profoundly during past centuries, such as by the embankment of intertidal flats and marshes (e.g. Lotze et al., 2006, Rippon, 2000) and by canalization of estuaries. These modifications have lead to the loss of estuarine water storage capacity, thereby stimulating the landward propagation and amplification of the estuarine tidal wave. Furthermore, deepening of estuarine stream channels by dredging activities for harbors, can reduce the bottom friction on the tidal wave propagation, and hence can contribute to further inland propagation and amplification of the tidal wave (Friedrichs

Chapter 1

and Aubrey, 1994). This human-induced increased tidal propagation and amplification can substantially increase the risk for flood events in the landward part of estuaries.

Flood risks along estuaries are especially caused by storm surges. Storm surges are the result of the passage of extreme storm depressions over a coastal sea, causing extreme heightening of the local sea level as a result of wind, waves and low atmospheric pressure effects (Resio and Westerink, 2008). The resulting storm surge also propagates as a long wave through estuaries, and similar amplification effects as for tidal waves will occur. Hence the above-mentioned human modifications to estuaries will also increase the landward propagation and amplification of storm surges and hence increase the risk for flood disasters along estuaries. For example, the last years major storm surges have caused severe flood disasters in New Orleans (by hurricane Katrina in 2005) and New York (by hurricane Sandy in 2012).

In order to reduce future flood risks, effective disaster management strategies need to be implemented (Nicholls et al., 2007). The best option to protect estuaries from storm surge flooding risks, is by tackling the problem before it reaches the urban environment (Jha et al., 2012). One of the solutions is by increasing water storage capacity of the estuary in the area downstream of the urban environment. This might be done by creating conveyance channels or restoring the floodplain. Multiple restoration projects of large river floodplains in Europe and North America are planned, executed or already finished (e.g. Mississipi (Louisiana), Emiquon (Illinois), Rhine (Western Europe), Danube (Central Europe) (Buijse et al., 2002), Humber estuary (UK) (Hemingway et al., 2008)). This study focuses on the Scheldt estuary which is also threatened by rising flood risks.

Introduction

Changes in the Scheldt Estuary The Scheldt river has its source in France (St.Quintin) and flows trough Belgium and the Southwestern part of the Netherlands to finally end up into the North Sea near Vlissingen. The estuarine part of the Scheldt river is close to 160 km long and is located between Ghent and Vlissingen (Fig. 1.1). The estuary is divided into two parts. The upstream Flemish part, from Ghent to the Belgian/Dutch border, is called the Sea Scheldt, while the more downstream Dutch part is called the Western Scheldt. The morphology of the Sea Scheldt consists of a single channel system that is bordered on a number of places by intertidal flats and marshes, while in the Western Scheldt the morphology transitions into a complex system of ebb and flood channels that flow between sandbanks and mudflats.

The estuary has a full salinity gradient (Fig. 1.2), ranging from a polyhaline zone (marine zone: > 18 PSU (Practical Salinity Unit)) (Vlissingen - Hansweert) to a mesohaline zone (brackish zone: 5-18 PSU) (Hansweert - Antwerp) and a oligohaline zone (freshwater zone: < 5 PSU) (Antwerp - Rupelmonde). The Scheldt estuary has a semidiurnal meso-tidal (2 - 4 m tidal range) to macrotidal ( > 4 m tidal range) regime with highest tidal ranges near Temse (freshwater part) (Fig. 1.2). The suspended sedimentation concentrations (SSC) of the upper part of the water column, which is the part of the water column that floods the marshes, changes along the estuary. The Western Scheldt has a SSC of 30 – 60 mg l-1. The part of the Sea Scheldt ranging from the Dutch-Belgian border to Temse has highest SSC (100-200 mg l-1), after which SSC decreases to 50-100 mg l-1 farther upstream (Temmerman et al., 2004b, Van Damme et al., 2001).

Chapter 1

Figure 1.1 Map of the Scheldt Estuary with indication of the polder area’s and salinity zones. Arrows indicate sampling places: the freshwater marsh “Notelaar”: Chapter 2, 4, 5, 6; the brackish marsh “Groot Buitenschoor”: Chapter 3; the salt marsh “Paulina marsh”: chapter 4

Introduction

Figure 1.2 Tidal range and Salinity along the estuary going from the mouth (Vlissingen) towards the end of the estuary at Ghent.

Figure 1.3 Historical changes in the embanked area along the Western Scheldt (De Kraker, 2004). The blue dotted line indicating the period before 1550 is an estimation of the embanked area.

Chapter 1

The high tidal range can be explained by the history of human induced modifications to the Scheldt (Coen, 2008, Mol, 1995). The estuarine morphology has undergone major changes during the last centuries (Meire et al., 2005). Especially the area of intertidal flats and marshes was originally very large (around 100 000 ha in the year 1200; (Mol, 1995)) and has been gradually reduced by embankments since the late Medieval period (Fig. 1.3).

A large part of the Western Scheldt estuary was already embanked (Fig. 1.3) before 1550, but multiple flooding events and insufficient dyke maintenance caused the loss of the major part of the embanked area. During the Eighty Year’s War (or Dutch War of Indepence (1564-1648)), large parts of the embanked area along the Westerscheldt was flooded (Fig. 1.3) by military inundation to prevent occupation of cities (e.g. Gent, Brugge, Antwerp) by the Spanish army. From 1600 onwards, the Western Scheldt estuary was gradually re-embanked with a gain of 100 000 ha of polders, and equal amount of loss of the intertidal area. Along the Sea Scheldt, the largest loss of intertidal areas occurred in the brackish part of the estuary downstream of Antwerp due to embankments (Mys, 1973) (Fig. 1.1). For the freshwater part of the estuary upstream of Antwerp, relatively less studies are available on the embankment history, but it is generally considered that embankments of the original river floodplain started around the 12th century (Mys et al., 1983). During the last century (1900 - 1980), another 4900 ha was embanked along the Western Scheldt. The largest embankments were located in Bath and surrounding areas (1510 ha) and the Braakman (1500 ha). Smaller embankments were at Hellegat (132 ha), Saefthinge (316 ha) and Sloe (681 ha) (Mol, 1995). Other modifications to the Scheldt estuary during the last century are dredging works and deepening of the shipping channel.

Introduction

Figure 1.4 The evolution of the tidal range over the last two millennia along the length of the Scheldt estuary. The full lines indicate measured tidal ranges, while the dotted lines represent only estimated tidal range reconstructions based on tidal data of numerous studies adapted from Coen, 2008.

The effect of the reduced intertidal area, as a consequence of the history of embankments and the more recent dredging, on the tidal range of the Scheldt estuary is shown in Fig. 1.4. The figure clearly shows the increase in tidal range, with highest tidal increase (5 m) at Temse, and the landward propagation of the tidal wave over the centuries (Vlissingen to Temse). The resulting high tidal range in combination with the sea level rise make the risks of flooding increase. The rise of inland water levels over the last century was about 1.5 m, while the sea level rise at the Belgian coast amounted only 0.2 m (Temmerman et al., 2004b). In Flanders, 0.8 million of people living at the coast or along the Scheldt estuary are threatened by this higher flooding risk (Lebbe et al., 2008). In order to reduce the risk of flooding, multiple floodplain

Chapter 1

restoration projects are (being) executed along the Scheldt. These projects are written down in the SIGMA plan (Broeckx et al., 2011, Meyvis et al., 2003) and LTVS (Long Term Vision Scheldt-estuary) (Zanting and ten Thij, 2001, ProSes, 2008).

The goal of these projects is to increase the water storage capacity of the Scheldt estuary, in order to reduce the inland propagation of storm surges. Two major storm surges happened in the last century. In February 1953, a large North Sea storm surge occurred due to a combination of spring tides, low atmospheric pressure and wind. This flooding disaster impacted the coastal areas of the UK, The Netherlands and Belgium and killed more than 2000 people, of which approximately 1800 lived in Zeeland (Nl) (Jonkman and Kelman, 2005, Meyvis et al., 2003). Later, in 1976, the combination of spring tide and a Northwestern storm induced a flood event, breaching dykes at several places in the Flemish part of the Scheldt estuary (e.g. Ruisbroek) (Meyvis et al., 2003).

Until now, effects of these modifications on flood risk reduction were estimated by simulation models (Broeckx et al., 2011). Though, multiple factors of these new water storage areas are difficult to assess, such as the effect of vegetation on the reduction of the tidal wave. Knowledge about the historical changes in estuarine water levels due to the large scale embankments of the last centuries can help to understand present and future changes in water levels along the estuary. In order to investigate the relationship between the changes in estuarine water levels and size of the intertidal area, empirical data is needed.

Sea level reconstructions using protists Holocene sea level reconstructions can be based on different types of methods (e.g. archaeological, geomorphologic, ecological, …). Ecological methods for the reconstruction of past sea level changes are based on

Introduction the use of bio-indicator species that have a good relationship with sea level and that preserve in the soil (fossils). It has been shown in multiple studies that certain protist species, which have shells that can fossilize, are particularly useful for sea level reconstructions. Sea level reconstruction studies are primarily based on benthic protist assemblages of salt marshes. The first step for making a sea level reconstruction is to study the relationship between modern protist assemblages and local mean sea level changes. This relationship is investigated by looking to what extend the spatial variation in species occurrence can be explained by the different environmental variables and in particular by elevation above mean high water level (MHWL). The more species variation is explained by elevation above mean high water level, the better the relationship. If the relationship is satisfying, a transfer function can be made. The transfer function consists of a regression between the modern species assemblages and corresponding elevation above MHWL, that will be calibrated in the next step. The second step implies the investigation of palaeo protist assemblages. Over time, protist assemblages are buried in the soil by deposition of sediments that are supplied by the tides and are deposited on the marsh surface. These old, fossil, protist assemblages stay preserved into the sediment profile and form, by their species composition, a memory of the original environment they lived in. The palaeo species assemblages are translated into reconstructed values of elevation above MHWL based on the the transfer function between modern protist assemblages and the marsh elevation above MHWL.

Transfer functions need to fulfill three assumptions in order to produce reliable reconstructions. Firstly, it is assumed that the relationship between the modern protist assemblage and the environmental variable has not changed over time. Secondly, the modern and palaeo habitat type should be analogous. And thirdly, it is assumed that the species assemblages do not change during fossilization (Sachs et al., 1977).

Chapter 1

The protist groups that are mainly studied for reconstructing past sea level changes are diatoms and foraminifera and to a lesser extent testate amoebae. These three protist groups have a couple of common characteristics that make them ideal organisms for studying environmental changes. Firstly, they have a short generation time, implying that they can react quickly to environmental changes. Secondly, they are species diverse. Thirdly, they have shells that remain in the soil after the death of the organism and that have species specific traits.

Salt marshes are the ideal coastal ecosystem to study protists assemblages for the reconstruction of past sea level changes, as they lie at the intersect between land and sea. Protists are investigated in the higher part of salt marshes where continuous sedimentation takes place at rates which are a reasonable proxy for rates of sea level rise (Leorri et al., 2010).

The first organisms that were used to study sea level changes are foraminifera. It was D.B. Scott that made his PhD thesis on foraminifera in relation to sea level changes and published together with F.S. Medioli in 1978 a letter to Nature about the vertical zonation of salt marsh foraminifera as accurate indicators of former sea levels (Scott and Medioli, 1978). From this moment on, multiple research studies have been undertaken all over the world to reconstruct sea level changes using foraminifera (e.g. Gehrels, 1994, Gehrels et al., 2001, 2002, Varekamp et al., 1992, Leorri et al., 2010, Woodroffe, 2009a). One of the advantages of working with foraminifera is that they are easy to study and to determine up to species level. There are also some limitations to foraminiferal sea level reconstructions, as they are bound to a salt environment and living foraminifera can still be found at a sediment depth of 30 cm (Berkeley et al., 2007). The large depth range

Introduction of living foraminifera, leads to a mixed foraminifera assemblage with fossil and living individuals at different sediment depths that might obscure the sea level signal. They can be used to reconstruct sea level changes with an accuracy of around 0.08 m (Gehrels et al., 2001).

The first sea level reconstruction studies (e.g. Haggart, 1986, Long, 1992, Shennan et al., 1996) that were based on diatoms, only used diatoms to confirm sea level index points. Sea level index points are points of which age, geographic location, altitude (related to former tidal level) and the tendency of sea level change is known (Edwards and Horton, 2000). Only in 1999, the first diatom studies using a transfer function based sea level reconstruction were published (Sherrod, 1999, Zong and Horton, 1999). The benefit of working with diatoms as sea level indicators is that they live in both fresh, brackish and saltwater habitats. Further, diatoms are autotrophic organisms, which implies that benthic diatoms only live and reproduce at the sediment surface level (as they are in need of sunlight), and do not live over a broader range of sediment depths as is the case for foraminifera. Nevertheless, diatoms are more species diverse and more difficult to determine up to species level. Furthermore, there is a high input of allochtonous diatom frustules that are supplied and deposited by the tide into marshes since also planktonic diatoms are highly abundant in sea water. Diatom based sea level reconstructions can indicate sea level changes with a precision of 0.05 m (Gehrels et al., 2001).

Testate amoebae appeared in sea level change reconstruction studies around the same period as diatom sea level studies. The first study appeared by Charman et al. (1998), who showed that testate amoebae assemblages from salt marsh sediments had a distinct spatial zonation that was related to elevation and tidal parameters. This study was used as a basis for the development of transfer functions for sea level

Chapter 1

reconstruction on salt marsh testate amoebae assemblages (e.g. Charman et al., 2002, Gehrels et al., 2006, Gehrels et al., 2001). Testate amoebae appeared to have the finest vertical marsh distribution, which can be translated to reconstructed sea level changes with a precision up to 0.04 m (Gehrels et al., 2006). In 2010, the first sea level reconstructions based on testate amoebae assemblages of salt marshes was published (Charman et al., 2010). All testate amoebae studies related to sea level change have been performed on salt marshes. So far, no brackish or freshwater marsh testate amoebae assemblages have been studied.

1.2. Testate amoebae

Testate amoebae phylogeny and taxonomy Testate amoebae are unicellular heterotrophic eukaryotic organisms that belong to the group of protozoa. The described phylogenetic tree (Fig. 1.5) of testate amoebae is based on the data of the Tree of life web project (Bass, 2009, Bass and Cavalier-Smith, 2009, Smirnov, 2008). The tree clearly shows that testate amoebae are a polyphyletic group consisting of all shelled amoebae. It includes a part of representatives of the Cercozoa and Amoebozoa. The main difference between both groups is the type of pseudopodia.

Species of the Lobosae (Amoebozoa) have lobose, thick, pseudopodia, while the species of the Euglyphida (Cercozoa; Filosa) and Thecofilosea (Cercozoa; Filosa) have filose pseudopodia. A third group, Gromiidea (Cercozoa; Endomyxa) represents the testate reticulate amoebae.

Introduction

Figure 1.5 Phylogenetic tree with testate amoebae phylogeny indicated in bold (based on Tree of life webproject).

Chapter 1

In total, approximately 2000 testate amoebae species are described worldwide. The distinction between species is based on morphological traits of the test (shell) (Mitchell et al., 2008a). Testate amoebae species determination is sometimes difficult as species descriptions are not always detailed and some species are described multiple times. Further, distinctions can be based on minor morphological differences, which results in distinct morphotypes rather than different species.

Testate morphology Testate amoebae are small organisms with an average size between 20 µm to 200 µm (Beyens and Meisterfeld, 2002). Their tests, which protect testate amoebae from desiccation and predation, exist in different shapes. Most of them are egg shaped, being round at the lowest half of the body and becoming smaller near the pseudostome region. The pseudostome, or mouth, is an opening in the test which is used by the amoeba to interact with the surrounding environment by stretching out pseudopodia for feeding, locomotion or reproduction. Variation in pseudostomes are described following Bonnet (1975), distinguishing 9 different pseudostome types (Fig. 1.6A). The other general morphological traits, which might apply to a part or to the whole test, are (partly) flattening or narrowing of the body and the abundance or lack of spines and teeth.

Testate amoebae tests are made from different materials. Some testate amoebae built their own test from self-secreted biogenic plates (idiosomes) embedded in an organic matrix. Other testate amoebae paste particles from the surrounding environment (xenosomes; e.g. mineral particles, organic detritus, diatom shells) together with organic cement. Apart from these two main groups, there are also testate amoebae that make their test from calcium and proteins or idiosomes embedded in a thick organic matrix (Fig. 1.6B) (Mitchell et al., 2008a).

Introduction

Figure 1.6 A/ Overview of the nine pseudostome types (Bonnet, 1975); B/Four testate amoebae types based on difference in test material composition, as described by Mitchell et al. (2008a). Pictures are from Ogden and Hedley (1980).

Chapter 1

Biogeography and Ecology Testate amoebae are found all over the world, inhabiting a variety of habitats including lakes, mosses, soils, sandy beaches, peat bogs, … (Beyens and Meisterfeld, 2002). Many of the described species are ubiquitous and/or cosmopolites. Though, there are some species that are restricted to one habitat type. Testate amoebae can encyst when environmental conditions are not favorable and by doing so can survive long time periods.

Testate amoebae eat bacteria, fungi, organic detritus and sometimes even other smaller testate amoebae. They are part of the lower trophic levels of the food web and consequently important for all organisms higher in the food web. Testate amoebae have probably an important function in nutrient recycling in soils. Although research on nutrient cycling by testate amoebae in soils is limited, there are some studies that show that testate amoebae might be nutrient hot spots for mycorrhizae, which in turn supply nutrients to forest plants (Ericaceae) (Vohník et al., 2008, Wilkinson and Mitchell, 2010). Further, they play a role in the nitrogen cycle by releasing immobilized nitrogen from bacteria (Schröter et al., 2003) and in the silicon cycle by the fast dissolution of siliceous testate amoebae tests in forest soils (Aoki et al., 2007). The role of testate amoebae within the silicon cycle of tidal marshes has not yet been looked into. Though, it is already known that tidal marshes play an important role in the Si cycle, as they deliver dissolved Si (DSi) to the estuary at times when DSi concentrations may be limiting for primary production by phytoplankton (diatoms) and thereby may prevent toxic algal blooms (Struyf et al., 2006). Until now, the main players that facilitate this buffering role are thought to be diatoms and higher plants (phytoliths).

Introduction

Testate amoebae assemblages are often studied in relation to ecological gradients. Testate amoebae assemblage composition seems to respond primarily to moisture (fen-bog transition, water table depth, sea/tidal level, … ) (Mitchell et al., 2008a). Though, testate amoebae assemblage structure can also change with other environmental variables, such as pH or nutrients with high range of variability. These modern relationships between testate amoebae assemblages and a specific environmental variable have led to the development of transfer functions (see above), which can be used to reconstruct the past variations in the environmental variable by investigating palaeo testate amoebae assemblages from sediment cores.

Palaeoecology The oldest found fossil remains of testate amoebae might belong to the Neoproterozoic (~ 1000 - 540 million years ago) and as such belong to one of the oldest organism groups on Earth (Porter and Knoll, 2000). Although testate amoebae fossils can preserve for a long time, most of the reconstructions based on testate amoebae investigate previous environmental changes over the much more recent Quaternary (~ last 2.5 million years) or Holocene (~ last 11 500 years) period (Payne and Mitchell, 2007, Turner et al., 2013, Wilmshurst et al., 2003).

The study of fossil testate amoebae need to take into account the taphonomy, or the study of processes and preservation and how they affect information in the fossil record (Behrensmeyer and Kidwell, 1985). The fact that fossil tests are found does not imply that all the testate amoebae species have the same preservation abilities. Since not all tests are built from the same material, it may be that not all testate amoebae remains are equally well preserved. Multiple studies have shown that the preservation of testate amoebae is dependent on different environmental factors (e.g. wet-dry cycles, decomposing bacteria, … ) (Booth and Jackson, 2003, Booth et al., 2004, Lousier and

Chapter 1

Parkinson, 1981, Mitchell et al., 2008b, Swindles and Roe, 2007). Though, most of these studies seem to agree that idiosomic testate amoebae tend to disappear quicker from the fossil record compared to xenosomic species (Mitchell et al., 2008b, Swindles and Roe, 2007, Wilmshurst et al., 2003).

3. Objectives and overview of the thesis.

Figure 1.7 Schematic overview of the different chapters. Purple boxes indicate chapters that contain both modern and sub-fossil data.

The general objective of this study is to investigate the potentials and limitations of using marsh testate amoebae as bio-indicators of past water level changes in the brackish and freshwater zones of the Scheldt estuary (Belgium) (Fig. 1.1, 1.2). Up to now, the potentials of testate amoebae as sea level indicators has only been demonstrated for coastal salt marshes and not yet for the more inland brackish and freshwater parts of estuaries. As described above, large human modifications to the morphology of the Scheldt estuary has taken place since the 16th century (Fig. 1.3), and it is generally expected that these morphological changes have had a serious impact on the tidal propagation and corresponding water level changes in the more inland part of estuary over the last centuries (Fig 1.4). However, detailed historical data on the

Introduction estuarine water level changes before the 20th century are lacking. With this study it is aimed to develop a potential new tool to reconstruct past water level changes since the 16th century based on testate amoebae. Such insights on past water level changes in response to human modifications to the estuary, can contribute to improved understanding of present and future water level changes and related flood risks as a result of human estuarine management.

This general objective is aimed for by following the next steps: 1) Studying the modern testate amoebae ecology in relation to environmental variables (Chapter 2, 3, 4). 2) Setting up a transfer function between modern testate amoebae assemblages and elevation relative to mean high water level (Chapter 2, 3). 3) Study of the preservation of (sub-)fossil testate amoebae (Chapter 2, 5). 4) Reconstruction of past water levels (Chapter 5). 5) As we found out that sub-fossil testate amoebae concentrations rapidly decreased with depth beneath the sediment surface, we investigated the role of testate amoebae in the Si cycle as an additional side-topic (Chapter 6).

The first step was to study the ecology of modern testate amoebae assemblages in the Scheldt estuary (Chapter 1, 2, 3, 4), mainly focusing on testate amoebae assemblages in relation to the mean high water level. The modern testate amoebae assemblages were studied in a freshwater tidal marsh (Chapter 2), a brackish tidal marsh (Chapter 3), and together with data of a salt marsh, a comparison of modern testate amoebae assemblages could be made along the salinity gradient (Chapter 4).

Chapter 1

The second step was to make a transfer function by use of the modern relationship between testate amoebae assemblages and marsh elevation relative to mean high water level. Testate amoebae assemblages variation was related to mean high water level for the freshwater tidal marsh and the brackish marsh, resulting in two separate transfer functions (Chapter 2, 3).

Then, as a third step, sub-fossil testate amoebae assemblages were studied in sediment cores to serve as input for reconstruction of past water level changes. The sub-fossil testate amoebae species composition was studied on the freshwater tidal marsh. Firstly, a number of samples of the old marsh were investigated on fossil testate amoebae assemblages to test the transfer function (Chapter 2). Secondly, four cores were sampled at the young part of the tidal freshwater marsh to study sub-fossil testate amoebae assemblages (Chapter 5, 6). Especially the taphonomy and selective preservation of testate amoebae was studied for these sediment cores (Chapter 5). The finding that the concentrations of idiosomic testate amoebae, made from BSi, decreased quickly with depth in the sediment cores lead us to investigate the role of testate amoebae in the Si cycle of a freshwater tidal marsh (Chapter 6).

The final step, the reconstruction of past water levels, by calibration of the transfer function with the sub-fossil testate amoebae assemblages is reported in Chapter 5. The resulting water level reconstruction was tested in this chapter by comparing the reconstructed water level changes with observed tide gauge data.

Chapter 2

Testate amoebae as estuarine water level indicators: modern distribution and the development of a transfer function from a freshwater tidal marsh (Scheldt estuary, Belgium)* Marijke Ooms, Louis Beyens, and Stijn Temmerman

Little is known about the century-scale response of water levels in inland estuaries to sea level change and human modifications to estuarine morphology. This study explored the ability to use testate amoebae (Protozoa, Rhizopoda) from sediments of a freshwater tidal marsh as indicators of water level in an inland estuary. The hypothesis was tested that modern testate amoeba assemblages change with surface elevation (~ duration of tidal flooding) within a freshwater tidal marsh. Variation in testate amoeba assemblages in relation to multiple environmental variables and sediment characteristics was studied through Redundancy Analysis. This demonstrated that a significant part of variation in modern testate amoeba assemblages could be explained by flooding frequency, surface elevation, organic content and particle size of the soil. Transfer functions, Partial Least Squares and Weighted Average regressions, were made to show that testate amoebae can be used for reconstruction of water level (with an accuracy of 0.05 Normalized Elevation). A preliminary test of application of the transfer function to palaeo testate amoeba assemblages showed promising results. Hence this study demonstrated that testate amoebae from a freshwater tidal marsh provide a potentially powerful new tool for estuarine water level reconstructions.

*Published as: Ooms M, Beyens L, & Temmerman S. 2011. Testate amoebae as estuarine water-level indicators: modern distribution and the development of a transfer function from a freshwater tidal marsh (Scheldt estuary, Belgium). Journal of Quaternary Science 26: 819-828. Testate amoebae as estuarine water-level indicators

2.1 Introduction

The ongoing and future climate change is a severe threat to densely populated areas along lowland coasts and estuaries (Miller and Douglas, 2004, Solomon et al., 2007). The acceleration of sea level rise (Church and White, 2006) and increase in frequency and intensity of storm surges (Webster et al., 2005) strongly enhances the risks of flood disasters. Along estuaries, these climate-induced flood risks are often further amplified by human-induced modifications to estuarine morphology. Especially the embankment of intertidal flats and marshes over the past centuries to millennia (e.g. Rippon, 2000) has led along many estuaries to a decrease of intertidal water storage capacity and hence to a significant additional rise of high water levels (e.g. Lane, 2004, van der Spek, 1997). For example in the Scheldt estuary (Belgium, The Netherlands), rise of mean high water level was up to 5 times faster in the inland estuary compared to the coast for the last century (Temmerman et al., 2004). Nevertheless, empirical reconstructions of estuarine water level changes over longer time-scales (centuries to millennia) are extremely scarce. Therefore, reconstructions of past estuarine water level changes are a potentially important reference for our understanding of present-day and future estuarine water level changes in response to global change.

The distribution of protist shells (especially foraminifera and diatoms) in salt marsh sediments has been used extensively to quantitatively reconstruct Holocene sea level changes all over the world (Campeau et al., 1999, Gehrels et al., 2001, Ghosh et al., 2009, Hassan et al., 2006, Horton et al., 2006, Kemp et al., 2009, Ng and Sin, 2003, Woodroffe and Long, 2009b). The most crucial aspect of this method is to establish a transfer function, which is based on the relationship between modern species assemblages and environmental variables. Once calibrated and

Chapter 2 validated, the transfer function is used to infer past environmental changes from palaeo species assemblages that are preserved in deeper, older sediment layers. For the reconstruction of past sea level changes, modern protist assemblages of salt marsh surface sediments are studied in relation to the elevation gradient (i.e. tidal inundation gradient). This relationship is translated in a transfer function in which elevation is expressed as soil elevation relative to sea level. Existing transfer functions for sea level reconstructions are very accurate with vertical errors of, for example, ± 0.08 m for diatoms (Horton et al., 2006) and ± 0.10 m for foraminifera (Leorri et al., 2010).

Rather recently a third group of protists, testate amoebae, has been studied in salt marsh sediments with the intention of using them as proxies for sea level reconstructions. (Charman et al., 1998) demonstrated that salt marsh testate amoebae appear to be related to elevation and tidal inundation. Since then, multiple studies in Britain and North America have proven that salt marsh testate amoebae assemblages can be used as accurate ( ± 0.10 m) sea level indicators (Charman et al., 2010, Charman et al., 2002, Gehrels et al., 2006, Gehrels et al., 2001, Riveiros et al., 2007).

Despite the fact that sea level change is often amplified further inland in estuaries, the potential of protists for inland estuarine water level reconstructions has not yet been investigated. The most inland zone of estuaries is characterized by the presence of freshwater tidal marshes. Existing transfer functions based on salt marsh protists may not be applied here. For freshwater tidal marshes, foraminifera cannot be used since this group of protists is bound to a marine environment. Diatoms may be expected in freshwater tidal marsh sediments, but high numbers of planktonic species may be buried in freshwater marsh sediments (e.g. Struyf et al., 2007), making it perhaps difficult to use diatoms as water level indicators in a freshwater tidal marsh. Testate amoebae have not

Testate amoebae as estuarine water-level indicators

been studied so far in freshwater tidal marshes, but they have been found to be good indicators of hydrological conditions in terrestrial freshwater wetlands such as peat bogs (e.g. Charman et al., 2007). Here, we hypothesize that testate amoebae are present in freshwater tidal marsh sediments in high abundances and that variations in species composition are related to variations in tidal flooding, so that a transfer function can be made and used for the reconstruction of estuarine water level changes, assuming good preservation of tests in stratigraphic records.

In this study, the spatial distribution of testate amoeba assemblages in a freshwater tidal marsh will be related to multiple environmental variables (soil elevation, flooding frequency, vegetation type, sediment particle size, organic matter content and bulk density). This study provides a transfer function between species composition and Elevation relative to mean high water level (MHWL), Normalized Elevation and Flooding frequency. The transfer functions are then applied to a limited number of palaeo testate amoeba assemblages from a sediment core, demonstrating that testate amoeba can be used in future studies for reconstructing past water level changes in the inland freshwater tidal zone of an estuary.

2.2 Material and Methods

Study site The estuarine part of the Scheldt river is situated in the Southwest of the Netherlands and the Northwest of Belgium (Fig. 2.1). The estuary is 160 km long, extending from its mouth near Vlissingen to its most upstream part at Gent. A full salinity gradient exists, mainly determined by the magnitude of the river discharge, covering a marine part (from the mouth up to Hansweert), brackish part (up to the tributary river Rupel), and freshwater part (up to Gent) (Meire et al., 2005) (Fig. 2.1).

Chapter 2

This estuary has a semi-diurnal tidal regime with a mean tidal range that varies along the estuary from 3.85 m at the mouth, reaching a maximum value of 5.39 m at Temse and decreasing again to 2.63 m at the most inland part in Gent (Taveniers and Mostaert, 2009).

Figure 2.1 A) The Scheldt estuary with indication of the Notelaar freshwater tidal marsh. B) Map of the Notelaar with indication of sampling locations and indication of vegetation zones. C) The marsh profile and vegetation of the Notelaar.

The study area, the Notelaar tidal marsh, is situated in the freshwater part (Salinity: 0 - 5 PSU) of the Scheldt estuary (Fig. 2.1). This freshwater tidal marsh has a surface area of 27 ha, over a length of 2 km and the marsh surface is cut by small tidal channels and creeks (Temmerman et al., 2003b). Fine sediments (clay, silt, and fine sand)

Testate amoebae as estuarine water-level indicators

are supplied and deposited on the marsh during tidal flooding, resulting in heightening of the marsh surface at a rate of 1 to 2 cm year-1 (Temmerman et al., 2003b, Temmerman et al., 2003a). The oldest part of the freshwater tidal marsh is visible on the Ferraris maps (1772 - 1779), the younger part is only established by plant colonization on a mudflat after 1944 (Hoffman, 1993). The sediment stratigraphy of the old and young marsh parts has been studied in detail, based on radiometric dating, sedimentological characterization and description of macroscopic plant remains in sediment cores (Temmerman et al., 2003b, Temmerman et al., 2004b). These studies showed that under the old marsh surface at least 4 meters of freshwater tidal marsh sediments exist (based on plant remains and silt content), which must be several hundreds of years old, since dating revealed a mean sediment deposition rate of 1.2 cm year-1. The young marsh part was accreting at a faster mean rate of 4.6 cm year-1 between 1944 and 1960, and is since then accreting at a mean rate of 1.8 cm year-1. The present-day marsh vegetation exhibits a vertical zonation (Fig. 2.1). The lowest part, bordering the tidal flat, is grown by dense reed (Phragmites australis) vegetation, that can reach a canopy height up to 4 m in summer (Temmerman et al., 2003a).The higher, older part is dominated by willow vegetation (Salix sp.), under which multiple herbaceous plants grow (e.g. Impatiens glandulifera, Urtica dioica, Convolvulus arvensis). At the highest parts of the marsh Populus canadensis trees are found.

Sampling method The surface sediment was collected at 54 sites during two sampling campaigns (Fig. 2.1C). During the first campaign (January of 2008), 44 sites were sampled of which six were situated on the mudflat in front of the marsh. The other 38 sites were sampled randomly over the elevation range within the marsh vegetation. Throughout the second campaign (September of 2009), 10 extra sites were sampled to obtain a

Chapter 2 more even distribution along the elevation range. Two surface sediment samples were collected at every site. The first one, for testate amoebae analysis, was obtained by pooling of five subsamples. These were taken in a grid of 20 x 20 cm, using a small sediment corer (diameter = 0.9 cm). The top 2 cm of the sediment surface was sampled, in order to recover sediment that has been deposited over about one year (Temmerman et al., 2004b) and hence to diminish potential seasonal effects in testate amoeba assemblages (Horton et al., 2006). The five subsamples were mixed into one bulk sample and fixated in 5 % formaldehyde solution. The second sample was collected for sediment analysis using a corer with a fixed volume of 84.5 cm³. Based on the most abundant plant species, each site was classified in three, non-exclusive, different vegetation types (Phragmites, Salix, herbaceous vegetation). GPS positions were recorded and sediment surface elevation was measured relative to the local Belgian Ordnance Level (m TAW) using a DGPS system and Total station (vertical accuracy of ± 1 cm). Palaeo-sediment samples were collected at the old part of the marsh, within the willow vegetation, using a gouge auger set (diameter = 2.5 cm). Every 5 cm a sample of 1 cm thick was taken for testate amoebae analysis. The selected samples for testate amoebae analysis were 5, 15, 25, 35, and 45 cm deep.

Preparation method for testate amoebae study The preparation method, used for modern and palaeo samples, was based on Hendon and Charman (1997). Before preparation, exotic Lycopodium spore tablets (batch 483216: Results of the calibration (5 tablets): X= 92914 ± 3820, Variance = ± 4.1 %) (Stockmarr, 1971) were added to the sample, in order to calculate testate amoeba concentrations. Aggregations of mineral particles and testate amoebae were separated while boiling the sample for 10 minutes. After that, the

Testate amoebae as estuarine water-level indicators

sample was sieved and the material between 10 µm and 300 µm was retained for analysis. On a glasslide two drops of testate amoebae solution were mixed with one droplet of glycer(ine)ol-Rose Bengal blend to stain the living tests. Total testate amoebae assemblages, combining living and dead testate amoebae, were used in the analysis . A coverslide was put on top and the edges were sealed with nail polish to protect against desiccation. Hundred fifty testate amoebae were counted per sample using an OLYMPUS BX50 microscope with Nomarski optics. A microscopic magnification of 400 was used to determine species. When less than 150 tests were found, a maximum of 15 slides was counted. All samples included in the analysis contained more than 100 testate amoebae. Modern samples from 42 sites were included. Samples of the mudflat were omitted, because they contained too low numbers of tests. Four out of five palaeo samples contained enough testate amoebae (150 tests), the sample at depth 35 cm was omitted from further analysis. Taxonomic works that were used for the determination of the testate amoeba species are Chardez, 1991, Decloitre, 1962, 1974, 1981, Deflandre, 1928, 1929, Foissner and Korganova, 2000, Grospietsch, 1964, 1965, 1983, Mazei and Tsyganov, 2006.

Sediment analysis First, the dry bulk density was measured by drying the sediment samples for 48 h at 105 °C, cooling them in an exsiccator with silicagel and weighing them. Secondly, the organic matter content was determined by loss on ignition (LOI), combusting samples during four hours at 550 °C. Thirdly, on a separate subsample, particle size distribution of the mineral fraction was determined, using the laser diffraction technique (Malvern particle size analyzer). The grain size classes were based on the Udden-Wenthworth scale (Blott and Pye, 2001) considering sand ( > 63 µm), silt (2 - 63 µm) and clay ( < 2 µm).

Chapter 2

Sediment analyses were only carried out on the modern sediment samples, in order to investigate the possible relationships between modern testate amoeba assemblages and sediment characteristics. No sediment analyses were carried out on the palaeo samples, since they were used to check the applicability of the obtained transfer function for marsh elevation. The sediment characteristics of deeper cores are described in Temmerman et al. (2003b) and are similar to the characteristics of surface sediments.

Calculating flooding frequency and Normalized Elevation High water level data of 2007 from the closest upstream and downstream tide gauges were used to interpolate high water levels for the Notelaar tidal marsh. Based on the distribution of high water levels, the MHWL and flooding frequency was computed for every sampling location as the percentage of high tides that flood a location. Normalized Elevation was calculated, in order to compare data with other studies, by the use of following formula:

Normalized Elevation = (Elevation - MTL)/(MHWS - MTL)

In which MTL stands for mean tide level and MHWS is the abbreviation of mean high water of spring tides (Zong and Horton, 1999, Charman, 2001). Tidal characteristics of the Notelaar tidal marsh are shown in Fig. 2.1.

Data analysis Relative abundances were calculated for the modern testate amoeba species of the remaining 42 sites. Species that have at least for one sample ≥ 2% relative abundance were used for further analysis. Analysis was performed with the total testate amoeba assemblages (dead + alive).

Testate amoebae as estuarine water-level indicators

Diversity indices (Shannon-Wiener index) and turn-over rate (Sørensen dissimilarity index) were calculated in R (R Development Core Team, 2009). The Sørensen dissimilarity index quantifies the difference in species composition between two successive samples along the elevation gradient. Since the distribution of the samples over the elevation gradient was not equal, a moving average of the Sørensen and Shannon-Wiener index values was calculated over a fixed but moving elevation range. The smallest elevation range was selected basing on at least two samples for every calculation. The elevation range was 26 cm for the Shannon-Wiener index and 16 cm for the Sørensen dissimilarity index.

Cluster analysis was performed using CONISS in Tg view (Tilia). In this case, square root transformation (Edwards & Cavalli-Sforza’s chord distance) was chosen. Different ordination analyses were carried out in the program CANOCO 4.5. Species data were square root transformed to give rare species more weight in the analysis, because they showed more variability in abundances between the different samples. First a Detrended Correspondence Analysis (DCA) was performed to find out whether a unimodal or monotonic (linear) model should be used. A gradient length of less than 2 Standard Deviations (SD) should indicate the use of a linear model (ter Braak, 1987a). Direct ordination analysis was applied to examine the relationship between species and the different environmental variables. Environmental variables were selected based on their Variance Inflation Factor (VIF) values. This factor was calculated for each environmental variable. Large VIF values ( > 20) indicated that the variable was highly correlated with other environmental variables, while VIF values of zero showed completely multicollinearity between environmental variables (ter Braak and Šmilauer, 1998). Environmental variables with VIF values of zero and more than 20 were left out since they did not contribute to the model. Furthermore, partial ordination

Chapter 2 analysis, using the Monte Carlo Permutation Test (999 permutations), was done to estimate the amount of species variation explained by the environmental variables.

Regression analyses were carried out in the program C2 version 1.6.3, in order to analyze the response of each separate species to the environmental variable (ter Braak, 1987a). A number of different regression models were used in order to test the robustness of the found relationships. As linear method, Partial Least Squares regression (PLS) was chosen (e.g. Woodland et al., 1998). Through this method multiple regressions are calculated, extracting underlying factors that explained the biggest amount of variation in the predictor (Testate amoeba assemblages) and response data (environmental variable) (SAS/STAT, 2009). Weighted Average (WA) was used as unimodal regression technique. This method was built on the assumption that a species was most abundant at sites with an environmental variable close to the species optimum (ter Braak and Juggins, 1993). Another method, WA-PLS, was used because it could calculate multiple weighted average regressions.

The first calculated component was the original weighted averageinverse of the environmental variable. The following components were the weighted averages for the residual of the environmental variable (ter Braak and Juggins, 1993). The accuracy of the regression analyses was tested by looking at the r² and Root Mean Square Error of Prediction (RMSEP) values. The RMSEP indicated the systematic differences in prediction errors, whereas the r² measured only the strength of the relationship between observed environmental variable and predicted environmental variable values (Horton et al., 2006). The jackknifing technique was applied to calculate r² and RMSEP values. In this technique multiple cycles were run. In every cycle the dataset was divided in a training set (for calibration of the transfer function) and a test set (for validation of the transfer

Testate amoebae as estuarine water-level indicators

function), but for every cycle one sample was left out of the analysis (Birks et al., 1990, ter Braak and Juggins, 1993). The ultimate goal was to use the regression analyses as part of a transfer function for soil elevation relative to mean high water level (MHWL), Normalized Elevation, and Flooding frequency.

2.3 Results

Testate amoebae & assemblages The modern testate amoeba assemblage of the freshwater tidal marsh was very species rich ( > 90 sp.; Appendix A ), covering 18 different genera. Almost half of the testate amoeba tests found belonged to the genus Trinema, with Trinema lineare as most abundant species (32 % of total counts). This taxon, together with Euglypha rotunda, Trinema enchelys, and Tracheleuglypha dentata occurred in all analyzed samples. Apart from freshwater species, some brackish and one marine interstitial species (Cyphoderia littoralis) were found.

Two major modern testate amoeba assemblages were distinguished based on CONISS cluster analysis (Fig. 2.2). The first major species assemblage (zone A), occurring on the highest elevations, was characterized by Cyclopyxis kahli, Arcella arenaria, Difflugia globulus and Difflugia globulosa. The second major species assemblage (zone B) was located in the lower part of the marsh and is characterized by Difflugia tenuis and Euglypha tuberculata. Three sub-assemblages were recognized in the lower marsh assemblage.

The highest of the three (zone BI) contained highest abundances of Trinema lineare var. truncatum and Centropyxis minuta. The middle sub-assemblage (zone BII) was defined by Hyalosphenia minuta, Difflugia lucida and Centropyxis aerophila var. sylvatica. Species as Difflugia pristis and Cyphoderia ampulla determined the lowest sub- assemblage (zone BIII).

Chapter 2

Figure 2.2 Testate amoeba species diagram, with indication of the 4 testate amoeba zones based on cluster CONISS-analysis. 49

Testate amoebae as estuarine water-level indicators

The short unconstrained DCA gradient length (1.702 SD ) indicated that a linear model was appropriate. Therefore, a Redundancy Analysis (RDA) was performed to investigate the relation between the testate amoeba assemblages and environmental variables (Flooding frequency, Salix, Phragmites, Herbaceous cover, LOI, Sand, Silt, Clay, Elevation, Bulk density). First, two environmental variables, Elevation (VIF > 20) and Sand (VIF = 0.00), were omitted from the model.

Figure 2.3 RDA analysis (species were square root transformed). The ellipse shows species that are negatively correlated with Flooding frequency and characteristic for zone A (Fig. 2.2). Abbreviations of species names are found in Fig. 2.2. Continuous variables are indicates with arrows, categorical variables with centroids.

Chapter 2

The RDA graph (Fig. 2.3) showed that Arcella arenaria, Trinema complanatum, Cyclopyxis kahli, Difflugia globulus and Difflugia globulosa formed a separate group, in agreement with the CONISS analysis (Fig. 2). These species were negatively correlated with flooding frequency (Fig. 2.3). Partial RDA analysis revealed that the environmental variables together, without Sand and Elevation, explained 39.3 % (p = 0.0010) of the species variance. The significant variables were Flooding frequency (p = 0.0330), loss on ignition (p = 0.0040), and Clay (p = 0.0030). Together, they made up a significant part of the total explained species variance (Table 2.1).

Table 2.1 Total and Partial explained species variation based on Redundancy Analysis Total variation Explainded variation 39 % Unexplained variation 61 % Partial variation Loss On Ignition 18 % Clay 17 % Flooding 13 % Intercorrelation 52 %

The four palaeo samples contained 29 testate amoeba species (10 genera) (Appendix B), of which two were not found in modern testate amoeba assemblages of this study area (Difflugia minuta, Euglypha strigosa var. glabra). In the palaeo testate amoeba assemblages was Trinema the dominant genera (42 %) and Trinema enchelys the most abundant species (19 % of total counts).

Testate amoebae as estuarine water-level indicators

Figure 2.4 Shannon-Wiener diversity index and Sørensen dissimilarity index plotted against Flooding frequency and Elevation (m). The line shows the moving average for Shannon-Wiener diversity index and Sørensen dissimilarity index

Chapter 2

Diversity analysis The moving average trend line fitted through the Shannon-Wiener species diversity data increased rather gradually with increasing marsh elevation (Fig. 2.4). The species diversity tended to be rather constant as long as the flooding frequency was higher than 8 %, while the diversity increased with decreasing flooding frequencies smaller than 8 % (Fig. 2.4). For the Sørensen dissimilarity index, there was an overall trend of decreasing dissimilarity with increasing elevation on the marsh. Hence on the low marsh there tended to be bigger changes in species composition between successive samples along the elevation gradient, while the higher on the marsh, the smaller changes in species composition between successive samples became. Especially at low flooding frequencies ( < 2 %) the dissimilarity decreased strongly.

Transfer function A transfer function was developed for Flooding frequency, Elevation (relative to MHWL) and Normalized Elevation. Since the gradient length of the unconstrained DCA (1.702 SD) was close to 2 SD, both linear and unimodal regression methods were applied. The WA-PLS regression was not useful for this dataset as the best component was component 1.

For Flooding frequency, the r² and RMSEP values of WA-TOLclassic (Table 2.2) seem to show the best results, but this method tended to overestimate low flooding frequencies and underestimate high flooding frequencies (Fig. 2.5). On the whole (Table 2.2 and Fig. 2.5), Weighted

Averageinverse showed the best fit for flooding frequency.

Testate amoebae as estuarine water-level indicators

Table 2.2 Jack-knifed cross validation results for Partial Least Squares, Weighted Average and Weighted Average-Partial Least Squares Regressions. The chosen regression methods values are put in bold. Regression Flooding Normalized model frequency Elevation (m) Elevation

RMSEP RMSEP r² (%) r² (m) r² RMSEP PLS (component 2) 0.54 11.0 0.69 0.15 0.69 0.05

WAclassic 0.60 11.8 0.70 0.17 0.70 0.05

WAinverse 0.59 10.3 0.69 0.16 0.69 0.05

WA-TOLclassic 0.64 9.8 0.53 0.23 0.53 0.07

WA-TOLinverse 0.63 11.8 0.51 0.20 0.51 0.06 WA-PLS (component 1) 0.59 10.3 0.69 0.16 0.69 0.05

For the Elevation, the cross validation plots were more or less comparable, apart from WA-TOLinverse that underestimated elevation of samples high on the marsh and WA-TOLclassic that had a bigger error range. Since there were no apparent differences between the other regression methods in cross validation, we chose PLS (component 2) as best regression method for Elevation. It had the lowest RMSEP (0.15 m). For the Normalized Elevation, the cross validation plots were comparable to those of Elevation, but in this case WAclassic was chosen since it showed a slightly better result for r² value (0.70).

Figure 2.5 (next page) Cross validation plots of PLS, WA, WA-PLS regression methods for Normalized Elevation (A), Flooding frequency (B) and Elevation (relative to MHWL) (C).

Chapter 2

Testate amoebae as estuarine water-level indicators

The chosen regression methods were applied as a transfer function on the palaeo testate amoeba assemblages. An overview of the results is shown in Table 2.3. Taking the error range (RMSEP) into account, there were no considerable changes in Flooding frequency, Elevation and Normalized Elevation between the different depths. The reconstructed values were close to the present values, suggesting that the transfer function works.

Table 2.3 Application of the transfer functions on palaeo-samples.

Transfer function Method Modern values RMSEP D05 D15 D25 D45 (error)

Flooding frequency (%) WAinverse 32 10.3 34.6 37.5 35.75 38.78

Elevation (m) PLS (comp. 2) 0.22 0.15 0.2 0.18 0.16 0.25

Normalized Elevation WAclassic 0.9 0.05 0.9 0.92 0.91 0.96

2.4 Discussion

Modern data This study demonstrates that testate amoebae from freshwater tidal marshes can be used as indicators for estuarine water levels with a relatively high accuracy. The precision of the presented transfer function for Normalized Elevation is similar to published sea level transfer functions based on foraminifera, diatoms, and testate amoebae from salt marshes (Table 2.4). Both r² and RMSEP of our transfer function lay within the range of the salt marsh transfer function values.

The different regression models that were used for the transfer functions for (Normalized) Elevation and Flooding frequency, all resulted in r² and RMSEP values that were in the same order of magnitude (Table 2.2). This demonstrates the robustness of the relationships that were found between testate amoeba assemblages and the environmental variables (Normalized) Elevation, Flooding frequency). Although these results are comparable, the precision of the reconstructions of the different

Chapter 2 regression methods is not. Some regression methods tend to under or overestimate the lowest and/or highest observed values of variables (Fig. 2.5). It is not surprisingly that for every environmental variable a different regression method was found to give the best fit, because there is no linear relationship between the environmental variables Flooding frequency is decreasing exponentially with increasing elevation (Fig. 2.2).

Although the obtained transfer functions appear to be robust and of similar precision as previously published transfer functions, they must be used cautiously. The relationship between testate amoeba assemblages and elevation or tidal flooding may differ greatly between marshes. Therefore, this study must be considered as a first step towards the use of testate amoebae as indicators of estuarine water levels, and data from more sites along the estuary should follow to enable the construction of a transfer function that is applicable on a more regional scale (for the whole estuary). Furthermore, variation in species assemblages is not solely related to the variation in elevation or flooding frequency. Other environmental variables such as sediment particle size and organic matter content play an important role in our freshwater tidal marsh (Fig. 2.3), as was also demonstrated for salt marshes (e.g. Charman et al., 2002).

Testate amoebae as estuarine water-level indicators

Table 2.4 Comparison of our study with studies on salt marshes of the UK. Table based on Gehrels et al., 2001.

Number Normalized Training of sampled Regression Normalized set samples range model r² RMSEP References

salt testate marshes(UK) amoebae 52 1.01-1.36 WA-TOL 0.44 0.076 Gehrels et al., 2001 freshwater tidal marsh testate (BE) amoebae 42 0.78-1.17 WA 0.70 0.05 This study salt marshes testate (N-Am) amoebae 29 ~0.65-1.18 WAclassic 0.85 0.054 Gehrels et al., 2010 salt marshes(UK) diatoms 94 0.73-1.33 WA-PLS 0.78 0.054 Gehrels et al., 2001 Zong & Horton 1999 salt marshes ~0.00- (as described in (UK) diatoms 88 1.40 WA-TOL 0.72 0.214 Gehrels et al., 2001) salt marshes(UK) foraminifera 92 0,73-1,21 PLS 0,38 0,08 Gehrels et al., 2001 Horton et al. (1999) salt marshes (as described in (UK) foraminifera 131 ~0,4-1,2 WA 0,67 0,116 Gehrels et al., 2001)

Although freshwater tidal marsh and salt marsh testate amoebae respond to the same environmental variables, there are a number of considerable differences. Firstly, testate amoeba concentrations of salt marshes (up to 65600 testate amoebae/cm³) (Charman et al., 1998) are much lower than freshwater tidal marsh testate amoeba concentrations (max. ± 300 000 testate amoebae/cm³). Secondly, testate amoebae inhabit a much smaller vertical range in salt marshes in comparison with freshwater tidal marshes (Table 2.4). Finally, testate amoebae species might act different in both marsh types. For example, Tracheleuglypha dentata is a good indicator species for sea-level changes in salt marshes (narrow vertical range) (Gehrels et al., 2006), while the species is found over the entire freshwater tidal marsh elevation range.

Chapter 2

Four different testate amoeba zones have been distinguished. In zone A (Fig. 2.2), the species assemblage consists of terrestrial soil testate amoebae (e.g. Cyclopyxis kahli, Difflugia globulus) (Chardez and Lambert, 1981). The soil testate amoebae Cyclopyxis kahli is also found in the highest, and therefore freshwater, part of the salt marsh in the Seymour-Belize Inlet Complex (Riveiros et al., 2007). Although a rather high number of testate amoeba species are found, Shannon-Wiener diversity numbers (Fig. 2.4) are predominantly quite low ( ≤ 2.5). Since the interpretation of Shannon-Wiener diversity index is difficult, the Shannon-Wiener categories following Patterson and Kumar (2002) and Riveiros et al. (2007) are applied. They state that the index indicates a stressed environment (Shannon-Wiener : 0.1 - 1.5), a transition environment (Shannon-Wiener: 1.5 - 2.5) or a stable environment (Shannon-Wiener: 2.5 - 3.5). This method shows that the whole sampled elevation gradient (zone A-BIII) lies within the category of an ‘environment in transition’. In the highest zone (zone A), the Shannon-Wiener index progressively approaches (but does not reach) a value characteristic for a ‘stable environment’. Here, less than 2 % flooding is required to find a rather stable (Sørensen dissimilarity < 0.1; Shannon-Wiener = 2.5; Fig. 2.4) terrestrial soil assemblage (zone A). Flooding frequencies of more than 2 % result in a ‘transition environment’ indicated by lower species diversity and a rise in dissimilarity between adjacent testate amoeba assemblages (zone B; Fig. 2.2). In zone B, two boundaries separate the three different sub- assemblages. The first boundary, between zone BI and BII, lies exactly on the MHWS level (i.e. mean high water level during spring tides), dividing testate amoebae in a supratidal zone (above MHWS) and an intertidal zone (between MHWS and MTL). At salt marshes, the MHWS forms the lower boundary of testate amoebae occurrence (MHWS = 1.0 Normalized Elevation; Table 2.4). The reason for the disappearance of salt marsh testate amoebae below MHWS is still not known for sure, but

Testate amoebae as estuarine water-level indicators

it is hypothesized that the testate amoebae distribution in salt marshes can be limited by salinity (Charman et al., 1998). In the freshwater tidal marsh a flooding frequency of 8 %, can be seen as a threshold value between a supratidal (zone A + BI; Fig. 2.2) and intertidal (zone BII + BIII; Fig. 2.2) environment. Trinema complanatum can be called a supratidal species, while Cyphoderia ampulla is found to be an intertidal species.

The intertidal testate amoebae zone contains two sub-assemblages (zone BII and zone BIII) which can be separated by the presence and absence of Phragmites australis. The boundary matches approximately (height difference of 0.05 m) with the natural upper boundary of the Phragmites australis vegetation. Thus, high abundances of Difflugia pristis and Cyphoderia ampulla, indicator species of zone BIII, might be seen as indicators for the presence of Phragmites australis. The fact that Cyphoderia ampulla was also only found within the range of Phragmites australis at the UK salt marshes of the Erme (transect 1& 2) and in Brancaster (Charman, 2001) supports this finding.

The lower elevation limit of testate amoebae appearance seems to be determined by the presence or absence of marsh vegetation. Within the marsh vegetation, testate amoebae were always found in high abundances (max. ± 300 000 testate amoebae/cm³). At the bare mudflat, however, testate amoebae were only present in much lower concentrations (max. ± 3700 testate amoebae/cm³). This is probably due to the fact that the sediment surface on the bare mudflat is much more mobile during flood tides compared to the marsh sediment surface that is stabilized and protected from erosion by the marsh vegetation (e.g. Bouma et al., 2005, Temmerman et al., 2003a).

Chapter 2

Application of the transfer function to palaeo samples Application of the transfer functions of Normalized elevation, Elevation and Flooding frequency to palaeo samples show the robustness of the function, since all reconstructed values lay close to the modern values. The fact that the reconstructed values do not show important differences in (Normalized) Elevation or Flooding frequency can be explained by the fact that the vertical sediment accretion in the marsh is keeping up with the rising water level, as reported in Temmerman et al. (2003b, 2004b). Since the marsh has been rising equally with the water level, the relative position of our samples to the water level should have stayed the same. Although the transfer functions seem to work up to a depth of 45 cm (i.e. sediments of about 40 years old; Temmerman et al., 2003a), it remains to be investigated whether samples from deeper sediments contain enough testate amoebae. Testate amoeba concentrations decreased considerably within the 45 cm depth interval (± 73 000 tests g-1 to 3 200 tests g-1), which may hamper the counting of test in deeper sections of the sediment profile. The same problems of decreasing test concentrations with depth were found in the study of Charman et al. (2010). A possible explanation for the poor preservation of testate amoebae in coastal settings is given in Roe et al. (2002), suggesting that partial dehydration of the fossil tests may influence testate amoebae preservation. This explanation is very unlikely for our freshwater tidal marsh, were the sediment is always saturated. Here we suggest another possible reason for poor preservation of testate amoeba fossils in estuarine settings. Freshwater tidal marshes play an important role in estuarine silica cycling (Struyf et al., 2006, Struyf et al., 2005a, Struyf et al., 2005b). Freshwater tidal marshes contain high amounts of biogenic silica in sediment and vegetation, which dissolves in the pore water and functions as an important source of dissolved silica to the river (Struyf et al., 2005a). The constant export of dissolved silica to the river with every tidal cycle, exponentially decreases the amount of

Testate amoebae as estuarine water-level indicators

biogenic silica stored in the marsh sediments (Struyf et al., 2007), retaining only 40 % of the deposited biogenic silica at a depth of 50 cm. Testate amoebae (together with diatoms) can be the source of biogenic silica in the freshwater tidal marsh sediments. Testate amoebae have shells (partly) made up by silica. This would also imply that there is a selective dissolution of testate amoebae and that testate amoeba assemblages might alter with depth.

2.5 Conclusion

This study demonstrated that testate amoebae from a freshwater tidal marsh can be used as indicators of water level in an inland estuary. The main conclusions are:

1/ Freshwater tidal marsh testate amoebae can be used in future to reconstruct estuarine inland water levels with an accuracy of 0.05 Normalized Elevation. The accuracy is comparable to sea level transfer functions based on salt marsh diatoms, foraminifera and testate amoebae.

2/ Testate amoeba assemblages of freshwater tidal marshes are controlled by the same environmental variables as testate amoeba assemblages of salt marshes, being flooding frequency, surface elevation, organic content and particle size of the soil. Though, there are big differences between both marsh types regarding testate amoeba concentrations, behavior of testate amoeba species and the elevation range over which testate amoeba species occur.

3/ The testate amoebae assemblage of the freshwater tidal marsh responds in the first place to flooding frequency, but in second place they seem to react to the presence (Phragmites australis) or absence (mudflat) of marsh vegetation.

Chapter 2

4/ The different transfer functions based on the modern testate amoeba assemblages give good preliminary results when applying to fossil testate amoeba assemblages. However it is not excluded that testate amoeba assemblages might be altered over time, because of selective preservation.

The presented transfer functions offer a potentially powerful new tool to reconstruct and investigate Holocene estuarine water level changes, in response to sea level change and human modifications of estuarine morphologies, assuming good preservation of testate amoebae in the sediments.

Acknowledgements

We thank the Flanders Hydraulic Research institute for providing tide gauge data of 2007. Research funded by a Ph.D. grant of the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT-Vlaanderen).

Chapter 3

Testate amoebae as proxy for water level changes in a brackish tidal marsh* Marijke Ooms, Louis Beyens, and Stijn Temmerman

Few studies have examined testate amoebae assemblages of estuarine tidal marshes. This study investigates the possibility of using soil testate amoebae assemblages of a brackish tidal marsh (Scheldt estuary, Belgium) as a proxy for water level changes. On the marsh surface an elevation gradient is sampled to be analyzed for testate amoebae assemblages and sediment characteristics. Further, vegetation, flooding frequency and soil conductivity have been taken into account to explain the testate amoebae species variation. The data reveal that testate amoebae are not able to establish assemblages at the brackish tidal marsh part with flooding frequencies equal to or higher than 36.5 %. Further, two separate testate amoebae zones are distinguished based on cluster analysis. The lower zone’s testate amoebae species composition is influenced by the flooding frequency (~ elevation) and particle size, while the species variability in the higher zone is related to the organic content of the soil and particle size. These observations suggest that the ecological meaning of elevation shifts over its range on the brackish tidal marsh. Testate amoeba assemblages in such a brackish habitat show thus a vertical zonation (RMSEP: 0.19 m) that is comparable to the vertical zonation of testate amoebae and other protists on freshwater tidal marshes and salt marshes.

*Published as: Ooms M, Beyens L, & Temmerman S. 2012. Testate amoebae as Proxy for Water Level Changes in a Brackish Tidal Marsh. Acta Protozoologica 51: 271-289. Testaceae of a brackish tidal marsh

3.1 Introduction

In the last two decades more and more research is done on protists from salt marshes all over the world (Charman et al., 2002, Gabriel et al., 2009, Horton et al., 2006, Sawai et al., 2002, Woodroffe, 2009c, Zong and Horton, 1999). The studies are mainly focusing on the potential of using soil protists (mainly diatoms, foraminifera and testate amoeba) as proxies for sea level, with the goal of reconstructing Holocene sea level changes from sediment cores in which protists are preserved. The reconstructed sea level changes can help to understand the ongoing present and future sea level changes resulting from global warming, as the reconstructions can be used to test sea level change models. These reconstructions are established by the use of a transfer function, in which the relationship between modern protist (assemblages) and an environmental variable (e.g. elevation) is used to infer past environmental variables from palaeo protist assemblages preserved in sediment profiles. The accuracy of the sea level reconstruction is dependent on the protists used. The highest precision is found in a study on testate amoeba giving an accuracy that might be up to 4 cm (Gehrels et al., 2006). Further, compared to foraminifera, testate amoebae show only limited infaunal distribution (3 cm deep) (Roe et al., 2002), and can be quicker to investigate than diatoms (Gehrels et al., 2001).

The effects of recent sea level rise may be amplified towards the more inland part of estuaries due to human modifications of the estuarine morphology. As a consequence of land reclamation, the reduced water storage capacity of the estuary may result in considerable enlargement of the tidal range. This means that areas lower than 10 m above present Mean Sea Level are vulnerable for future inundations (McGranahan et al., 2007). Among these areas are large coastal cities like e.g. New York, Miami, Shanghai and Mumbai. The fact that inland water levels are

Chapter 3

highly influenced by the sea level rise, makes it interesting to reconstruct estuarine water level changes. This might help to reach a better understanding of the relation between sea level rise and estuarine tidal water level changes. For example, sea level rise at the Belgian coast was about 3 mm year-1 during the past century, while mean high water level rise was up to 15 mm year-1 in the inland part of the Scheldt estuary (Temmerman et al., 2004b). Though, there is only one study that investigated the use of protists (testate amoeba) as proxy for water level changes in the inland part of an estuary (Ooms et al., 2011). While that study was conducted at a freshwater marsh, present study investigates the modern testate amoebae assemblages of a brackish tidal marsh, expecting different species assemblages.

Brackish tidal marshes have a unique character by their occurrence at the transition from the freshwater to marine parts of an estuary. The average salinity (5 - 18 PSU) of these brackish areas is intermediate between that of the freshwater and marine parts and is highly variable during the year. The major causes of salinity variations in estuaries are freshwater discharge from the upstream and tributary rivers and the mixing of fresh and sea water through wind and tidal action (Peterson, 2007). These salinity variations occur over large distances within the estuary. The salinity instability, in combination with tidal currents and high sediment load, makes it more difficult for species to settle or adapt in the brackish part of an estuary in comparison to the more stable salinity conditions in the marine or freshwater zones (Little, 2000). Therefore, brackish environments are typically species poorer than marine and freshwater ecosystems (Remane and Schlieper, 1971). Yet the study of Więski (2010) discusses that plant diversity, primary production and nutrient recycling within brackish marshes is equal to freshwater tidal marshes and exceeds those of salt marshes. Furthermore, species densities and biomass might be bigger in brackish marshes. Species living in this environment are mostly marine or

Testaceae of a brackish tidal marsh freshwater species that have adapted to this habitat, while species that only occur in the brackish zone are rare (Little, 2000).

Here, we focus on the protozoan group of testate amoebae. Testate amoebae are shelled amoebae with an average size between 20 – 200 µm (Hendon and Charman, 1997). They can be found in a variety of moist to wet environments. Until now, around 2000 species have been described (Mitchell et al., 2008a). Testate amoebae have been studied in marine salt marshes in the UK and North America as proxies for Holocene sea level changes (Charman et al., 2010, Charman et al., 1998, 2002, Gehrels et al., 2006, Gehrels et al., 2001, Riveiros et al., 2007). So far, only one study is known of a freshwater tidal marsh (Scheldt estuary, Belgium) (Ooms et al., 2011). To our knowledge, no studies have been reported on modern testate amoebae assemblages of brackish marshes yet.

The objectives of this study are to investigate the modern testate amoebae species composition of a brackish tidal marsh, in relation to environmental variables (elevation, flooding frequency, particle size, vegetation, … ). Further, a transfer function will be developed to assess if the relationship between testate amoebae and elevation is comparable to that of other marsh types (salt and freshwater tidal marsh).

3.2 Material and Methods

Study area The Scheldt river is a rain-fed lowland river that has its source in France (St. Quentin) and streams through Belgium and The Netherlands to flow into the North Sea at Vlissingen. The tidal part of the Scheldt river is 160 km long, stretching between Ghent (Belgium) and the North Sea (Fig. 1.1). The Scheldt estuary has a semidiurnal meso- to macrotidal

Chapter 3

regime and is characterized by a full salinity gradient with a freshwater, brackish and marine part (Fig. 1.1). The tidal marshes and mudflats of Groot Buitenschoor are located in the mesohaline, brackisch part (5 - 18 PSU) of the Scheldt estuary at the border of Belgium and The Netherlands (Meire et al., 2005) (Fig. 1.1; Fig. 3.1). The tidal difference between mean high water and mean low water is here 4.98 m (Taveniers and Mostaert, 2009).

Figure 3.1 A. Map of the Scheldt estuary with indication of Groot Buitenschoor and the tide gauge of Bath (detailed map: Fig. 1.1). B. Map of the brackish tidal marsh Groot Buitenschoor with indication of vegetation zones and the elevation transects that are sampled. C. Photos of the two sampled transects; Photo 1. from Salix sp. to outer edge of Phragmites australis vegetation. Photo 2; from Phragmites australis to outer edge of Scirpus maritimus.

Testaceae of a brackish tidal marsh

The intertidal area of Groot Buitenschoor is 226 ha ( ± 2.3 km long and max. 1200 m broad). Groot Buitenschoor consists of a vegetated marsh with a mud flat in front. The marsh sediments consist of fine grained particles ranging from clay to fine sands. The vegetation zones on this marsh are, from high to low elevation, willows (Salix sp.), herbaceous vegetation (Urtica dioca, Elytrigia atherica), reed (Phragmites australis) and pioneer vegetation (Scirpus maritimus). At some parts of the marsh, the marsh edge is eroding, creating a cliff with a height up to about 1 m between the high marsh vegetation and the pioneer vegetation (Scirpus maritimus) on the mudflat (Fig. 3.1).

Sampling method Samples were taken at the Dutch part of Groot Buitenschoor along the biggest altitudinal gradient (-1.2 - 2.31 m ~ MHWL) (Fig. 3.1). The sampling campaign took place in April of 2010. The samples were collected along two altitudinal gradients, one running from the highest point, with Salix vegetation, to the transition between Phragmites australis and Scirpus maritimus, and the second, at a location with a smoothly sloping elevation gradient, from the transition zone between Phragmites australis and Scirpus maritimus to the lowest reach of the Scirpus maritimus vegetation on the mudflat. In this way, the whole elevation gradient, spanning all vegetation communities, was sampled. The sampling started at the highest point and for every elevation decrease of 5 cm, determined with a laser level, a sampling location was marked. In this way, 80 sampling locations were indicated along the elevation gradients. The accurate elevation of each sample location was measured with a Total station Sokkia set 5X10 (vertical error of ± 2 mm). The elevation measurements were relative to a benchmark (ALTI- Hd34) from the NGI (National Geography Institute). Further, two sediment samples (testate amoebae analysis + sediment properties) were collected from each sample location.

Chapter 3

The first sediment sample was a mixed sample for testate amoebae analysis, taken by the pooling method. For this method, a grid of 20 x 20 cm was laid down on the ground and in each corner and in the middle of the grid, a sediment sample of 2 cm deep was taken with a small corer (diameter= 0.9 cm). Pooling of the 5 cores smoothed out local variation of species assemblages. Considering a local marsh sedimentation rate of about 2 cm per year (Temmerman et al., 2003a,b), the soil sampling of 2 cm deep comprised the testate amoebae assemblage over about a one-year period.

The second sediment sample was taken by pushing a metal ring of fixed volume (84.5 cm³) into the ground within the pooling grid. The sediment sample was used for analyzing sediment properties (bulk density, loss on ignition, particle size and soil conductivity).

Preparation method for testate amoebae Testate amoebae samples were fixated with alcohol (95 %) directly after field sampling. Samples were dried in the oven at 30 °C for 36 hours. Only 2.5 g of sediment was used for testate amoebae preparation. Before preparation, Lycopodium spores (batch 177745: Results of the calibration (5 tablets): X= 92918 ± 1853, Variance = ± 2.0 %) (Stockmarr, 1971) were added to each sample for estimating testate amoebae concentrations. The preparation method was based on Hendon and Charman (1997), but samples were stained with Rose Bengal instead of Safranine. Analyses of testate amoebae assemblages were done on an OLYMPUS BX50 microscope with Nomarski optics. If the first slide of a sample contained at least 10 testate amoebae, the sample was further counted until 150 tests were found (appendix C). A number of 150 testate amoebae gave enough precision for this study following Patterson and Fishbein (1989) (Error calculations are in appendix D). Samples with less than 10 testate amoebae per slide were omitted from further analysis, because of low

Testaceae of a brackish tidal marsh testate amoebae concentrations. Differentiation between death, empty, tests and living testate amoebae, tests with amoebae inside, was made during counting, but total assemblages were used in the analyses.

Measuring and calculating environmental data Bulk density and loss on ignition were measured using a standard analysis method (Last and Smol, 2001). Fresh sediment material was oxidized with H2O2 before particle size analysis by the laser diffraction technique (Malvern 1000). Conductivity was measured using the standard protocol of Tucker & Beatty (1974), by mixing of 5 g of soil with 25 ml of distilled water for 60 minutes and settling for 30 minutes before measuring electrical conductivity.

The measured elevations were transferred to elevations relative to mean high water level by subtracting the mean high water level (MHWL). The MHWL was calculated from nearby tide gauge data of Bath from the period of the 1st of January 2010 to the 1st of November 2010. These data were downloaded from the online database “waterbase” from the Dutch Rijkswaterstaat. The dataset contained measurements of the water level in 10 minute intervals. The MHWL was determined by calculating the average of all the highest water levels for each tide. Flooding frequency was calculated by ranking the highest water levels for each tide from highest to lowest. Every highest water level was cumulatively numbered according to the number of times the flooding height was reached. The highest water level got value 1 and the lowest water level got the value corresponding to total water level values. Based on this ranking, flooding frequency was computed by dividing the number of ranking by total water level values. The normalized elevation was calculated using the formula of Gehrels et al. (2001) to be able to compare the performance of our transfer function with those published for different locations.

Chapter 3

Normalized Elevation = (Absolute Elevation–Mean Tidal Level)/(Mean High Water of Springtides-Mean Tidal Level).

Data analysis Relative abundances of testate amoebae species were used in all analyses. Furthermore, all species that never reached 2 % of relative abundance in one of the samples were deleted from the analyses. Patterson and Fishbein (1989) stated that for 150 specimens counted, only rare species of at least 2 % relative abundance could be used to help distinguish environments that differ by 2 % abundance. The species values were square root transformed for cluster and ordination analyses, to give more weight to less dominant species.

Cluster analysis was performed to distinguish different testate amoebae zones. The analysis was done with the dissimilarity coefficient Edwards and Cavalli-Sforza’s chord distance in the Tilia program CONISS (Grimm, 2004). Ordination analyses were carried out, in the program CANOCO 4.5 (ter Braak and Šmilauer, 1998), to study the relationship between testate amoebae species and environmental variables. First, an unconstrained Detrended Canonical Correspondance Analysis (DCCA) was run to find out if a linear or unimodal ordination method was appropriate. A DCCA gradient length smaller than 2 Standard Deviations (gradient length = 1.448 SD) indicated the use of a linear ordination model. Therefore, direct linear gradient analysis was performed, called Redundancy Analysis (RDA). This type of analysis uses the environmental variables to explain the species data. The RDA calculates fitted values by performing a multiple regression for each species on the environmental variables (ter Braak and Šmilauer, 1998). To limit the number of variables, a forward selection procedure was carried out.

Testaceae of a brackish tidal marsh

Figure 3.2 Environmental variables (flooding frequency, particle sizes, organic content, bulk density, soil conductivity, vegetation) and testate amoebae concentrations over the elevation gradient (m ~ MHWL) of the brackish tidal marsh. The four indicated zones are based on cluster analyses of testate amoebae assemblages (see Fig. 3.3).

Chapter 3

Testaceae of a brackish tidal marsh

Figure 3.3 (previous page) Relative abundances of testate amoebae over the elevation range (m ~ MHWL) and division of testate amoebae in four different bio-zones based on cluster analysis (right side of graph). The species’ names are followed by the abbreviation used in the RDA analysis.

The selection method ranks environmental variables following their importance for determining the species data (ter Braak and Šmilauer, 1998). Here, only significant (p value ≤ 0.05) environmental variables were used in the redundancy analysis. The partial variance was determined in Total, with all significant variables, and separately for Elevation and the subset of other significant variables using the Monte Carlo Permutation Test (999 permutations).

Transfer functions were made for elevation and normalized elevation, using multiple regression models (Partial least squares, Weighted averaging, tolerance Weighted averaging and Weighted averaging- Partial least squares (WA-PLS). More information about the regression models can be found in (ter Braak, 1987a,b; ter Braak and Juggins, 1993). The regression models were calculated in C2 (Juggins, 2007). The outcome of the regression models was tested by the jack-knifing technique. This technique is related to the bootstrapping technique, but leaves one sample out of the dataset with every cycle. The dataset is run for multiple cycles and is every time divided in a calibration and validation set (Birks et al., 1990, ter Braak and Juggins, 1993). The resulted r2 and root mean square error of prediction (RMSEP) values can be used to test the robustness of the models. The r2 value indicated the strength of the correlation between the observed modern data and the inferred data, and RMSEP is a measure of the prediction error (Horton et al., 2006).

Chapter 3

After determination of the best suited regression model, analysis of outliers was performed on the full transfer function model by checking for residual values that were higher than the standard deviation of the environmental variable (Edwards et al., 2004).

3.3 Results

Environmental variables The environmental variables that were taken into account in this study are shown in Figure 3.2. The horizontal lines showed the bio-zonation based on cluster analyses of the testate amoebae assemblages (Fig. 3.3). Two major bio-zones (A and B) were separated, which were each splitted into two smaller bio- zones (A1, A2, B1, B2). The A-zones were separated from the B-zones by the very low or lacking flooding frequency. The characteristic differences for dividing zone A1 and A2 are related to the high amount of organic matter of zone A2 and the low amount of clay in zone A1. Zone B is separated in the low zone B2 with highest flooding frequency and zone B1 with highest amount of sand and highest bulk density. The particle size distribution follows the changes in flooding frequency really well. For elevations with a tidal flooding frequency of more than 10 %, the grain size distribution is characterized by a high silt content (60 – 75 %), moderately high clay content (16 – 22 %), and low sand content ( < 10 %), which is typical for low-energy tidal deposition in tidal marshes in the Scheldt estuary (e.g., Temmerman et al. 2004; Ooms et al. 2011). For higher elevations with less than 10 % flooding frequency, the silt and clay contents gradually decrease while the sand content increases, which can be explained because the higher elevations are only rarely flooded by very shallow water depths, so that few or no tidal deposition occurs anymore and a transition towards a terrestrial soil takes place.

Testaceae of a brackish tidal marsh

Testate amoebae species composition The whole vegetated marsh elevation range was sampled to investigate the testate amoebae boundary of occurrence. Following the elevation range, the 50 highest samples were investigated on testate amoebae content. Forty, mostly elevated, samples were fully counted (150 testate amoebae individuals/sample) and used for analyses. The other 10, mostly lower samples, did not bare enough testate amoebae to count the whole sample ( < 10 testate amoebae slide-1) and were only used for calculating testate amoebae concentrations (see appendix C). This lack of testate amoebae in lower samples and the knowledge of decreasing testate amoebae concentrations with decreasing elevation on the marsh (Charman et al., 2002), gave reason to expect insufficient testate amoebae numbers in the remaining 30 samples, which is why they were not counted. The lower boundary of testate amoebae assemblage occurrence was set at an elevation of ± 20 cm above MHWL. In total, 43 testate amoebae species were found within the 40 samples. The most abundant species were Tracheleuglypha dentata, Trinema enchelys and Difflugia globulus. They were found along with Trinema lineare and Euglypha rotunda in all analyzed samples. Difflugia elegans var. parva was found in high relative abundances in the analyzed samples with insufficient testate amoebae numbers (pioneer vegetation). Most of the found species were known from freshwater biotopes, but there were also some marine interstitial species (e.g. Cyphoderia littoralis, Pseudocorythion acutum, P. wailesi) present (Golemansky, 1971, Golemansky and Todorov, 2004). The testate amoebae species belonged to 19 taxa. The genus Trinema represented more than 30 % of the found testate amoebae. The most abundant species, however, was Tracheleuglypha dentata, which accounted for more than 20 % of all counted testate amoebae. The genera of Campascus, Centropyxiella and Paulinella were only found once.

Chapter 3

On Figure 3.4, a for us unknown testate amoebae species/type was found that is called for this study Pseudohyalosphenia sp. 1. This species had the appearance of Hyalosphenia, but with a broad collar. The found specimens had a length of ± 50 µm, a width of ± 40 µm and a pseudostome of ± 20 µm. The body was round, only the pseudostome was flattened.

Figure 3.4 Photographs of unidentified species, called Pseudohyalosphenia sp.1

Testate amoebae assemblages After deleting species with low relative abundance ( < 2 %), 27 testate amoebae species were kept for further analyses. They are all shown in Fig. 3.3. The cluster analyses shown in the same figure made a division in two distinctive testate amoebae zones (A and B), which could both be subdivided in two separate zones (A1, A2, B1, B2). The two cluster zones A and B followed the environmental separation of intertidal and supratidal zone. Since these two zones represented different environmental conditions (tidal flooding or not), they were threaten separately in the Redundancy Analysis.

Testaceae of a brackish tidal marsh

Figure 3.5 Result of the Redundancy Analysis both on the intertidal zone and the supratidal bio-zone. Sub-zones were indicated by circles. Species abbreviations were used, full names can be found in Fig. 3.3.

Chapter 3

The species variation of intertidal zone B was partly explained by the environmental variables Elevation, Flooding, Clay and Silt (47.7 % ; p = 0.0010) (Fig. 3.5). Elevation is not a real ecological measure, but rather an anthropogenic one. This means that Elevation does not explain much species variation by itself. It are the real environmental variables (vegetation, flooding, conductivity, LOI, particle size, … ) that are linked to and change with elevation that explain the species variation. Partial RDA was calculated (Fig. 3.6) to understand the relationship between the environmental variables and Elevation. It showed that the biggest part of the species variation caused by elevation could be assigned to the relation with flooding and particle size (clay, silt). Flooding and particle size are only loosely linked to Elevation, because 30.5 % of the species variation was explained by Flooding, Clay and Silt.

The RDA of the supratidal zone A was executed without the variables Flooding, Herbs, Reed and Scirpus, as they were not of interest. The result of the RDA (Fig. 3.5) showed that Elevation, Loss on Ignition and Clay explained a significant part of the species variation of zone A (26.1 % ; p = 0.0150). Partial RDA (Fig. 3.6) showed that the biggest part of species variation due to Elevation could be explained by LOI and amount of clay (6.7 % in common). Apart from the interaction with Elevation, LOI and Clay explain together another 15.7 % of the species variation. The lowest zone (zone B2) (0.2 – 0.8 m ~ MHWL) was characterized by the presence of Pseudohyalosphenia sp. 1 , Pseudocorythion acutum, Cyphoderia ampulla, Cyphoderia littoralis and highest abundances of Difflugia pristis, Trinema lineare and Trinema lineare var. truncatum. The species variation of this zone was mainly determined by the particle size (Fig. 3.5). However, Difflugia pristis seemed to be more influenced by the amount of flooding. Zone B1 (0.8 – 1.35 m ~ MHWL) was represented by a higher number of species, namely Euglypha polyepsis, Cyclopyxis kahli, Difflugia lucida, Plagiopyxis declivis, Centropyxis ecornis, Centropyxis aerophila var. aerophila, Arcella catinus,

Testaceae of a brackish tidal marsh

Tracheleuglypha dentata and Centropyxis laevigata. Most of these species can be found in the left lower corner of the RDA (Fig. 3.5). These clustered species were not influenced by Flooding or Elevation, but were negatively correlated with Silt and Clay. Tracheleuglypha dentata showed a good relation with Elevation.

Figure 3.6 Results of the partial Redundancy Analyses for both intertidal (zone B) and supratidal (zone A) bio-zones. The values in the intersection of the circle are the common variation explained by the two variables.

Chapter 3

The highest middle zone A2 (1.35 – 1.65 m ~ MHWL) had high abundances of Centropyxis elongata, Euglypha rotunda and Difflugiella oviformis, which were positively correlated with the amount of organic matter in the ground (LOI) (Fig 3.5). Centropyxis laevigata was very abundant in the highest part of this zone. This was related to a small rise in clay. Zone A1 of the marsh (1.65 – 2.3 m ~ MHWL) was determined by the presence of Trinema complanatum, Euglypha strigosa var. glabra, Assulina muscorum and Trinema penardi, which were all correlated with Elevation (Fig. 3.5).

Transfer function The best regression method for Elevation and Normalized elevation was WA-PLS (component 2) with both highest r² and lowest RMSEP (Table 3.1; Fig. 3.7).

The analysis on outliers within the residualjack values (standard deviation environmental variable > residualjack value), resulted in the omission of three samples (S07, S11 ,S46; see appendix C). These samples had deviated species composition, probably related to the high flooding frequency (S46) or (antropogenic) disturbance of the environment (S07, S11). After removing the outliers, the model was highly improved (Fig. 3.7). This resulted in an increase of 0.14 for the r² value and decrease of the prediction error (RMSEP) with ± 0.08 m (Table 3.1).

Testaceae of a brackish tidal marsh

Table 3.1 Jack Knifed Results of the WA-PLS method for the environmental variables Elevation and Normalized Elevation. The table is split in three parts: the WAPLS model with complete data set (N= number of samples), after removing outliers and the partial transfer function for the intertidal zone.

WA-PLS (component 2)

Complete data set (N=40) r²jack RMSEPjack

Elevation 0.66 0.23 m ~ MHWL

Normalized Elevation 0.66 0.27 m

After removing outliers (N=37) Elevation 0.8 0.24 m ~ MHWL

Normalized Elevation 0.8 0.19 m

Partial Transfer function (zone B)(N=19) Elevation 0.67 0.17 m ~ MHWL Normalized Elevation 0.67 0.14 m

Based on cleaned transfer function, a partial transfer function was built using only the intertidal data (zone B) (Table 3.1; Fig. 3.7). The results showed lowest RMSEP values (RMSEPNormalized Elevation = 0.14), while the r² value was comparable to that of the complete data set transfer function. The partial transfer function underestimated the highest marsh levels of the intertidal zone (Fig. 3.7).

Chapter 3

Figure 3.7 Graphs of observed versus estimated Elevation and Normalized elevation, predicted by the transfer function based on Jack- knifed WA-PLS (component 2) for the complete dataset, after the removing of outliers and for the partial dataset.

Testaceae of a brackish tidal marsh

3.4 Discussion

Two testate amoebae species boundaries have been discovered on this brackish marsh. Firstly, the boundary separating the testate amoebae assemblage zones from the testate amoebae poor zone. Secondly, the border between intertidal (B1 and B2) and supratidal testate amoebae assemblages (A1 and A2). The different zones divided by these boundaries will be discussed below, starting with the lowest, the testate amoebae poor zone.

Testate amoebae poor zone (lower than 0.2 m ~ MHWL) This testate amoebae poor zone contains the whole pioneer vegetated Scirpus maritimus zone and also the lowest part of the Phragmites australis zone. All the investigated samples of the Scirpus maritimus zone have average testate amoebae concentrations of approximately 450 tests g-1. The Phragmites australis vegetation close to the marsh edge has slightly higher testate amoebae concentrations (average ± 970 tests g-1). These very low testate amoebae numbers together with the fact that only dead testate amoebae are found, might indicate that the few found tests are possibly allochtonous. This might also imply that the environment in the pioneer vegetation is too harsh for testate amoebae to survive or settle. The tidal inundation and salinity stress are possibly preventing testate amoebae to live within this zone. The boundary between poor testate amoebae densities and the presence of testate amoebae assemblages is set at a flooding frequency of 36.5 %, which corresponds with the appearance of countable testate amoebae concentrations (± 6560 tests g-1 (Table 3.2)).

Chapter 3

The intertidal testate amoebae zone (Zone B) This zone (0.2 - 1.35 m ~ MHWL) reaches from 36.5 % to 1 % of flooding frequency. The testate amoebae inhabitants of this zone are mainly related to the particle size of the soil and the flooding frequency (~ elevation) (Fig. 3.5 & 3.6). Since the two subzones (B1 and B2) have different testate amoebae assemblages and dominant environmental variables (Fig. 3.2 & 3.3), they will be discussed separately.

The lowest zone B2 (0.2 - 0.8 m ~ MHWL) starts with the highest flooding frequency (36.5 % - 3.5 %), at which testate amoebae could form an assemblage. The high flooding frequency with brackish water makes it possible for marine interstitial species to establish. These testate amoebae species (Pseudocorythion spec., Cyphoderia littoralis) (Fig. 3.3) are almost exclusively related to this zone. Apart from the fact that their appearance is associated with high flooding frequency, they seem to be more related to the interstitial space made by the particle size of the soil (Fig. 3.5). The interstitial space of this zone is rather small by the high concentration of clay and silt in the soil. This results in a testate amoebae assemblage of small species (e.g. Difflugia pristis). Looking at Table 3.2 with mean species numbers for each zone, zone B2 shows the highest species number, but also bears the lowest testate amoebae concentrations. The fact that there are many species (almost all had living representatives) but low concentrations might be explained in multiple ways. A couple of the possible explanations are: Firstly, the high species number might be explained by the tidal inundation. Every flooding gives the opportunity for allochtonous tests, carried through the river, to come ashore. The high flooding frequency in this zone facilitates the immigration. However, this would also mean that testate amoebae can be picked up in this zone and carried away with the tide. This might be the reason for low testate amoebae concentrations.

Testaceae of a brackish tidal marsh

A second explanation might be that the salty environmental conditions in this zone are very stressful for the organisms. Testate amoebae should osmoregulate to maintain or restore their cell volume. These high energy costs might slow down growth and therefore extend the generation time of the species. It is suggested in the study of Mbugua (2008) on marine Gymnamoebae that these amoebae have optimum growth at lower salinity levels because of the saving on energetic costs involved in osmoregulation. Following this hypothesis, a number of species may colonize the site, but the stress prevents them from reproducing rapidly and reaching large populations. The low concentration of testate amoebae implies low competition between species, facilitating a diverse species assemblage to establish. Further Cyphoderia ampulla was only found within this brackish marsh in the Phragmites australis vegetated zone. This finding is consistent with the exclusive presence of Cyphoderia ampulla within the reed zone of salt and freshwater marshes (Charman, 2001, Ooms et al., 2011). Possibly, it is more related to the environmental conditions in which Phragmites australis occurs, than to the presence of Phragmites australis, as Cyphoderia ampulla is also found within the deeper parts of lakes (e.g.Schönborn, 1962) and in moss (Todorov et al., 2009).

The second zone (zone B1) (0.8 - 1.35 m ~ MHWL) has a very low flooding frequency (between 3.5 % and 1 %) compared to the B1 zone. It also differs from the zone B2 by its high amount of Sand, which also results in high bulk density values. This sandy soil facilitates the occurrence of bigger species, like Centropyxis species (Charman et al., 2002), since interstitial space is bigger. Therefore, it is not surprising that this zone has the highest percentage over the four zones of Centropyxis (15 %) and Difflugia (26 %). Probably due to competition with bigger species, the concentrations of Trinema species are halved in

Chapter 3

this zone (Fig. 3.3). Most of the dominant species of this zone are not influenced by the flooding frequency at all (Fig. 3.5). For this zone, the tidal inundation (~ elevation) has no determining effect on the species assemblage.

The supratidal zone (zone A) Flooding frequency has a negligible effect on this zone (1.35 - 2.30 m ~ MHWL), as there is only one occurrence known of flooding in this zone for the past five years (highest water level: 1.49 m ~ MHWL). The environmental variables that influence the species composition of the testate amoebae assemblages in this supratidal zone are the amount of organic matter in the soil (LOI), particle size (Clay) and also the Elevation. This zone will also be separated in two sub-zones A1 and A2 for discussion.

Table 3.2 Mean and total mean species numbers and testate amoebae concentrations over the different zones. Species Mean number concentration zone A1 14.53846 27 867.03 zone A2 13 41 231.93 zone B1 14.36364 8 264.44 zone B2 17.375 6 560.11 Total Mean 14.79487 20 169.13

Zone A2 has a very high level of organic matter content of the soil. The highest concentrations of testate amoebae are also found within this zone. The testate amoebae concentration curve and the LOI curve (Fig. 3.2) follow the same pattern in this zone, indicating that they might be linked. The testate amoebae concentration peaks when organic content

Testaceae of a brackish tidal marsh is highest. Although the testate amoebae concentration is highest in this layer, the number of species (Table 3.2) is lowest. This low number of testate amoebae species might be explained by (selective) predation, competition between testate amoebae and/or between the different groups of protozoa. It is also possible that the yearly accumulating plant litter creates an uncolonized new habitat. Hence it could mean that only pioneer species have had the chance to colonize this new habitat.

Following Figure 3.5, the highest zone (A1) is highly influenced by the elevation on the marsh. Here, species like Assulina muscorum and Trinema penardi who like dry areas appear (Decloitre, 1981, Charman et al., 2000). This zone has lower species concentrations than the zone below. This is probably related to the fact that this zone might be too dry in comparison with the third zone. The species number rises slightly compared to the lower zone.

Comparison between intertidal and supratidal zone Although there is a clear boundary between the intertidal and the supratidal zone based on the cluster analysis (Fig. 3.3), there are a lot of species that occur in both zones. The exceptions are the marine interstitial species for the intertidal zone and Trinema complanatum, Assulina muscorum and Trinema penardi for the supratidal zone. These species inhabit mainly the lowest (B2) and highest zone (A1). This might be explained by the fact that these two outer zones are more stable. Zones B2 and A1 are either regularly flooded or not flooded at all. The intermediate zones B1 and A2 undergo a more irregular or occasional flooding and suffer more from the salinity variations throughout the year, as they appear as a sudden event. Therefore, the specific species assemblage inhabiting this zone is adapted to this (extreme)

Chapter 3

environment, which might explain the lack of high species variations in these two zones.

The anthropogenic measure Elevation has a different ecological meaning for the intertidal and supratidal zones (Fig. 3.6). The variable Elevation has no direct ecological meaning, since it pools together a number of environmental variables. For the intertidal and supratidal zone it can clearly be pointed out that the ecological meaning of the variable Elevation can change over the elevation gradient. The Elevation of the intertidal zone of the marsh (zone B) is mainly linked with Flooding and particle size, while in the supratidal zone (zone A) the Elevation greatly covers the differences in amount of organic matter and particle size.

Transfer function The transfer function on the full dataset show large prediction errors (RMSEP ± 0.30 m ~ MHWL) for Elevation and Normalized Elevation compared to published transfer functions of freshwater tidal marshes and salt marshes (Table 3.3) and also compared to transfer functions based on other protists. After removing outliers, the regression model has a comparable r² and RMSEP value as to other published studies. Still, two things need to be kept in mind. Firstly, the high accuracy of this transfer function should be treated with caution, as the sample number of the transfer function model is rather low (37 samples). However, it has been shown that transfer functions with around 40 samples can give good results (Ooms et al., 2011, Patterson et al., 2012). The partial intertidal transfer function has good results, but is in need of extra samples to become useful for actual palaeo water level reconstructions.

Testaceae of a brackish tidal marsh

Table 3.3 Table with an overview of made transfer functions on different marsh types and protists (based on the original table of Gehrels et al. 2001).

Number Normalized Training of sampled Regression Normalized set samples range model r² RMSEP References salt testate marshes(UK) amoebae 52 1.01-1.36 WA-TOL 0.44 0.076 Gehrels et al., 2001 freshwater tidal testate marsh (BE) amoebae 42 0.78-1.17 PLS 0.69 0.06 Ooms et al., 2011 brackish tidal marsh testate (BE) amoebae 37 0.16-1.91 WA-PLS 0.80 0.19 This study salt marshes(UK) diatoms 94 0.73-1.33 WA-PLS 0.78 0.054 Gehrels et al., 2001 Zong & Horton 1999 salt marshes ~0.00- (as described in (UK) diatoms 88 1.40 WA-TOL 0.72 0.214 Gehrels et al,. 2001)

Salt-marshes Hamilton & Shennan (Alaska) diatoms 154 ~0.4-2.4 WA-PLS 0.75 0.22 2005 salt marshes(UK) foraminifera 92 0.73-1.21 PLS 0.38 0.08 Gehrels et al., 2001

Horton et al. (1999) salt marshes (as described in (UK) foraminifera 131 ~0.4-1.2 WA 0.67 0.116 Gehrels et al., 2001) Salt marshes Leorri et a.l (Morbihan (FR) foraminifera 43 ~1.45-2.15 PLS 0.52 0.12 (2010) Salt marshes Leorri et a.l (Basque(S) foraminifera 59 ~1.3-2.25 WA-PLS 0.77 0.13 (2010) Salt marshes Leorri et a.l (Minho-Lima(P) foraminifera 49 ~0.5-2.0 WA-PLS 0.39 0.42 (2010) Salt marshes Leorri et a.l (Sado (P) foraminifera 22 ~1.45-2.10 PLS 0.22 0.14 (2010)

Secondly, as pointed out above, the ecological meaning of Elevation for testate amoebae assemblages has changed over the elevation gradient, as the intertidal zone is linked to the flooding frequency and supratidal zone to soil organic matter content. This raises the question whether the variable Elevation is useful for the reconstruction of water level changes if samples from higher on the marsh are included, which is often the case in testate amoebae studies. 92

Chapter 3

For now, we can say that the results of the transfer function give us the indication that the vertical testate amoebae assemblages of a brackish marsh have a range comparable to other protist species vertical marsh assemblages.

3.5 Conclusion

There are multiple environmental variables that influence the soil testate amoebae assemblages along an elevation gradient in a brackish tidal marsh. From this study we can conclude that:

1/ The lowest boundary of testate amoebae assemblage establishment (6560 tests g -1) is found at flooding frequencies of 36.5 % (± 0.2 m ~ MHWL). Below 0.2 m ~ MHWL, within pioneer vegetated Scirpus maritimus and Phragmites australis vegetation, testate amoebae concentrations are very poor (average 450 tests g -1).

2/ Within the altitudinal gradient, two testate amoebae zones (A and B), an intertidal and supratidal one, could be distinguished. Both testate amoebae zone assemblages varied with particle size of the soil. Also, the intertidal testate amoebae zone was related to the flooding frequency (from 36.5 % to < 1 %) and the supratidal zone to the amount of organic matter in the soil.

3/ Two intermediate intertidal zones (B1 and A2) are weakly related to Elevation and have a specific species assemblage that is adapted to the more extreme irregular environment of occasional flooding.

4/ The antrophogenic measure Elevation does not explain much species variation by itself. The species variation that Elevation explains is more related to the ecological variables that differentiate with elevation. For

Testaceae of a brackish tidal marsh this study, the Elevation was linked with flooding and particle size for the intertidal zone and was linked with organic matter and particle size for the supratidal zone, indicating that the ecological meaning of elevation changed with elevation. This emphasizes that caution is needed when using soil protists from tidal marshes as a proxy of elevation, and thus of mean sea level. Sea level studies should be based on the protist assemblages occurring in the intertidal zone of tidal marshes, which is regularly flooded by the tides, and should better not include the supratidal zone, which is only flooded during very rare extreme conditions.

5/ Transfer functions for Elevation and Normalized Elevation are made. The results point out that the testate amoebae of brackish water marshes show comparable vertical zonation as to salt and freshwater tidal marshes.

Acknowledgements

Research funded by a Ph.D. grant of the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT- Vlaanderen). We like to thank Gert Vanlerberghe for his help with correcting English grammar and vocabulary.

Chapter 4

Comparison of testate amoebae assemblages of the different salinity zones within the Scheldt estuary. Marijke Ooms, Louis Beyens en Stijn Temmerman

The two factors that may change the salinity gradient within an estuary are sea level rise and changes in freshwater discharge of the river. The present climate change influences both factors, possibly changing the salinity gradient throughout the estuary. Testate amoebae assemblages have been studied in salt and estuarine marshes as indicators for sea level changes. Until now, no research has been done on the changes in testate amoebae assemblages along a salinity gradient. This study presents a comparison of modern testate amoebae assemblages from salt, brackish and freshwater tidal marshes along the Scheldt estuary. No testate amoebae assemblages were found at the salt marsh, probably related to insufficient marsh elevation. Though, comparison of our testate amoebae assemblage data from fresh- and brackish marshes with salt marsh testate amoebae assemblage data of other studies made it clear that testate amoebae have high species numbers in salt and freshwater marshes and low in brackish marshes. Testate amoebae assemblages of freshwater tidal marshes change continuously with marsh elevation, while brackish testate amoebae assemblages change abruptly one into another. Although brackish and freshwater tidal marsh testate amoebae assemblages share a significant part of the testate amoebae species, their optima and tolerances are different for both locations. Further, with decreasing salinity, testate amoebae can be found at lower marsh locations subject to higher flooding frequencies.

Comparison of testate amoebae assemblages along a salinity gradient

4.1 Introduction

The focus of testate amoebae studies on salt marshes and estuarine marshes lies on the use of testate amoebae as sea level indicators, with the aim of reconstructing past sea level changes. Multiple studies have investigated modern testate amoebae assemblages in relation to marsh elevation relative to mean high water level with the purpose of making a transfer function (Charman et al., 2002, Gehrels et al., 2006, Gehrels et al., 2001). This transfer function describes the relationship between modern testate amoebae assemblages and an environmental variable (here elevation relative to MHWL) and can be used to infer past environmental variables from palaeo testate amoebae assemblages. Thus, palaeo testate amoebae assemblages have been studied to calibrate the transfer function for the reconstruction of past sea level changes (Charman et al., 2010, Roe et al., 2002).

For estuaries, it is likely that, together with sea level change, the salinity gradient within estuaries will change. This gradient is determined by the tidal-driven input of sea water into the estuary and the freshwater river discharge. Both sea level and freshwater discharge are expected to change by the ongoing climate change (Bindoff et al., 2007). This additional effect of sea level change on changing salinity gradient in an estuary is generaly not taken into account in the reconstructions of past sea level changes.

Therefore, this study will investigate the changes in modern testate amoebae assemblages over the salinity gradient of an estuary. Modern testate amoebae assemblages of a freshwater tidal marsh, brackish and salt marsh along the Scheldt estuary (South West of The Netherlands) will be compared.

Chapter 4

4.2 Material and Methods

Study sites The Scheldt estuary stretches from Ghent (Belgium) to the North Sea (Netherlands) (Fig. 1.1, 4.1). This estuary, with a length of 160 km, has a semidiurnal and meso- to macrotidal regime. The estuary spans the whole salinity gradient, ranging from a polyhaline zone (Vlissingen – Hansweert (NL)), to a mesohaline zone (Hansweert (NL) - Antwerpen (BE) and a oligohaline zone (Antwerpen – Ghent (BE)) (Meire et al., 2005) (Fig. 1.1). The Paulina marsh (Fig. 4.1, 4.2) is located in the polyhaline zone of the Western Scheldt (NL). Its vegetation composition varies with the age of the marsh. The old, high, marsh is covered with herbaceous saltmarsh vegetation (e.g. Atriplex portulacoides, Limonium vulgare) and the young, low, marsh is dominated by pioneer vegetation (Spartina anglica) (Temmerman et al., 2003a). The Groot Buitenschoor (Fig. 4.1, 4.2) is located in the mesohaline zone at the Dutch-Belgian border. This marsh shows also different stages of marsh development, as the highest locations are grown by willow trees, indicating very low salinity levels, and the lowest by pioneer vegetation (Scirpus maritimus). In between, a reed zone (Phragmites australis) and some herbaceous vegetations (Urtica dioica, Elytrigia atherica) are found (Ooms et al., 2012). The Notelaar marsh (Fig. 4.1, 4.2) is located in the oligohaline part of the estuary near Temse. This marsh is located at the place of highest tidal range. Also this marsh has different vegetation zones, but lacks pioneer vegetation. The oldest, highest part of the marsh is vegetated with willow trees, the younger, lower, part is grown with Phragmites australis. Herbaceous vegetation (Impatiens glandulifera, Urtica dioica, Convolvulus arvensis) grows beneath the willow trees and at the borders of the Phragmites australis vegetated zone (Ooms et al., 2011, Temmerman et al., 2003a).

Comparison of testate amoebae assemblages along a salinity gradient

Figure 4.1 A. Map of the Scheldt estuary with indication of the 3 studied marshes. B, C, D. Overview of the individual marshes with indication of the vegetation zones and the sampling locations: salt marsh Paulina (B), Brackish marsh Groot Buitenschoor (C), freshwater tidal marsh Notelaar (D).

Chapter 4

Figure 4.2 Average salinity gradient along the Scheldt estuary, going from the mouth stream upwards to Ghent, with indication of the study sites (Taverniers et al., 2011).

Sampling method The sampling for testate amoebae analysis was done by collecting the upper 2 cm of sediments by pushing a cork corer of 0.9 cm diameter into the ground for five times using the pooling method on a grid of 20 x 20 cm. The five resulting sediment samples were mixed and fixated with formaldehyde (Notelaar samples) or ethanol (Groot buitenschoor and Paulina samples).

At all three marshes, the whole elevation gradient was sampled ranging from the highest to the lowest marsh locations. Further details on the sampling method on the freshwater tidal marsh of the Notelaar and brackish marsh of Groot buitenschoor can be found in Ooms et al. 2011 (Chapter 2) and Ooms et al. 2012 (Chapter 3).

The sampling campaign of the salt marsh of Paulina was performed in April of 2011. The whole elevation gradient on the marsh was sampled by following a transect from the highest locations near the dyke towards the lowest marsh locations near the Scheldt stream channel. The highest samples were taken at some elevated ridges located within the marsh, which are remnants of human-made dams that served in the past as high water refuge for grazing sheep (until some decades ago). These ridges were constructed decades ago from locally excavated marsh 99

Comparison of testate amoebae assemblages along a salinity gradient sediments, are now about 0.3 m higher than the surrounding marsh and are covered with natural marsh vegetation. Samples were taken along the transect with elevation intervals of 5 cm. This resulted in a sampled elevation gradient from -0.18 m to +0.5 m relative to mean high water level (based on the MHWL of Terneuzen from 2010 (waterbase, Rijkswaterstaat NL) ; 2.28 m NAP).

The sediment samples were prepared for testate amoebae analysis following the method of Hendon and Charman (1997), adapted as described in Ooms et al. (2012) with adding of lycopodium spore tablets (Stockmarr, 1971) to calculate testate amoebae concentrations. Testate amoebae counting was performed at a magnification of 400 times using a Olympus BX50 microscope with Nomarski optics.

Calculation of mean high water level and flooding frequency The mean high water level and flooding frequency of the salt marsh of Paulina was calculated with tidal data of Terneuzen from the year 2010. These data were accessed at “waterbase”, an online database provided by the Dutch Rijkswaterstaat. The MHWL was calculated as the average of the highest water levels of each tide over the period of a whole year. The flooding frequency was calculated by ranking the high water level of each tide from high to low. Each high water level was given a cumulative number corresponding to the number of times this water level was exceeded. The flooding frequency was determined by dividing the cumulative number trough the total number of high water levels.

Data analysis: Total merged dataset A merged dataset was made for the comparison of species composition. First, all species data of both brackish and freshwater tidal marshes was put together in one dataset. This dataset was further processed by selecting the species that had at least for one sample a relative abundance of > 2 %.

100

Chapter 4

A second merged dataset was made for the environmental variables. Soil conductivity was left out of the merged dataset, as only samples of the brackish marsh were measured. An extra categorical environmental variable, namely salinity (i.e. salt, brackish or freshwater), was put into the dataset to be used in further data analyses in order to test for species assemblages differences related to salinity.

Data analysis: Appendix E Appendix E shows the number of species per samples, counted on the original datasets (no removal of species with less than 2 % of relative abundance) of both freshwater tidal marsh and brackish marsh. The Shannon-Wiener index was calculated for the brackish marsh samples in the same way as for the freshwater tidal marsh samples (Chapter 2).

Data analysis: Ordination analysis (RDA) Ordination analysis was performed on the total merged dataset to study species variation and their relationship with environmental variables. The analyses were performed in CANOCO 4.5 (ter Braak and Šmilauer, 1998). To check if data should be treated with a linear (gradient length < 2 SD) or unimodal ordination technique (gradient length > 4 SD), a DCCA was executed. The gradient length of the DCCA was 1.371 SD, which indicated that the use of a RDA analysis was appropriate. First, an automatic forward selection procedure was run in order to select the environmental variables that explained a significant part of the species variation. The significance (p value < 0.05) was tested with a Monte Carlo permutation test on the full model with 999 permutations. The significant environmental variables were used in the RDA analysis.

Data analysis: Optima and tolerances Optima and tolerances of testate amoebae were calculated for each location seperately, on the subset of species that were abundant enough to be in the merged dataset. The calculations were done in C2 (Juggins,

101

Comparison of testate amoebae assemblages along a salinity gradient

2007). Optima and tolerances were calculated using Weighted Average regression with jack knifing cross validation. The optima and tolerances were calculated for each species as the average and standard deviation of the environmental variable (here flooding frequency) for the samples in which the species occured, weighted by the relative abundance of the species for each sample (Stevenson et al., 2006). The jack knifing technique is a cross validation technique that recalculates species optima and tolerances for multiple cycles under different scenarios (a random chosen sample is left out at each run).

4.3 Results

Salt marsh testate amoebae assemblages Testate amoebae analysis of the highest located samples revealed very low concentrations of testate amoebae (500 tests g-1). These low concentrations indicate that there is no formation of testate amoebae assemblages at this marsh, since highest testate amoebae concentrations were expected to be found in highest located samples. There were only two species found within the samples, namely Difflugiella oviformis and Paraquadrula irregularis.

Since the investigated elevation gradient on the salt marsh did not bear testate amoebae assemblages, salt marsh data will not take part in the comparison of species assemblages throughout the estuary.

Comparison of testate amoebae species composition of a freshwater- and a brackish tidal marsh The total species numbers that were found at both locations differ greatly. More than twice as much species were found in the freshwater tidal marsh (95 species) compared to the brackish marsh (42 species). Appendix E shows that there are more species per sample in the freshwater tidal marsh compared to the brackish marsh at equal

102

Chapter 4

elevation. In general, the Shannon-Wiener diversity index is not that different between both locations.

Though, a great part of these freshwater tidal marsh species were rare. In order to make a better comparison between both marsh species compositions, a total merged dataset was made. This dataset consisted of 48 species of which only four testate amoebae species were found within all counted modern samples of both marshes. These species were Tracheleuglypha dentata, Euglypha rotunda, Trinema enchelys and Trinema lineare. Together, they accounted for 60 % relative abundance calculated over all samples. Trinema lineare was the most abundant species with 23.3 % of relative abundance calculated over all samples.

Thirty one species, out of 48, of the merged dataset appeared in both the brackish and freshwater tidal marsh. This implies that 17 species were only found in one of both locations. Fifteen species were only found within the freshwater tidal marsh. Though, these species were not very abundant as, together, they accounted for only 2 % relative abundance of the whole dataset. The other way around, two species (Pseudocorythion acutum, Pseudohyalosphenia sp. (see short description Chapter 3)) were only found within the brackish tidal marsh.

Testate amoebae assemblages and environmental variables Testate amoebae assemblages were investigated in relation to environmental variables in order to see if freshwater and brackish species assemblages show the same relationship with environmental variables. First, forward selection was used to determine the significant environmental variables. In total, seven environmental variables were selected (Table 4.1), which together explain 40.6 % of the species variation. The results of the RDA clearly show that there is a distinction

103

Comparison of testate amoebae assemblages along a salinity gradient

Figure 4.3 RDA diagram of merged dataset of freshwater and brackish samples. Each plotted point represents a sample assemblage and its corresponding name consists of numbers, related to the sample elevation (relative to MHWL), and abbreviation b (brackish marsh) or f (freshwater tidal marsh). Continuous environmental variables are indicated with an arrow and categorical ones with a circle. The thick black line indicates the division into freshwater and brackish samples, thin dotted lines indicate separate species groups. between the brackish and freshwater tidal marsh samples (Fig. 4.3). The position of the samples is mainly orientated along the line of Elevation, with highest samples located in the lower right corner. The freshwater tidal marsh samples make up one cloud, indicating a more continuous

104

Chapter 4

change in species assemblages. The brackish marsh samples are more divided in separate species assemblage zones, which corresponded to the division of the brackish samples by cluster analysis (Chapter 3).

Table 4.1 Significant environmental variables selected with forward selection and the partial species variation explained by the environmental variables. Forward selection procedure partial Variables (p value) RDA (%) Elevation 0.001 7.7 Salinity 0.001 7.6 Sand 0.001 5 Willow 0.003 3.1 Flooding 0.04 2.5 LOI 0.024 2.4 Herbs 0.034 2.2

Lowest boundary of testate amoebae occurrence Testate amoebae assemblages (≥± 6500 tests g-1) were not found over the entire elevation gradient within a tidal marsh. At the freshwater marsh, the lowest elevated testate amoebae assemblages are found slightly below MHWL ( -0.18 m ~ MHWL) at the boundary between vegetated marsh and mudflat (flooding frequency of 69 %) (Chapter 2). At the brackish tidal marsh, testate amoebae assemblages were observed starting from 0.20 m above MHWL, corresponding to a flooding frequency of 36.5 % (Chapter 3). This first occurrence does not correspond to the boundary between vegetated marsh and mudflat, but is located within the Phragmites australis vegetation above the pioneer Scirpus maritimus vegetated zone. No testate amoebae assemblages were found at the Paulina salt marsh with the highest elevation at 0.49 m above MHWL (flooding frequency of 13.6 %).

105

Comparison of testate amoebae assemblages along a salinity gradient

Testate amoebae optima and tolerances Both freshwater and brackish tidal marshes share a significant number of species. Calculation of testate amoebae optima and tolerance (Fig. 4.4) shows that the position of a species on the marsh surface is dependent on the salinity (freshwater vs. brackish). Species that occur at both locations have an optima and tolerance at higher flooding frequency for the freshwater tidal marsh compared to their occurrence on the brackish marsh.

Figure 4 (next page) A. Optima and tolerances of the abundant species of the freshwater and brackish marsh for flooding frequency (i.e. % of total number of tides that a certain location is flooded). The black squares represent values of brackish species (Groot Buitenschoor), the white dots show freshwater testate amoebae species (Notelaar). B. Optima and tolerances of abundant species of UK salt marshes based on the study of Gehrels et al. (2001) calculated on % of time flooded or ( i.e. flooding duration; % of time that a certain location is flooded).

106

Chapter 4

107

Comparison of testate amoebae assemblages along a salinity gradient

4.4 Discussion

Salt marsh testate amoebae Almost no testate amoebae were found at the Paulina salt marsh. This was to be expected as the lowest testate amoebae assemblages of other salt marsh studies were located in the upper half between mean high water of spring tides and highest astronomical tide level (Charman et al., 2002, Gehrels et al., 2006, Gehrels et al., 2001). The highest elevations measured at the Paulina marsh only laid 20 cm above mean high water of spring tides (2.58 m NAP Terneuzen).

Though, one can discuss whether the testate amoebae assemblages that are found on salt marshes in other studies represent real salt marsh testate amoebae assemblages or rather form a salt tolerant terrestrial testate amoebae assemblage, as flooding frequency is so low near Highest Astronomical Tide (i.e. the highest tide in the so-called 18.6 year nodal tidal cycle). Most of the species that have been found in previous salt marsh studies of (Charman et al., 2002, Gehrels et al., 2006, Gehrels et al., 2001, Roe et al., 2002) are originally described as freshwater species. So far 129 marine interstitial testate amoebae of the supralittoral zone (i.e. splash zone; lying above the high water line) were described globally (Golemansky and Todorov, 2004).

Further, it is not clear if the supralittoral zone consists of salt, brackish or freshwater sediments. In Riveiros et al. (2007), it is shown that with increasing elevation the salinity decreases. Therefore salt marsh assemblages, which have only been found in extremely high salt marshes close to Highest Astronomical Tide, might be considered as freshwater terrestrial testate amoebae assemblages as they are only very rarely and very shallowly flooded by salt seawater. The exhibition of these salt marsh testate amoebae assemblages to saline water is really low as all species optima are located at a flooding duration of less

108

Chapter 4

than 1 %, where flooding duration equals the summation of submergence time of all tidal cycles over a one year period divided by the total time. The representation of square rooted flooding duration gives impression of higher flooding duration than actually occurring ( e.g. species with a sqrt (% time flooded) of 0.5 experience only a 0.25 % of time flooded (average of 50 minutes in a 14-days tidal cycle).

Comparison between marshes along the estuary Total Testate amoebae species numbers are lowest in the brackish marsh (42 species) compared to the freshwater tidal marsh (95 species) and even the salt marshes (48 species; Gehrels et al., 2001). This was also reflected in the fact that freshwater tidal marsh samples generally had more species per sample compared to brackish tidal marsh samples of the same elevation above MHWL (Appendix E). The low number of testate amoebae species within the brackish marsh is consistent with findings in many studies on the diversity of animals and plants along estuarine salinity gradients. Generally, the brackish number of species is less than the number of species in freshwater or sea water (McLusky and Elliot, 2009). This is related to the fact that all estuarine species originate either from sea or freshwater species and salinity is the determining factor for the penetration of the species into the estuary. Therefore, the brackish marsh is inhabitated by the few freshwater and or sea water species that are resistant to the changes in salinity along the estuary. True estuarine species, mostly animals, are rare (McLusky and Elliot, 2009).

The testate amoebae assemblage of the brackish zone seems to exist mainly of species that are described as freshwater species. Though, some marine interstitial species such as Pseudocorythion acutum (Golemansky, 1971) and Cyphoderia littoralis (Todorov et al., 2009), are found. Further, the origin of the un-identified species Pseudohyalosphenia sp. is unclear.

109

Comparison of testate amoebae assemblages along a salinity gradient

The highest recorded freshwater tidal and salt marsh testate amoebae concentrations along the elevation range on the marsh (max. 300 000 tests/cm³ and 76 440 tests/cm³ (Charman et al., 2002) respectively) are higher than the highest brackish testate amoebae concentrations (max. 41 300 tests/cm³ ). Freshwater tidal marsh testate amoebae concentrations were higher than brackish tidal marsh testate amoebae concentrations (Appendix E). This does not follow the general finding that the low species numbers in the brackish zone of estuaries is compensated by having higher abundance of individual species (McLusky and Elliot, 2009). These low brackish testate amoebae concentrations might indicate that environmental stress is high at brackish marshes. The brackish zone of an estuary is typically subject to frequent shifts in salinity level related to the tidal water movement and to fluctuations in the freshwater discharge of the river (Van Damme et al., 2005), which makes it more difficult for species to establish assemblages.

The testate amoebae assemblages of the freshwater tidal and brackish marsh are different, as is shown in the RDA figure (Fig. 4.3). The main variation is due to local environmental differences. There are numerous local effects that might explain the discrepancy (e.g. food availability, predation, pH, … ). Though, we think that the clear separation between freshwater and brackish species assemblages might be related to the difference in salinity.

The first appearance of testate amoebae assemblages along the elevation gradient of the marshes was not the same for all locations. Testate amoebae assemblages existed only below a flooding frequency of 69 % at the freshwater marsh, while they appeared only below 36.5 % flooding frequency at the brackish tidal marsh. The salt marsh did not even contain testate amoebae assemblages at a flooding frequency of 13.5 %. It seems that the lowest occurrence of testate amoebae

110

Chapter 4

assemblages on the tidal marshes occurs at decreasing flooding frequency. The difference in lowest occurrence of testate amoebae assemblages was also noticeable in the species optima and tolerances. Species that habitated both freshwater tidal and brackish marshes (Fig. 4.4A) had higher optima and tolerances for flooding frequency in the freshwater tidal marsh compared to the brackish tidal marsh. Further, the freshwater species have a broader tolerance range, which makes that the testate amoebae assemblage composition changes continuously over the elevation gradient. This is confirmed in the RDA figure (Fig. 4.3, 4.4A), where no distinct testate amoebae assemblage groups could be distinguished. The brackish testate amoebae species have narrow tolerance ranges that either overlap or are separate from other species, indicating separate testate amoebae assemblages (Fig. 4.3, 4.4A).

This difference in arrangement of testate amoebae assemblages over the elevation gradient makes a difference in the precision when used in a transfer function. A transfer function was made for both freshwater and brackish tidal marsh testate amoebae assemblages (Chapter 2 and 3). Comparison of both shows that a continuous change in species assemblages as on the (freshwater tidal marsh) gives a better precision (RMSEP 0.06 Normalized Elevation) compared to the separate grouping of brackish marsh assemblages (RMSEP 0.19 Normalized Elevation). This can be explained by the fact that small changes in species assemblages will be noticed in the continuously changing assemblages, while they might be assigned to the same group in the separate assemblage arrangement.

The testate amoebae species that were exclusively found within the brackish marsh and had the highest tolerance for flooding frequency were all salt tolerant, marine interstitial species (e.g. Cyphoderia littoralis, Pseudocorythion acutum, Pseudohyalosphenia) (Golemansky and Todorov, 2004, Todorov et al., 2009). The most tolerant testate

111

Comparison of testate amoebae assemblages along a salinity gradient amoebae species of the freshwater tidal marsh ( e.g., Difflugia angulostoma, … ) are species that are generally found in pools (Gauthier-Liévre and Thomas, 1958).

The study of Gehrels et al. (2001) also contained a table with optima and tolerances for their abundant salt marsh species (Fig. 4.4B). Since they calculated optima and tolerances in relation to the square root of flooding duration, it is not possible to compare actual numbers with our analyses. Though, the low flooding times shown in the figure 4.4B suggest that these species have optima and tolerances of lower flooding frequencies than the brackish species, as these salt marsh species are only found very high within the marshes (lowest occurrence is above mean high water level of spring tides). A comparison may be made between salt marsh, brackish and freshwater tidal marsh testate amoebae optima and tolerances by looking at the ranking of species within the figure. The ranking gives the indication that Cyphoderia ampulla is one of the more resistant species to high flooding frequency/duration for both freshwater, brackish and salt marshes.

The separation in species optima and tolerances for different locations along the estuary is an important finding for the use of transfer functions regarding the reconstruction of sea level changes and the effect on salinity changes in estuaries. In order to reconstruct reliable sea level changes, it is important to use a local modern dataset in the transfer function. Further, the transfer function should be calibrated with palaeo testate amoebae assemblages of equal salinity of the modern transfer function dataset. Therefore, it is important that the study site has a stabile history of salinity to gain reliable reconstructions. A regional testate amoebae dataset, containing a combined dataset of testate amoebae assemblages of multiple marshes along the estuary,

112

Chapter 4

might be used to make up a transfer function to reconstruct past changes in salinity.

4.5 Conclusions

1/ Testate amoebae species numbers follow similar trends as reported for diversity patterns of plants and animals within our estuary, having high species numbers in the salt or freshwater zone and low numbers in the brackish part of an estuary.

2/ Testate amoebae assemblages behave differently in the brackish and freshwater tidal marsh with increasing elevation. Changes in testate amoebae assemblages occur continuously at the freshwater tidal marsh, while distinct testate amoebae assemblages follow up each other at the brackish marsh.

3/ Testate amoebae species that occur at the freshwater tidal marsh and brackish marsh have different optima and tolerance for flooding frequency. Species optima and tolerances for flooding frequency lie higher in the freshwater tidal marshes compared to the brackish marsh.

4/ The elevation of the lowest testate amoebae assemblage occurrence increases with increasing salinity along the estuary.

113

Chapter 5

Sea level reconstruction using sub-fossil testate amoebae: implications of selective preservation in sediment cores from a freshwater tidal marsh. Marijke Ooms, Louis Beyens, and Stijn Temmerman (Submitted in Palaeogeography, Palaeoclimatology, Palaeoecology)

Testate amoebae from tidal marshes have been used in palaeoecological studies for the reconstruction of Holocene sea level changes. However, little is known about the preservation or loss of fossil testate amoebae in tidal marsh sediments. This study investigated the taphonomy of testate amoebae in sediment cores from a freshwater tidal marsh along the Scheldt estuary (Belgium), and resulting implications for tidal water level reconstructions. Testate amoebae assemblages and concentrations were determined on samples from four cores. Firstly, the preservation was different for different testate amoebae groups based on test composition and morphotype. Especially idiosomic testate amoebae disappeared faster. Secondly, selective dissolution of the testate amoebae groups and its effects on the performance of a previously published transfer function for tidal water level reconstruction was simulated, showing that loss of idiosomic testate amoebae decreased the prediction capacity of the transfer function. Thirdly, the transfer function was used to reconstruct historical (20th century) changes in Mean High Water Level from the palaeo- testate amoebae data. The reconstructed water level changes were compared with observed tide gauge data, indicating a good fit between observed and reconstructed data until a depth of 50 cm.

Sea level reconstruction and selective preservation of testaceae

5.1 Introduction

Protozoa are used worldwide as bio-indicators for multiple environmental variables (Foissner, 1999). The fact that some protozoa have shells which preserve after death of the organism and get buried over time, makes it possible to extend their use as proxy to reconstruct past environmental changes. These reconstructions are based on transfer functions that utilize the modern relationship between a protozoan assemblage and an environmental variable, to infer the past environmental variable from a palaeo-protozoan assemblage. In order to obtain a reliable reconstruction, three assumptions need to be fulfilled. Firstly, it is assumed that the relationship between the protozoan assemblage and the environmental variable does not change with time; secondly, that the modern environment is analogue to the palaeo environment; and thirdly, that the protozoan assemblage is not altered during fossilization (Sachs et al., 1977).

Taphonomy concerns the transition from living to non-living organic material and the geological (both natural and cultural) formation processes during this transition (Lyman, 2010). The alteration of a modern (living) assemblage into a fossil (empty shell) assemblage (Fig. 5.1) coincides with the downward movement of sediment as a consequence of burial by sedimentation (Berkeley et al., 2007). This downward movement triggers a taphonomic loss of shells (Fig. 5.1), meaning a proportional loss over time (Berkeley et al., 2007). The taphonomic loss is due to processes such as (re)burial, (bio)turbation, decay and diagenesis (Berkeley et al., 2007, Lyman, 2010). The final fossil assemblage, or residual assemblage (Fig. 5.1), might also be influenced by initial surface and sub-surface protozoan production (Berkeley et al., 2007). Apart from above mentioned processes, one has to keep in mind that the sub-fossil assemblages might represent an environment of which no modern analogue was sampled, for example

116

Chapter 5 due to different nutrient availability, temperature regime, etc. in the past. This study focuses on the taphonomic processes, especially the decay and (selective) preservation of testate amoebae in sediment cores from a freshwater tidal marsh.

Figure 5.1 Overview of taphonomic processes that occur during fossilization of protists (scheme from Berkeley et al. 2007).

Protists, such as foraminifera, diatoms, and more recently testate amoebae, have been studied in salt marshes to reconstruct sea-level changes (Charman et al., 2010, Charman et al., 1998, 2002, Gehrels et al., 2006, Gehrels et al., 2001, Roe et al., 2002). These studies show that modern testate amoebae assemblages have a strong relationship with salt marsh elevation relative to mean high water level (MHWL), giving them a high potential for sea level reconstructions. Similarly, in a study on a freshwater tidal marsh it was found that modern testate amoebae assemblages exhibit a strong relationship with marsh elevation relative to MHWL, which is used to make a transfer function to reconstruct past estuarine water level changes (Ooms et al., 2011). Unfortunately, the application of these transfer functions may be difficult as only low concentrations of palaeo testate amoebae are found in both tidal salt marshes (Roe et al., 2002) and tidal freshwater marshes 117

Sea level reconstruction and selective preservation of testaceae

(Ooms et al., 2011). This raises the question if these poor palaeo- protozoan assemblages affect the reliability of sea/inland water level reconstructions. Until now, little is known about the taphonomic processes of testate amoebae in tidal marsh sediments. To our knowledge, only one article reports on the taphonomy of testate amoebae palaeo-assemblages from a salt marsh (Roe et al., 2002). That study concluded that the relative abundance of testate amoebae decreased significantly below 18 cm, but the palaeo-assemblage composition of these first 18 cm still resembled the contemporary surface assemblage. Though, the preservation of the testate amoebae species was variable and generally poor, possibly caused by partial dehydration. The preservation of testate amoebae might be influenced by multiple physical, biological and chemical processes and the interactions between them. One of the physical processes that possibly increases testate amoebae decay is shifting moisture content (wet – dry cycles) (Couteaux, 1992, Roe et al., 2002). Another potential mechanism is temperature increase, which positively affects the decomposition rate of testate amoebae (Meisterfeld and Heisterbaum, 1986) in an unsterile environment (Couteaux, 1992). Effects of moisture and temperature are biologically related by influencing the activity of testate amoebae decomposing bacteria (Meisterfeld and Heisterbaum, 1986). Accordingly Lousier and Parkinson (1981) have found that testate amoebae decomposition is largely linked to bacteria that eat the organic cement of tests. Other biological factors like predation might speed up decomposition, as broken shells decompose faster (Couteaux, 1992, Ogden and Couteaux, 1989). Some processes may however favor preservation, like the acidity of peaty environments helps to preserve testate amoebae (Wilmshurst et al., 2003).

118

Chapter 5

Now the question arises if taphonomic processes influence all testate amoebae species equally. This is important, as selective decay of certain species may alter the preserved testate amoebae assemblage and therefore affect the reliability of palaeo-environmental reconstructions. Apart from the article of Meisterfeld and Heisterbaum (1986), most articles agree on the fact that selective preservation of tests can occur (Booth and Jackson, 2003, Booth et al., 2004, Lousier and Parkinson, 1981, Mitchell et al., 2008b, Swindles and Roe, 2007, Wilmshurst et al., 2003). A tendency has been found for faster decomposition of transparent (Wilmshurst et al. 2003), containing biogenic silica (Mitchell et al. 2008b), or having plated tests (Lousier and Parkinson 1981) testate amoebae. In accordance with these findings, Swindles and Roe (2007) reported that Euglypha is prone to severe dissolution. Also Trinema species are regarded as faster decomposing species and have therefore been excluded in moisture reconstruction studies in peat bogs (Booth & Jackson 2003, Booth et al. 2004).

To summarize, it is important to understand (selective) preservation of testate amoebae, as changes in species composition during fossilization might influence the outcome of reconstructions. The decomposition of testate amoebae in salt marsh sediments has only been studied by Roe et al. (2002), while there is no information about the preservation in freshwater tidal marshes. Therefore, the objective of this study is to explore the preservation and taphonomy of testate amoebae in a freshwater tidal marsh in the Scheldt estuary (Belgium). In a previous study at the same study area, a transfer function for inland water level changes due to sea level change was developed based on the relationship between modern testate amoebae and MHWL (Ooms et al. 2011). In the present study, (1) it is first investigated how the numbers and species composition of preserved fossil testate amoebae are changing in four freshwater tidal marsh cores over a depth range of 100 cm. (2) Secondly, the effect of selective loss of testate amoebae

119

Sea level reconstruction and selective preservation of testaceae compositional and morphological based groups are simulated in order to estimate the effects on the performance of the transfer function. (3) Thirdly, the transfer function is applied to the fossil assemblages of the sediment cores to reconstruct past changes in MHWL, and the reconstruction is compared to an independent historical time series of MHWL measurements to assess the reliability of the reconstruction method.

Figure 5.2 Study site the Notelaar marsh. A. Indication of position of the Notelaar freshwater marsh along the Scheldt estuary (detailed map: Fig. 1.1). B. Indication of modern surface sample locations (rounds) and sediment cores (squares). C. Cros-section of the marsh surface with indication of vegetation and waterlevel.

120

Chapter 5

5.2 Material and Methods

Study area The Notelaar tidal marsh (Fig. 1.1, 5.2) is located in the freshwater part (salinity 0 - 5 Practical Salinity Units (PSU)) of the macro tidal Scheldt estuary. The local tidal regime is semi-diurnal with a mean tidal range of around 5.39 m (nearest tide gauge Temse). The high sediment load of the Scheldt water results in vertical accretion of the marsh surface by fine sediments (clay, silt, fine sand) at a rate of 1 - 2 cm year-1 (Temmerman et al., 2003b, Temmerman et al., 2003a). The marsh can be divided in an old and young part. The old marsh was already visible on the Ferraris maps (1772 - 1779), while the young marsh developed after 1944 by plant colonization on the bare mudflat (Hoffman, 1993). At present, different vegetation zones can be distinguished (Fig. 5.2B). At the border with the mudflat, a dense reed (Phragmites australis) vegetated young marsh is found. The highest, and old part of the marsh is dominated by willow vegetation (Salix sp.). Underneath and between the willows and at the edges of the the reed vegetation, multiple herbs grow (e.g. Impatiens glandulifera, Urtica dioica, Convolvulus arvensis). The eastern part of the marsh, near the dyke, is grown with Populus canadensis trees. More information on the modern ecology of testate amoebae assemblages in the study area, and testate amoebae transfer function to reconstruct water level changes, are described in Ooms et al. (2011).

Sampling locations Two coring locations, with an interdistance of less than 10 m, were selected within the young marsh. One coring location was situated at a natural levee alongside a tidal channel, the other location at a lower depression further away from the tidal channel (Fig. 5.2), close to the coring locations of Temmerman et al. (2003). Two replicate cores (with an interdistance of less than 1 m) were taken at both locations by

121

Sea level reconstruction and selective preservation of testaceae pushing a thick-walled pvc tube (with 10 cm inner diameter) into the ground. Replicate cores were sampled to check for local differences in testate amoebae assemblages. At least a depth of 100 cm was sampled. In the laboratory, the cores were cut open lengthwise. Sample slices were taken with a thickness of one centimeter of which the outer one centimeter of sediment alongside the tube walls and along the lengthwise cut was removed to avoid potential contamination. A small corer (inner diameter of 2 cm) was used to sub-sample a fixed volume from the sediment slices for sediment analyses. The sub-samples and remaining material of the sediment core were stored in refrigerated conditions (4 °C).

Chronostratigraphy of the depression coring location The chronostratigraphy of the coring location in the depression was studied in detail in Temmerman et al. (2003b, 2004b). These studies quantified the sediment accretion history at the depression location by dating of the sediment cores. Dating was based on two methods. First, macroscopic plant remains (stems, leaves) in the cores were determined and compared with the marsh vegetation succession that was reconstructed from a time series of aerial photographs ranging from 1944 up to 1998. The photographs made it possible to date the age of transition between successive vegetation types (i.e., mudflat (1944), Scirpus maritimus (1951), Phragmites australis (1965), and plant remains of these successive vegetation types could be well recognized in the cores (Fig. 5.3). Secondly, also 137Cs profiles were measured in the cores (Temmerman et al., 2004a), which allowed the dating of a horizon at around 1963 and a second at 1986. Furthermore, a sedimentation model, MARSED (Temmerman et al., 2004a) was developed to simulate the sediment accretion over time (Fig. 5.3).

122

Chapter 5

Figure 5.3 Chronostratigraphy of the coring location in the marsh depression based on the work of Temmerman et al. (2003b, 2004b) The evolution of yearly mean high water level (MHWL; bold line) at the coring location is interpolated from time series for the two nearest upstream and downstream tide gauge stations. Time-elevation points of the marsh surface are based on observations (symbols) and modeling (dashed line).

The MARSED sedimentation model is used to estimate the age of the deposited sediments at different depths throughout the cores sampled in the depression. First, the sampled core depths were recalculated to previous surface levels by accounting for the core compression (10 % and 13 % in depression core 1 and 2, respectively). Core compression was calculated as the length of the sampled core divided by the total depth of the core hole, assuming that the compaction of sediments was equally spread over the whole core. Subsequently core depths were recalculated in order to correct for the compaction of sediments, and core depths were calculated relative to MHWL (mean high water level).

123

Sea level reconstruction and selective preservation of testaceae

In this way the resulting previous surface level elevations could be linked to previous surface sediment deposition age using the MARSED model. The resulting chronostratigraphy of our sediment cores closely matches the chronostratigraphy determined by Temmerman et al. (2003b, 2004b), as in our cores the transition from Scirpus to Phragmites plant remains (at depths of 80 and 84 cm, recalculated to compensate for compression) was dated at 1955 and 1956, which falls within the time range determined from the aerial photos.

Testate amoebae preparation and counting Testate amoebae assemblages were determined for depth intervals of 5 cm for the upper 40 cm of the sediment core, and for depth intervals of 10 cm in the lower part of the sediment core (40 - 100 cm depth). Testate amoebae preparation was done on oven dried samples (30 °C for 24h). Further preparation was based on Hendon and Charman (1997). Two and a halve gram of dried sample was soaked in distilled water together with 5 tablets of Lycopodium spores (batch 177745: Results of the calibration (5 tablets): X= 92918 ± 1853, Variance = ± 2.0 %) (Stockmarr, 1971). Then, the solution was boiled for 10 minutes and sieved to retain the material between 10 µm and 300 µm. One or two drops of concentrate were put on a glass slide. A cover slip was added and sealed with nail polish to prevent desiccation. A fixed number of Lycopodium spores (85 spores), in order to count a fixed amount of sample, were counted per sample. The adding and counting of Lycopodium spores made it possible to calculate testate amoebae concentrations. The counts were made using an Olympus BX50 microscope with Nomarski optics at a magnification of 400.

Data analysis First, the difference between species composition of the four cores was tested by a PCA between-group analysis. This type of 3D analysis accounts for the species composition, core depth and location. The

124

Chapter 5 analysis was performed in R 2.10.2, using the ade4 package (Thioulouse et al., 1997). The outcome of the PCA between-group analyses was permutated in a Monte Carlo analysis (999 cycles) to test for significant differences between the cores. The outcome was interpreted from the p value and the “observation”, which is equal to the explained inertia.

Reconstruction of past water level changes and simulations of the effects of selective loss of testate amoebae were calculated with the local transfer function of Ooms et al. (2011) in C2 (Juggins, 2007). This transfer function is based on 42 modern testate amoebae assemblages for which species with < 2 % relative abundance for one sample were removed) in relation to elevation, covering an elevation range of -0.18 m to +1.0 m relative to MHWL. The transfer function is based on Partial Least Squares regression (PLS), a linear regression method that had the best outcome following the results of the jack knifing technique. This technique, also called “leave- one-out”, divided the dataset in a training and test set, leaving out one sample each run. Then, the environmental variable, here elevation ( ~ to MHWL), was reconstructed with the test set serving as modern dataset and training set being the palaeo species dataset. This procedure was run for many cycles (Birks et al., 1990, ter Braak and Juggins, 1993). The Root Mean Square Error of Prediction (RMSEP) and the r² values were used to interpret results. The RMSEP value gave the systematic errors on the predictions expressed in meter (Horton et al., 2006). The r² value (coefficient of determination) gave an indication of the fit between the observed and the prediction values (Horton et al., 2006). The closer r² reached one, the better the fit. The palaeo-testate amoebae dataset was prepared for application of the transfer function by taking the square root of the relative abundances of the palaeo- testate amoebae, which down weights dominant species (ter Braak and Šmilauer, 1998). There was no selection of species with less than 2 % relative abundance for the palaeo-dataset.

125

Sea level reconstruction and selective preservation of testaceae

The reconstructed MHWL, were calculated relative to a fixed reference level (m TAW). The resulting past MHWL values were then compared to the observed MHWL. These observed MHWL were calculated from tidal data of the two nearest tide gauge stations (Schelle and Temse) for the period between 1901 and 2007, provided by the Flanders Hydraulics research center.

Both modern and palaeo testate amoebae dataset were partitioned into groups and morphotypes in order to study respectively the effect of selective loss on transfer function performance by data simulation and the preservation of testate amoebae. There were two types of partitioning. The first was based on Mitchell et al. (2008b) in which testate amoebae were divided in groups based on the composition of the test material:

Idiosomes. All testate amoebae that make their test from self- secreted silicious plates or calcite without embedding them in a thick organic matrix. Here, following taxa were included: Trinema, Tracheleuglypha, Paraqudrula, Euglypha, Corythion, Cyphoderia.

Protein + Calcium. Testate amoebae that build their own test from material that is different from biogenic silica and calcite. Taxa from this type, found in the freshwater tidal marsh, are Arcella and Hyalosphenia.

Xenosomes. All testate amoebae taxa that recycle existing environmental particles to make their test. Both organic and mineral particles, even from prey (e.g. diatom frustules), are used and glued together with organic cement. This group consists of Centropyxis, Cyclopyxis, Difflugia, Cryptodifflugia, Difflugiella and Heleopera.

126

Chapter 5

The fourth group, Idiosomes + Organic (tests with idiosomes embedded in a thick organic matrix, described in Mitchell et al. (2008b) was not included as there were no Assulina species found.

The second partitioning was based on testate amoebae morphotypes, particularly on pseudostome and test shape, following the article of Bonnet (1975). This study divided testate amoebae into nine different morphotypes (Fig. 5.4). A list with species and corresponding test material based group and morphotype can be found in Appendix F.

To check if proportions of different testate amoebae groups indicate selective preservation (Fig. 5.7), proportions of these different testate amoebae groups were calculated for modern samples of the reed zone of the marsh combined with surface samples of the four cores (19 samples). Further, a non-parametric one-sided Wilcoxon signed rank test was performed in R to check if idiosomes or xenosomes decay faster. The test was performed on the difference between xenosome and idiosome remaining % values (Fig. 5.7). The Wilcoxon signed rank test checks if the values are bigger or smaller than a median value. This test was performed with median value of zero and with indication of the alternative hypothesis to be greater than zero.

The transfer function’s performance was tested for selective preservation of testate amoebae by simulating selective loss. This selective loss of testate amoebae groups and types was simulated by down weighting one particular group or type in steps of 10 % loss. Data down weighting was performed on the absolute counting, where after they were put into relative abundances and square rooted. These altered modern testate amoebae data sets were used to make new transfer functions using the PLS regression and jack knifing technique. The performance of the resulting transfer functions was estimated by the RMSEP and r² values.

127

Sea level reconstruction and selective preservation of testaceae

Figure 5.4 Overview of the different morphotypes as described in Bonnet (1975).

128

Chapter 5

Not all analyses were performed on the same dataset. The data of the four cores was used for studying the fossil testate amoebae assemblages and possible selective preservation. Only data of the two replicate cores of the depression location were used to reconstruct average water level changes. Simulation of selective loss of testate amoebae was performed on the modern dataset of the transfer function described in Ooms et al. (2011).

5.3 Results

Palaeo-testate amoebae assemblages In total, 41 testate amoebae species were identified. An overview of the testate amoebae species counts over the depth of each core is shown in Fig. 5.5. The between-group analysis revealed a significant difference (p-value 0.01) between the four cores (Table 5.1), more specifically between the levee and the marsh depression cores (p-value 0.002). Replicate cores were not significantly different. Therefore, data of both coring locations will be analyzed separately.

Table 5.1 Results of the between group analyses of the four cores. Between cores p-value observation all cores 1, 2, 3, 4 0.01 7.34 % levee-depression 1+2, 3+4 0.002 3.42 % 2 cores of edge 1, 2 0.209 3.97 % 2 cores of depression 3, 4 0.181 4.12 %

Testate amoebae concentrations for both locations throughout the sediment profile are similar (Fig. 5.6), showing a quick decrease within the first 40 cm of the core. Below 40 cm, testate amoebae concentrations decreased at a slower rate. The average concentration of testate amoebae at a depth of 100 cm, calculated over the four cores, was around 1486 tests g-1 ± 840 SD.

129

Sea level reconstruction and selective preservation of testaceae

Figure 5.5 Species diagram of the four cores; species concentrations (x-axis) in relation to the depth of the core (Y-axes). After the species name follows an indication of the species morphotype (see Fig. 5.4).

130

Chapter 5

Figure 5.6 Testate amoebae concentration (tests g-1) for all cores in relation to core depth. Further, sedimentation rate (based on the MARSED model Temmerman et al. (2003b, 2004b)) and vegetation is indicated for the depression location.

The rates of decrease of testate amoebae concentrations were not the same for different species groups (Fig. 5.7, 5.8). Figure 5.7 shows the testate amoebae concentrations for the three test composition groups for the levee and depression cores. Both levee and depression cores contain more idiosomic (72 % and 60 % for the levee cores and 42 % and 41 % for the depression cores) and xenosomic testate amoebae (27 % and 38 % for the levee cores and 57 % and 56 % for the depression cores) compared to protein and calcium shelled testate amoebae (0.7 % and 0.6 % for the levee cores and 0.3 % for both depression cores). Protein and Calcium composed testate amoebae seem to disappear completely around a depth of 35 cm for both coring locations, apart from one appearance at a depth of 70 cm at the levee location. The more abundant idiosome and xenosome groups decline steadily in concentrations over the depth of the core, but were still found in very low concentrations at a depth of 100 cm. 131

Sea level reconstruction and selective preservation of testaceae

Figure 5.7 A. Concentrations of Protein and Calcium testates, Idiosomes and Xenosomes in relation to core depth for each core; B Proportions of the testate amoebae groups. The vertical lines indicate the highest measured proportion of xenosomes/idiosomes at the sediment surface, as found in 19 modern testate amoebae samples. Lines with bleu circles indicate the first core of each location, lines with green squares indicate the second core of each location.

132

Chapter 5

Figure 5.8 Concentrations of testate amoebae morphotypes, xenosome morphotypes and idiosome morphotypes in relation to core depth. Lines with bleu circles indicate the first core of each location, lines with green squares indicate the second core of each location.

133

Sea level reconstruction and selective preservation of testaceae

They differed in the fact that the idiosome shells showed a steep decrease within the highest 20 cm, while the amount of Xenosome shells stayed rather constant over the first 15 cm.

The performance of a one-sided Wilcoxon signed rank test indicated that idiosomic testate amoebae decreased faster compared to xenosome testate amoebae within the cores of both locations (Wilcoxon test Levee: p = 0.015 ; Wilcoxon test Depression: p = 0.00012). Though, proportions of xenosomes versus idiosomes were not higher than was recorded in modern highest proportions (vertical line in Fig. 5.7B) for the levee location. In contrast, for the depression location was the proportion of xenosomes versus idiosomes already higher compared to modern highest proportions from a depth of 10 cm on, possibly indicating faster disappearance of idiosomes compared to xenosomes.

An alternative grouping based on the test shape and pseudostome type was analyzed (Fig. 5.8) for both coring locations. Figure 5.8 shows the division of the total species assemblages into seven morphotype groups (Fig. 5.8A), morphotypes of the idiosome testate amoebae (Fig. 5.8B), and morphotypes of the xenosome testate amoebae (Fig. 5.8C). This was not done for the protein and calcium group because this group contained very low numbers. Fig. 5.8A showed that not all test shapes were equally abundant, though the change within different test shapes over the core depth was comparable for both locations. The Arched Acrostome type (AA) was the first that completely disappeared (deepest occurrence = 30 cm). The Simple Plagiostome type (SP) vanished secondly after a depth of 70 cm. Both AA and SP types lost 50 % of their surface numbers after 5 cm depth. Contradictory, it was noticed that the Simple Acrostome (SA) type did not change within the first 15 cm. This trend was found for SA types within both the idiosome and xenosome group, and the numbers decreased at similar rates below 15 cm. The most numerous group, Plagiostome with Visor (PV) had

134

Chapter 5 members in both the idiosome and xenosome group. Idiosomic PV concentrations were very low compared to xenosomic PV concentrations after 20 cm depth for the depression location. For the levee location, xenosomic PV concentrations seem to be rather stable for the first 25 cm, though the idiosomic PV testate amoebae concentrations are decreasing already directly beneath the surface.

Figure 5.9 Reconstructed water level changes based on testate ameobae and the observed water level changes. The RMSEP values are added as error bars.

Application of the transfer function for sea level reconstruction Eleven species were found in the sediment cores without, or too rare, modern representatives and were not included in the transfer function (see Ooms et al. (2011)). A total of 30 palaeo-species were used in the transfer function to reconstruct past water level changes.

135

Sea level reconstruction and selective preservation of testaceae

The results of the water level reconstructions (Fig. 5.9) showed that the PLS regression method based reconstructed water level changes corresponded well with the observed water level changes until the year 1964. The performance of the PLS transfer function decreased when going further back in time, as the reconstructed water level changes were underestimating the observed water level changes considerably.

Simulations of selective loss of testate amoebae and effects on the performance of the transfer function

The RMSEPjack of the simulated transfer functions with down weighted testate amoebae groups and morphologic types is shown in Fig. 5.10.

The selective removal of idiosomes made the RMSEPjack value bigger (~ 0.16 - 0.21 m), while for the xenosomes the opposite trend was observed (~ 0.16 - 0.13 m) (Fig. 5.10A). The effect of selective loss of testate amoebae morphotypes was rather small. Only the selective loss of PV typed testate amoebae increased the RMSEPjack value with 2 cm (Fig. 5.10B).

The transfer function performance was investigated by scatter plots of the observed versus predicted elevations (Fig. 5.11). For the simulations of selective loss of idiosomes, the transfer function based on the complete dataset (0 % loss of idiosomes) underestimated the observed elevation (Fig. 5.11B). The complete (100 %) loss of idiosomes results in an overestimation of the MHWL, which is rather constant over the entire elevation range, independently of the relative abundance of idiosomes. The prediction capacity of the transfer function was also affected by the simulated loss of xenosomes (Fig. 5.11A). Complete (100 %) loss of xenosomes resulted in larger underestimation of the observed elevation for high elevation (> 0.5 m ~ MHWL), while for lower elevations complete loss did not affected the predicted elevations highly. The latter is related to the decrease in relative abundance of xenosomes with decreasing elevation in the observed modern dataset. The simulated

136

Chapter 5

Figure 5.10 Simulation of selective loss of testate amoebae: RMSEPjack values of test material groups (A) and morphotypes (B). loss of testate amoebae in the protein and calcium group, and other minor represented morphotype groups (CA, SA, AA, SP, A, ARC (Fig. 5.11 D, F, G, H, I, J), had a negligible effect on the performance of the transfer function. Selective loss of the PV morphotype affected the prediction capacity of the transfer function in such a way that it is better to include all or none rather than half of the initial amount (Fig. 5.11E).

137

Sea level reconstruction and selective preservation of testaceae

Figure 5.11 (continues on next page) Simulation of selective loss of testate amoebae: relative abundance graph and observed versus predicted elevation graph for test material groups and morphotypes.

138

Chapter 5

139

Sea level reconstruction and selective preservation of testaceae

5.4 Discussion

Taphonomic processes and transfer function simulations of selective loss This study investigated the decrease of testate amoebae concentrations with depth beneath the sediment surface, and the effect of selective preservation of testate amoebae shells on reconstructions of water level changes. Consistent with earlier studies (Charman et al., 2002, Ooms et al., 2011, Roe et al., 2002), testate amoebae concentrations decreased substantially in the first 50 cm in all sediment cores (Fig. 5.6). In particular, idiosome testate amoebae concentrations decreased faster with depth than other species groups (e.g. xenosome group) (Fig. 5.7). This might indicate selective preservation, which can be explained in multiple ways. Firstly, idiosome testate amoebae build their own plates out of biogenic Silica which dilutes easier compared to the lithogenic Silica particles of xenosome shells (Van Cappellen, 2003). Secondly, idiosome shells are built out of flat plates ( < 10 µm large; < 1 mm thick (Aoki et al., 2007)); while xenosome shells consist out of more thicker and circular particles. Flat idiosomic plates have a larger surface area that is in contact with the surrounding environment, which might facilitate dissolution. Thirdly, idiosome shells might be more prone to shell breakage, as broken tests disappear faster (Ogden and Couteaux, 1987). The idiosome test plates are embedded in a very thin organic matrix, which might be weaker compared to other testate amoebae organic matrices (e.g. Assulina sp).

The proportion of xenosomes versus idiosomes at the levee location stays within the boundaries of the modern maximal proportions, which indicates that there is probably no selective preservation at this location. Though, it is possible that the proportion of xenosomes versus

140

Chapter 5 idiosomes is naturally different from that of depression locations and that the maximal modern proportions, which were mostly based on depression testate amoebae data, are not valid for the levee location. Apart from the test material, test morphology (after Bonnet, 1975) can play an important role in the rigidness of the test to withstand mechanical pressure and overcome breakage. There might be selective preservation of xenosomes compared to idiosomes at both locations (Fig. 5.8B, C) for the PV type. Idiosomic PV testate amoebae concentrations declined from the surface on compared to the xenosomic PV fraction.

One has to keep in mind that it might be that the surface production of sub-fossil testate amoebae assemblages was lower compared to nowadays surface testate amoebae concentrations, as testate amoebae concentrations decrease at lower marsh levels (Ooms et al., 2011). Apart from (selective) dissolution, the observed decreases in testate amoebae concentrations with depth may be facilitated by, a combination of, other processes like sedimentation, bioturbation and different environmental conditions during initial surface production (changing salinity, nutrients, temperature regime,…).

The cores of the depression location contained remains of Scirpus maritimus vegetation at depths deeper than 74 cm (Fig. 5.6). Surface sediments within this pioneer vegetation typically contain low modern testate amoebae concentrations ( < 6000 tests g-1). The reason for low testate amoebae concentrations in the pioneer zone might be related to high sedimentation. This high sedimentation rate can have two effects. Firstly, fast sedimentation makes it more difficult to establish modern assemblages with high species concentrations, as there is limited time before burial. Secondly, this fast accretion of sediments makes that testate amoebae assemblages pass quickly through the aerobic layer with high decomposing bacteria activity, which leads to better

141

Sea level reconstruction and selective preservation of testaceae preservation of the initial assemblages. The sedimentation rate slowed down over the years as the marsh surface reached MHWL elevation (Fig. 5.6). This resulted in a longer time spent within the aerobic bacterial decomposing layer, thus facilitating testate amoebae disappearance.

Reworking of the sediment layers by bioturbation is performed mainly by invertebrates. Low bioturbation facilitates preservation as time for movement trough the oxygen rich layer, with biodegradation by oxygen dependent bacteria, is short (Berkeley et al., 2007). Bioturbation within freshwater tidal marsh sediments is fairly unknown, only the article of Beauchard et al. (2012) states that there is few bioturbation at high marsh levels. Since the transfer function for water level changes worked rather well, it can be assumed that environmental conditions did not change that much over the period until 1965, thus that the sub-fossil testate amoebae assemblages were formed under conditions that were similar to present- day surface conditions.

Based on the discussion above, a ranking of species or taxa from worst to better preservation is made (Fig. 5.12). Caution is needed when using this figure, as it might be possible that changes in species assemblages are reflected rather than dissolution or preservation. Especially species morphotypes with only one representative have to be interpreted carefully (e.g. Arched Acrostome). It seems that test material and test morphology are two important factors regarding the selective preservation of testate amoebae of a freshwater tidal marsh.

142

Chapter 5

Figure 5.12 Range of testate amoebae groups/taxa from bad to relatively better preservation capacity.

Transfer function simulations with selective loss of testate amoebae Fig. 5.12 indicates the rate of testate amoebae disappearance from the palaeo testate amoebae record. This knowledge can be used to interpret the simulation of the selective loss of testate amoebae and the effect on the transfer function performance (Fig.5.10, 5.11). The idiosome group, which seems to preserve badly , is a big concern for the use of transfer function as they make up a large part of the (modern) testate amoebae assemblages. This is confirmed by the fact that the prediction error

(RMSEPjack = 0.21) increases with 5 cm with a total loss of idiosomes (Fig. 5.10A), especially loss of the idiosome PV group (Fig. 5.10B). This is in contrast with other environments such as peaty environments where the idiosome fraction has been omitted from the transfer function

143

Sea level reconstruction and selective preservation of testaceae to gain a better model for reconstruction (Booth and Jackson 2003, Booth 2004).

Oppositely, the simulation of selective loss of xenosomes would increase the prediction capacity of the transfer function with 2 cm, but is undone by the overall underestimation of elevation (Fig. 5.11), especially when xenosomes where highly abundant.

Reconstruction of former water levels and comparison with measured tidal data The PLS reconstruction of past water levels agrees fairly well with the measured tide data for the period from about 1965 (depth of 50 cm) to present (Fig. 5.9). This time step around 1965 correspond to a testate amoebae concentration of ± 7500 test g-1 or an average of 17 counted testate amoebae (Fig. 5.6).

Although testate amoebae concentration does not change much below 50 cm (Fig. 5.6), the transfer function does not perform well any more (Fig. 5.9). This might be related to the low number of species, below 40 cm depth less than 10 different taxa (except 4070; Appendix G) per sample took part in the transfer function. The effect of a small number of taxa and low testate amoebae counts is that one individual test can reach high relative abundance. As a result, rare species that have better preservation capacity can make out a high relative abundance, while abundant species with bad preservation can get an underestimated relative abundance.

Another reason for bad transfer function performance before 1965 is that all our modern testate amoebae samples are taken within the Phragmites australis and willow vegetation, since Scirpus maritimus only occurred in 1951 (aerial photographs). Phragmites first appeared on the picture of 1965, maybe explaining the reconstruction boundary of 1965.

144

Chapter 5

Samples of Scirpus vegetation of a brackish tidal marsh (Ooms et al., 2012) did not bare high testate amoebae concentrations ( < 450 tests g- 1). Further, some samples of Scirpus vegetation at the small freshwater tidal marsh of Appels (further upstream) had also rather poor testate amoebae concentrations ( < 6000 tests g-1). Possible explanation for low testate amoebae concentrations within Scirpus vegetation is the tidal inundation stress.

It can be useful to keep in mind that palaeo testate amoebae concentrations become very rare in Scirpus maritimus layers. For reconstruction studies, it might be best to avoid the presence of Scirpus maritimus in cores. Therefore, we suggest to sample locations without Scirpus maritimus vegetation.

5.5 Conclusion

This study investigated the taphonomy and selective preservation of testate amoebae of a freshwater tidal marsh and the implications for water level reconstructions.

1/ Our results clearly demonstrate that testate amoebae concentrations rapidly decrease within the upper 40 cm of the sediment profile, indicating bad preservation. Furthermore, the results suggest that there is selective preservation of testate amoebae, at least for the depression location. In accordance with previous studies, idiosome testate amoebae are found to be more vulnerable to dissolution compared to other testate amoebae groups. Based on the alternative division of testate amoebae into the morphotype groups of Bonnet (1975), we were able to refine the scale of testate amoebae vulnerability to dissolution. This resulted in a ranking from bad to relatively better preservation capacity of testate amoebae.

145

Sea level reconstruction and selective preservation of testaceae

2/ Simulations with selective loss of testate amoebae groups pointed out that loss of idiosome testate amoebae enlarges the prediction error of the transfer function previously published in Ooms et al. (2011). Therefore, it is not recommendable for water level reconstructions to remove the idiosome species out of the transfer function.

3/ The comparison between predicted and observed water level changes revealed that the transfer function of Ooms et al. (2011) was robust enough to reconstruct water level changes until around 1965, which corresponded to a depth in the sediment cores of ca. 50 cm beneath the surface or testate amoebae concentrations of ± 7500 tests g-1.

4/ This study demonstrated that it is possible to use palaeo assemblages of testate amoebae for reliable reconstructions of water level changes in a freshwater tidal marsh, if adapted number of conditions are fulfilled. First, the cores should only contain sediments that deposited in mature marsh vegetation, and preferably no pioneer vegetation (as testate amoebae concentrations are naturally low in the pioneer zone). Secondly, more than 10 testate amoebae taxa should be present in the palaeo assemblage to obtain reliable results with the transfer function of Ooms et al. (2011). In the present study, these conditions were only fulfilled in the upper 50 cm of the sediment profile.

146

Chapter 5

Acknowledgements

We would like to thank Marijn van den broeck and Wim Clymans for their help with coring. We would also like to acknowledge Flanders Hydraulics center for providing tide gauge data of Schelle and Temse. This research was funded by a PhD grant of the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT-Vlaanderen).

147

Chapter 6

Testate amoebae: high-turnover processors of Si in freshwater tidal marshes Ooms M.1, Struyf E.1, Smis A.1, Van Braeckel A.2, Beyens L.1, Meire P. and Temmerman S.1 (Submitted in Aquatic Biology)

Tidal marshes play an important role in the silicon cycle by providing dissolved Si (DSi) to the estuary in times of DSi limitation, in order to prevent toxic algal blooms. The source of tidal marsh DSi is thought to be mainly originating from diatoms and phytoliths. Until now, the contribution of testate amoebae to the Si cycle was not explored. This study has investigated the dissolution of (idiosomic) testate amoebae and diatoms in relation to biogenic Si (BSi) within sediment profiles of a depression and natural levee at the freshwater tidal marsh (Notelaar). Results show that, especially idiosomic, testate amoebae concentrations are highly correlated with BSi throughout the sediment profiles. Therefore, we conclude that (idiosomic) testate amoebae are important sinks and processors of Si in freshwater tidal marshes for both depressions and natural levees.

Role of testate amoebae in Si cycle

6.1 Introduction

Tidal marshes play an important role in the silicon (Si) cycle in the estuarine and coastal zone. Si is an essential nutrient to sustain phytoplanktonic diatom communities, which are a basic and vital part at the base of the food web in estuarine and coastal waters (Sullivan and Moncreiff, 1990). Tidal marshes form large biogenic Si (BSi; amorphous hydrated form of Si or SiO2nH2O) reservoirs, and tidal flux measurements between marshes and the adjacent estuary mostly point out that there is a net import and deposition of BSi in tidal marshes (Norris and Hackney, 1999, Struyf et al., 2005a, Vieillard et al., 2011). The large availability of BSi in tidal marsh sediments induces an enrichment of dissolved Si (DSi, silicic acid or H4SiO4) in marsh sediment pore water (Hackney et al., 2000, Struyf et al., 2005b), compared to adjacent estuarine waters. At spring tides, this can cause a DSi enrichment of exported ebb water. Especially in periods of DSi- limited primary production, this DSi export is potentially important to sustain estuarine diatom communities (Struyf et al., 2006, Vieillard et al., 2011) and to prevent toxic algal blooms. Toxic algal blooms take place when DSi depletion occurs in combination with high availability of phosphorus (P) and nitrogen (N) (Conley et al., 1993, Lancelot, 1995). DSi depletion may be caused by DSi uptake by several protists (i.e. diatoms, radiolarians, testate amoebae) and plants, which form their skeletons or phytoliths of BSi. Until now, the BSi stock of tidal marshes was thought mainly to originate from allochtonous diatom BSi deposition during high-tides (Struyf et al., 2006), from autochthonous diatom reproduction and from litterfall of BSi rich macrophytes such as reed (Struyf et al., 2007, Struyf et al., 2005a, Struyf et al., 2005b) (Fig. 6.1). Up to now testate amoebae have not been considered yet to play a role in this biogeochemical Si cycle in tidal marshes. Research on the relative contribution of different BSi reservoirs to the total marsh BSi content is virtually non-existent.

150

Chapter 6

Figure 6.1 Schematic overview of the Si cycle involving diatoms and phytoliths of a freshwater tidal marsh.

Yet, testate amoebae are highly abundant on tidal marshes (Charman et al., 1998, 2002, Gehrels et al., 2006, Gehrels et al., 2001, Ooms et al., 2011, 2012, Riveiros et al., 2007, Roe et al., 2002), especially in freshwater marshes where they can reach concentrations up to 300 000 testate amoebae/cm³ (Ooms et al., 2011). Further, Aoki et al. (2007) showed that testate amoebae can play an important role in the Si cycle in forests, as they act similarly in the terrestrial Si cycle as higher plants, playing a role as both Si consumers and suppliers.

Testate amoebae shells consist of different kinds of materials depending on the species. In general, two main groups can be distinguished. One is the idiosomic testate amoebae group, containing testate amoebae species that form their shells by making biogenic Si plates pasted together with organic cement. The second group is the xenosomic testate amoebae group, having shells made out of various allochthonous materials pasted into an organic matrix. This allochthonous material can be clay or silt particles (lithogenic Si), diatom shells (biogenic Si) or organic detritus.

In multiple sedimentary environments, testate amoebae shells are well- preserved within the sediment profile and are therefore often used to reconstruct paleo-environmental conditions (Charman et al., 2012,

151

Role of testate amoebae in Si cycle

Langdon et al., 2012, Mitchell et al., 2008a). However, in tidal marsh sediments, where testate amoebae have been used to reconstruct sea level changes, it has been found that concentrations of testate amoebae strongly decrease with depth within the sediment profile (Charman et al., 2010, Ooms et al., 2012, Roe et al., 2002). This poor preservation of testate amoebae demonstrates a fast dissolution in marsh sediments, which might give a first indication of the role of testate amoebae in the DSi delivery from tidal marshes to the pelagic waters. However, although marshes are known as important Si processors, internal processing by different potential BSi accumulators, and associated recycling rates, are yet to be quantified.

Here, we have studied, for the first time, the occurrence of testate amoebae and their contribution to BSi in freshwater tidal marsh sediments. We collected four deep sediment cores in the Scheldt estuary (Belgium), and analyzed the relative proportions of testate amoebae, diatoms, and plant organic matter concentrations explaining the variations in BSi concentrations with depth throughout the sediment cores. Two replicate cores were collected on a highly elevated natural levee, and two in a lower marsh depression, in order to test for the potential effects of differences in groundwater level fluctuations on BSi dissolution. Our results are the first to indicate that testate amoebae are key organisms in tidal marsh Si cycling.

152

Chapter 6

6.2 Material and Methods

Study area

Figure 6.2 A. Overview of the Scheldt estuary with indication of the Notelaar freshwater tidal marsh. B. Detailed view of the Notelaar freshwater tidal marsh with indication of vegetation zones and coring locations. The groundwater level measurements took place at the same locations as the coring locations.

The Scheldt river stretches from its source in St. Quentin (France), passing trough Belgium, to its mouth in the North Sea near Vlissingen (The Netherlands). The estuarine part of the Scheldt is 160 km long from Vlissingen (NL) up to Ghent (BE) (Fig. 6.2A) and has a semidiurnal meso-to macrotidal regime. The estuary has a large salinity gradient, ranging from salt water (at Vlissingen (NL) to brackish (at Hansweert (NL) and fresh water (at the tributary river Rupel (BE) (Fig. 6.2A). The study was conducted on a freshwater tidal marsh, the Notelaar (Fig. 6.2B) (Meire et al., 2005). The tidal range of the Scheldt reaches its maximum near this marsh (5.39 m). The marsh exists of an old (high- elevated) and a young (lower elevated) part. The old part is vegetated with willow vegetation. The young part has developed from around 1950

153

Role of testate amoebae in Si cycle

onwards, by succession from a mudflat to Scirpus maritimus (Bolbosschoenus maritimus) vegetated marsh towards a Phragmites australis vegetated marsh (Temmerman et al., 2004a).

The chronostratigraphical history of the coring location in the marsh depression was already extensively studied in Temmerman et al. (2003b, 2004b). These studies reconstructed the sediment accretion rates of the marsh surface based on dating of several sediment cores in combination with modeling, and found that the upper 1.4 meters of the sediment profile was deposited since ca. 1945 up to present (Fig. 6.3). The cores were dated by linking dated aerial photographs of the marsh vegetation succession (from bare mudflat to Scirpus maritimus and Scirpus maritimus to Phragmites australis), to macroscopic plant remains found within the cores. Additionally, two levels in the sediment cores were dated using the 137Cs method (Temmerman et al., 2004b). The obtained time series of marsh surface levels was compared to the MARSED model simulating the increase in marsh surface elevation by sedimentation (Fig. 6.2). The outcome shows that the marsh surface sedimentation is not steady over the years. There was a higher sedimentation rate in the period 1945 - 1970, because marsh surface level was still lower than MHWL (Mean High Water Level), resulting in more frequent, deeper and longer tidal inundations and therefore faster sedimentation rate. The rate of sedimentation lowered when surface sediment levels reached MHWL levels (Fig. 6.3).

154

Chapter 6

Figure 6.3 Observed and modeled marsh surface sediment accretion over time (after Temmerman et al., 2003b, 2004b), with indication of the observed macro-remain transitions, 137Cs peaks and the topographic surveying of the marsh surface. The triangle indicates the transition from Phragmites australis to Scirpus maritimus of the depression cores.

Coring location The sampling took place in the young marsh within the present-day P. australis zone. The elevation of the sampling locations corresponds with an average flooding frequency of 30 % (calculation flooding frequency: Ooms et al., (2011) of the high tides, which means that the sampling locations are flooded on average 8 - 9 times within a 14-days spring- neap tidal cycle with an average duration of each flood event between 1 and 2 hours (Temmerman et al., 2003a). Two different sample sites were selected (Fig. 6.2B); the first is situated at a natural levee alongside a tidal channel, the other is half a meter lower and situated in the depression further away from the channel.

155

Role of testate amoebae in Si cycle

These two sites were selected, in order to test for the effects of groundwater level fluctuations on BSi dissolution: we hypothesized that groundwater level fluctuations would be larger at the levee than in the depression, due to deeper groundwater drainage at low tide at the levee (closer to the channel), potentially resulting in alterations in BSi turnover at the levee site. Two replicate cores were collected at both the levee and depression locations. Cores were drilled by pushing high pressure water tubes into the ground until at least one meter of marsh sediments was sampled. The tubes were transported as a whole to the lab, where they were cut lengthwise. The disturbed sediment along the longitudinal cut and along the border of the tube was removed before sub-sampling. Sub-sample slices with a thickness of 1 cm were taken to analyze for testate amoebae and diatom concentrations. BSi and particle size analysis were performed on a small sub-sample with fixed volume that was taken from the core slices. The first 40 cm was sampled with intervals of 5 cm, as largest changes in BSi content were expected within this upper zone based on previous analyses by Struyf et al. 2007. Below 40 cm depth, one sample was taken every 10 cm. In total 15 samples were collected for each core (at depths of 0, 5, 10, 15, 20, 25, 30, 35, 40, 50, 60, 70, 80, 90, 100 cm).

Groundwater data and soil particle size distribution In order to test for the potential influence of differences in groundwater level fluctuations on BSi dissolution between the two sampling locations (high natural levee versus lower depression), the Flemish institute INBO (Research Institute for Nature and Forest) provided groundwater level data from the two sampling locations. The measurements have been conducted in a piezometer tube with a pressure sensor logger (Eijkelkamp CERADiver) measuring groundwater levels every 5 minutes. The measurements took place from June to August of 2009.

156

Chapter 6

Particle size was measured on H2O2 digested samples through laser diffraction using a Mastersize 2000 instrument. Particle size classes are clay (0 - 2 µm), silt (2 - 63 µm) and sand (63 - 1000 µm).

Figure 6.4 Groundwater level fluctuations at the two coring locations. relative to the sediment surface level.

BSi content BSi content of the sediment samples was measured using the alkaline

(Na2CO3) extraction method which was developed for marine sediments by DeMaster (1981) and applied to tidal marsh sediments by Struyf et al. (2005b). First, sediment was dried for 24 h at 105 °C. Then, the sediment samples were ground with a mortar and pestle. A total of 30 mg of ground sediment was weighted for BSi extraction. The sediment was digested in 25 ml of 0.1 M Na2CO3 solution in a shaking water bath of 80 °C. Subsamples of 1 ml were taken at four, five and six hours and were diluted in 5 ml distilled water. DSi concentration was measured colorimetrically using a SKALAR SA 1500 colorimeter. The BSi concentration of the sediment was calculated by linear extrapolation of the three subsampled DSi concentrations to the start of the extraction, adjusting for the release of DSi by mineral dissolution (DeMaster, 1981).

157

Role of testate amoebae in Si cycle

Preparation and analysis of testate amoebae samples Testate amoebae samples were prepared following the method of Hendon and Charman (1997): 2.5 g of dried sediment was diluted with distilled water and 5 Lycopodium spore tablets (batch 177745: Results of the calibration (5 tablets): X= 92918 ± 1853, Variance = ± 2.0 %) (Stockmarr, 1971) were added for calculation of testate amoebae concentrations. Then, the sample was boiled for 10 minutes to loosen aggregates of clay and testate amoebae. The boiled solution was sieved twice to retain the part between 10 µm and 300 µm. Rose Bengal solution was added to color the living testate amoebae present in the superficial layers of the cores. Microscopic slides were made by putting a drop of glycerol and a drop of testate amoebae solution on a glass slide, covering it with a cover slide and fixing it with nail polish. Testate amoebae analysis was done using an Olympus BX50 microscope with Nomarski optics at magnification of 400. In total, 85 lycopodium spores were counted per sample, which corresponded to at least 150 testate amoebae shells at the surface level. Total testate amoebae numbers, combining dead and living testate amoebae, as well as numbers of testate amoebae species belonging to the idiosome group (i.e. with tests made up of BSi) were counted.

The modern testate amoebae dataset of the Notelaar marsh, published in Ooms et al. (2011), has been used in this study to calculate the average and variation of testate amoebae concentrations observed at the present-day marsh surface, both for total testate amoebae and the idiosomic testate amoebae fraction. The average testate amoebae concentrations at the present surface were used as a reference for comparison with the concentrations found with increasing depth in the cores. This is done to make a sound judgment of testate amoebae dissolution when concentrations in the cores were lower than the

158

Chapter 6

observed variation in average concentrations at the surface level. Average (idiosomic ) testate amoebae concentrations and variation (95 % confidence interval) were calculated on 23 samples; the present-day surface samples belonging to the Phragmites australis zone of the marsh and lying above MHWL (19 samples) and the (idiosome) testate amoebae surface concentrations of the four cores. Further, the lowest present-day testate amoebae and idiosome concentration within the 95 % confidence interval was obtained by subtracting the variation from the average (idiosomic) testate amoebae concentration. The obtained average and lowest surface concentrations of testate amoebae and idiosomes is used to compare with the concentrations within the first 50 cm depth of the cores as this part contains sediments from marsh surface levels above MHWL (Fig. 6.3). The deeper part of the cores was deposited when the marsh surface was lower than MHWL, and therefore sediments deeper than 50 cm are not directly comparable to the present-day surface samples.

Preparation of diatom samples The preparation of diatom samples was done using the extraction method of Van der Werf (1955). For diatom preparation, half a gram of dried sediments was saturated with H2O2 and heated until most of the organic matter was oxidized. For further oxidation, MnO4 was added until a black precipitate was formed. The precipitate was dissolved with

HCl and H2O2. After that, 20 lycopodium spore tablets (Stockmarr, 1971) were added. Samples were centrifuged 3 times for 10 minutes at 3500 rpm and rinsed in between each centrifugation cycle. The resulting diatom solution was used to make microscopic slides. The cover slide was put on a heater and was covered with a layer of distilled water in which a drop of diatom solution was added to equally spread over the cover slip. The cover slip was heated until all water was evaporated. The cover slip was put onto a glass slide with a layer of

159

Role of testate amoebae in Si cycle

Naphrax and heated until the resin was fluently enough to spread evenly between glass slide and cover slip. Slides were left then to harden. Diatom concentrations were counted at a magnification of 1000 on an Olympus BX50 microscope with Nomarski optics, using emersion oil. Only the number of frustules and Lycopodium spores were counted. In total, 500 frustules were counted for each sample.

Organic matter content The organic matter content of the sediment was determined on the subsamples with fixed volume, as proxy for phytoliths. These samples were first dried at 105 °C and then combusted at 550 °C for four hours to calculate total organic matter content (Loss on Ignition (LOI)). Total organic matter content is interpreted here as proportional to plant organic matter preserved in the marsh sediments, and hence as proportional to BSi in the form of phytoliths (e.g. present in Phragmites australis litter; Struyf et al., 2005b). Further, the cores were investigated on macroscopic plant remains, because based on the chronostratigraphy of the depression coring location (Temmerman et al., 2003b), a transition in plant material is expected from Phragmites australis dominated plant remains in the upper part of the cores to Scirpus maritimus dominated plant remains in the lower part of the cores (Fig. 6.3).

Data analysis The data analysis was performed in R 2.10.1 using the MASS library (R Development Core Team, 2009). A paired t-test was performed to check if the difference between the groundwater levels at levee and depression were significant. Further, correlations between concentrations of total testate amoebae, idiosome testate amoebae, diatoms, LOI and BSi were tested. The Shapiro-Wilk normality test, to check normality of the data, showed that

160

Chapter 6

the datasets were not normal (W < 0.95). Therefore a non-parametric spearman rank correlation test was performed.

6.3 Results

Groundwater level and soil particle size distribution Groundwater levels are significantly higher (paired t test, p value < 2.2e-16) at the lower depression compared to the natural levee (Fig. 6.4). The indication of the lowest groundwater level relative to surface level for both locations shows that groundwater at the natural levee is on average 0.40 m lower than at the depression location (Fig. 6.4, 6.5). In general, particle size distributions are similar at the two coring locations, demonstrating that sediment deposition is similar at both locations (Fig. 6.5). Only the lowest 20 cm of the natural levee cores contains more sand relative to the depression, which is related to the fact that this location is situated at the edge of a channel.

BSi content Both at the levee and in the depression, a clear downward decrease in BSi concentration is observed (Fig. 6.5). Minimum BSi at the depression sites is found at shallower depths (of 35 - 40 cm) compared to the natural levee (minimum values at 100 cm and 60 cm). At both locations, BSi level rises slightly again in the deepest part of the sediment profile.

Testate amoebae and diatom concentrations Concentrations of (idiosomic) testate amoebae and diatoms are declining with depth of the sediment profile (Fig. 6.5). Table 6.1 shows an overview of testate amoebae, idiosome and diatom concentrations found at the present-day sediment surface as well as a number of critical depths in the cores (see explanations in Table 6.1). The table also contains an indication of the percentage of testate amoebae and idiosomes that have remained in the core at a depth of 50 cm. The

161

Role of testate amoebae in Si cycle

Figure 6.5 Overview of the changes in average concentrations of testate amoebae, diatoms and idiosome testate amoebae, BSi, sediment grain size fractions and vegetation type over the depth of the natural levee and depression cores. An extra graph is added for the depression cores indicating the sedimentation rate (cm year-1) over the depth of the core (after Temmerman et al., 2003b).

162

Chapter 6

remaining concentrations were calculated relative to the average surface concentrations (see Materials and Methods). The last column indicates the highest core depth that reached concentrations below the lowest surface (idiosomic) testate amoebae concentrations.

A high percentage of testate amoebae has clearly dissolved within the sediment profile. Only 13.8 % or less of the average surface testate amoebae concentration remains at a depth of 50 cm. Further, from a depth of 20 cm on testate amoebae concentrations are lower than the lowest surface testate amoebae concentrations indicating dissolution. Testate amoebae concentrations further decrease towards a depth of 100 cm. Comparing concentrations of total testate amoebae and idiosomes, shows that idiosomes, i.e. with tests made up of BSi, make up the largest part of the total testate amoebae concentrations. Idiosomes dissolve easily, as it is clearly shown for both locations that idiosomes are completely lacking at some core depths (core 2 at 100 cm; core 3 at 60, 90 cm) (Fig. 6.4).

Surface concentrations of diatoms are comparable at the levee and in the depression. Still, there are some differences between both locations. Highest diatom concentrations do not occur at the surface level for the depression location, but at a depth of 70 to 100 cm. At the levee location, highest diatom concentrations are situated close to or at the surface level. Further, diatom concentrations at a depth of 50 cm are relatively high (around 50 % of surface concentration) compared to remaining testate amoebae concentrations (max. 13.8 % of average present-day testate amoebae concentration).

163

Role of testate amoebae in Si cycle

Table 6.1 Overview of important concentrations and core depths of testate amoebae, idiosomes and diatoms. Indication of remaining percentage, equal to the proportion of concentrations at 50 cm depth and average surface concentrations for (idiosomic) testate amoebae multiplied by 100, and depth where concentrations are lower than lowest present-day concentration of (idiosomic) testate amoebae, based

on the calculation of average present-day surface concentrations.

tration

Concentration 50 cmat depth

Location Numbercore Surfaceconcentration Lowest concentration depth Lowest concentration Highest concentration cm 50 depth Highest concen Remainingpercentage (%) Concentrationthan lower lowest modern concentration Testate amoebae

levee 1 49959 80 1706 0 49959 4114 11,5 15

levee 2 29415 100 419 0 29415 4945 13,8 15

depression 3 45841 90 419 5 51749 437 1,2 20

depression 4 72399 90 1410 0 72399 4493 12,5 20

Idiosomes

levee 1 41632 100 1200 0 41632 2286 8,8 15

levee 2 19756 100 0 5 33059 2060 7,9 10

depression 3 24557 60,90 0 5 31134 437* 1,7 10

depression 4 48266 90 470 0 48266 3145 12,1 5

Diatoms

levee 1 20862796 60 6414803 15 24794889 10829760

levee 2 19181383 100 5707141 0 19181383 10498607

depression 3 23958157 60 4304633 100 24495130 18311221

depression 4 21721953 60 3781122 70 27317868 7188972

* concentration calculated on 1 idiosome testate amoebae counted

164

Chapter 6

Organic matter content The organic matter content of the lower depression and the natural levee does not change much over the depth of the core (Fig. 6.5). The organic matter content at the natural levee is rather stable around 13 %, while the organic matter content of the depression site lies around 18% until a depth of 70 cm and shifts to 11 % at 100 cm depth (Fig. 6.4). The shift in organic matter content corresponds to a change in macroscopic plant remains found within the depression cores from Phragmites australis for the upper 80 cm to Scirpus maritimus deeper than 80 cm (Fig. 6.5). The shift in vegetation cover corresponds with the chronostratigraphy of the coring location previously reported in Temmerman et al. (2003b, 2004b).

Correlations The variations in BSi content are best correlated with the concentrations of idiosome testate amoebae (highest correlation coefficient of 0.89 at the levee; Table 6.2). Diatom concentrations have a lower correlation (0.40 - 0.71) with BSi content than idiosome and total testate amoebae concentrations (0.60 - 0.82). The correlation between LOI (as a proxy for phytoliths content) and BSi is low, ranging from 0.5 - 0.6 for the levee location and even lacking significance for the depression location.

Table 6.2 Non-parametric Spearman Rank correlation results of testate amoebae, idiosomic testate amoebae and diatoms correlated with BSi.

BSi- BSi- BSi- test p- BSi- BSi-idio BSi- diatom BSi- LOI p BSi- correlation value test p-value idiosomes pvalue diatom value LOI levee 0.0018 0.7507 0.0000 0.7731 0.0011 0.5674 0.0040 0.5104 levee average 0.0003 0.8179 0.0000 0.8912 0.0036 0.7179 0.0204 0.6000 depression 0.0005 0.5961 0.0001 0.6644 0.0290 0.3990 0.0590 0.3487 depression average 0.0111 0.6464 0.0005 0.7824 0.0506 0.5179 0.2949 0.2893

165

Role of testate amoebae in Si cycle

6.4 Discussion

Correlation of testate amoebae, idiosomes, diatoms and phytoliths with BSi The investigated sediment cores from a freshwater tidal marsh show that testate amoebae concentrations, and especially idiosomic testate amoebae concentrations, are better correlated with variations in BSi content in the marsh sediments than diatom concentrations. There was no good correlation between LOI, as indicator for phytoliths, and the amount of BSi, indicating that the sediment BSi-stock is dominated by diatoms and testate amoebae, rather than phytoliths. Plant litter BSi dissolves quickly after litter deposition on the marsh surface, and almost 100 % of all reed BSi dissolved within one year in a litter bag experiment (Struyf et al., 2007b). Overall, testate amoebae concentrations were the best indicators for BSi variations observed along our marsh sediment profiles.

BSi fluctuations and dissolution of diatoms and testate amoebae The slight rise in BSi values at deeper levels of the cores ( > 0.6 m depth; especially in the depression; Fig. 6.5) is explained in Struyf et al. (2007) as the result of higher sedimentation rates at the time of deposition of these deeper sediment layers (Temmerman et al., 2003b) (Fig. 6.5). Fast sedimentation causes testate amoebae and diatom fossils to reach depth levels faster at which the dissolution rate is lowered. Likely this is due to reduced contact with infiltrating flood water and the formation of anaerobic conditions with reduced bacterial activity at larger depth. Berkeley et al. (2007) also found that foraminifera preserve better if they transfer quicker through the oxygen rich upper sediment layer of a marsh. Our study shows that mainly diatom concentrations correspond to this pattern and remain high throughout the sediment profile (especially in

166

Chapter 6

the depression; Fig. 6.5), while the testate amoebae concentrations are minimal in the deeper part of the sediment profile.

It is clear that dissolution of testate amoebae is quicker and more complete compared to the dissolution of diatoms (Table 6.1). One potential explanation for the robustness of diatom frustules to dissolution is described in Martin-Jézéquel et al. (2000): after the quick dissolution of the finest diatom frustule structures, a more robust diatom frustule with a lower surface area remains, slowing down further dissolution. A second explanation is that diatom shells are known to form aggregates. These aggregates have, again, a smaller surface that is prone to dissolution (Moriceau et al., 2007). The fast dissolution of testate amoebae shells might be explained by the fact that their shells have on organic matrix keeping all particles (idiosomes and xenosomes) together. Shell breakage fastens dissolution of testate amoebae (Ogden and Couteaux, 1987). Therefore, the preservation time of testate amoebae is linked to the rigidness of the organic matrix: decomposition rates are usually high in tidal marshes due to the strong dynamics and frequent pore water refreshing in the upper sediment layers. Frequent wetting and drying of the upper sediment profile likely stimulates shell breakage, and dissolution of BSi in broken testate amoebae shells subsequently fastens as the exchange surface with the surrounding environment increases (Couteaux, 1992, Roe et al., 2002).

It is shown in many studies that idiosomes dissolve more easily compared to other testate amoebae (Lousier and Parkinson, 1981, Mitchell et al., 2008b, Wilmshurst et al., 2003). The results of Table 6.1 also indicate faster disappearance of idiosome testate amoebae compared to other testate amoebae types. The slower dissolving xenosome testate amoebae likely contribute relatively little to Si

167

Role of testate amoebae in Si cycle

(re)cycling compared to idiosomes , as they mostly consist of lithogenic Si particles which have a much lower dissolution rate compared to amorphous BSi particles.

Both cores at the depression site contained remains of Scirpus maritimus vegetation in the deeper layers (Fig. 6.5). This agrees with an earlier study showing that this site developed from a mudflat to tidal marsh, with Scirpus maritimus as the pioneer vegetation (Temmerman et al., 2003b). During this stage with pioneer vegetation and the early Phragmites australis stage, sedimentation rates were substantially higher compared to now (Temmerman et al., 2003b) (Fig. 6.5). This is reflected in a strong peak in BSi and diatom concentrations at 70 and 80 cm depths, and to a much lesser degree in the testate amoebae concentrations at a depth of 70 cm.

The cores at the depression site contained the lowest testate amoeae concentrations at a depth deeper than 80 cm, which is probably related to the fact that these sediments were deposited at the time that Scirpus maritimus vegetation was present (see Fig. 6.5). Present-day testate amoebae concentrations were found to be very low within Scirpus maritimus vegetation ( ± 6000 tests g-1) (Ooms et al., 2012). In contrast, the highest diatom concentrations in the depression core are found within this part of the sediment profile that contains Scirpus maritimus remains (70 - 100 cm depth) (Fig. 6.4). Nowadays, microbial diatom mats can be observed between Scirpus maritimus pioneer plants. Therefore, the sampling of a (part) of a microbial diatom mat might explain the high diatom concentrations at the deep levels of the depression sediment profiles. This finding makes it more difficult to discuss dissolution of diatoms in the marsh sediments Therefore, only diatom concentrations of the upper 60 cm of the sediment profile, which were deposited in Phragmites australis vegetation, will be discussed. Although, diatom concentrations

168

Chapter 6

for both levee and depression site do not show such a clear declining trend over the depth of the core as (idiosomic) testate amoebae do (Fig. 6.5), their concentrations were highly variable between both locations at a depth of 50 cm (Table 6.1). This big difference might be related to initial concentration differences between the original surface diatom assemblages.

The lowest measured BSi values for both the depression and levee location are observed close to the lowest groundwater level (Fig. 6.5). Above the lowest groundwater level, the marsh sediments form a more unstable environment as this zone of the sediment profile is subject to strong tidal fluctuations in the groundwater level (Fig. 6.4) and to infiltration of surface flooding water. This constantly changing environment facilitates bacterial degradation as oxic – anoxic conditions alternate. At the same time dissolution of BSi towards the sediment pore water is facilitated since the pore water is regularly refreshed with relatively Si poor flooding water. Beneath the minimum groundwater level, the marsh sediments are constantly water-saturated forming a stabile, anoxic environment that once it is saturated with DSi will probably buffer further BSi dissolution. This is the first clear evidence that frequent flooding with Si poor water is the main driving force for dissolution of BSi in tidal marsh sediments.

Based on this study, the schematic overview of the freshwater tidal marsh cycle is adapted, displaying the role of testate amoebae (Fig. 6.6).

169

Role of testate amoebae in Si cycle

Figure 6.6 New schematic overview of the freshwater tidal marsh Si cycle, involving diatoms, testate amoebae and phytoliths.

The use of protists as indicator for BSi trends Since (idiosomic) testate amoebae are highly correlated with BSi in the soil, it might be possible to use their concentrations to infer trends of BSi rise or decline for present-day studies. Thought, this potential use of (idiosomic) testate amoebae as proxy for BSi should be further investigated.

This study shows that diatom concentrations are a less good proxy for BSi. This is probably related to the fact that for calculation of diatom concentrations, equal weight is given to each counted individual diatom frustule. Though, individual diatom scales of one species can highly vary in Si content (4 fold differences are reported) (Brzezinski et al., 1990, Martin-Jézéquel et al., 2000). The amount of Si in the cell wall is linked to the duration of the cell wall synthesis phase, which is influenced by external factors (e.g. light, temperature, DSi availability, … ). Also, large diatom species have more BSi in their shells than small species (Martin- Jézéquel et al., 2000).

170

Chapter 6

6.5 Conclusion

1/ Testate amoebae, and more specifically idiosome testate amoebae, are important sinks and processors of Si in freshwater tidal marshes. Their concentration shows the best correlation with total biogenic Si concentration, better than diatoms and organic matter. Relatively quick dissolution implies a strong capacity of testate amoebae to contribute to DSi enrichment of water exported from marshes, which is important to sustain estuarine planktonic diatom communities during times of DSi limitation.

2/ Cycling of testate amoebae was not different at a natural levee and depression: at both locations dissolution of testate amoebae was almost complete in layers deeper than 0.4 m. Preservation of diatoms was better in the deeper sediment layers. The sediment depth range in which BSi is dissolved and exported as DSi is reaching up to the lowest groundwater level at a particular marsh location. At natural levees close to channels, the lowest groundwater level and depth range for BSi dissolution was found to be deeper than in depressions further away from channels.

Acknowledgements

We would like to thank Marijn van den broeck and Wim cleymans for their help with drilling the cores. This research was funded by a PhD grant of the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT-Vlaanderen).

171

Chapter 7

General Discussion

The results of modern testate amoebae assemblages and fossil testate amoebae assemblages are discussed using two conceptual diagrams. Further, the potentials and limitations of the use of testate amoebae in sea level reconstruction studies are summarized. At the end of this chapter, recommendations for future sea level reconstruction studies and silica studies are made.

General Discussion

7.1 Modern testate amoebae assemblages

The results of the chapters handling on modern testate amoebae assemblages (Chapter 2, 3 and 4) were merged together in a general conceptual diagram (Fig. 7.1).

Figure 7.1 Conceptual diagram of modern testate amoebae assemblages on freshwater, brackish and salt marshes, based on modern testate amoebae data from the Scheldt estuary.

174

Chapter 7

The diagram was arranged in a graph setting with the two most important axes related to variation in testate amoebae assemblages. The X–axis or flooding frequency was shown to be the most important variable on estuarine marshes for testate amoebae assemblages. This variable indicated variation in testate amoebae assemblages within each location. The Y-axis or salinity, showed the effect of the location of the marsh along the estuary and indicated variation in testate amoebae assemblages between locations.

The flooding frequency could be seen as an environmental stress factor for testate amoebae, with decreasing flooding frequency indicating a decreasing stress or more stable environment. Flooding frequency determined if testate amoebae species assemblages could or could not establish on the marsh surface. The maximal amount of flooding frequency at which a testate amoebae assemblage could occur depended on the location of the marsh within the estuary (Chapter 4; Fig. 7.1). This maximal flooding frequency at which a testate amoebae assemblage established decreased going further downstream the estuary (i.e. with increasing salinity) (Chapter 4). At the freshwater tidal marsh testate amoebae assemblages were found at high flooding frequencies of up to 69 % (Chapter 2). The maximal flooding frequency of testate amoebae assemblage occurrence at the brackish marsh lied already much lower (36.5 %) (Chapter 3). The salt marsh location was not even elevated enough to have locations with such low flooding frequency that testate amoebae assemblages could be found (lowest flooding frequency 13.5 %) (Chapter 4).

Like mentioned in the paragraph above, salinity influenced, in combination with flooding frequency, the occurrence of testate amoebae assemblages at the marsh. The conceptual diagram showed that the vegetation communities of marshes also altered along the estuary. In the fresh water marsh study it appeared that the occurrence of testate

175

General Discussion amoebae assemblages coincided with the presence (Phragmites australis) or absence (mud flat) of vegetation (Chapter 2). Looking to the appearance of testate amoebae assemblages at other marshes in relation to vegetation cover (Fig. 7.1), the above mentioned assumption should be adjusted. In the brackish marsh and salt marsh, vegetation could settle at higher flooding frequencies than that testate amoebae could establish their assemblages. This made it difficult to address the relationship between occurrence of testate amoebae and the vegetation. However, we can conclude that testate amoebae will not establish assemblages on the bare mudflat or within pioneer vegetated zones in brackish or salt marshes (Scirpus maritimus, Spartina anglica) (Chapter 2, 3 and 4).

Testate amoebae assemblages of the freshwater tidal marsh were discussed in Chapter 2 in separate testate amoebae zones dividing testate amoebae assemblages into supratidal and intertidal assemblages. This distinction was also made for the testate amoebae assemblages of the brackish marsh (Chapter 3). Though, the distinction between these different zones was not so clear for the freshwater tidal marsh assemblages, as the species followed each other up in a continuous way with changing marsh elevation (Chapter 4; Fig. 7.1). In contrast, the brackish marsh testate amoebae species did show a separate grouping, forming separate assemblages for each zone (Chapter 4).

Comparison of testate amoebae species numbers along the salinity gradient (Chapter 4; Gehrels et al. 2001) showed that testate amoebae species numbers were highest in either salt- or freshwater tidal marshes, and lower in the brackish marsh. The influence of salinity on the lowest occurrence of testate amoebae assemblages on a marsh continues on the higher elevated testate amoebae assemblages, as was noticed in the changed optima and tolerances of testate amoebae

176

Chapter 7

species that occurred in both freshwater- and brackish tidal marsh. An example is given in figure 7.1, where Cyphoderia ampulla and Cyclopyxis kahli are indicated at both freshwater- and brackish marsh. Both species are found at the outer ends of the elevation gradient that is inhabitated by testate amoebae assemblages. Cyphoderia ampulla can be found at highest flooding frequencies, though there is a discrepance of around 25 % of flooding frequency between the species optima for both locations (i.e. Cyphoderia ampulla can inhabite freshwater tidal marsh with 25% more flooding frequency compared to the brackish marsh). At lowest flooding frequencies, Cyclopyxis kahli can be found. This species optima has also shifted along the salinity gradient with around 10% of flooding frequency (i.e. Cyclopyxis kahli occurred at the freshwater marsh in a zone with 10% more flooding frequency compared to the brackish marsh).

Although flooding frequency was generally the most important environmental variable for the variation in testate amoebae assemblages of marshes, it played a minor role in explaining the species variation of the supratidal testate amoebae assemblages. This highest zone of the marsh is rarely flooded and species variation is more related to the organic matter content of the soil. This finding had repercussions for the use of transfer functions made for reconstructing sea level changes. It is better not to use testate amoebae assemblages of the supratidal zone, as they might indicate changes in organic matter, rather than differences in flooding frequency.

7.2 Sub- fossil testate amoebae assemblages and their role in the Si cycle

A general overview of the results of the studies on sub-fossil testate amoebae and their role in the Si cycle is summarized in a second conceptual diagram (Fig. 7.2).

177

General Discussion

Figure 7.2 Schematic overview of the process of dissolution of testate amoebae in the marsh soil.

Investigation of multiple sediment profiles of different locations within the freshwater tidal marsh indicated that concentrations of testate amoebae decreased fast with depth in the sediment profile (Chapter 2, Chapter 5). At around a depth of 40 cm, most of the testate amoebae already disappeared. There are multiple explanations for the lower amounts of testate amoebae with increasing depth within the cores (e.g. dry-wet cycles, bacterial decomposition, effect of temperature, or initially less high concentrated testate amoebae assemblage; Chapter 5).

The decrease of testate amoebae species throughout the sediment profiles was investigated, finding evidence of selective dissolution of idiosomic testate amoebae (Fig. 7.2; Chapter 5). They disappeared much faster from the sub-fossil record compared to other testate amoebae groups (xenosomic and calcium and protein shelled testate amoebae).

The fact that there were very low testate amoebae numbers left at depths below 40 cm and that selective preservation of testate amoebae 178

Chapter 7

occured will make it impossible to reconstruct sea level changes over long time spans. This was confirmed by reconstructing tidal water levels on the sediment profiles and comparing them to real measured tide gauge data resulting in a good water level reconstruction up to a sediment depth of 50 cm (until 1965), but increasing deviation between reconstructed and measured water levels at deeper parts of the sediment profile (chapter 5).

The fast disappearing idiosomic testate amoebae have shells made from self-secreted biogenic silica plates pasted together into an organic matrix (Mitchell et al., 2008b). This feature made us look into the role of (idiosomic) testate amoebae in the silicon cycle (chapter 6).

It is known that freshwater tidal marshes buffer the amount of dissolved silica in the estuarine river in times of DSi depletion (Struyf et al., 2005a). This means that with each tidal cycle dissolved marsh pore water that is rich in dissolved Si exported to the estuary. It is mainly BSi that is dissolved and exported. Sources of BSi in marshes are diatoms, phytoliths and testate amoebae (Chapter 6). Our comparative study involving diatom concentrations, (idiosomic) testate amoebae and organic matter content (as proxy for phytoliths) showed that (idiosomic) testate amoebae were important sinks and processors of Si in freshwater tidal marshes (Chapter 6).

7.3 Potentials and limitations of testate amoebae for sea level reconstruction studies.

Testate amoebae fulfilled all of the requirements to be a good MODERN bio-indicator for sea level change. Following Holt and Miller (2011), the requirement to be a good bio-indicator are: being abundant and common, having a good indicator ability and being well studied.

179

General Discussion

Testate amoebae were abundant and common, in a limited elevation range, within estuarine marshes The modern testate amoebae assemblages of estuarine marshes had a high concentration and diversity of testate amoebae species (Chapter 2 and 3; Charman et al., 2002, Gehrels et al., 2001, Roe et al., 2002). Though, testate amoebae assemblages were not common within the whole marsh area. They had only a limited distribution along the marsh surface elevation gradient. This limitation in distribution was most probably caused by high flooding frequency and/or salinity. Freshwater tidal and brackish marshes had a rather big elevation range inhabited by testate amoebae assemblages spanning from respectively just below and just above MHWL to HAT, compared to salt marsh testate amoebae assemblages (MHWS - HAT) (Charman et al., 2002, Gehrels et al., 2006, Gehrels et al., 2001, Riveiros et al., 2007). Since the transfer function is based on the relationship between testate amoebae assemblages and elevation (relative to MHWL) for the reconstruction of water level changes, the application of this transfer function is limited to the elevation range of testate amoebae assemblage occurrence. Thus, testate amoebae assemblages of freshwater tidal marshes can be used to reconstruct the largest water level changes.

Testate amoebae have good indicator ability Different testate amoebae assemblages followed each other up along the marsh elevation gradient (Chapter 2, 3 and 4). The variation in testate amoebae assemblages was mainly explained by flooding frequency (Fig. 7.1, Chapter 2, 3 and 4; Charman et al., 2002). Since flooding frequency changes with elevation along the marsh, elevation relative to MHWL was used for the transfer function. The transfer function resulting from the modern datasets of freshwater and brackish marshes showed comparable accuracy with transfer functions of other testate amoebae, foraminifera and diatom studies

180

Chapter 7

(Chapter 2 and 3; Charman et al., 2002, Gehrels et al., 2001, Hamilton and Shennan, 2005, Leorri et al., 2010, Zong and Horton, 1999). Since testate amoebae species of different marshes along the estuary have different optima and tolerances for flooding frequency (Chapter 4), it is recommended to use a local transfer function for the reconstruction of water level changes.

Testate amoebae are well studied The fact that most of the testate amoebae could have been identified within this study indicates that the testate amoebae are well enough studied for the use as bio-indicators of estuarine marshes. There was only one species with significant abundance ( > 2 % relative abundance) that was not found in testate amoebae literature. This species was given the name of Pseudohyalosphenia sp. (Chapter 3).

Testate amoebae do not fulfill the FOSSIL requirements that are needed for reliable reconstructions. The requirements of fossil species to be used in reconstructions will be discussed, based on the three assumptions made by Sachs et al., 1977.

The relationship between the modern protist assemblage and the environmental variable has not changed over time It is expected that fossil testate amoebae assemblages react in the same way to environmental changes as modern testate amoebae assemblages do (Sachs et al., 1977). The young sub-fossil testate amoebae assemblages (first 50 cm of the core) exhibited the same relation with elevation relative to MHWL as the modern testate amoebae assemblages. Evidence for this is the reliable reconstruction of water level changes until 1965 (Chapter 5).

181

General Discussion

The modern and palaeo habitat type should be analogous Since species assemblages variate with multiple environmental variables, the modern and fossil environments should be analogous to have a good basis for the reconstruction. It is discussed in the study of Wilson and Lamb (2012) that there are often no good modern analogues to past estuarine environments. In our study, we encountered also a fossil environment without modern analogue. Namely, two sediment profiles contained remains of the pioneer vegetation Scirpus maritimus, which is not found any more in the modern freshwater tidal marsh. For this specific case, it is suggested to take cores within old marshes, which had their pioneer vegetated stage in much earlier times.

Species assemblages do not change during fossilization The transfer function is based on modern testate amoebae assemblage data, so fossil testate amoebae assemblages should exist of the same species as the modern testate amoebae assemblages in order to gain reliable reconstructions. For this study, 30 species were found both in modern and fossil testate amoebae assemblages. Though, testate amoebae assemblages did not preserve well in marsh sediments. Testate amoebae assemblages showed a steep decrease in concentrations within the first 40 cm of the studied sediment profiles (Fig. 7.2; Chapter 5 and 6). Additionally, investigation of different testate amoebae groups (based on test material; Mitchell et al., 2008b) and types (based on pseudostome; Bonnet, 1975) showed that there is selective preservation of testate amoebae within the marsh sediments. Especially the concentration of idiosomic testate amoebae was decreasing fast with depth compared to other testate amoebae groups. This means that there is a difference between the fossil testate amoebae assemblage and the original modern testate amoebae assemblage. Altered fossil testate amoebae assemblage might be translated during reconstruction in a false environmental signal and can therefore not be used in reconstruction studies.

182

Chapter 7

Summarizing, testate amoebae are good bio-indicators for sea level reconstruction if adequate, unchanged fossil testate amoebae assemblages are found. Further, the sediment profiles need to contain environments with modern analogues. The reconstruction is best performed based on a local transfer function.

7.4 Recommendations for future studies

Recommendation for future sea level reconstruction studies When using testate amoebae, the problems with sea level reconstruction are mainly related to the fossil assemblages. A possible alternative for water level reconstruction is the use of diatoms as bio-indicator species. The investigation of the sediment profiles on diatom concentrations (Chapter 6) showed that diatoms preserve better in the marsh soil, as high concentrations were still found at a depth of 1 m. Thus, diatoms might be more suited as bio-indicator for sea level reconstructions. Other advantages of working with diatoms for sea level reconstruction studies would be that they are not bound to a specific salinity or flooding frequency; they are found within the whole tidal salinity range (Gehrels et al., 2001). Disadvantages of working with diatoms are the high load of allochtonous diatoms that are found on the marshes (e.g. planktonic species washing ashore with the tides). High species diversity, together with a constantly changing species taxonomy leads to high number of hours/sample (Gehrels et al., 2001).

Apart from the problems related to the bio-indicator species, some other factors need to be taken into account while reconstructing sea level changes:

Firstly, the relationship between environmental variables and elevation has to be looked into. Since elevation is not an ecological measure, it

183

General Discussion reflects the changes in environment along the elevation. In the intertidal zone, a significant part of the species variation was related to the flooding frequency which altered with elevation (Chapter 2, 3). However, the organic content of the soil explained the changes in species assemblages over the elevation at the supratidal zone of the brackish marsh (Chapter 3). Since the organic content of the soil is not related to the water level, the use of supratidal testate amoebae assemblages is not recommended for sea level change transfer functions.

Secondly, the history of coring locations should be studied carefully for flawes in the chronology of the marsh sedimentation. Cores to be investigated on fossil assemblages should best contain a sediment record of equal salinity; as it is possible, like shown for testate amoebae, that optima and tolerances of species shift with salinity and therefore might be translated into erroneous reconstructions. Further, the vegetation history of the coring location should be studied to be sure that modern analogue environments exists.

Recommendations for future BSi studies It might be interesting to investigate testate amoebae dissolution rates and corresponding BSi values at brackish marshes, as it is known that dissolution of BSi fastens with salinity. Perhaps more saline conditions will trigger faster dissolution of diatoms, while testate amoebae might disappear so quickly that most of them might be dissolved before a depth of 0.4 m. Will it still be testate amoebae that are providing dissolved silica for the estuary? Or will diatoms have higher dissolution rates within more saline conditions?

184

Chapter 7

7.5 Conclusion

This study investigated the modern testate amoebae assemblages of tidal marshes along the Scheldt estuary. The modern testate amoebae assemblages species composition had a strong relationship with changing flooding frequency along the elevation gradient at a tidal marsh. Further, salinity gradient along the estuary influenced the lowest occurrence of testate amoebae assemblages on the marsh surface and changed species optima and tolerances.

The potential use of testate amoebae for the reconstruction of water level changes within the estuary was only valid for the reconstruction of water level changes over a short time span (until 1965). The insufficient and selective preservation of testate amoebae shells lied at the basis for this limitation. This insufficient preservation of testate amoebae shells was investigated by looking into their role in the Si cycle. We found that testate amoebae shells dissolved quickly in the marsh sediments, providing dissolved Si to the estuary.

We recommend that further studies investigate the potential use of diatoms as bio-indicators for the reconstruction of water level changes within the estuary. Other studies can be undertaken to explore the role of testate amoebae in the Si cycle in more saline estuarine environments.

185

References

Aoki Y, Hoshino M, & Matsubara T. 2007. Silica and testate amoebae in a soil under pine-oak forest. Geoderma 142: 29-35.

Bass D. 2009. Thecofilosea. Version 06 May 2009. http://tolweb.org/Thecofilosea/124280/2009.05.06

Bass D & Cavalier-Smith T. 2009. Cercozoa. Version 22 March 2009. http://tolweb.org/Cercozoa/121187/2009.03.22

Beauchard O, Ciutat A, Gerino M, Munoz T, Jacobs S, Tackx M, Stora G, & Meire P. 2012. Spatiotemporal bioturbation patterns in a tidal freshwater marsh. Estuarine Coastal and Shelf Science 96: 159-169.

Behrensmeyer AK & Kidwell SM. 1985. Taphonomy: a key to paleoenvironmental interpretation based on fossil preservation. Palaios 1: 207-227.

Berkeley A, Perry CT, Smithers SG, Horton BP, & Taylor KG. 2007. A review of the ecological and taphonomic controls on foraminiferal assemblage development in intertidal environments. Earth-Science Reviews 83: 205-230.

Beyens L & Meisterfeld R. 2002. Protozoa: Testate Amoebae. In: Tracking Environmental Change Using Lake Sediments, Smol, J, Birks, HJ, Last, W, et al. (eds.).Springer Netherlands: Dordrecht.

Bindoff NL, Willebrand J, Artale V, Cazenave A, Gregory J, Gulev S, Hanawa K, Le Quéré C, Levitus S, Nojiri Y, Shum CK, Talley LD, & Unnikrishnan A. 2007. Observations:Oceanic Climate Change and Sea Level. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Solomon, S, Qin, D, Manning, C, et al. (eds.).Cambridge University Press: Cambridge, United Kingdom and New York, NY, USA.

Birks HJB, Line JM, Stevenson AC, & ter Braak CJF. 1990. Diatoms and pH reconstruction. Philosophical Transactions of the Royal Society of London B Biological Sciences 263-278.

Blott SJ & Pye K. 2001. Gradistat: a grain size distribution and statistics package for the analysis of unconsolidated sediments. Earth Surface Processes and Landforms 1237-1248.

Bonnet L. 1975. Types morphologiques, écologie et évolution de la Thèque chez les thécamoebiens. Protistologica 11: 363-378.

References

Booth RK & Jackson ST. 2003. A high resolution record of late-Holocene moisture variability from a Michigan raised bog, USA. Holocene 13: 863- 876.

Booth RK, Jackson ST, & C.E.D. G. 2004. Paleoecology and high- resolution paleohydrology of a kettle peatland in upper Michigan. Quaternary Research 61: 1-13.

Bouma TJ, De Vries MB, Low E, Kusters L, Herman PMJ, Tanczos IC, Temmerman S, Hesselink A, Meire P, & van Regenmortel S. 2005. Flow hydrodynamics on a mudflat and in salt marsh vegetation: identifying general relationships for habitat characterisations. Hydrobiologia 540: 259-274.

Broeckx S, Smets S, Liekens I, Bulckaen D, & De Nocker L. 2011. Designing a ong-term flood risk management plan for the Scheldt estuary using a risk-based approach. Natural Hazards 57: 245-266.

Brzezinski MA, Olson RJ, & Chisholm SW. 1990. Silicon availability and cell-cycle progression in marine diatoms. Marine Ecology Progress Series 67: 83-96.

Buijse AD, Coops H, Staras M, Jans LH, Van Geest GJ, Grifts RE, Ibelings BW, Oosterberg W, & Roozen FCJM. 2002. Restoration strategies for river floodplains along large lowland rivers in Europe. Freshwater Biology 47: 889-907.

Campeau S, Pienitz R, & Hequette A. 1999. Diatoms as quantitative paleodepth indicators in coastal areas of the southeastern Beaufort Sea, Arctic Ocean. Palaeogeography Palaeoclimatology Palaeoecology 146: 67-97.

Chardez D. 1991. Le genre Cyphoderina Schlumberger, 1845 (Protozoa: Rhizopoda: Testacea). Acta Protozoologica 30: 49-53.

Chardez D & Lambert J. 1981. Thécamoebiens indicateurs biologiques (Protozoa: Rhizopoda testacea). Bulletin des Recherches Agronomiques de Gembloux 16: 181-204.

Charman DJ. 2001. Biostratigraphic and palaeoenvironmental applications of testate amoebae. Quaternary Science Reviews 20: 1753- 1764.

Charman DJ, Blundell A, & Members A. 2007. A new European testate amoebae transfer function on ombrotrophic peatlands. Journal of Quaternary Science 22: 209-221.

188

References

Charman DJ, Gehrels WR, Manning C, & Sharma C. 2010. Reconstruction of recent sea-level change using testate amoebae. Quaternary Research 73: 208-219.

Charman DJ, Hendon D, & Woodland WA. 2000. The identification of testate amoebae (Protozoa: Rhizopoda) in peats. Quaternary Research Association: London. 147 pp.

Charman DJ, Hohl V, Blundell A, Mitchell F, Newberry J, & Oksanen P. 2012. A 1000-year reconstruction of summer precipitation from Ireland: Calibration of a peat-based palaeoclimate record. Quaternary International 268: 87-97.

Charman DJ, Roe HM, & Gehrels WR. 1998. The use of testate amoebae in studies of sea-level change: a case study from the Taf Estuary, south Wales, UK. Holocene 8: 209-218.

Charman DJ, Roe HM, & Gehrels WR. 2002. Modern distribution of saltmarsh testate amoebae: regional variability of zonation and response to environmental variables. Journal of Quaternary Science 17: 387-409.

Church JA, Gregory JM, White NJ, Platten SM, & Mitrovica JX. 2011. Understanding and projecting sea level change. Oceanography 24: 130- 143.

Church JA & White NJ. 2006. A 20th century acceleration in global sea- level rise. Geophysical Research Letters 33: L01602.

Coen I. 2008. De eeuwige Schelde? Ontstaan en ontwikkeling van de Schelde. Waterbouwkundig Laboratorium 1933 - 2008. Waterbouwkundig Laboratorium: Borgerhout. 112 pp.

Conley DJ, Schelske CL, & Stoermer EF. 1993. Modification of the biogeochemical cycle of silica with eutrophication. Marine Ecology Progress Series 101: 179-192.

Constanza R, Perez-Maqueo O, Martinez ML, Sutton P, Anderson SJ, & Mulder K. 2008. The value of coastal wetlands for hurricane protection Ambio 37: 241-248.

Couteaux MM. 1992. Decomposition of cells and empty shells of Testate Amebas (Rhizopoda, Testacea) in an organic-acid soil sterilized by propylene-oxide fumigation, autoclaving, and gamma-ray irradiation. Biology and Fertility of Soils 12: 290-294.

189

References

De Kraker AMJ. 2004. Reconstructie van het stroomgebied van de Westerschelde tussen 1550 en 2000 aan de hand van kaarten en bedijkingsgeschiedenis. RIKZ: Amsterdam. 72 pp.

Decloitre L. 1962. Le genre Euglypha Dujardin. Archiv für Protistenkunde 106: 51-100.

Decloitre L. 1974. Le genre Euglypha. Compléments à jour ou 31 décembre 1974 de la monographie du genre parue en 1962. Archiv für Protistenkunde 118: 18-33.

Decloitre L. 1981. Le Genre Trinema DUJARDIN, 1841. Révision à jour au 31 XII. 1979 Archiv für Protistenkunde 124: 193-218.

Deflandre G. 1928. Le genre Arcella Ehrenberg. Morphologie-Biologie. Essai phylogénetique et systématique. Archiv für Protistenkunde 152- 287.

Deflandre G. 1929. Le genre Centropyxis Stein. Archiv für Protistenkunde 67: 322-375.

DeMaster DJ. 1981. The supply and accumulation of silica in the marine environment. Geochimica et Cosmochimica Acta 45: 1715-1732.

Edwards RJ & Horton BP. 2000. Reconstructing relative sea-level change using UK salt-marsh foraminifera. Marine Geology 169: 41-56.

Edwards RJ, van de Plassche O, Gehrels WR, & Wright AJ. 2004. Assessing sea-level data from Connecticut, USA, using a foraminiferal transfer function for tide level. Marine Micropaleontology 51: 239-255.

Foissner W. 1999. Soil protozoa as bioindicators: pros and cons, methods, diversity, representative examples. Agriculture Ecosystems & Environment 74: 95-112.

Foissner W & Korganova GA. 2000. The Centropyxis aerophila Complex (Protozoa: Testacea). Acta Protozoologica 39: 257-273.

Friedrichs CT & Aubrey DG. 1994. Tidal propagation in strongly convergent channels. Journal of Geophysical Research 99: 3321-3336.

Gabriel JJ, Reinhardt EG, Peros MC, Davidson DE, van Hengstum PJ, & Beddows PA. 2009. Palaeoenvironmental evolution of Cenote Aktun Ha (Carwash) on the Yucatan Peninsula, Mexico and its response to Holocene sea-level rise. Journal of Paleolimnology 42: 199-213.

190

References

Gauthier-Liévre L & Thomas R. 1958. Les genres Difflugia, Pentagonia, Maghrebia et Hoogenraadia (Rhizopodes testacés) en Afrique. Archiv für Protistenkunde 103: 241-370.

Gehrels WR. 1994. Determining Relative Sea-Level Change from Salt- Marsh Foraminifera and Plant Zones on the Coast of Maine, U.S.A. Journal of Coastal Research 10: 990-1009.

Gehrels WR, Hendon D, & Charman DJ. 2006. Distribution of testate amoebae in salt marshes along the North American East Coast. Journal of Foraminiferal Research 36: 201-214.

Gehrels WR, Roe HM, & Charman DJ. 2001. Foraminifera, testate amoebae and diatoms as sea-level indicators in UK saltmarshes: a quantitative multiproxy approach. Journal of Quaternary Science 16: 201-220.

Gehrels WR, Roe HM, & Charman DJ. 2002. Foraminifera, testate amoebae and diatoms as sea-level indicators in UK saltmarshes: a quantitative multiproxy approach (vol 16, pg 201, 2001). Journal of Quaternary Science 17: 285-285.

Ghosh A, Saha S, Saraswati PK, Banerjee S, & Burley S. 2009. Intertidal foraminifera in the macro-tidal estuaries of the Gulf of Cambay: Implications for interpreting sea-level change in palaeo-estuaries. Marine and Petroleum Geology 26: 1592-1599.

Golemansky V. 1971. Taxonomical and Zoogeographical Notes on the Testate Amoebae (Rhizopoda, Testacea) of the Coastal Interstitial Waters of the Soviet Far-Eastern Coast (The Sea of Japan) and the Western Coast of Canada (The Pacific Ocean). Archiv für Protistenkunde 113: 235-249.

Golemansky V & Todorov M. 2004. Shell morphology, biometry and distribution of some marine interstitial testate amoebae (Sarcodina : rhizopoda). Acta Protozoologica 43: 147-162.

Griffies SM & Greatbatch RJ. 2012. Physical processes that impact the evolution of global mean sea level in ocean climate models. Ocean Modelling 51: 37-72.

Grimm E. 2004. Tilia 2.2, software.

Grospietsch T. 1964. Die Gattungen Cryptodifflugia and Difflugiella (Rhizopoda testacea). Zoologische Anzeiger 172: 243-257.

Grospietsch T. 1965. Monographische Studie der Gattung Hyalosphenia Stein (Rhizopoda, Testacea). Hydrobiologia 26: 211-241.

191

References

Hackney CT, Cahoon LB, Preziosi C, & Norris A. 2000. Silicon is the link between tidal marshes and estuarine fisheries: A new paradigm. In: Concepts and Controversies in Tidal Marsh Ecology, Weinstein, MP,Kreeger, DA (eds.).Springer: Dordrecht.

Haggart BA. 1986. Relative sea-level change in the Beauly Firth, Scotland. Boreas 15: 191-207.

Hamilton S & Shennan I. 2005. Late Holocene relative sea-level changes and the earthquake deformation cycle around upper Cook Inlet, Alaska. Quaternary Science Reviews 24: 1479-1498.

Hassan GS, Espinosa MA, & Isla FI. 2006. Modern diatom assemblages in surface sediments from estuarine systems in the southeastern Buenos Aires province, Argentina. Journal of Paleolimnology 35: 39-53.

Hemingway KL, Cutts NC, & Pérez-Dominguez R. 2008. Managed Realignment in the Humber estuary, UK. Institute of Estuarine & Coastal Studies (IECS), University of Hull, UK. Report produced as part of the European Interreg IIIB HARBASINS project.

Hendon D & Charman DJ. 1997. The preparation of testate amoebae (Protozoa: Rhizopoda) samples from peat. Holocene 7: 199-205.

Hoffman M. 1993. Vegetatiekundig-ecologisch onderzoek van de buitendijkse gebieden langs de Zeeschelde met vegetatiekartering. Laboratorium voor Morfologie, Systematiek en Ecologie van de Planten 214.

Holt EA & Miller SW. 2011. Bioindicators: Using organisms to Measure Environmental impacts. Nature Education Knowledge 3.

Horton BP, Corbett R, Culver SJ, Edwards RJ, & Hillier C. 2006. Modern saltmarsh diatom distributions of the Outer Banks, North Carolina, and the development of a transfer function for high resolution reconstructions of sea level. Estuarine Coastal and Shelf Science 69: 381-394.

Jha AK, Bloch R, & Lamond J. 2012. Cities and Flooding. A Guide to Integrated Urban Flood Risk Managment for the 21th Century. The World Bank: Washington DC. 638 pp.

Johnson GC & Wijffels SE. 2011. Ocean density change contributions to sea level rise. Oceanography 24: 112-131.

Jonkman SN & Kelman I. 2005. Deaths during the 1953 North Sea Storm Surge. In: Solutions to Coastal Disasters 2005, Wallendorf, L, Ewing, L, Rogers, S, et al. (eds.).American Society of Civil Engineers.

192

References

Juggins S. 2007. C2 Version 1.5 User guide. Software for ecological and palaeoecological data analysis and visualisation. . UK. 73 pp.

Kemp AC, Horton BP, Corbett DR, Culver SJ, Edwards RJ, & van de Plassche O. 2009. The relative utility of foraminifera and diatoms for reconstructing late Holocene sea-level change in North Carolina, USA. Quaternary Research 71: 9-21.

Krauss KW, Doyle TW, Doyle TJ, Swarzenski CM, From AS, Day RH, & Connor WH. 2009. Water level observations in mangrove swamps during two hurricanes in Florida. Wetlands 29: 142-149.

Lancelot C. 1995. The mucilage phenomenon in the continental coastal waters of the North-Sea. Science of the Total Environment 165: 83-102.

Lane A. 2004. Bathymetric evolution of the Mersey Estuary, UK, 1906- 1997: causes and effects. Estuarine Coastal and Shelf Science 59: 294- 293.

Langdon PC, Brown AG, Caseldine CJ, Blockley SPE, & Stuijts I. 2012. Regional climate change from peat stratigraphy for the mid- to late Holocene in central Ireland. Quaternary International 268: 145-155.

Lanzoni S & Seminara G. 1998. On tide propagation in convergent estuaries. Journal of Geophysical Research 103: 30793-30812.

Last WM & Smol JP. 2001. Tracking environmental change using lake sediments. Vol. 2:Physical and geochemical methods. Kluwer Academic Publishers: Dordrecht, Boston, London. 548 pp.

Lebbe L, Van Meir N, & Viaene P. 2008. Potential implications of sea- level rise for Belgium. Journal of Coastal Research 24: 358-366.

Leorri E, Gehrels WR, Horton BP, Fatela F, & Cearreta A. 2010. Distribution of foraminifera in salt marshes along the Atlantic coast of SW Europe: Tools to reconstruct past sea-level variations. Quaternary International 221: 104-115.

Little C. 2000. The Biology of Soft Shores and Estuaries. Oxford University Press: Oxford. 252 pp.

Long AJ. 1992. Coastal responses to changes in sea level in the East kent Fens and southeast England, Uk over the last 7500 years. Proceedings of the Geologists' Association 103: 187-199.

193

References

Lotze HK, Lenihan HS, Bourque BJ, Bradbury RH, Cooke RG, Kay MC, Kidwell SM, Kirby MX, Peterson CH, & Jackson BC. 2006. Depletion, degradation, and recovery potential of estuaries and coastal seas. Science 312: 1806-1809.

Lousier JD & Parkinson D. 1981. The disappaerance of the tests of litter- and soil-testate amoebae (Testacea Rhizopoda Protozoa). Archiv für Protistenkunde 124: 312-336.

Lyman LR. 2010. What Taphonomy Is, WHat it Isn't, and Why Taphonomists Should Care about the Difference. Journal of Taphonomy 8: 1-16.

Martin-Jézéquel V, Hildebrand M, & Brzezinski MA. 2000. Silicon metabolism in diatoms: implications for growth. Journal of Phycology 36: 821-840.

Mazei YA & Tsyganov AN. 2006. Freshwater Testate Amoebae KMK: Moscow. 300 pp.

Mbugua MW. 2008. Characterization of Unusual Gymnamoebae Isolated from the Marine Environment. Marshall University: 139 pp.

McGranahan G, Balk D, & Anderson B. 2007. The rising tide: assessing the risks of climate change and human settlements in low elevation coastal zones. Environment and urbanization 19: 17-37.

McLusky DS & Elliot M. 2009. The Estuarine Ecosystem: ecology, threats and management. Oxford University Press: Oxford. 214 pp.

Meire P, Ysebaert T, Van Damme S, Van den Bergh E, Maris T, & Struyf E. 2005. The Scheldt estuary: A description of a changing ecosystem. Hydrobiologia 540: 1-11.

Meisterfeld R & Heisterbaum M. 1986. The decay of empty tests of testate amoebae (Rhizopoda, Protozoa). Symposia Biologica Hungarica 33: 285-290.

Meyvis L, Graré W, & Dauwe W. 2003. Actualisatie van het SIGMAPLAN. Ministerie van de Vlaamse Gemeenschap AWZ-afdeling Zeeschelde: 12 pp.

Miller L & Douglas BC. 2004. Mass and volume contributions to twentieth-century global sea level rise. Nature 428: 406-409.

194

References

Mitchell EAD, Charman DJ, & Warner BG. 2008a. Testate amoebae analysis in ecological and paleoecological studies of wetlands: past, present and future. Biodiversity and Conservation 17: 2115-2137.

Mitchell EAD, Payne RJ, & Lamentowicz M. 2008b. Potential implications of differential preservation of testate amoeba shells for paleoenvironmental reconstruction in peatlands. Journal of Paleolimnology 40: 603-618.

Mol G. 1995. De Westerschelde: een resultaat van menselijke ingrepen. Rapport RIKZ-95.030. Rijksinstituut voor Kust en Zee: 102 pp.

Moriceau B, Garvey M, Ragueneau O, & Passow U. 2007. Evidence for reduced biogenic silica dissolution rates in diatom aggregates. Marine Ecology Progress Series 333: 129-142.

Mys M. 1973. De landschapsgeschiedenis van de Scheldepolders ten noorden van Antwerpen. Tijdschrift van de Belgische Vereniging voor Aardrijkskundige Studies 17: 39-124.

Mys M, Gullentops F, Janssens P, Wuytack MJ, & H. S. 1983. De holocene evolutie van de ulluviale vlakte van de Beneden-Schelde. Tijdschrift van de Belgische Vereniging voor Aardrijkskundige Studies 30: 217-255.

Ng SL & Sin FS. 2003. A diatom model for inferring sea level change in the coastal waters of Hong Kong. Journal of Paleolimnology 30: 427- 440.

Nicholls RJ & Cavenave A. 2010. Sea-Level Rise and Its Impact on Coastal Zones. Science 38: 1517-1520.

Nicholls RJ, Hanson S, Herweijer C, Patmore N, Hallegatte S, Corfee- Morlot J, Chateau J, & Muir-Wood R. 2007. Ranking of the world's cities most exposed to coastal flooding today and in the future. Executive Summary. In: OECD Environmental Working Paper No. 1 (ENV/WKP(2007)1), OECD

Norris AR & Hackney CT. 1999. Silica Content of a Mesohaline Tidal Marsh in North Carolina. Estuarine, Coastal and Shelf Science 49: 597- 605.

Ogden CG. 1983. Observations on the systematics of the genus Difflugia in Britain (Rhizopoda, Protozoa). Bulletin of the British Museum of Natural History (Zoology) 44: 1-73.

195

References

Ogden CG & Couteaux MM. 1987. The biology and ultrastructure of the soil testate amoebae Tracheuglypha dentate (Rhizopoda: Euglyphidae). . European Journal of Protistology 2328-2342.

Ogden CG & Couteaux MM. 1989. Interstitial Marine Rhizopods (Protozoa) from Littoral Sands on the East Coast of England. European Journal of Protistology 24: 281-290.

Ogden CG & Hedley RH. 1980. An atlas of freshwater testate amoebae. British museum (Natural History) and Oxford University Press: London and Oxford. 222 pp.

Ooms M, Beyens L, & Temmerman S. 2011. Testate amoebae as estuarine water-level indicators: modern distribution and the development of a transfer function from a freshwater tidal marsh (Scheldt estuary, Belgium). Journal of Quaternary Science 26: 819-828.

Ooms M, Beyens L, & Temmerman S. 2012. Testate amoebae as Proxy for Water Level Changes in a Brackish Tidal Marsh. Acta Protozoologica 51: 271-289.

Patterson RT & Fishbein E. 1989. Re-Examination of the Statistical Methods Used to Determine the Number of Point Counts Needed for Micropaleontological Quantitative Research. Journal of Paleontology 63: 245-248.

Patterson RT & Kumar A. 2002. A review of current testate rhizopod (thecamoebian) research in Canada. Palaeogeography Palaeoclimatology Palaeoecology 180: 225-251.

Patterson RT, Roe HM, & Swindles GT. 2012. Development of an Arcellacea (testate lobose amoebae) based transfer function for sedimentary Phosphorus in lakes. Palaeogeography, Palaeoclimatology, Palaeoecology 348-349: 32-44.

Payne RJ & Mitchell EAD. 2007. Ecology of Testate Amoebae from Mires in the Central Rhodope Mountains, Greece and Development of a Transfer Function for Palaeohydrological Reconstruction. Protist 158: 159-171.

Peterson DH. 2007. Seasonal/Yearly Salinity variations in San Francisco bay. http://sfbay.wr.usgs.gov/hydroclimate/sal_variations/index.html.

Porter SM & Knoll AH. 2000. Testate amoebae in the Neoprotozoic Era: evidence from vase-shaped microfossils in the Chuar Group, Grand Canyon. Paleobiology 26: 360-385.

196

References

ProSes. 2008. Ontwikkelingsschets 2010 Schelde estuarium. http://www.ontwikkelingsschets2010.nl/www/scripts/index.php.

R Development Core Team. 2009. R: A Language and Environment for Statistical Computing.

Raper SCB & Braithwaite RJ. 2006. Low sea level rise projections from mountain glaciers and icecaps under global warming. Nature 439: 311- 313.

Remane A & Schlieper C. 1971. Biology of Brackish Water. E. Schweizerbart'sche Verlagbuchhandlung (Nägele u. Obermiller): Stuttgart. 371 pp.

Resio DT & Westerink JJ. 2008. Modelling the physics of storm surges. Physics Today 61: 33-38.

Rippon S. 2000. The transformation of coastal wetlands. Exploitation and management of marshland landscapes in North West Europe during the Roman and medieval periods. University Press: Oxford. 332 pp.

Riveiros NV, Babalola AO, Boudreau REA, Patterson RT, Roe HM, & Doherty C. 2007. Modern distribution of salt marsh foraminifera and thecamoebians in the Seymour-Belize Inlet Complex, British Columbia, Canada. Marine Geology 242: 39-63.

Roe HM, Charman DJ, & Gehrels WR. 2002. Fossil testate amoebae in coastal deposits in the UK: implications for studies of sea-level change. Journal of Quaternary Science 17: 411-429.

Sachs HM, Webb T, & Clark DR. 1977. Paleoecological transfer- functions. Annual Review of Earth and Planetary Sciences 5: 159-178.

SAS/STAT. 2009. SAS/STAT 9.2 User's Guide, Second Edition. http://support.sas.com/documentation/cdl/en/statug/63033/HTML/defa ult/viewer.htm#titlepage.htm.

Savenije HHG. 2001. A simple analytical expression to describe tidal damping or amplification. Journal of Hydrology 243: 205-215.

Sawai Y, Nasu H, & Yasuda Y. 2002. Fluctuations in relative sea-level during the past 3000 yr in the Onnetoh estuary, Hokkaido, northern Japan. Journal of Quaternary Science 17: 607-622.

Schönborn W. 1962. Die Ökologie der Testaceen im oligotrophen See, dargestellt am Beispiel des Großen Stechlinsees. Limnologica 2: 111- 182.

197

References

Schröter D, Wolters V, & De Ruiter PC. 2003. C and N mineralisation in the decomposer food webs of a European forest transect. Oikos 102: 294-308.

Scott DB & Medioli FS. 1978. Vertical zonations of marsh foraminifera as accurate indicators of former sea-levels. Letters to nature 272: 528- 531.

Shennan I, Long AJ, Rutherford MM, Green FM, Innes JB, Lloyd JM, Zong Y, & Walker KJ. 1996. Tidal marsh stratigraphy, sea-level change and large earthquakes, I: a 5000 year record in Washington, USA. Quaternary Science Reviews 15: 1023-1059.

Shepard CC, Agostini VN, Gilmer B, Allen T, Stone J, Brooks W, & Beck MW. 2012. Evaluating alternative future sea-level rise scenarios. Natural Hazards 63: 1223-1224.

Sherrod BL. 1999. Gradient analysis of diatom assemblages in a Puget Sound salt marsh: can such assemblages be used for quantitative paleoecological reconstructions? Palaeogeography, Palaeoclimatology, Palaeoecology 149: 213-226.

Smirnov A. 2008. Lobosea. Lobose amoebae. Version 02 September 2008. http://tolweb.org/Lobosea/121174/2008.09.02

Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, & Miller HL. 2007. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, 2007. Cambridge University Press: Cambridge, United kingdom and New York, NY, USA. 996 pp.

Stevenson JR, Pan Y, Kociolek PJ, & Kingston JC. 2006. Advances in Algal Biology: A Commemoration of the Work of Rex Lowe (Developments in Hydrobiology). Springer: Dordrecht (The Netherlands). 257 pp.

Stockmarr J. 1971. Tablets with spores used in absolute pollen analysis. Pollen Spores 13: 615-621.

Struyf E, Dausse A, Van Damme S, Bal K, Gribsholt B, Boschker HTS, Middelburg JJ, & Meire P. 2006. Tidal marshes and biogenic silica recycling at the land-sea interface. Limnology and Oceanography 51: 838-846.

Struyf E, Temmerman S, & Meire P. 2007. Dynamics of biogenic Si in freshwater tidal marshes: Si regeneration and retention in marsh sediments (Scheldt estuary). Biogeochemistry 82: 41-53.

198

References

Struyf E, Van Damme S, Gribsholt B, Bal K, Beauchard O, Middelburg JJ, & Meire P. 2007b. Phragmites australis and silica cycling in tidal wetlands. Aquatic Botany 87: 41-53.

Struyf E, Van Damme S, Gribsholt B, & Meire P. 2005a. Freshwater marshes as dissolved silica recyclers in an estuarine environment (Schelde estuary, Belgium). Hydrobiologia 540: 69-77.

Struyf E, Van Damme S, Gribsholt B, Middelburg JJ, & Meire P. 2005b. Biogenic silica in tidal freshwater marsh sediments and vegetation (Schelde estuary, Belgium). Marine Ecology-Progress Series 303: 51- 60.

Sullivan MJ & Moncreiff CA. 1990. Edaphic algae are an important component of salt-marsh food-webs: Evidence from multiple stable isotope analyses. Marine Ecology-Progress Series 62: 149-159.

Swindles GT & Roe HM. 2007. Examining the dissolution characteristics of testate amoebae (Protozoa : Rhizopoda) in low pH conditions: Implications for peatland palaeoclimate studies. Palaeogeography Palaeoclimatology Palaeoecology 252: 486-496.

Taveniers E & Mostaert F. 2009. MONEOS-jaarboek monitoring WL 2008: Overzicht monitoring hydrodynamiek en fysische parameters zoals door WL in 2008 in het Zeescheldebekken gemeten. Versie 4.0 WL rapporten, projectnr 883_07. Waterbouwkundig Laboratorium.

Taverniers E, Vereecken H, & Mostaert F. 2011. MONEOS-jaarboek monitoring WL 2010: Overzicht monitoring hydrodynamiek en fysische parameters zoals door WL in 2010 in het Zeescheldebekken gemeten. Versie 4.0. WL rapporten, 833_07. Waterbouwkundig Laboratorium

Temmerman S, Govers G, Meire P, & Wartel S. 2003b. Modelling long- term tidal marsh growth under changing tidal conditions and suspended sediment concentrations, Scheldt estuary, Belgium. Marine Geology 193: 151-169.

Temmerman S, Govers G, Meire P, & Wartel S. 2004a. Simulating the long-term development of levee-basin topography on tidal marshes. Geomorphology 63: 39-55.

Temmerman S, Govers G, Wartel S, & Meire P. 2003a. Spatial and temporal factors controlling short-term sedimentation in a salt and freshwater tidal marsh, Scheldt estuary, Belgium, SW Netherlands. Earth Surface Processes and Landforms 28: 739-755.

199

References

Temmerman S, Govers G, Wartel S, & Meire P. 2004b. Modelling estuarine variations in tidal marsh sedimentation: response to changing sea level and suspended sediment concentrations. Marine Geology 212: 1-19. ter Braak CJF. 1987a. Calibration. In: Data analysis in community and landscape ecology, Jongman, RHG, ter Braak, CJF,van Tongeren, OFR (eds.).Pudoc Wageningen. ter Braak CJF. 1987b. Ordination. In: Data analysis in community and landscape ecology, Jongman, RHG, ter Braak, CJF,van Tongeren, OFR (eds.).Pudoc Wageningen. ter Braak CJF & Juggins S. 1993. Weighted averaging partial least squares regression (WA-PLS): an improved method for reconstructing environmental variables from species assemblages. Hydrobiologia 485- 502. ter Braak CJF & Šmilauer P. 1998. CANOCO Reference Manual and User's Guide to Canoco for Windows: Software for Canonical Community Ordination (version 4). Microcomputer Power: Ithaca, NY, USA. 352 pp.

Thioulouse J, Chessel D, Dolédec S, & Olivier J-M. 1997. ADE-4: a multivariate analysis and graphical display software. Statistics and Computing 7: 75-83.

Todorov M, Golemansky V, Mitchell EAD, & Heger TJ. 2009. Morphology, Biometry, and Taxonomy of Freshwater and Marine Interstitial Cyphoderia (Cercozoa: Euglyphida). Journal of Eukaryotic Microbiology 56: 279-289.

Tucker BM & Beatty HJ. 1974. pH, conductivity and chlorides. . In: Methods for Analysis of Irrigated Soils. , Loveday, J (eds.).Commonwealth Bureau of Soils: Farnham Royal, England.

Turner TE, Swindles GT, Charman DJ, & Blundell A. 2013. Comparing regional and supra-regional transfer functions for palaeohydrological reconstruction from Holocene peatlands. Palaeogeography, Palaeoclimatology, Palaeoecology 369: 395-408.

Van Cappellen P. 2003. Biomineralization and global biogeochemical cycles. Reviews in Mineralogy and Geochemistry 54: 357-381.

Van Damme S, De Winder B, Ysebaert T, & Meire P. 2001. Het 'bijzondere' van de Schelde: de abiotiek van het Schelde-estuarium. De Levende Natuur 102: 37-39.

200

References

Van Damme S, Struyf E, Maris T, Ysebaert T, Dehairs F, Tackx M, Heip C, & Meire P. 2005. Spatial and temporal patterns of water quality along the estuarine salinity gradient of the Scheldt estuary (Belgium and The Netherlands): results of an integrated monitoring approach. Hydrobiologia 540: 29-45. van der Spek AJF. 1997. Tidal asymmetry and long term evolution of Holocene tidal basins in The Netherlands: simulation of palaeo-tides in the Schelde estuary. Marine Geology 141: 71-90.

Van der Werf A. 1955. A new method of concentrating and cleaning diatoms and other organisms. Verhandlungen der Internationalen Vereinigung für Theoretische und Angewandte Limnologie 276-277.

Varekamp JC, Thomas E, & Plassche O. 1992. Relative sea-level rise and climate change over the last 1500 years. Terra Nova 4: 293-304.

Vieillard AM, Fulweiler RW, Hughes ZJ, & Carey JC. 2011. The ebb and flood of Silica: Quantifying dissolved and biogenic silica fluxes from a temperate salt marsh. Estuarine, Coastal and Shelf Science 95: 415- 423.

Vohník M, Burdíková Z, & Vosátka M. 2008. Testate amoebae (Arcellinida and Euglyphida) vs. ericoid mycorrhizae and DSE fungi: a possible novel interaction in the mycorrhizosphere of ericaeous plants? Microbial Ecology 57: 203-214.

Wamsley TV, Cialone MA, Smith JM, Atkinson JH, & Rosati JD. 2010. The potential of wetlands in reducing storm surge. Ocean Engineering 37: 59-68.

Webster PJ, Holland GJ, Curry JA, & Chang HR. 2005. Changes in tropical cyclone number, duration, and intensity in a warming environment. Science 309: 1844-1846.

Więski K, Guo H, Craft CB, & Pennings SC. 2010. Ecosystem Functions of Tidal Fresh, Brackish, and Salt Marshes on the Georgia Coast. Estuaries and Coasts 33: 161-169.

Wilkinson DM & Mitchell EAD. 2010. Testate Amoebae and Nutrient Cycling with Particular Reference to Soils. Geomicrobiology Journal 27: 520-533.

Wilmshurst JM, Wiser SK, & Charman DJ. 2003. Reconstructing Holocene water tables in New Zealand using testate amoebae: differential preservation of tests and implications for the use of transfer functions. Holocene 13: 61-72.

201

References

Wilson GP & Lamb AL. 2012. an assessment of the utility of regional diatom-based tidal-level transfer functions. Journal of Quaternary Science 27: 360-370.

Woodland WA, Charman DJ, & Sims PC. 1998. Quantitative estimates of water tables and soil moisture in Holocene peatlands from testate amoebae. Holocene 8: 261-273.

Woodroffe SA. 2009a. Recognising subtidal foraminiferal assemblages: implications for quantitative sea-level reconstructions using a foraminifera-based transfer function. Journal of Quaternary Science 24: 215-223.

Woodroffe SA. 2009c. Testing models of mid to late Holocene sea-level change, North Queensland, Australia. Quaternary Science Reviews 28: 2474-2488.

Woodroffe SA & Long AJ. 2009b. Salt marshes as archives of recent relative sea level change in West Greenland. Quaternary Science Reviews 28: 1750-1761.

Zanting HA & ten Thij F. 2001. Langetermijnvisie Schelde-estuarium. Projectbureau LTV: 50 pp.

Zong YQ & Horton BP. 1999. Diatom-based tidal-level transfer functions as an aid in reconstructing Quaternary history of sea-level movements in the UK. Journal of Quaternary Science 14: 153-167.

202

Appendices

Appendix A Absolute testate amoebae counts of modern freshwater tidal marsh.

S01 S23 S07 S1 S04 S12 S02 s03 s40 s34 Sample number n n n 7 n n n n n n Arcella arenaria 1 0 0 0 0 0 0 0 10 0 Arcella arenaria var. sphagnicola 0 0 0 0 0 0 0 0 0 0 Arcella atava 0 1 0 0 0 0 0 0 0 0 Arcella catinus 0 0 0 1 0 0 0 1 0 0 Arcella discoides 0 0 0 0 0 0 0 0 0 0 Arcella discoides var. foreosa 0 0 0 0 0 0 0 9 3 3 Arcella discoides var. scutelliformis 0 0 0 0 0 0 0 0 0 0 Arcella gibbosa 0 0 0 0 0 0 0 0 0 0 Arcella hemisphaerica 0 0 1 0 0 0 0 0 0 0 Arcella sp 0 0 0 0 0 0 1 1 0 0 Arcella sp1 0 1 0 0 0 0 0 0 0 0 Arcella vulgaris 0 0 0 0 0 0 0 0 0 0 Arcella vulgaris var. penardi 0 0 0 0 0 0 0 0 0 0 Assulina muscorum 0 0 0 0 0 0 0 0 0 0 Assulina seminulum 0 0 0 0 0 0 0 0 0 0 Campascus sp 0 0 0 0 0 1 0 0 0 0 Centopyxis cryptostoma 0 1 0 0 0 0 0 0 0 0 Centropyxis aculeata 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata var. oblonga 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila var. aerophila 9 1 7 3 3 9 5 9 0 11 Centropyxis aerophila var. sphagnicola 0 0 9 0 0 1 0 0 1 1 Centropyxis aerophila var. sylvatica 4 0 0 0 0 3 0 0 0 0 Centropyxis arcelloides 0 0 0 0 0 0 0 0 0 0 Centropyxis constricta 0 1 0 0 0 0 0 0 0 0 Centropyxis ecornis 0 1 0 0 0 0 0 0 0 0 Centropyxis elongata 0 0 0 0 0 0 0 0 0 0 Centropyxis eurystoma 0 0 0 0 0 0 1 4 0 0 Centropyxis gibba 0 0 0 0 0 0 0 0 0 0 Centropyxis laevigata 2 0 3 0 1 0 0 0 0 2 Centropyxis minuta 0 0 0 0 0 0 0 0 0 0 Centropyxis plagiostoma 0 0 2 0 0 0 0 0 0 0 Centropyxis plagiostoma var. terricola 0 0 1 0 0 0 0 0 0 0 Centropyxis platystoma 0 0 0 0 0 0 0 0 0 0 Centropyxis sp 0 0 0 0 0 0 0 0 0 0 Cryptodifflugia compressa 0 0 0 0 3 0 4 0 0 0 Cyclopyxis kahli 0 0 0 0 0 0 0 0 19 0 Cyclopyxis kahli var. cyclostoma 0 0 0 0 0 1 0 0 0 0 Cyphoderia ampulla 0 0 1 0 1 0 1 0 0 2 Cyphoderia ampulla var. vitrae 0 0 0 0 0 0 0 0 0 0 Cyphoderia laevis 0 0 0 0 0 0 0 0 0 0 Cyphoderia littoralis 0 0 0 0 0 0 0 0 0 0 Difflugia amphoralis 0 0 0 0 0 0 0 0 0 0 Difflugia ampullula 1 0 0 0 0 0 0 0 0 0 Difflugia angulostoma 1 1 0 0 0 0 0 0 0 0

Appendix A

Difflugia bryophila 0 0 0 0 2 0 0 0 0 0 Difflugia decloitrei 0 0 0 0 0 0 0 0 0 0

s32 s35 s43 s33 s36 s14 s39 s05 s19 s44 Sample number n n n n n n n n n n Arcella arenaria 0 0 0 0 0 0 0 0 0 0 Arcella arenaria var. sphagnicola 0 0 0 0 0 0 0 0 0 0 Arcella atava 0 0 0 0 0 0 0 0 0 0 Arcella catinus 0 0 0 3 2 0 1 0 0 1 Arcella discoides 0 0 0 0 2 1 0 0 0 0 Arcella discoides var. foreosa 4 0 0 0 0 0 0 0 0 0 Arcella discoides var. scutelliformis 0 0 1 0 0 0 0 0 0 0 Arcella gibbosa 0 0 0 0 0 0 0 0 0 0 Arcella hemisphaerica 0 1 0 0 0 0 0 0 0 0 Arcella sp 0 1 0 0 0 0 0 0 0 0 Arcella sp1 0 0 0 0 0 0 0 0 0 0 Arcella vulgaris 0 0 1 0 0 0 0 0 0 0 Arcella vulgaris var. penardi 0 0 0 0 0 0 0 1 0 0 Assulina muscorum 0 0 1 0 0 0 1 0 0 0 Assulina seminulum 0 0 0 0 0 0 0 0 0 0 Campascus sp 0 0 0 0 0 0 0 0 0 0 Centopyxis cryptostoma 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata var. oblonga 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila var. aerophila 10 8 12 6 2 2 6 4 7 10 Centropyxis aerophila var. sphagnicola 1 1 0 0 1 2 2 3 0 0 Centropyxis aerophila var. sylvatica 0 0 0 0 0 0 0 0 7 2 Centropyxis arcelloides 0 0 0 0 0 0 0 0 0 1 Centropyxis constricta 0 0 0 1 0 0 0 1 0 0 Centropyxis ecornis 0 0 0 0 1 0 2 0 0 1 Centropyxis elongata 0 0 0 0 0 0 0 0 0 0 Centropyxis eurystoma 0 0 0 0 0 0 0 0 0 0 Centropyxis gibba 0 0 0 0 0 0 0 0 0 1 Centropyxis laevigata 0 1 0 0 0 0 0 0 0 0 Centropyxis minuta 0 0 1 0 0 0 0 0 0 0 Centropyxis plagiostoma 0 0 0 0 0 0 0 0 0 0 Centropyxis plagiostoma var. terricola 0 0 0 0 0 0 0 0 0 0 Centropyxis platystoma 0 0 0 0 0 0 0 0 0 0 Centropyxis sp 0 0 0 0 0 0 0 1 0 0 Cryptodifflugia compressa 0 0 0 1 0 0 0 0 2 0 Cyclopyxis kahli 0 0 0 0 2 2 2 0 0 0 Cyclopyxis kahli var. cyclostoma 0 0 0 0 0 0 0 0 0 0 Cyphoderia ampulla 0 0 0 1 0 0 0 0 1 0 Cyphoderia ampulla var. vitrae 0 3 0 1 0 0 0 0 0 0 Cyphoderia laevis 0 0 0 1 0 0 0 0 0 0 Cyphoderia littoralis 0 0 0 0 1 0 0 0 0 0 Difflugia amphoralis 0 0 0 0 0 0 0 0 0 0 Difflugia ampullula 0 0 0 0 0 0 0 0 0 0 Difflugia angulostoma 0 0 0 0 0 0 0 0 0 0 Difflugia bryophila 0 0 0 0 0 0 0 0 0 0 Difflugia decloitrei 0 0 0 0 0 0 0 0 0 0

204

Appendix A

s20 s06 s18 s11 s16 s08 s09 s13 s15 s10 Sample number n n n n n n n n n n Arcella arenaria 0 0 0 0 0 0 0 0 0 0 Arcella arenaria var. sphagnicola 0 0 0 0 0 6 0 0 0 0 Arcella atava 0 0 0 0 0 0 0 0 0 0 Arcella catinus 2 3 0 2 0 0 0 1 0 0 Arcella discoides 0 0 0 0 0 0 0 0 0 0 Arcella discoides var. foreosa 0 0 0 0 0 0 0 0 0 0 Arcella discoides var. scutelliformis 1 0 0 0 1 0 0 0 0 0 Arcella gibbosa 0 0 0 0 0 0 0 0 0 0 Arcella hemisphaerica 0 0 2 0 0 0 0 0 0 0 Arcella sp 0 0 0 0 0 0 0 0 0 0 Arcella sp1 0 0 0 0 0 0 0 0 0 0 Arcella vulgaris 0 0 0 0 0 0 0 0 0 0 Arcella vulgaris var. penardi 0 0 0 0 0 0 0 0 0 0 Assulina muscorum 0 0 0 0 0 0 0 0 0 0 Assulina seminulum 0 0 1 0 1 1 0 0 1 0 Campascus sp 0 0 0 0 0 0 0 0 0 0 Centopyxis cryptostoma 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata 0 0 0 0 0 0 0 0 0 1 Centropyxis aculeata var. oblonga 0 1 0 0 0 0 0 1 0 0 Centropyxis aerophila var. aerophila 5 5 5 6 3 15 11 5 1 4 Centropyxis aerophila var. sphagnicola 2 4 0 2 1 0 2 0 0 3 Centropyxis aerophila var. sylvatica 0 0 0 0 0 0 0 0 0 0 Centropyxis arcelloides 0 0 0 0 0 0 0 0 0 0 Centropyxis constricta 0 0 0 0 2 0 0 0 0 0 Centropyxis ecornis 4 0 2 2 0 2 0 1 0 5 Centropyxis elongata 0 0 0 0 0 0 0 0 0 0 Centropyxis eurystoma 0 0 0 0 0 0 0 0 0 0 Centropyxis gibba 0 0 0 0 0 0 0 0 0 0 Centropyxis laevigata 0 0 0 0 0 0 0 0 0 0 Centropyxis minuta 0 0 0 0 0 2 0 0 0 0 Centropyxis plagiostoma 0 0 0 0 0 0 0 0 0 0 Centropyxis plagiostoma var. terricola 0 0 0 0 0 0 0 0 0 0 Centropyxis platystoma 0 0 0 0 0 0 0 0 0 0 Centropyxis sp 0 0 0 0 0 0 0 0 0 0 Cryptodifflugia compressa 2 0 1 0 0 0 0 0 0 0 Cyclopyxis kahli 0 1 0 0 1 1 0 0 0 0 Cyclopyxis kahli var. cyclostoma 0 0 0 0 0 0 0 0 0 0 Cyphoderia ampulla 0 0 0 0 0 2 2 2 4 7 Cyphoderia ampulla var. vitrae 0 0 0 0 0 0 0 0 0 0 Cyphoderia laevis 0 0 0 0 0 0 0 0 0 2 Cyphoderia littoralis 0 0 0 0 0 0 0 0 0 0 Difflugia amphoralis 0 0 1 0 0 0 0 0 0 0 Difflugia ampullula 0 0 0 0 0 0 0 0 0 0 Difflugia angulostoma 0 0 0 0 0 0 0 0 0 0 Difflugia bryophila 0 0 0 0 0 0 0 0 0 0 Difflugia decloitrei 0 0 0 0 0 0 1 0 0 0

205

Appendix A

Sample number s28n s37n s29n s27n s51 s49 s48 s50 s47 s52 Arcella arenaria 0 0 0 0 2 1 0 2 5 5 Arcella arenaria var. sphagnicola 0 0 0 0 0 0 0 0 0 0 Arcella atava 0 0 0 0 0 0 0 0 0 0 Arcella catinus 0 0 0 0 1 0 0 0 0 3 Arcella discoides 0 0 0 0 0 0 0 0 0 0 Arcella discoides var. foreosa 0 0 0 0 0 0 0 0 0 0 Arcella discoides var. scutelliformis 0 0 0 0 0 0 0 0 0 0 Arcella gibbosa 0 0 0 0 0 0 0 0 0 0 Arcella hemisphaerica 4 1 0 0 0 0 0 0 0 0 Arcella sp 0 0 0 0 0 0 0 0 0 0 Arcella sp1 0 0 0 0 0 0 0 0 0 0 Arcella vulgaris 0 0 0 0 0 0 0 0 0 0 Arcella vulgaris var. penardi 0 0 0 0 0 0 0 0 0 0 Assulina muscorum 0 0 0 0 0 0 0 0 0 0 Assulina seminulum 0 0 0 0 0 0 0 0 0 0 Campascus sp 0 0 0 0 0 0 0 0 0 0 Centopyxis cryptostoma 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata var. oblonga 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila var. aerophila 6 7 7 0 4 10 5 9 20 6 Centropyxis aerophila var. sphagnicola 0 1 0 0 14 5 11 6 0 6 Centropyxis aerophila var. sylvatica 0 0 0 0 0 0 0 0 1 0 Centropyxis arcelloides 0 0 0 0 0 0 0 0 0 0 Centropyxis constricta 0 0 0 0 0 0 0 0 0 0 Centropyxis ecornis 0 7 0 0 0 0 3 0 2 2 Centropyxis elongata 0 0 0 0 0 0 0 0 0 0 Centropyxis eurystoma 0 1 0 0 8 0 0 3 0 8 Centropyxis gibba 0 0 0 0 0 0 0 0 0 0 Centropyxis laevigata 0 0 0 0 0 0 0 0 0 1 Centropyxis minuta 0 1 0 0 0 8 0 0 0 0 Centropyxis plagiostoma 0 0 0 0 0 0 0 0 0 0 Centropyxis plagiostoma var. terricola 0 0 0 0 0 0 0 0 0 0 Centropyxis platystoma 0 0 0 0 0 0 0 2 0 0 Centropyxis sp 0 0 0 0 0 0 0 0 0 0 Cryptodifflugia compressa 0 0 0 0 0 0 0 0 0 0 Cyclopyxis kahli 0 0 1 0 6 3 0 0 6 10 Cyclopyxis kahli var. cyclostoma 0 0 0 0 0 0 0 0 0 0 Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 0 Cyphoderia ampulla var. vitrae 0 0 0 0 0 0 0 0 0 0 Cyphoderia laevis 0 0 0 0 0 0 0 0 0 0 Cyphoderia littoralis 0 2 0 0 0 0 0 0 0 0 Difflugia amphoralis 0 1 1 0 0 0 0 0 0 0 Difflugia ampullula 0 0 0 0 0 0 0 0 0 0 Difflugia angulostoma 0 8 0 0 0 0 0 0 0 0 Difflugia bryophila 0 0 0 0 0 0 0 0 0 0 Difflugia decloitrei 0 0 0 0 0 0 0 0 0 0

206

Appendix A

Sample number s53 s55 s54 s56 Arcella arenaria 3 7 3 4 Arcella arenaria var. sphagnicola 0 0 0 0 Arcella atava 0 0 0 0 Arcella catinus 3 0 1 1 Arcella discoides 0 0 0 0 Arcella discoides var. foreosa 0 0 0 0 Arcella discoides var. scutelliformis 0 0 0 0 Arcella gibbosa 0 0 1 0 Arcella hemisphaerica 0 0 0 0 Arcella sp 0 0 0 0 Arcella sp1 0 0 0 0 Arcella vulgaris 0 0 0 0 Arcella vulgaris var. penardi 0 0 0 0 Assulina muscorum 0 0 1 2 Assulina seminulum 0 0 0 0 Campascus sp 0 0 0 0 Centopyxis cryptostoma 0 0 0 0 Centropyxis aculeata 0 0 0 0 Centropyxis aculeata var. oblonga 0 0 0 0 Centropyxis aerophila var. aerophila 9 6 15 6 Centropyxis aerophila var. sphagnicola 2 1 4 0 Centropyxis aerophila var. sylvatica 0 0 0 0 Centropyxis arcelloides 0 0 0 0 Centropyxis constricta 0 0 0 0 Centropyxis ecornis 9 2 2 0 Centropyxis elongata 0 0 0 0 Centropyxis eurystoma 13 1 1 3 Centropyxis gibba 0 0 0 0 Centropyxis laevigata 0 2 0 0 Centropyxis minuta 0 0 0 1 Centropyxis plagiostoma 0 0 0 0 Centropyxis plagiostoma var. terricola 0 0 0 0 Centropyxis platystoma 0 1 0 2 Centropyxis sp 0 0 0 0 Cryptodifflugia compressa 0 0 0 0 Cyclopyxis kahli 5 1 4 5 Cyclopyxis kahli var. cyclostoma 0 0 0 0 Cyphoderia ampulla 0 0 0 0 Cyphoderia ampulla var. vitrae 0 0 0 0 Cyphoderia laevis 0 0 0 0 Cyphoderia littoralis 0 0 0 0 Difflugia amphoralis 0 0 0 0 Difflugia ampullula 0 0 0 0 Difflugia angulostoma 0 0 0 0 Difflugia bryophila 0 0 0 0 Difflugia decloitrei 0 0 0 0

207

Appendix A

S01 S23 S07 S04 S12 S02 s03 s40 s34 Sample number n n n S17 n n n n n n Difflugia glans 0 0 0 0 1 0 0 0 0 0 Difflugia globulosa 1 3 4 3 0 1 0 0 23 4 Difflugia globulus 11 3 9 5 11 12 9 8 45 15 Difflugia lucida 0 0 0 0 0 0 0 0 0 0 Difflugia mamillaris 0 0 0 0 0 0 0 0 0 0 Difflugia manicata 1 0 0 0 0 0 2 9 0 0 Difflugia maxilabiosa var. minima 0 0 0 0 0 0 0 0 0 0 Difflugia mica 0 0 0 0 0 0 0 0 0 0 Difflugia pristis 10 0 14 7 7 2 11 2 5 2 Difflugia pulex 0 0 0 0 1 0 0 0 0 0 Difflugia stautii 0 0 0 0 0 0 0 0 0 0 Difflugia tenuis 0 0 0 0 1 1 0 0 0 2 Difflugiella oviformis 3 0 5 13 6 6 8 1 2 4 Euglypha ciliata 0 1 2 0 0 0 0 0 0 0 Euglypha ciliata var. glabra 0 1 0 0 1 0 0 0 0 0 Euglypha cristata 0 0 2 0 0 0 0 0 0 0 Euglypha cristata var. lanceolata 0 0 0 0 0 0 0 0 0 0 Euglypha cristata var. major 0 0 2 0 0 0 0 0 0 0 Euglypha denticulata 0 1 0 0 0 0 0 0 0 0 Euglypha dolioliformis 0 0 0 0 0 0 0 0 0 0 Euglypha filifera var. spinosa 0 0 1 0 0 0 0 0 0 0 Euglypha laevis 0 3 2 3 0 0 0 0 0 1 Euglypha polylepsis 0 0 0 0 0 0 0 0 2 0 Euglypha rotunda 8 1 5 3 4 16 14 9 5 19 Euglypha rotunda var. obliqua 0 0 0 0 0 0 0 0 0 0 Euglypha rotunda var.dorsalis 0 0 0 0 0 0 0 0 0 0 Euglypha sp 0 0 0 0 0 0 0 0 0 0 Euglypha tuberculata 0 4 3 0 0 2 0 0 0 1 Euglypha tuberculata var. minor 0 0 0 0 0 0 0 0 0 0 Euglypha tuberculata var. subcylindricae 0 0 0 0 0 4 4 0 0 0 Heleopera petricola 0 0 0 0 0 0 4 0 0 0 Heleopera sylvatica 0 0 0 0 0 0 0 0 0 0 Hyalosphenia elegans 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 4 2 0 0 Hyalosphenia ovalis 0 0 0 1 0 0 0 0 0 0 Hyalosphenia sp 1 0 0 0 0 0 1 1 0 0 0 Paraquadrullella irregularis 1 0 0 1 1 1 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 Placosista glabra var. minima 1 0 0 0 0 0 0 0 0 0 Plagiopyxis penardi 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 24 0 5 24 8 18 5 10 7 6 Tracheleuglypha sp1 0 0 0 0 0 0 0 0 0 0 Trinema chardezii 0 0 0 0 0 0 0 0 0 0 Trinema complanatum 0 0 0 0 0 0 0 0 2 0 Trinema enchelys 45 4 47 35 30 70 26 26 12 24 Trinema lineare 24 4 32 51 69 5 50 59 15 53 Trinema lineare var. truncatum 3 0 0 0 0 0 0 0 0 0 Trinema penardi 0 0 0 0 0 0 0 0 0 0 Valkanovia elegans 0 0 0 0 0 0 0 0 0 0 Total counts 150 32 157 150 150 154 150 150 151 150

208

Appendix A

s32 s35 s43 s33 s36 s14 s39 s05 s19 s44 Sample number n n n n n n n n n n Difflugia glans 0 1 0 0 0 0 0 0 0 0 Difflugia globulosa 2 6 3 3 3 5 3 2 2 0 Difflugia globulus 16 9 7 15 8 17 4 3 12 6 Difflugia lucida 0 1 0 0 0 0 0 0 0 0 Difflugia mamillaris 0 0 0 0 0 0 0 0 0 0 Difflugia manicata 0 0 0 0 0 0 0 0 0 0 Difflugia maxilabiosa var. minima 0 0 0 0 0 0 0 0 0 0 Difflugia mica 4 3 0 0 1 0 0 0 0 0 Difflugia pristis 5 0 0 0 1 3 10 1 3 8 Difflugia pulex 0 0 0 0 0 0 0 0 0 0 Difflugia stautii 0 0 0 0 0 0 0 0 0 0 Difflugia tenuis 4 4 1 0 0 0 2 4 5 0 Difflugiella oviformis 11 10 4 5 18 3 5 3 2 12 Euglypha ciliata 0 0 0 0 0 0 0 0 0 0 Euglypha ciliata var. glabra 0 0 0 0 0 0 0 0 0 0 Euglypha cristata 0 0 0 0 0 0 0 0 0 0 Euglypha cristata var. lanceolata 0 0 0 0 0 0 0 0 0 2 Euglypha cristata var. major 0 0 0 0 0 0 0 0 0 0 Euglypha denticulata 0 0 0 0 0 0 0 0 0 0 Euglypha dolioliformis 0 0 0 0 0 0 0 0 0 0 Euglypha filifera var. spinosa 0 0 0 0 0 0 0 0 0 0 Euglypha laevis 0 0 0 0 0 0 0 0 0 0 Euglypha polylepsis 0 0 0 0 0 0 1 0 0 0 Euglypha rotunda 15 9 16 9 13 16 6 9 11 20 Euglypha rotunda var. obliqua 0 0 0 0 0 0 0 0 0 0 Euglypha rotunda var.dorsalis 0 0 0 0 0 0 0 0 0 0 Euglypha sp 0 0 0 0 0 0 0 0 0 0 Euglypha tuberculata 2 0 2 1 4 4 0 5 0 3 Euglypha tuberculata var. minor 0 0 0 2 3 4 0 0 1 0 Euglypha tuberculata var. subcylindricae 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 0 Heleopera sylvatica 0 0 0 0 0 0 0 0 0 0 Hyalosphenia elegans 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 1 1 10 1 1 0 0 1 3 0 Hyalosphenia ovalis 0 0 0 0 0 0 0 0 0 0 Hyalosphenia sp 1 0 0 0 0 0 0 0 0 0 0 Paraquadrullella irregularis 0 2 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 Placosista glabra var. minima 0 0 0 0 0 0 0 0 0 0 Plagiopyxis penardi 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 9 28 24 15 12 20 21 9 8 11 Tracheleuglypha sp1 0 0 0 0 1 0 1 0 0 0 Trinema chardezii 0 0 0 1 0 0 0 0 0 0 Trinema complanatum 0 0 0 0 0 0 0 0 0 0 Trinema enchelys 7 1 35 20 10 8 24 22 13 16 Trinema lineare 60 60 33 64 61 60 17 74 70 55 Trinema lineare var. truncatum 0 0 0 0 3 3 0 7 3 3 Trinema penardi 0 0 0 0 0 0 0 0 0 0 Valkanovia elegans 0 0 0 0 0 0 0 0 0 0 Total counts 151 150 151 150 150 150 108 150 150 152 209

Appendix A

s20 s06 s18 s11 s16 s08 s09 s13 s15 s10 Sample number n n n n n n n n n n Difflugia glans 0 0 0 0 0 0 1 0 0 0 Difflugia globulosa 2 2 1 3 1 6 1 1 2 0 Difflugia globulus 6 4 2 13 0 8 11 8 1 3 Difflugia lucida 2 0 0 0 0 0 0 0 0 0 Difflugia mamillaris 0 0 0 0 0 0 0 0 0 0 Difflugia manicata 0 1 0 0 0 1 0 0 0 0 Difflugia maxilabiosa var. minima 0 0 0 0 0 0 0 0 0 0 Difflugia mica 0 0 3 0 0 0 0 0 0 0 Difflugia pristis 4 5 18 7 3 11 20 13 27 19 Difflugia pulex 1 0 0 0 0 0 0 0 0 0 Difflugia stautii 0 0 0 1 0 0 0 0 0 0 Difflugia tenuis 0 0 6 6 0 7 0 2 1 9 Difflugiella oviformis 4 1 9 2 2 10 5 0 8 6 Euglypha ciliata 0 0 0 0 0 0 0 0 0 0 Euglypha ciliata var. glabra 0 0 0 0 0 0 0 0 0 0 Euglypha cristata 0 0 0 0 0 0 0 0 0 0 Euglypha cristata var. lanceolata 0 0 0 0 0 0 0 0 0 0 Euglypha cristata var. major 0 0 0 0 0 0 0 0 0 0 Euglypha denticulata 0 1 0 0 0 0 0 0 0 0 Euglypha dolioliformis 0 0 0 6 0 1 2 0 0 0 Euglypha filifera var. spinosa 0 0 0 0 0 0 0 0 0 0 Euglypha laevis 0 0 0 0 0 0 0 0 0 0 Euglypha polylepsis 3 2 0 0 0 0 0 0 0 0 Euglypha rotunda 9 15 8 19 10 10 16 15 17 5 Euglypha rotunda var. obliqua 0 0 0 0 0 0 0 0 0 0 Euglypha rotunda var.dorsalis 0 0 0 0 0 0 0 0 0 0 Euglypha sp 0 0 0 0 0 0 0 0 0 0 Euglypha tuberculata 2 2 2 1 1 0 0 0 1 1 Euglypha tuberculata var. minor 2 0 0 0 0 0 0 0 0 0 Euglypha tuberculata var. subcylindricae 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 0 Heleopera sylvatica 0 0 0 0 0 0 0 0 0 0 Hyalosphenia elegans 0 0 1 0 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 2 0 0 0 Hyalosphenia ovalis 0 0 0 0 0 0 0 0 0 0 Hyalosphenia sp 1 0 0 0 0 0 0 0 0 0 0 Paraquadrullella irregularis 1 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 1 Placosista glabra var. minima 0 0 0 0 0 0 0 0 0 0 Plagiopyxis penardi 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 16 16 6 19 14 8 14 19 15 15 Tracheleuglypha sp1 0 0 0 0 0 0 0 0 0 0 Trinema chardezii 0 0 0 0 0 0 0 0 0 0 Trinema complanatum 0 0 0 3 0 0 0 0 0 0 Trinema enchelys 21 14 1 17 12 6 23 18 6 14 Trinema lineare 61 73 80 40 96 53 39 59 65 55 Trinema lineare var. truncatum 0 0 1 1 2 0 0 5 1 0 Trinema penardi 0 0 0 0 0 0 0 0 0 0 Valkanovia elegans 0 0 0 0 0 0 0 0 0 0 Total counts 150 150 150 150 150 150 150 150 150 150

210

Appendix A

Sample number s28n s37n s29n s27n s51 s49 s48 s50 s47 s52 Difflugia glans 0 0 0 0 0 0 0 0 0 0 Difflugia globulosa 0 0 5 0 5 0 1 5 11 3 Difflugia globulus 2 5 5 0 21 19 6 17 31 11 Difflugia lucida 0 0 0 0 0 0 5 3 0 0 Difflugia mamillaris 0 1 0 0 0 0 0 0 0 0 Difflugia manicata 0 0 0 0 0 0 0 0 0 0 Difflugia maxilabiosa var. minima 0 0 0 0 0 0 0 0 0 0 Difflugia mica 0 0 0 0 0 0 0 0 0 0 Difflugia pristis 7 19 4 1 3 1 11 1 1 3 Difflugia pulex 0 0 0 0 0 0 0 0 0 0 Difflugia stautii 0 0 0 0 0 0 0 0 0 0 Difflugia tenuis 1 5 2 0 0 1 0 0 0 0 Difflugiella oviformis 0 7 1 0 5 6 4 4 1 0 Euglypha ciliata 0 0 0 0 0 0 0 0 0 0 Euglypha ciliata var. glabra 0 0 0 0 0 0 0 0 0 0 Euglypha cristata 0 0 0 0 0 0 0 0 0 0 Euglypha cristata var. lanceolata 0 0 0 0 0 0 0 0 0 0 Euglypha cristata var. major 0 0 0 0 0 0 0 0 0 0 Euglypha denticulata 0 0 0 0 0 0 0 0 0 0 Euglypha dolioliformis 1 0 0 0 0 0 0 1 0 3 Euglypha filifera var. spinosa 0 0 0 0 0 0 0 0 0 0 Euglypha laevis 0 0 0 0 0 0 0 0 0 0 Euglypha polylepsis 0 0 0 0 2 0 2 2 0 0 Euglypha rotunda 6 15 11 0 5 10 11 4 2 7 Euglypha rotunda var. obliqua 0 0 0 0 0 0 0 2 0 0 Euglypha rotunda var.dorsalis 0 0 0 0 0 0 0 1 1 1 Euglypha sp 0 0 0 0 0 1 0 0 0 0 Euglypha tuberculata 0 3 0 0 0 0 1 0 0 0 Euglypha tuberculata var. minor 0 0 0 0 0 0 0 0 0 0 Euglypha tuberculata var. subcylindricae 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 0 Heleopera sylvatica 0 0 0 0 0 0 0 0 0 1 Hyalosphenia elegans 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 1 1 0 0 1 0 Hyalosphenia ovalis 0 0 0 0 0 0 0 0 0 0 Hyalosphenia sp 1 0 0 0 0 0 0 0 0 0 0 Paraquadrullella irregularis 0 2 0 0 1 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 Placosista glabra var. minima 0 0 0 0 0 0 0 0 0 0 Plagiopyxis penardi 0 0 0 0 0 0 0 4 3 4 Tracheleuglypha dentata 16 17 25 2 14 7 16 4 10 11 Tracheleuglypha sp1 0 0 0 0 0 0 0 0 0 0 Trinema chardezii 0 0 0 0 0 0 0 0 0 0 Trinema complanatum 0 0 0 0 11 3 0 4 6 14 Trinema enchelys 9 5 13 0 13 30 33 38 19 44 Trinema lineare 98 41 71 0 32 32 37 34 29 5 Trinema lineare var. truncatum 0 1 4 0 2 12 4 4 1 2 Trinema penardi 0 0 0 0 0 0 0 0 0 0 Valkanovia elegans 0 0 0 0 0 0 0 0 0 0 Total counts 150 150 150 3 150 150 150 150 150 150 211

Appendix A

Sample number s53 s55 s54 s56 Difflugia glans 0 0 0 0 Difflugia globulosa 3 0 18 2 Difflugia globulus 15 5 26 12 Difflugia lucida 0 1 0 0 Difflugia mamillaris 0 0 0 0 Difflugia manicata 0 0 0 0 Difflugia maxilabiosa var. minima 0 0 0 1 Difflugia mica 0 0 0 0 Difflugia pristis 3 4 2 1 Difflugia pulex 0 0 0 0 Difflugia stautii 0 0 0 0 Difflugia tenuis 0 1 0 0 Difflugiella oviformis 0 2 4 0 Euglypha ciliata 0 0 0 0 Euglypha ciliata var. glabra 0 0 0 0 Euglypha cristata 0 0 0 0 Euglypha cristata var. lanceolata 0 0 0 0 Euglypha cristata var. major 0 0 0 0 Euglypha denticulata 0 0 0 0 Euglypha dolioliformis 2 0 3 0 Euglypha filifera var. spinosa 0 0 0 0 Euglypha laevis 0 0 0 0 Euglypha polylepsis 4 0 0 0 Euglypha rotunda 13 11 6 14 Euglypha rotunda var. obliqua 0 0 1 1 Euglypha rotunda var.dorsalis 0 0 0 0 Euglypha sp 0 0 0 0 Euglypha tuberculata 0 0 0 1 Euglypha tuberculata var. minor 0 0 0 0 Euglypha tuberculata var. subcylindricae 0 0 0 0 Heleopera petricola 0 0 2 0 Heleopera sylvatica 0 0 0 0 Hyalosphenia elegans 0 0 0 0 Hyalosphenia minuta 0 0 1 0 Hyalosphenia ovalis 0 0 0 0 Hyalosphenia sp 1 0 0 0 0 Paraquadrullella irregularis 1 0 0 0 Paulinella chromatophora 0 0 0 0 Placosista glabra var. minima 0 0 0 0 Plagiopyxis penardi 0 0 0 1 Tracheleuglypha dentata 16 10 19 12 Tracheleuglypha sp1 0 0 0 0 Trinema chardezii 0 0 0 0 Trinema complanatum 4 2 8 0 Trinema enchelys 18 43 3 26 Trinema lineare 25 50 23 55 Trinema lineare var. truncatum 0 0 0 0 Trinema penardi 1 0 0 0 Valkanovia elegans 1 0 2 0 Total counts 150 150 150 150

212

Appendix B

Appendix B Absolute palaeo testate amoebae counts of the old marsh of the freshwater tidal marsh (Notelaar).

Depth(cm) 25 15 5 45

Arcella arenaria 0 0 0 2

Centropyxis aculeata var. oblonga 7 3 3 2

Centropyxis aerophila aerophila 14 12 18 14

Centropyxis aerophila sphagnicola 0 3 4 1

Centropyxis constricta 0 3 1 0

Centropyxis ecornis 3 2 5 8

Centropyxis elongata 0 1 0 0

centropyxis laevigata 2 1 0 1

Centropyxis minuta 0 1 0 2

Centropyxis platystoma 6 16 1 0

cyclopyxis kahli 1 1 1 2

cyphoderia ampulla 0 0 1 0

Difflugia globularis 0 1 3 0

Difflugia globulosa 0 0 0 6

Difflugia globulus 14 8 11 10

Difflugia lucida 0 0 0 5

Difflugia manicata 5 1 5 0

Difflugia minuta 0 0 0 1

difflugia pristis 3 7 8 41

Difflugia tenuis 4 0 8 2

Difflugiella oviformis 4 2 5 17

Euglypha rotunda 7 6 2 5

Euglypha strigosa var. glabra 0 1 0 2

Heleopera petricola 0 1 1 3

Paraquadrulla irregularis 0 1 0 0

Tracheleuglypha dentata 10 8 10 15

Trinema enchelys 52 42 21 3

Trinema lineare 20 27 41 8

Trinema lineare var.truncatum 0 2 1 0

Total counts 152 150 150 150

213

Appendix C

Appendix C Absolute testate amoebae counts of Modern brackish samples of Groot Buitenschoor.

Sample number s01 s03 s07 s09 s11 s13 s15 s17 s19 s21 s22 s24 Arcella catinus 0 0 1 0 0 0 0 0 0 0 0 0 Arcella arenaria 0 0 0 0 0 0 0 0 0 0 1 0 Assulina muscorum 3 0 0 1 0 0 0 0 1 0 0 0 Campascus minutus 0 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 6 2 8 2 26 1 3 1 7 0 5 4 Centropyxis aerophila sphagnicola 13 0 0 3 0 2 1 12 5 5 9 15 Centropyxis elongata 0 0 1 0 0 0 0 0 0 0 0 0 Centropyxis ecornis 0 0 1 0 0 0 0 0 2 0 0 2 Centropyxis laevigata 1 0 0 0 2 0 1 3 0 1 9 0 Centropyxis platystoma 0 0 0 0 0 0 0 0 1 0 0 0 Centropyxiella arenaria 0 0 0 0 0 0 0 0 0 0 0 0 Corythion dubium 1 0 0 1 1 0 1 0 0 0 1 0 Cyclopyxis kahli 0 4 2 1 0 0 1 6 4 4 2 1 Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 0 0 0 Cyphoderia littoralis 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia elegans parva 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia globularis 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia globulosa 4 2 3 16 2 0 9 11 7 12 1 11 Difflugia globulus 25 28 40 47 10 39 26 17 24 27 28 16 Difflugia lucida 0 1 0 0 0 0 3 0 0 0 0 0 Difflugia pristis 0 0 1 0 0 0 0 0 0 0 0 0 Difflugia tenuis 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia sp1 0 0 0 0 0 0 0 0 0 0 0 0 Difflugiella oviformis 0 1 5 1 13 0 0 2 3 0 5 3 Euglypha dolioliformis 0 0 0 0 0 0 0 1 0 0 0 0 Euglypha polyepsis 2 0 0 5 0 0 3 0 3 0 0 5 Euglypha strigosa var. Glabra 3 5 0 0 1 0 2 1 0 0 0 0 Euglypha rotunda 3 8 10 3 19 15 8 3 1 5 3 5 Euglypha cristata 0 0 0 0 0 0 0 1 0 0 0 0 Heleopera petricola 0 0 0 1 0 0 0 0 0 1 0 0 Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0 1 0 Paraquadrula irregularis 0 0 0 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 0 0 Plagiopyxis declivis 0 10 1 1 1 2 2 5 10 4 5 2 Pseudocorythion accutum 0 0 0 0 0 0 0 0 0 0 0 0 Pseudocorythion wailesi 0 0 0 0 0 0 0 0 0 0 0 0 Pseudohyalosphenia 0 0 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 15 20 25 14 39 18 24 24 37 39 32 33 Trinema complanatum 19 33 11 10 3 31 17 27 11 12 13 9 Trinema enchelys 24 21 38 31 22 33 28 17 17 32 24 14 Trinema lineare 20 15 5 13 19 7 21 18 17 7 11 30 Trinema lineare var. truncatum 0 0 0 0 1 2 0 2 0 1 0 0 Trinema penardi 11 0 0 0 0 0 0 0 0 0 0 0 Total counts 150 150 152 150 159 150 150 151 150 150 150 150

214

Appendix C

Sample number s27 s29 s28 s30 s31 s32 s33 s34 s35 s36 s37 s38 Arcella catinus 0 0 0 0 0 0 0 0 1 1 0 0 Arcella arenaria 0 0 0 0 0 0 0 0 0 0 0 0 Assulina muscorum 0 0 0 0 0 0 0 0 0 0 0 0 Campascus minutus 0 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 3 12 6 6 5 0 4 3 3 4 2 13 Centropyxis aerophila sphagnicola 6 4 9 0 2 3 3 1 0 0 2 10 Centropyxis elongata 0 0 0 0 0 0 4 0 0 0 0 0 Centropyxis ecornis 0 0 0 0 0 2 0 0 1 0 0 1 Centropyxis laevigata 2 10 15 0 3 0 6 7 6 5 2 0 Centropyxis platystoma 0 0 0 0 0 0 0 0 0 0 0 0 Centropyxiella arenaria 0 0 0 0 0 0 0 0 0 0 0 0 Corythion dubium 0 0 0 0 0 0 1 0 0 0 0 1 Cyclopyxis kahli 0 2 0 0 4 0 0 0 1 0 1 2 Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 0 0 0 Cyphoderia littoralis 0 0 1 0 0 0 0 0 0 0 0 0 Difflugia elegans parva 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia globularis 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia globulosa 3 0 2 0 15 0 0 2 6 11 3 10 Difflugia globulus 23 23 15 4 48 6 17 9 22 30 34 43 Difflugia lucida 0 0 0 1 0 0 1 0 0 0 0 3 Difflugia pristis 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia tenuis 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia sp1 0 0 0 0 0 0 0 0 0 0 0 0 Difflugiella oviformis 0 18 1 6 8 2 8 3 0 0 4 0 Euglypha dolioliformis 0 0 0 0 0 1 0 0 0 0 0 0 Euglypha polyepsis 2 0 1 5 0 0 0 0 7 0 0 0 Euglypha strigosa var. Glabra 0 0 0 0 0 0 2 1 0 1 0 0 Euglypha rotunda 4 3 12 23 4 39 21 27 7 4 13 6 Euglypha cristata 0 0 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 1 0 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0 0 0 Paraquadrula irregularis 0 1 1 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 0 0 Plagiopyxis declivis 1 4 2 0 0 0 2 3 15 4 5 11 Pseudocorythion accutum 0 0 0 0 0 0 0 0 0 0 0 0 Pseudocorythion wailesi 0 0 0 0 0 0 0 0 0 0 0 0 Pseudohyalosphenia 0 0 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 34 40 39 27 7 35 39 57 51 68 48 31 Trinema complanatum 14 5 3 6 6 18 1 3 0 1 0 1 Trinema enchelys 54 14 22 56 34 15 17 14 18 18 19 8 Trinema lineare 2 14 20 16 14 29 28 19 12 3 17 10 Trinema lineare var. truncatum 1 0 1 0 0 0 1 1 0 0 0 0 Trinema penardi 0 0 0 0 0 0 0 0 0 0 0 0 Total counts 150 150 150 150 150 150 155 150 150 150 150 150

215

Appendix C

Sample number sx s39 s40 s41 s42 s43 s16 s08 s10 s02 s23 s06 Arcella catinus 1 4 0 0 0 0 0 0 0 0 0 0 Arcella arenaria 0 0 0 0 0 0 0 0 0 0 0 0 Assulina muscorum 0 0 0 0 0 0 0 0 0 0 0 0 Campascus minutus 0 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 8 16 32 11 23 17 2 0 8 9 6 6 Centropyxis aerophila sphagnicola 8 3 0 7 13 5 11 0 0 2 3 1 Centropyxis elongata 0 0 0 0 0 1 0 0 0 0 0 0 Centropyxis ecornis 0 0 0 5 0 0 1 0 0 0 0 0 Centropyxis laevigata 5 15 6 0 6 6 0 1 3 3 0 4 Centropyxis platystoma 0 0 0 1 0 0 0 0 0 0 0 0 Centropyxiella arenaria 0 0 0 0 0 0 0 0 0 0 0 0 Corythion dubium 0 0 1 1 0 0 0 0 1 1 0 1 Cyclopyxis kahli 0 0 0 1 1 4 10 0 1 0 0 1 Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 3 0 3 Cyphoderia littoralis 0 0 0 0 0 0 0 0 0 1 0 0 Difflugia elegans parva 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia globularis 0 0 0 0 0 0 0 0 0 1 0 0 Difflugia globulosa 5 12 7 9 18 12 2 2 13 8 8 3 Difflugia globulus 23 30 14 27 35 18 5 1 7 2 4 0 Difflugia lucida 0 1 1 1 3 4 0 1 2 0 0 0 Difflugia pristis 2 0 1 1 1 0 1 0 2 0 2 0 Difflugia tenuis 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia sp1 0 0 0 0 0 0 0 0 0 0 0 0 Difflugiella oviformis 4 1 3 3 2 8 4 0 1 0 3 1 Euglypha dolioliformis 0 0 0 0 0 1 0 0 0 0 2 0 Euglypha polyepsis 3 0 1 6 0 0 11 0 0 2 0 0 Euglypha strigosa var. Glabra 2 0 0 0 1 0 0 1 0 0 0 4 Euglypha rotunda 11 3 17 8 1 5 15 5 11 2 10 4 Euglypha cristata 0 0 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 1 0 0 0 0 0 Paraquadrula irregularis 0 0 0 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 0 0 Plagiopyxis declivis 1 14 0 25 3 12 16 3 7 3 6 11 Pseudocorythion accutum 0 0 0 0 0 0 0 0 1 3 10 11 Pseudocorythion wailesi 0 0 0 0 0 0 0 0 0 0 1 0 Pseudohyalosphenia 0 0 0 0 0 0 0 0 9 0 0 1 Tracheleuglypha dentata 45 37 49 31 34 41 41 11 52 37 30 29 Trinema complanatum 0 0 1 1 0 1 0 0 0 0 0 0 Trinema enchelys 12 8 6 7 7 1 7 4 12 27 33 10 Trinema lineare 20 6 12 6 2 14 22 4 21 47 32 60 Trinema lineare var. truncatum 0 0 0 0 0 0 1 0 0 0 0 0 Trinema penardi 0 0 0 0 0 0 0 0 0 0 0 0 Total counts 150 150 151 151 150 150 150 33 151 151 150 150

216

Appendix C

Sample number s05 s14 s04 s12 s25 s44 s45 s51 s47 s46 s56 s62 Arcella catinus 1 0 2 0 0 0 0 0 0 3 0 0 Arcella arenaria 0 1 0 0 0 0 0 0 0 0 0 0 Assulina muscorum 0 0 0 0 0 0 1 0 0 0 0 0 Campascus minutus 1 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 5 1 7 0 0 0 2 0 0 2 2 0 Centropyxis aerophila sphagnicola 6 0 6 0 0 0 0 0 0 8 0 0 Centropyxis elongata 0 0 0 0 0 0 0 0 0 0 0 0 Centropyxis ecornis 0 0 0 0 0 0 1 0 0 3 0 0 Centropyxis laevigata 0 2 2 0 0 1 0 0 0 0 0 0 Centropyxis platystoma 0 0 0 0 0 0 0 0 0 0 0 0 Centropyxiella arenaria 0 1 0 0 0 0 0 0 0 0 0 0 Corythion dubium 2 2 2 0 0 0 0 0 0 0 0 0 Cyclopyxis kahli 1 0 1 0 0 0 0 0 0 1 0 0 Cyphoderia ampulla 6 16 4 0 1 0 0 0 0 0 0 0 Cyphoderia littoralis 0 7 0 0 0 0 0 0 0 0 0 0 Difflugia elegans parva 0 0 0 0 0 0 0 1 0 0 1 0 Difflugia globularis 0 0 0 0 0 0 0 0 0 0 0 0 Difflugia globulosa 7 2 1 0 0 0 0 0 0 16 0 0 Difflugia globulus 3 5 7 0 1 0 0 0 0 36 1 0 Difflugia lucida 0 0 2 0 0 0 0 0 0 0 0 0 Difflugia pristis 3 2 0 0 0 0 1 0 0 4 0 0 Difflugia tenuis 0 0 1 0 0 0 0 0 0 0 0 0 Difflugia sp1 1 0 0 0 0 0 0 0 0 0 0 0 Difflugiella oviformis 2 2 2 0 0 0 0 0 1 2 0 0 Euglypha dolioliformis 0 0 1 0 0 0 0 0 0 0 0 0 Euglypha polyepsis 0 0 0 0 0 0 0 0 0 0 0 0 Euglypha strigosa var. Glabra 0 0 1 0 0 0 0 0 0 0 0 0 Euglypha rotunda 11 14 11 0 0 0 0 0 0 1 0 0 Euglypha cristata 0 0 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0 0 0 Paraquadrula irregularis 0 0 0 0 0 0 0 0 0 1 0 0 Paulinella chromatophora 1 0 0 0 0 0 0 0 0 0 0 0 Plagiopyxis declivis 3 4 7 1 0 0 0 0 0 6 0 0 Pseudocorythion accutum 0 8 1 0 0 0 0 3 0 0 0 0 Pseudocorythion wailesi 0 1 0 0 0 0 0 0 0 0 0 0 Pseudohyalosphenia 0 1 3 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 19 10 16 1 0 0 1 0 0 32 0 0 Trinema complanatum 0 0 1 0 0 0 0 0 0 7 0 0 Trinema enchelys 40 30 30 3 0 0 1 0 1 12 0 0 Trinema lineare 35 41 39 3 0 0 0 0 0 18 0 0 Trinema lineare var. truncatum 3 0 3 0 0 0 0 0 0 0 0 0 Trinema penardi 0 0 0 0 0 0 0 0 0 0 0 0 Total counts 150 150 150 8 2 1 7 4 2 152 4 0

217

Appendix D

Appendix D Error calculation on testate amoebae counts of brackish marsh (Patterson & Fishbein 1989. equation 2).

Sample number s01 s03 s07 s10 s13 s15 s17 s21 s22 s23 s24

Arcella catinus 0 0 0.007 0 0 0 0 0 0 0 0

Arcella arenaria 0 0 0 0 0 0 0 0 0.007 0 0

Assulina muscorum 0.011 0 0 0 0 0 0 0 0 0 0

Campascus minutus 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 0.016 0.009 0.018 0.018 0.007 0.011 0.007 0 0.015 0.016 0.013 Centropyxis aerophila sphagnicola 0.023 0 0 0 0.009 0.007 0.022 0.015 0.019 0.011 0.024

Centropyxis elongata 0 0 0.007 0 0 0 0 0 0 0 0

Centropyxis ecornis 0 0 0.007 0 0 0 0 0 0 0 0.009

Centropyxis laevigata 0.007 0 0 0.011 0 0.007 0.011 0.007 0.019 0 0

Centropyxis platystoma 0 0 0 0 0 0 0 0 0 0 0

Centropyxiella arenaria 0 0 0 0 0 0 0 0 0 0 0

Corythion dubium 0.007 0 0 0.007 0 0.007 0 0 0.007 0 0

Cyclopyxis kahli 0 0.013 0.009 0.007 0 0.007 0.016 0.013 0.009 0 0.007

Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 0 0

Cyphoderia littoralis 0 0 0 0 0 0 0 0 0 0 0

Difflugia globularis 0 0 0 0 0 0 0 0 0 0 0

Difflugia globulosa 0.013 0.009 0.011 0.023 0 0.019 0.021 0.022 0.007 0.018 0.021

Difflugia globulus 0.03 0.032 0.036 0.017 0.036 0.031 0.026 0.031 0.032 0.013 0.025

Difflugia lucida 0 0.007 0 0.009 0 0.011 0 0 0 0 0

Difflugia pristis 0 0 0.007 0.009 0 0 0 0 0 0.009 0

Difflugia tenuis 0 0 0 0 0 0 0 0 0 0 0

Difflugia sp1 0 0 0 0 0 0 0 0 0 0 0

Difflugiella oviformis 0 0.007 0.014 0.007 0 0 0.009 0 0.015 0.011 0.011

Euglypha dolioliformis 0 0 0 0 0 0 0.007 0 0 0.009 0

Euglypha polyepsis 0.009 0 0 0 0 0.011 0 0 0 0 0.015 Euglypha strigosa var. Glabra 0.011 0.015 0 0 0 0.009 0.007 0 0 0 0

Euglypha rotunda 0.011 0.018 0.02 0.021 0.024 0.018 0.011 0.015 0.011 0.02 0.015

Euglypha cristata 0 0 0 0 0 0 0.007 0 0 0 0

Heleopera petricola 0 0 0 0 0 0 0 0.007 0 0 0

Hyalosphenia minuta 0 0 0 0 0 0 0 0 0.007 0 0

Paraquadrula irregularis 0 0 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 0

Plagiopyxis declivis 0 0.02 0.007 0.017 0.009 0.009 0.015 0.013 0.015 0.016 0.009 Pseudocorythion accutum 0 0 0 0.007 0 0 0 0 0 0.02 0 pseudocorythion wailesi 0 0 0 0 0 0 0 0 0 0.007 0

Pseudohyalosphenia 0 0 0 0.019 0 0 0 0 0 0 0

Tracheleuglypha dentata 0.024 0.028 0.03 0.039 0.027 0.03 0.03 0.036 0.033 0.033 0.034

Trinema complanatum 0.027 0.034 0.021 0 0.033 0.026 0.031 0.022 0.023 0 0.019

Trinema enchelys 0.03 0.028 0.035 0.022 0.034 0.032 0.026 0.033 0.03 0.034 0.024

Trinema lineare 0.028 0.024 0.014 0.028 0.017 0.028 0.026 0.017 0.021 0.033 0.033 Trinema lineare var. truncatum 0 0 0 0 0.009 0 0.009 0.007 0 0 0

Trinema penardi 0.021 0 0 0 0 0 0 0 0 0 0 218

Appendix D

Sample number s28 s30 s32 s31 s33 s34 s35 s36 s37 s38

Arcella catinus 0 0 0 0 0 0 0.007 0.007 0 0

Arcella arenaria 0 0 0 0 0 0 0 0 0 0

Assulina muscorum 0 0 0 0 0 0 0 0 0 0

Campascus minutus 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 0.016 0.016 0 0.015 0.013 0.011 0.011 0.013 0.009 0.023 Centropyxis aerophila sphagnicola 0.019 0 0.011 0.009 0.011 0.007 0 0 0.009 0.02

Centropyxis elongata 0 0 0 0 0.013 0 0 0 0 0

Centropyxis ecornis 0 0 0.009 0 0 0 0.007 0 0 0.007

Centropyxis laevigata 0.024 0 0 0.011 0.015 0.017 0.016 0.015 0.009 0

Centropyxis platystoma 0 0 0 0 0 0 0 0 0 0

Centropyxiella arenaria 0 0 0 0 0 0 0 0 0 0

Corythion dubium 0 0 0 0 0.006 0 0 0 0 0.007

Cyclopyxis kahli 0 0 0 0.013 0 0 0.007 0 0.007 0.009

Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 0

Cyphoderia littoralis 0.007 0 0 0 0 0 0 0 0 0

Difflugia globularis 0 0 0 0 0 0 0 0 0 0

Difflugia globulosa 0.009 0 0 0.024 0 0.009 0.016 0.021 0.011 0.02

Difflugia globulus 0.024 0.013 0.016 0.038 0.025 0.019 0.029 0.033 0.034 0.037

Difflugia lucida 0 0.007 0 0 0.006 0 0 0 0 0.011

Difflugia pristis 0 0 0 0 0 0 0 0 0 0

Difflugia tenuis 0 0 0 0 0 0 0 0 0 0

Difflugia sp1 0 0 0 0 0 0 0 0 0 0

Difflugiella oviformis 0.007 0.016 0.009 0.018 0.018 0.011 0 0 0.013 0

Euglypha dolioliformis 0 0 0.007 0 0 0 0 0 0 0

Euglypha polyepsis 0.007 0.015 0 0 0 0 0.017 0 0 0 Euglypha strigosa var. Glabra 0 0 0 0 0.009 0.007 0 0.007 0 0

Euglypha rotunda 0.022 0.029 0.036 0.013 0.027 0.031 0.017 0.013 0.023 0.016

Euglypha cristata 0 0 0 0 0 0 0 0 0 0

Heleopera petricola 0 0 0 0 0 0 0 0 0 0

Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0

Paraquadrula irregularis 0.007 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0

Plagiopyxis declivis 0.009 0 0 0 0.009 0.011 0.024 0.013 0.015 0.021 Pseudocorythion accutum 0 0 0 0 0 0 0 0 0 0 pseudocorythion wailesi 0 0 0 0 0 0 0 0 0 0

Pseudohyalosphenia 0 0 0 0 0 0 0 0 0 0

Tracheleuglypha dentata 0.036 0.031 0.035 0.017 0.035 0.04 0.039 0.041 0.038 0.033

Trinema complanatum 0.011 0.016 0.027 0.016 0.006 0.011 0 0.007 0 0.007

Trinema enchelys 0.029 0.039 0.024 0.034 0.025 0.024 0.027 0.027 0.027 0.018

Trinema lineare 0.028 0.025 0.032 0.024 0.031 0.027 0.022 0.011 0.026 0.02 Trinema lineare var. truncatum 0.007 0 0 0 0.006 0.007 0 0 0 0

Trinema penardi 0 0 0 0 0 0 0 0 0 0

219

Appendix D

Sample number s39 s40 s41 s42 s43 s46 sx s19 s09 s11

Arcella catinus 0.013 0 0 0 0 0.011 0.007 0 0 0

Arcella arenaria 0 0 0 0 0 0 0 0 0 0

Assulina muscorum 0 0 0 0 0 0 0 0.007 0.007 0

Campascus minutus 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 0.025 0.033 0.021 0.029 0.026 0.009 0.018 0.017 0.009 0.029 Centropyxis aerophila sphagnicola 0.011 0 0.017 0.023 0.015 0.018 0.018 0.015 0.011 0

Centropyxis elongata 0 0 0 0 0.007 0 0 0 0 0

Centropyxis ecornis 0 0 0.015 0 0 0.011 0 0.009 0 0

Centropyxis laevigata 0.024 0.016 0 0.016 0.016 0 0.015 0 0 0.009

Centropyxis platystoma 0 0 0.007 0 0 0 0 0.007 0 0

Centropyxiella arenaria 0 0 0 0 0 0 0 0 0 0

Corythion dubium 0 0.007 0.007 0 0 0 0 0 0.007 0.006

Cyclopyxis kahli 0 0 0.007 0.007 0.013 0.007 0 0.013 0.007 0

Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 0

Cyphoderia littoralis 0 0 0 0 0 0 0 0 0 0

Difflugia globularis 0 0 0 0 0 0 0 0 0 0

Difflugia globulosa 0.022 0.017 0.019 0.027 0.022 0.025 0.015 0.017 0.025 0.009

Difflugia globulus 0.033 0.024 0.031 0.035 0.027 0.034 0.029 0.03 0.038 0.019

Difflugia lucida 0.007 0.007 0.007 0.011 0.013 0 0 0 0 0

Difflugia pristis 0 0.007 0.007 0.007 0 0.013 0.009 0 0 0

Difflugia tenuis 0 0 0 0 0 0 0 0 0 0

Difflugia sp1 0 0 0 0 0 0 0 0 0 0

Difflugiella oviformis 0.007 0.011 0.011 0.009 0.018 0.009 0.013 0.011 0.007 0.022

Euglypha dolioliformis 0 0 0 0 0.007 0 0 0 0 0

Euglypha polyepsis 0 0.007 0.016 0 0 0 0.011 0.011 0.015 0 Euglypha strigosa var. Glabra 0 0 0 0.007 0 0 0.009 0 0 0.006

Euglypha rotunda 0.011 0.026 0.018 0.007 0.015 0.007 0.021 0.007 0.011 0.026

Euglypha cristata 0 0 0 0 0 0 0 0 0 0

Heleopera petricola 0 0 0 0 0 0 0 0 0.007 0

Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0

Paraquadrula irregularis 0 0 0 0 0 0.007 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0

Plagiopyxis declivis 0.024 0 0.03 0.011 0.022 0.016 0.007 0.02 0.007 0.006 Pseudocorythion accutum 0 0 0 0 0 0 0 0 0 0

pseudocorythion wailesi 0 0 0 0 0 0 0 0 0 0

Pseudohyalosphenia 0 0 0 0 0 0 0 0 0 0

Tracheleuglypha dentata 0.035 0.038 0.033 0.034 0.036 0.033 0.037 0.035 0.024 0.034

Trinema complanatum 0 0.007 0.007 0 0.007 0.017 0 0.021 0.02 0.011

Trinema enchelys 0.018 0.016 0.017 0.017 0.007 0.022 0.022 0.026 0.033 0.027

Trinema lineare 0.016 0.022 0.016 0.009 0.024 0.026 0.028 0.026 0.023 0.026 Trinema lineare var. truncatum 0 0 0 0 0 0 0 0 0 0.006

Trinema penardi 0 0 0 0 0 0 0 0 0 0

220

Appendix D

Sample number s02 s27 s06 s05 s16 s04 s14 s29

Arcella catinus 0 0 0 0.007 0 0.009 0 0

Arcella arenaria 0 0 0 0 0 0 0.007 0

Assulina muscorum 0 0 0 0 0 0 0 0

Campascus minutus 0 0 0 0.007 0 0 0 0 Centropyxis aerophila aerophila 0.019 0.011 0.016 0.015 0.009 0.017 0.007 0.022 Centropyxis aerophila sphagnicola 0.009 0.016 0.007 0.016 0.021 0.016 0 0.013

Centropyxis elongata 0 0 0 0 0 0 0 0

Centropyxis ecornis 0 0 0 0 0.007 0 0 0

Centropyxis laevigata 0.011 0.009 0.013 0 0 0.009 0.009 0.02

Centropyxis platystoma 0 0 0 0 0 0 0 0

Centropyxiella arenaria 0 0 0 0 0 0 0.007 0

Corythion dubium 0.007 0 0.007 0.009 0 0.009 0.009 0

Cyclopyxis kahli 0 0 0.007 0.007 0.02 0.007 0 0.009

Cyphoderia ampulla 0.011 0 0.011 0.016 0 0.013 0.025 0

Cyphoderia littoralis 0.007 0 0 0 0 0 0.017 0

Difflugia globularis 0.007 0 0 0 0 0 0 0

Difflugia globulosa 0.018 0.011 0.011 0.017 0.009 0.007 0.009 0

Difflugia globulus 0.009 0.029 0 0.011 0.015 0.017 0.015 0.029

Difflugia lucida 0 0 0 0 0 0.009 0 0

Difflugia pristis 0 0 0 0.011 0.007 0 0.009 0

Difflugia tenuis 0 0 0 0 0 0.007 0 0

Difflugia sp1 0 0 0 0.007 0 0 0 0

Difflugiella oviformis 0 0 0.007 0.009 0.013 0.009 0.009 0.027

Euglypha dolioliformis 0 0 0 0 0 0.007 0 0

Euglypha polyepsis 0.009 0.009 0 0 0.021 0 0 0 Euglypha strigosa var. Glabra 0 0 0.013 0 0 0.007 0 0

Euglypha rotunda 0.009 0.013 0.013 0.021 0.024 0.021 0.024 0.011

Euglypha cristata 0 0 0 0 0 0 0 0

Heleopera petricola 0 0.007 0 0 0 0 0 0

Hyalosphenia minuta 0 0 0 0 0.007 0 0 0

Paraquadrula irregularis 0 0 0 0 0 0 0 0.007 Paulinella chromatophora 0 0 0 0.007 0 0 0 0

Plagiopyxis declivis 0.011 0.007 0.021 0.011 0.025 0.017 0.013 0.013 Pseudocorythion accutum 0.011 0 0.021 0 0 0.007 0.018 0 pseudocorythion wailesi 0 0 0 0 0 0 0.007 0

Pseudohyalosphenia 0 0 0.007 0 0 0.011 0.007 0

Tracheleuglypha dentata 0.035 0.034 0.032 0.027 0.036 0.025 0.02 0.036

Trinema complanatum 0 0.024 0 0 0 0.007 0 0.015

Trinema enchelys 0.031 0.039 0.02 0.036 0.017 0.033 0.033 0.024

Trinema lineare 0.038 0.009 0.04 0.035 0.029 0.036 0.036 0.024 Trinema lineare var. truncatum 0 0.007 0 0.011 0.007 0.011 0 0

Trinema penardi 0 0 0 0 0 0 0 0

221

Appendix E

Appendix E Number of species per sample, concentrations and Shannon Wiener index corresponding to elevation m ~MHWL for both freshwater- and brackish tidal marsh (Notelaar and Groot buitenschoor respectively) samples. Samples with elevation differences of max. 2 cm are compared between both marshes and highest values are indicated in red. Elevation Number of species Concentration (m ~MHWL) sample-1 (tests /cm³) Shannon-Wiener index Groot Groot Groot buitenscho buitensc Groot Notelaar buitenschoor Notelaar or Notelaar hoor Notelaar buitenschoor -0,18 17 10566 2,25 0,05 16 48487 2,06 0,07 21 17578 2,32 0,09 21 55544 2,08 0,1 13 22621 1,83 0,11 16 45594 2,03 0,11 15 98982 2 0,12 17 175498 1,65 0,14 18 61424 1,9 0,14 15 93383 2,1 0,15 19 50518 1,74 0,15 14 134680 1,76 0,16 17 169174 1,79 0,18 18 44036 2,27 0,18 21 27249 2,07 0,18 13 69183 1,75 0,2 0,2 17 16 26793 2938 2,16 2,22 0,21 10 197666 1,31 0,22 18 15406 2,07 0,22 19 35450 1,94 0,22 15 72986 1,3 0,23 14 295388 1,93 0,24 20 115813 2,08 0,26 14 30569 1,97 0,26 15 98022 2,14 0,27 17 59318 1,82 0,3 18 59873 1,85 0,3 16 134682 2,03 0,34 15 70611 2,04 0,38 15 35797 2,24 0,4 16 122963 1,99 0,43 0,44 19 23 25962 3135 1,98 2,35 0,52 18 65429 1,94 0,57 0,59 17 16 66657 2511 2,32 1,97 0,65 0,63 21 14 120705 6202 2,3 2,09 0,72 0,71 18 16 36996 1234 2,29 2,16 0,72 18 14870 2,33 0,79 19 26690 2,5 0,87 21 15987 2,48 0,95 0,93 20 15 25071 3176 2,55 2,11 1,03 23 16957 2,47 1 18 4329 2,33 1,05 1,06 14 14 18906 8948 2,17 1,98 0,84 16 3296 2,25 0,9 16 2646 2,3 1,17 14 4297 2,13 1,32 13 6251 2,06 0,47 19 3790 2,16 0,55 19 3657 2,14

222

Appendix E

Number of species Concentration Elevation sample-1 (tests/cm³) Shannon-Wiener index Groot Groot Groot Groot Notelaar buitenschoor Notelaar buitenschoor Notelaar buitenschoor Notelaar buitenschoor 1,22 12 5986 1,91 1,09 13 5052 2,17 0,68 16 6503 1,93 1,86 13 7421 2,03 1,96 17 5746 2,33 1,26 12 9138 1,71 2,19 15 7274 2,05 1,75 14 12384 2,25 1,69 14 5423 1,84 1,36 14 3060 1,94 1,13 15 8900 2,21 2 16 11757 2,22 1,49 12 4508 2,02 2,15 16 12982 2,02 2,24 13 13931 2,15 1,82 16 16133 2,21 1,91 16 10540 2,29 1,61 16 12992 2,2 1,44 16 4684 2,17 1,63 13 10864 2,17 2,04 10 41392 1,84 2,1 14 23511 2,09 2,32 15 35713 2,32 1,46 10 3849 1,82 1,55 10 4099 1,83

223

Appendix F

Appendix F Species list with indication of corresponding test material based group and morphotype for each species.

Test material based groups (Mitchell et al. Morphotypes Bonnet Species (2008b) 1975 Arcella arenaria Protein + Calcium Arcella Arcella arenaria sphagnicola Protein + Calcium Arcella Arcella catinus Protein + Calcium Arcella Arcella discoides foreosa Protein + Calcium Arcella Arcella hemisphaerica Protein + Calcium Arcella Centropyxis aculeata aculeata Xenosome Simple Plagiostome Centropyxis aculeata minima Xenosome Simple Plagiostome Centropyxis aculeata oblonga Xenosome Simple Plagiostome Centropyxis aerophila aerophila Xenosome Plagiostome with Visor Centropyxis aerophila sphagnicola Xenosome Plagiostome with Visor Centropyxis aerophila sylvatica Xenosome Plagiostome with Visor Centropyxis constricta Xenosome Plagiostome with Visor Centropyxis ecornis Xenosome Plagiostome with Visor Centropyxis elongata Xenosome Plagiostome with Visor Centropyxis eurystoma Xenosome Plagiostome with Visor Centropyxis laevigata Xenosome Simple Plagiostome Centropyxis minuta Xenosome Simple Plagiostome Corythion dubium Idiosome Plagiostome with Visor Cryptodifflugia compressa Xenosome Compressed Acrostome Cyclopyxis kahli Xenosome Axial Cyphoderia ampulla Idiosome Arched Acrostome Cyphoderia ampulla vitrae Idiosome Arched Acrostome Difflugia angulostoma Xenosome Simple Acrostome Difflugia avellana Xenosome Simple Acrostome Difflugia cylindricus Xenosome Simple Acrostome

224

Appendix E

Test material based groups (Mitchell et al. Morphotypes Bonnet Species (2008b) 1975 Difflugia globularis Xenosome Simple Acrostome Difflugia globulosa Xenosome Simple Acrostome Difflugia globulus Xenosome Simple Acrostome Difflugia lucida Xenosome Compressed Acrostome Difflugia manicata Xenosome Simple Acrostome Difflugia mica Xenosome Simple Acrostome Difflugia pristis Xenosome Simple Acrostome Difflugia tenuis Xenosome Simple Acrostome Difflugiella oviformis Xenosome Simple Acrostome Euglypha laevis Xenosome Compressed Acrostome Euglypha dolioliformis Idiosome Compressed Acrostome Euglypha polyepsis Idiosome Compressed Acrostome Euglypha rotunda Idiosome Compressed Acrostome Euglypha strigosa glabra Idiosome Compressed Acrostome Euglypha tuberculata Idiosome Simple Acrostome Euglypha tuberculata minor Idiosome Simple Acrostome Euglypha tuberculata subcylindricae Idiosome Simple Acrostome Heleopera petricola Xenosome Compressed Acrostome Protein + Hyalosphenia minuta Calcium Compressed Acrostome Paraquadrula irregularis Idiosome Simple Acrostome Plagiopyxis penardi Xenosome Simple Cryptostome Tracheuglypha dentata Idiosome Simple Acrostome Trinema complanatum Idiosome Plagiostome with Visor Trinema enchelys Idiosome Plagiostome with Visor Trinema lineare Idiosome Plagiostome with Visor Trinema lineare truncatum Idiosome Plagiostome with Visor

225

Appendix G

Appendix G Absolute testate amoebae counts of core samples of a young freshwater tidal marsh.

sample 1000 1005 1010 1015 1020 1025 1030 1035 1040 1050 1060 Arcella arenaria 0 1 0 0 0 0 1 0 0 0 0 Arcella catinus 0 0 0 0 0 1 0 0 0 0 0 Centropyxis aculeata aculeata 0 0 0 0 0 0 0 0 0 1 0 Centropyxis aculeata minima 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata oblonga 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 10 6 22 5 3 8 8 1 0 0 0 Centropyxis aerophila sphagnicola 0 1 0 0 0 0 0 0 0 0 0 Centropyxis aerophila sylvatica 0 0 0 0 0 0 0 0 0 0 0 Centropyxis constricta 0 0 4 0 4 0 0 0 0 0 0 Centropyxis ecornis 0 2 0 1 0 0 0 0 0 0 0 Centropyxis elongata 0 0 0 1 0 1 0 0 0 0 0 Centropyxis laevigata 0 0 0 0 0 1 0 0 0 0 0 Centropyxis minuta 0 1 0 0 0 0 0 1 0 0 0 Corythion dubium 1 0 0 0 0 0 0 0 0 0 0 Cryptodifflugia compressa 0 0 0 0 0 0 0 0 0 0 0 Cyclopyxis kahli 0 0 2 0 2 0 1 0 1 0 1 Cyphoderia ampulla 0 1 0 0 1 0 1 0 0 0 0 Difflugia avellana 0 0 0 0 0 0 0 0 0 0 0 Difflugia cylindricus 0 0 0 0 0 0 0 0 0 0 0 Difflugia globularis 1 0 3 0 0 0 0 0 0 0 1 Difflugia globulosa 0 0 0 1 0 0 0 0 0 0 0 Difflugia globulus 4 2 1 5 3 1 3 0 1 1 1 difflugia lucida 0 0 0 0 0 0 0 0 0 0 0 Difflugia manicata 0 0 0 0 0 0 0 0 0 0 0 Difflugia mica 0 0 0 0 0 0 0 0 0 0 0 Difflugia pristis 1 1 1 0 0 0 2 0 0 0 0 Difflugia tenuis 0 1 5 0 0 0 1 0 0 2 0 Difflugiella oviformis 3 0 0 1 0 0 3 0 0 0 1 Euglypha dolioliformis 0 0 0 0 0 0 0 0 0 0 0 Euglypha polyepsis 1 0 0 0 0 1 0 0 0 0 0 Euglypha rotunda 0 6 5 1 0 1 2 0 0 0 2 Euglypha strigosa glabra 0 0 0 0 0 0 0 0 0 0 0 Euglypha tuberculata 0 0 0 0 0 0 3 0 0 0 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0 0 Paraquadrulla irregularis 0 0 1 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 7 10 21 15 18 7 15 0 0 1 1 Trinema enchelys 25 7 14 9 9 11 20 2 11 2 2 Trinema lineare 61 32 20 4 10 3 26 3 0 2 1 Trinema lineare truncatum 0 0 1 2 1 1 3 0 0 0 0 Total counts 114 71 99 43 50 35 86 7 13 9 10

Appendix G

107 108 109 110 200 200 201 201 202 202 203 sample 0 0 0 0 0 5 0 5 0 5 0 Arcella arenaria 1 0 0 0 1 3 0 0 0 0 0 Arcella catinus 0 0 0 0 0 0 0 1 0 1 0 Centropyxis aculeata aculeata 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata minima 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata oblonga 0 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 7 0 2 0 8 1 7 7 2 10 1 Centropyxis aerophila sphagnicola 0 0 0 0 0 0 2 0 2 0 0 Centropyxis aerophila sylvatica 0 0 0 0 0 0 0 0 0 0 0 Centropyxis constricta 0 0 0 0 0 2 4 5 6 0 0 Centropyxis ecornis 0 0 0 0 0 0 1 0 0 0 0 Centropyxis elongata 0 0 0 0 0 0 0 0 0 0 1 Centropyxis laevigata 0 0 0 0 0 0 0 0 0 0 0 Centropyxis minuta 0 0 0 0 1 0 0 0 0 0 0 Corythion dubium 0 0 0 0 0 0 0 0 0 0 0 Cryptodifflugia compressa 0 0 0 0 0 0 0 0 0 0 0 Cyclopyxis kahli 0 0 0 0 1 0 1 0 0 0 0 Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 0 0 Difflugia avellana 0 0 0 0 0 0 0 0 0 0 0 Difflugia cylindricus 0 0 0 0 0 0 0 0 0 0 0 Difflugia globularis 0 0 0 0 0 0 0 0 0 0 0 Difflugia globulosa 0 0 0 0 2 0 2 0 1 2 0 Difflugia globulus 0 0 0 1 3 1 0 2 7 0 0 difflugia lucida 0 0 0 0 0 0 0 0 0 0 0 Difflugia manicata 0 0 0 0 0 0 0 0 0 0 0 Difflugia mica 0 0 0 0 0 0 0 0 0 0 0 Difflugia pristis 0 0 0 0 1 3 0 2 0 3 0 Difflugia tenuis 0 0 0 1 0 1 0 1 0 0 0 Difflugiella oviformis 1 0 0 0 3 1 1 3 3 2 0 Euglypha dolioliformis 0 0 0 0 0 1 0 1 0 0 0 Euglypha polyepsis 0 0 0 0 0 0 0 0 0 0 0 Euglypha rotunda 0 0 2 0 3 5 5 1 2 0 0 Euglypha strigosa glabra 0 0 0 0 0 0 0 0 0 0 0 Euglypha tuberculata 0 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 0 0 0 0 1 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0 0 Paraquadrulla irregularis 0 0 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 1 0 0 0 0 0 0 Tracheleuglypha dentata 4 0 0 0 7 9 10 4 5 3 1 Trinema enchelys 1 1 0 1 5 21 9 4 6 9 6 Trinema lineare 3 3 7 2 30 40 36 17 11 3 5 Trinema lineare truncatum 0 0 0 0 0 0 1 0 1 0 1 Total counts 17 4 11 5 67 88 79 48 46 33 15

227

Appendix G

sample 2035 2040 2050 2060 2070 2080 2090 2100 3000 3005 Arcella arenaria 0 0 0 0 0 0 0 0 0 0 Arcella catinus 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata aculeata 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata minima 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata oblonga 0 0 0 0 0 0 0 0 2 0 Centropyxis aerophila aerophila 0 1 0 1 0 0 0 0 23 14 Centropyxis aerophila sphagnicola 0 0 0 2 0 0 0 0 0 1 Centropyxis aerophila sylvatica 0 0 2 0 0 0 0 0 3 0 Centropyxis constricta 0 0 0 0 0 0 0 0 0 4 Centropyxis ecornis 0 1 1 1 0 0 0 0 1 0 Centropyxis elongata 1 0 2 0 0 0 0 1 0 0 Centropyxis laevigata 1 0 0 0 0 0 0 0 0 1 Centropyxis minuta 0 0 0 0 0 0 0 0 2 0 Corythion dubium 0 0 0 0 0 0 0 0 0 1 Cryptodifflugia compressa 2 0 0 0 0 0 0 0 0 0 Cyclopyxis kahli 0 0 0 1 0 1 1 0 0 0 Cyphoderia ampulla 0 0 0 0 0 0 0 0 2 5 Difflugia avellana 0 0 0 0 0 0 0 0 0 0 Difflugia cylindricus 0 0 0 0 0 0 0 0 0 0 Difflugia globularis 0 0 0 0 0 0 0 0 0 0 Difflugia globulosa 0 0 0 0 0 0 0 0 7 2 Difflugia globulus 2 0 1 0 0 0 0 0 5 14 difflugia lucida 0 0 0 0 0 0 0 0 1 0 Difflugia manicata 0 0 0 0 0 0 0 0 2 0 Difflugia mica 0 0 0 0 0 0 0 0 0 0 Difflugia pristis 1 0 0 0 0 0 0 0 4 8 Difflugia tenuis 0 0 1 0 0 0 1 0 1 3 Difflugiella oviformis 3 0 0 1 0 3 0 0 1 2 Euglypha dolioliformis 0 0 0 0 0 0 0 0 0 0 Euglypha polyepsis 0 0 0 0 0 0 0 0 1 0 Euglypha rotunda 0 0 0 1 0 1 1 0 7 3 Euglypha strigosa glabra 0 0 0 0 0 0 0 0 0 2 Euglypha tuberculata 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0 Paraquadrulla irregularis 0 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 4 7 1 3 1 3 0 0 12 14 Trinema enchelys 2 7 4 1 3 1 0 0 10 9 Trinema lineare 3 9 0 0 2 9 0 0 24 39 Trinema lineare truncatum 1 0 0 0 0 0 0 0 4 1 Total counts 20 25 12 11 6 18 3 1 112 123

228

Appendix G

sample 3010 3015 3020 3025 3030 3035 3040 3050 3060 3070 Arcella arenaria 0 0 0 0 0 0 0 0 0 0 Arcella catinus 0 0 0 0 1 0 0 0 0 0 Centropyxis aculeata aculeata 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata minima 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata oblonga 0 0 2 0 1 0 0 0 0 0 Centropyxis aerophila aerophila 12 11 1 2 4 8 4 0 0 0 Centropyxis aerophila sphagnicola 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila sylvatica 2 0 0 0 0 0 0 0 0 0 Centropyxis constricta 2 7 1 0 2 0 0 0 0 0 Centropyxis ecornis 1 0 2 1 0 0 0 0 0 1 Centropyxis elongata 1 0 0 0 0 0 0 0 0 0 Centropyxis laevigata 1 0 0 1 0 0 1 0 2 0 Centropyxis minuta 0 0 0 0 0 0 0 0 0 0 Corythion dubium 0 0 0 0 0 0 0 0 0 0 Cryptodifflugia compressa 0 0 0 0 0 0 0 0 0 0 Cyclopyxis kahli 1 1 0 0 0 0 0 0 0 0 Cyphoderia ampulla 0 0 0 0 0 0 0 0 0 0 Difflugia avellana 0 0 0 0 0 0 0 0 0 0 Difflugia cylindricus 0 0 0 0 0 0 0 0 0 0 Difflugia globularis 0 0 0 2 0 0 0 0 0 0 Difflugia globulosa 3 4 0 0 0 0 2 0 0 0 Difflugia globulus 12 14 5 4 5 1 2 0 0 0 difflugia lucida 0 0 0 0 3 0 0 0 0 1 Difflugia manicata 2 0 0 0 0 0 0 0 0 0 Difflugia mica 0 0 0 0 0 0 0 0 0 0 Difflugia pristis 1 2 0 1 1 2 1 0 0 0 Difflugia tenuis 0 1 1 0 3 2 1 0 2 0 Difflugiella oviformis 2 1 0 3 1 7 1 0 0 0 Euglypha dolioliformis 0 0 0 0 0 0 0 0 0 0 Euglypha polyepsis 0 0 0 0 0 0 0 0 0 0 Euglypha rotunda 2 7 1 1 2 2 0 0 0 0 Euglypha strigosa glabra 0 0 0 1 0 0 0 0 0 0 Euglypha tuberculata 0 0 0 0 0 0 0 0 0 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 0 Hyalosphenia minuta 1 0 0 0 0 0 0 0 0 0 Paraquadrulla irregularis 0 0 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 4 8 6 7 6 2 2 0 0 3 Trinema enchelys 16 17 2 1 1 1 0 1 0 0 Trinema lineare 14 7 1 2 3 2 0 0 0 0 Trinema lineare truncatum 0 0 0 0 0 0 0 0 0 0 Total counts 77 80 22 26 33 27 14 1 4 5

229

Appendix G

sample 3080 3090 3100 4000 4005 4010 4015 4020 4025 4030 Arcella arenaria 0 0 0 0 0 0 0 0 0 0 Arcella catinus 0 0 0 1 0 2 0 0 0 0 Centropyxis aculeata aculeata 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata minima 0 0 0 0 0 0 0 0 0 0 Centropyxis aculeata oblonga 0 0 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 1 0 0 23 17 17 14 2 11 5 Centropyxis aerophila sphagnicola 0 0 0 0 0 1 0 2 0 0 Centropyxis aerophila sylvatica 0 0 0 0 0 0 0 0 0 0 Centropyxis constricta 0 0 0 0 0 5 5 2 2 0 Centropyxis ecornis 0 0 0 2 1 0 1 0 1 1 Centropyxis elongata 0 0 0 0 0 0 0 0 0 0 Centropyxis laevigata 0 0 0 5 4 2 2 0 0 0 Centropyxis minuta 0 0 0 0 0 0 0 0 2 0 Corythion dubium 0 0 0 1 0 0 0 0 0 0 Cryptodifflugia compressa 0 0 0 0 0 0 0 0 0 0 Cyclopyxis kahli 0 0 1 0 0 0 1 0 0 0 Cyphoderia ampulla 0 0 0 5 2 0 1 0 0 0 Difflugia avellana 0 0 0 0 0 0 0 0 0 0 Difflugia cylindricus 0 0 0 0 0 0 0 1 0 0 Difflugia globularis 0 0 0 1 0 0 1 1 0 0 Difflugia globulosa 0 0 0 0 0 1 3 0 0 0 Difflugia globulus 0 1 0 10 3 7 15 7 13 3 difflugia lucida 0 0 0 0 2 0 1 0 0 0 Difflugia manicata 0 0 0 1 0 0 0 0 0 0 Difflugia mica 0 0 0 0 0 0 0 0 0 0 Difflugia pristis 1 0 0 8 7 2 4 0 5 2 Difflugia tenuis 0 0 0 5 15 3 5 1 5 2 Difflugiella oviformis 0 0 0 1 0 3 4 0 4 6 Euglypha dolioliformis 0 0 0 0 0 0 0 0 0 0 Euglypha polyepsis 0 0 0 2 1 0 1 0 0 0 Euglypha rotunda 0 0 0 21 6 3 9 2 1 1 Euglypha strigosa glabra 0 0 0 0 0 0 0 0 0 0 Euglypha tuberculata 0 0 0 0 1 0 1 0 1 0 Heleopera petricola 0 0 0 0 0 0 0 0 0 1 Hyalosphenia minuta 0 0 0 0 0 0 0 0 0 0 Paraquadrulla irregularis 0 0 0 1 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 1 0 0 5 5 3 9 4 8 1 Trinema enchelys 0 0 0 2 3 2 20 4 5 3 Trinema lineare 0 0 1 77 23 4 60 3 2 0 Trinema lineare truncatum 0 0 1 0 0 0 1 0 0 0 Total counts 3 1 3 171 90 55 158 29 60 25

230

Appendix G

sample 4035 4040 4050 4060 4070 4080 4090 4100 Arcella arenaria 0 0 0 0 0 0 0 0 Arcella catinus 0 0 0 0 0 0 0 0 Centropyxis aculeata aculeata 0 0 0 0 0 0 0 0 Centropyxis aculeata minima 0 0 0 1 0 0 0 0 Centropyxis aculeata oblonga 0 0 0 0 0 0 0 0 Centropyxis aerophila aerophila 6 7 1 6 5 0 0 0 Centropyxis aerophila sphagnicola 1 0 0 0 2 0 0 0 Centropyxis aerophila sylvatica 0 0 0 0 0 0 0 0 Centropyxis constricta 0 1 0 0 0 0 0 2 Centropyxis ecornis 3 0 0 1 0 0 0 0 Centropyxis elongata 2 0 0 0 0 0 0 0 Centropyxis laevigata 1 0 0 0 0 0 0 0 Centropyxis minuta 0 0 0 0 0 0 0 0 Corythion dubium 0 0 0 0 0 0 0 0 Cryptodifflugia compressa 0 0 0 0 1 0 0 0 Cyclopyxis kahli 0 0 0 0 1 0 0 0 Cyphoderia ampulla 0 0 0 0 0 0 0 0 Difflugia avellana 0 0 1 0 0 0 0 0 Difflugia cylindricus 0 0 0 0 0 0 0 0 Difflugia globularis 0 0 0 0 0 0 0 0 Difflugia globulosa 0 0 0 0 1 0 0 0 Difflugia globulus 0 1 1 2 0 0 1 0 difflugia lucida 0 0 0 0 0 0 0 0 Difflugia manicata 0 0 0 2 0 0 0 0 Difflugia mica 0 0 0 1 0 0 0 0 Difflugia pristis 3 0 0 0 3 0 1 0 Difflugia tenuis 0 1 0 0 3 0 0 1 Difflugiella oviformis 1 4 0 0 1 1 0 0 Euglypha dolioliformis 0 0 0 0 0 0 0 0 Euglypha polyepsis 0 1 0 0 0 0 0 0 Euglypha rotunda 2 0 0 0 2 1 0 0 Euglypha strigosa glabra 0 0 0 0 0 0 0 0 Euglypha tuberculata 2 0 0 0 1 0 0 0 Heleopera petricola 1 0 0 1 0 0 0 0 Hyalosphenia minuta 0 0 0 0 0 0 0 0 Paraquadrulla irregularis 0 0 0 0 0 0 0 0 Paulinella chromatophora 0 0 0 0 0 0 0 0 Tracheleuglypha dentata 4 4 6 2 8 2 1 1 Trinema enchelys 1 0 0 0 3 0 0 0 Trinema lineare 0 0 1 0 12 1 0 1 Trinema lineare truncatum 0 0 0 0 1 0 0 0 Total counts 27 19 10 16 44 5 3 5

231

Abbreviations

General abbreviations

A Testate amoebae type; Axial AA Testate amoebae type; Arched Acrostome Arc Testate amoebae type; Arcella BSi Biogenic Silica CA Testate amoebae type; Compressed Acrostome Cot Testate amoebae type; Cotylostome DCA Detrended Correspondence Analysis DCCA Detrended Canonical Correspondence Analysis DSi Dissolved Silica HAT Highest Astronomical Tide LOI Loss On Ignition LTVS Long Term Vision Scheldt-estuary MARSED (zero-dimensional time-stepping) MARsh SEDimentation model (Temmerman et al., 2003a) MHWL Mean High Water Level ~ MHWL relative to Mean High Water Level MTL Mean Tide Level N Nitrogen NAP Nieuw Amsterdams Peil (Dutch Ordnance Level) P Phospor PLS Partial Least Squares regression method PSU Practical Salinity Unit PV Testate amoebae type; Plagiostome with Visor RDA Redundancy Analysis RMSEP Root Mean Square Error of Prediction SA Testate amoebae type; Simple Acrostome SC Testate amoebae type; Simple Cryptostome SD Standard Deviations SP Testate amoebae type; Simple Plagiostome SSC Suspended sediment concentration

Abbreviations

TAW Tweede Algemene Waterpassing (Belgian Ordnance Level) VIF Variation Inflation Factor WA Weighted Average regression method WA-PLS Weighted Average-Partial Least Squares regression

Testate amoebae species abbreviations

arcare Arcella arenaria Greef 1866 arcares Arcella arenaria sphagnicola Deflandre 1928 arc(c)at Arcella catinus Penard 1980 arcdisf Arcella discoides foveosa Playfair 1918 archem Arcella hemisphaerica Perty 1852 assmus Assulina muscorum Greef 1888 cenacu Centropyxis aculeate Stein 1857 cenacuo Centropyxis aculeata oblonga Deflandre 1929 cenaeae Centropyxis aerophila aerophila Deflandre 1929 cenaeasy Centropyxis aerophila sylvatica Deflandre 1929 cenaesp Centropyxis aerophila sphagnicola Deflandre 1929 cencon Centropyxis constricta Deflandre 1929 ceneco Centropyxis ecornis Leidy 1879 cenelo Centropyxis elongate Thomas 1959 ceneur Centropyxis eurystoma Deflandre 1929 cenlae Centropyxis laevigata Penard 1890 cenmin Centropyxis minuta Deflandre 1929 crycom Cryptodifflugia compressa Penard 1902 cycka Cyclopyxis kahli Deflandre 1930 cypamp Cyphoderia ampulla Leidy 1879 cypampv Cyphoderia ampulla virtae Wailes & Penard 1911 cyplit Cyphoderia littoralis Golemansky 1973 difang Difflugia angulostoma Gauthier-Lièvre & Thomas 1958 difglo(b) Difflugia globulus Cash & Hopkinson 1909 difglo(b)a Difflugia globulosa Dujardin 1837

234

Abbreviations

difluc Difflugia lucida Penard 1980 difman Difflugia manicata Penard 1902 difmic Difflugia mica Frenzel 1892 difovi Difflugiella oviformis Penard 1980 difpri Difflugia pristis Penard 1902 difsten Difflugia tenuis Ogden 1983 eugdol Euglypha dolioliformis Bonnet 1959 euglae Euglypha laevis Perty 1849 eugpol Euglypha polylepsis Decloitre 1962 eugrot Euglypha rotunda Wailes 1915 eugtub Euglypha tuberculata Dujardin 1841 eugstrgl Euglypha strigosa glabra Wailes 1898 eugtubm Euglypha tuberculata minor Taranek 1882 eugtubs Euglypha tuberculata subcylindrica Decloitre 1962 helpet Heleopera petricola Leidy 1879 hyamin Hyalosphenia minuta Cash 1891 parirr Paraqudrula irregularis Deflandre 1932 pauchro Paulinella chromatophora Lauterborn 1895 pladec Plagiopyxis declivis Thomas 1958 pscorac Pseudocorythion acutum pshyal Pseudohyalsophenia sp. plagpen Plagiopyxis penardi Thomas 1958 traden Tracheleuglypha dentata Deflandre 1938 tricom Trinema complanatum Penard 1890 trienc Trinema enchelys Leidy 1878 trilin Trinema lineare Penard 1890 trilint Trinema lineare truncatum Chardez 1964 tripen Trinema penardi Thomas & Chardez 1958

235

Dankwoord

Tijdens het uitwerken van mijn master thesis, en door het enthousiasme van mijn thesisbegeleiders Prof. Dr. Stijn Temmerman en Prof. Dr. Bart Van de Vijver, kreeg ik de microbe voor onderzoek te pakken. Ik was dan ook in de wolken toen ik enkele maanden later hoorde dat ik steun kreeg van de Universiteit Antwerpen, en erna van het IWT, om een doctoraatsonderzoek uit te voeren. Ondertussen zijn er vijf jaar verstreken en het onderzoek is afgerond. Nu kan ik met trots terugblikken op de weg die ik heb afgelegd, want een doctoraat haal je niet zonder slag of stoot en zeker niet zonder de steun van jullie allemaal.

Ik wil dan ook in de eerste plaats Prof. Dr. Stijn Temmerman bedanken om mij wegwijs te maken in de schorren- en Scheldekunde. Verder wil ik jou bedanken voor alle inspanningen die je geleverd hebt om thuis te worden in de wereld van de thecamoeben om mij zo goed mogelijk te kunnen begeleiden. Dankzij jou inspiratie, inzichten, input en uren verbeterwerk had ik dit doctoraat nooit kunnen afronden. BEDANKT!!

Ook wil ik Prof. em. Dr. Louis Beyens bedanken voor al de energie en werk dat u gestopt heeft in dit doctoraat. Uw kritische bedenkingen, opmerkingen en tekstuele verbeteringen hebben mij en dit proefschrift enorm voorruit geholpen.

Thank you, Andrey! I want to express my gratitude for helping me with the determination of some obscure testate amoebae and for the multiple discussions regarding statistical techniques and results. You were the ideal scientific, Sheldonesc , roommate at university!

Dankwoord

Ik ben Dr. Eric Struyf erg dankbaar voor de samenwerking tijdens het onderzoeken naar de rol van thecamoeben in de Si cyclus. Jou passie en oneindige kennis van Si maakt iedereen enthousiast om onderzoek naar Si (in bier ) uit te voeren.

Verder wil ik ook alle andere mede auteurs bedanken voor de input die jullie geleverd hebben voor het tot stand brengen van hoofdstuk 5 zijnde Adriaan Smis, Alexander Van Braeckel en Prof. Dr. Patrick Meire.

Mijn ex-PLG collega’s Wouter, Jean, Alexandra, Chen en Sven; dank u wel voor het helpen met veldwerk, GIS-kaartjes maken, leuke babbels en de feestjes of uitstappen buiten de werkuren. Jullie maakten van ’t unief mijn tweede thuis!

Annick, Lucie en Magda, een dikke merci voor al jullie hulp met praktische zaken en voor al het werk dat jullie uitvoeren achter de schermen om de onderzoeksgroep goed te laten draaien.

Frans, dank u wel voor het assisteren bij het veldwerk en voor al die sappige verhalen tijdens de koffiepauzes.

Ook kijk ik met pretoogjes terug naar al die dinsdag lunchen met Stefi, Fana en Ann, waarin veel liefde en leed gedeeld werd en artikels gevierd werden met een zakje M&M’s.

Dank u wel aan alle ECOBE’rs, in het bijzonder de Schelde groep, voor jullie raad en advies bij het uitvoeren, verwerken en interpreteren van data.

238

Dankwoord

Ook wil ik mijn familie en schoonfamilie bedanken voor het vertrouwen dat jullie in mij hadden om dit doctoraat tot een goed einde te brengen.

Lieve Stijn, dank je wel om mij altijd te steunen tijdens de hoogtes en laagtes van de afgelopen jaren, om mijn gezaag aan te horen en mij met mijn beide voeten terug op de grond te zetten als ik weer eens teveel aan’t piekeren was.

Lien, flinke dochter, jij bent mijn zonnetje en laat mij genieten van de kleine dingen in het leven, wat mij de nodige energie gaf om dit doctoraat af te werken.

239