Naturalis - Research plan

Below is an excerpt of the research plan, entitled Palaeo-ecological excavation Mare aux Songes, Mauritius; Scenario for the second field campaign July/August 2007 (5th version), composed by K. Rijsdijk & T. Vernimmen, with contributions from field teammembers and other experts.

The scenario for the first excavation at Mare aux Songes (MAS) in 2006 was created based on the geophysical and borehole investigations carried out in the field during 2006 and analysis of information of test trench excavated in October 2005. Based on the experiences of 2006 fieldwork it the scenario for the upcoming 2007 campaign is adapted.

During the time that passed after the last campaign (2006) there have been evaluations by the team members, project leader and the science institutions involved, to ensure that the field experience of 2006 can be put into practice efficiently with the help of a scenario. After the last fieldwork especially, we know much more about the nature, quality and extent of the fossil layer at MAS (i.e. the quantity and value of palaeontological and palaeo-ecological finds in this layer). This means we can adapt previously formulated research questions to the actual situation. This has consequences, amongst other things, for the way we are going to sample. The scoop method was also reviewed and a new strategy was designed by which we intend to excavate more or less stratigraphically several small areas that will be made dry artificially for the time of the dig.

The current scenario is based on an accepted type of scenario for archaeological excavations (University of Amsterdam, The Netherlands) because we are largely following archaeological methodology when excavating and treating wet organic materials. However, the exceptional nature of this natural site and the presence of such an abundance of bones of (extinct) animals, the remains of - possibly - a buried forest and the challenging hydrology of the site makes it necessary to adapt our strategy likewise and implement both paleontological and sedimentological fieldtechniques and approaches.

This scenario will act as a document that describes the lines of approach, responsibility and relevant logistic information. It will be used to base decisions upon during the 2007 campaign and will be presented relevant organisations relevant to the excavation activity. Also, this scenario will serve as a day-to-day guide for all members of the field team.

This fieldwork is part of the Dodo Research Programme (DRP). The main aim of the DRP is reconstructing the world of the dodo and determining the factors of its demise. In order to reconstruct the world of the dodo, the fossil layer at Mare aux Songes will be investigated. The investigation is a multidisciplinary activity and involves various disciplines ranging from paleontology, aDNA to hydrogeology and paleoecology.

Below firstly the aims and results of the 2006 excavation are presented. Then the aims and approach of the 2007 excavation are presented.

A team of more than 15 scientists from UK, MAU en NL carried out the excavation at Mauritius. Firstly a geophysical survey was carried out using Georadar system. Total surveyed length was approx. 4 km. After the survey 12 cores were drilled by means of a mechanical drilling system. The cores have a combined length of 80 m. Based on the geophysical and core survey three locations were identified for excavation pits.

The excavation was carried out with a mechanical digger. The pits ranged in size by 2m*3m, 2m*2m and 2m*2 m and provided a window of the stratigraphy of the top 2 m of the succession, sampling from the scoop (ex situ sampling) and sampling from the bulk (sieving). Scoop samples were accurately described from within the scoop (sedimentology, stratigraphy). Also scoops were obtained that were sieved in the following sieving fractions 10 cm, 5 mm. Samples were take of sediments finer than 5 mm. Further subsamples were taken for research by the following disciplines: geochemistry (MSIRI), microbiology (MSIRI), ancient DNA (UOXF), isotope dating (UGRO), grainsize (TNO), micromorphology (TNO), micropalaeontology (TNO).

In 2006 based on assessing the geohydrological conditions at Mare aux Songes we concluded that it is not possible given the equipment and finances available to drain (parts of ) the swamp in order to obtain dry exposures. Therefore the excavation was carried out under sub-aquatic conditions. Samples were obtained by a mechanical digger as in October 2005. They provided acceptable sample quality for geological, sedimentological and in situ sampling purposes.

Based on the experiences and analyses of the 2006 excavation, new aims have been identified for the 2007 excavation:

By means of sheet piles, pumps and mechanical digger a dry excavation pit will be realised. All material that will be removed in order to generate the excavation pit will be stored in sample bags and sieved. All rock material that cover the fossil layer will be dumped back. All excess unused sieve residue will be dumped back in the ditches and excavation pits.

We observed that fossils finer than 5 mm are preserved at MAS they include for example fine bird bones, bat bones and teeth and mollusca. These smaller organisms must have formed a major part of the ecosystem and it is therefore crucial to obtain a representative sample of these.

In order to understand the biostrationomic processes (transport, disarticulation, scatter, modification) that led to the formation of the bone layer it is crucial to map the bone layer in detail. Measuring the 3d orientations of the fossils will provide insight in the organisation of the bone layer and will help to resolve the question what processes led to the bone concentration at MAS and ultimately what was the death cause of the animals.

We will pay special attention into assessing the effects of post syn mortem transportation. Transportation may be minimal where vertebrates are in marine or lacustrine sedimentary basins.

Organisms may also be transported considerable distances by predators, scavengers, fluvial systems, or marine currents. Other taphonomic aspects of the fossils will be assessed especially macroscopic aspects of bone modification will be described. Bone modification includes any form if alteration to individual bones, including cracking, surface marks, abrasion, polishing and breakage.

Damage to bone may occur at various stages throughout the taphonomic history of a vertebrate bone. They can be divided into: pathological and traumatic damage sustained during life, including diseased bone and natural fractures, post-mortem damage due to predation, scavenging, trampling, or transport and post-burial damage sustained due to compaction, cracking or dissolution. Damage sustained during life may show evidence of healing, including mew bone growth. Postmortem damage will not show new bone growth and it may be difficult to distinguish from damage sustained during compaction.

In order to understand the conditions of preservation of the fossils including collagen and aDNA we will take soil samples to study in situ geochemical and microbiological conditions. Further subsamples will be taken to study molluscs, insects and microfossils.

In 2006 sieving was focussed on the fraction > 5 mm, in 2007 we will focus on obtaining bulk samples finer than 5 mm. From these first undisturbed samples for insect, molluscs and fine flora studies will be taken. Then the rest of the bulk samples will be sieved for the fraction residue will be sorted out on-site.

Combining archaeological, paleontological and sedimentological approaches the 3 dimensional composition and continuity of the fossil layer will be mapped (anatomy). Fossil bones and wood remains will be mapped focussing on fractions > 2 cm. A grid 50 by 50 cm will be constructed on a platform above the fossil layer that allow the coding of the positions of the exposed fossils (fig 1a).

Depending on the vertical and lateral variability of the fossils within the fossil layer, maps will be made of horizontal surfaces (fig 1b).

When relevant the orientation of fossils, tree stems, rootlets will be measured with a 3d compass and data will be registered in a database and plotted in polar plots (fig 1c).

Orientation of fossil long axes will help to reconstruct taphonomic processes. Post mortem transportation of bones may result in specific orientation of skeletons, in most cases post mortem stable positions of fossil are when a-b axes (a is long axis of symmetry, b is second longest axis of symmetry, c is smallest axis of symmetry) are horizontal, with specific sides up or down. Trampling, bioturbation, plant growth and soil movement can disturb this pattern into near vertical positions. Fossil skeletons in life position indicate rapid burial.

The focus will be on bone elements of which orientations can be related to the organism death position. Long axes will be measured, axes of symmetry, dip orientation of proximal bone and wood ends. The positions of stems and rootlets will be measured, also here distal and proximal ends will recorded and for rootlets way upward criteria will be mapped. The orientations will be mapped in polar plots.

As convention for proximal orientations is taken for bones pointing towards the heart and for wood pointing towards the stem base (fig 2). From skulls orientations of longest axis of symmetry are measured (beak to back end skull) and from tortoise shields (head to tail) idem ditto. Skulls, shields, bones including vertebrae, tree rootlets that lie upside down will be denoted with a `-` sign and will be plotted as an indented line on polar plots.

Figure 2. showing proximality criteria for bone orientations and long axis (A) of symetry of skull.

3d orientations will be measured with Leica compasses that contain besides a polar needle a dip needle. The compass will aligned parallel to long axes of the fossil and the orientation of the long axis will be measured (proximality of the fossil will firstly be determined), then the compass will be tilted and the compass long axis will be aligned parallel to long axis dip of the fossil.

Bones and wood fragments that are articulated will be sampled as one sample. They will be mapped in black. Unarticulated bones and wood fragments, will be bulk sampled pro grid segment. They will be mapped in white. Small articulated specimens will be mapped with a high resolution 5 by 5 cm subgrid.

Grid coordinates will be adapted accordingly (using 1 to 6 for y coordinates and UVWXYZ for x coordinates). For example: Trench 4 A1/U1 denotes upper left cel U1 of grid sector A1.

Depending on depth variation within the fossil layer one or more features (sub layers) will be mapped. Each layer will be photographed (from fixed position in top).

Taphonomy can be broken down into three groups: necrolysis: the death and decomposition of and organism, biostratonomy: the sedimentary history of the fossil and fossil diageneses: the chemical and mechanical alterations within the sediment.

All items mentioned above are undisturbed samples, not meant for processing on-site.

It would be best to sample sediment samples rather than sieved samples. There is more probability ofgetting contamination in sieved samples.

The best material to send would be samples from the debris rich layers and some small gyttja samples just to see what is in them. I would like large samples (10 kg+) if possible. The fauna should be quite diverse and larger samples will collect the rare taxa. If you can't send samples this large then perhaps save some for future work and store them. Ideally, if I find good material I would like to go to Mauritius next year to sample larger samples and collect the modern fauna.

Based on the 2006 excavation and geohydrological requirements for the 2007 excavation the position of the excavation pit is determined. The pit will be position on a location with a rich fossil concentration and where a thick (>0.5 m) and continuous impermeable layer underlying the fossil layer is present. Borehole data indicate that central in the basin these conditions are fullfilled.

Our findings so far indicate that Mare aux Songes is a globally exceptionally fossil depository and may be indicated as a fossil Lagerstatten. The fossil site is therefore extremely valuable scientifically.

As the fossil layer contains remains of the world famous dodo (Raphus cucullatus) the site is also culturally very important. The excavation activities and continue press releases around the dodo will provide free advertisement to Mauritius as a tropical paradisiacal island where once the dodo lived.

The state of Mauritius declared a special status to Mare aux Songes and is now listed as a national heritage site, that is only accessible for research when permissions are onbtained by the Nation Heritage Fund and Mauritian government.

The fossil layer extends for 36 ha with highest concentrations of fossils in ´basin A´.

Excavations are not extensive (probably only one pit of 9 M2 in total and only one feature) and thus are not expected to pose a problem to the sites new protected status.

Our research focuses to reconstruct the world of the dodo and determining the factors of its demise. It will provide a scientifically based substantiation of the claim of its uniqueness and completeness. We strife to minimise the damage done on the fossil layer and record all data in detail in maps, films, photographs and databases. All data collected and analysed will be shared with the community through museum displays and publications.

We will follow ethical guidelines of Archaeological and Paleontological disciplines that encompass not to collect fossils without scientific justification or to inflict unnecessary damage to the fossil site.

Beginning of the second field campaign on 29 July 2007, for the duration of 1 month.