Summary

The ADEPD project compiled biological and biogeochemical data from a range of international and national deep sea projects in the data information system PANGAEA. Diverse data sets were brought together in an uniform data format and are made available to a wider public. More than one hundred biogeochemical variables and 114,000 published and unpublished data sets were compiled in the last two years in PANGAEA. A new and very simple approach to the data base via the world wide web was implemented. Now for the first time a large deep sea data base is easily accessible for the general public on: http://www.pangaea.de/Projects/ADEPD

It became obvious that many data are clustered in some very well investigated areas of the Atlantic, but large regions are devoid of biological and biogeochemical data. This applies in particular for the Mid Atlantic Ridge, parts of the South Atlantic and Southern Ocean, while the the eastern part of the Atlantic, northern North Atlantic and Arctic regions are well covered. Most deep sea research projects did not carry out geochemical and biological studies at the same locations. Therefore, the number of data pairs suitable for the investigation of empirical correlations is very low despite the high total amount of data gathered. Different methods employed for the determination of one parameter add further restrictions to the comparability of data sets. ADEPD applied empirical correlations to convert biomass measurements of small organisms to a uniform variable and to extend the data base available for regional analysis.

For the northern North Atlantic and some areas in the South Atlantic relationships between primary production and benthic fluxes of nutrients and oxygen have been established (this analysis was done as part of other projects using the ADEPD data base). In these regions the coverage of geochemical data is good enough to obtain a spatial resolution and define benthic provinces.

Two very different approaches to estimate total carbon flux and oxgen consumption at the deep sea floor of the world ocean were compared. A value of 5.33x10¹³ mol O₂ y^-1 is estimated from surface productivity and vertical carbon flux relationship while a value of 5.43x10¹³ mol O₂ y^-1 is estimated from sea floor benthic oxygen flux and organic carbon burial rate compilation. This close agreement may be fortuitous, but the estimates support each other and suggest that a reasonable assessment of global seafloor oxygen flux has been achieved. Regional differences in both assessments point to methodological shortcomings by one or the other method and to gaps in data coverage as well as gaps in knowledge in respect to carbon fluxes along continental margins.

Evaluation of the data collection helped to identify gaps in data availability as well as systematic patterns and problems in deep sea biogeochemical research. It resulted in recommendations for further data collections and analysis of deep sea biogeochemical data.

1 Introduction

The global cycling of carbon and associated elements through the world’s oceanic systems is one cornerstone of the understanding of the linkage between climate and oceanic processes as well as the role of the long-term reservoirs of the deep waters and the bottom sediments. It is still one of the major goals of global ocean research to reconcile rates of surface water production and rates of vertical export with data on benthic turnover to arrive at a full description of transport, burial and turnover of matter within ocean basins. Furthermore, the deep sea ocean fluxes, albeit much smaller than those in surface waters, can be measured directly at a physical boundary and are less subjected to annual variability or short term variation. Therefore, they represent average flux rates and mirror, with some aberrations, average surface water productivity. It is to be expected that different surface water productivity and export in the biogeochemical oceanic “provinces” as defined by Longhurst (1995) influence the benthic regions and give rise to different biological and geochemical characteristics at the sea floor. Topography of the sea floor, deep sea currents and proximity to continental margins are likely to be further forcing functions for deep sea processes.

The deep sea floor is furthermore increasingly at the focus for potential exploitation. Oil drilling is now technically feasible in water depth up to 2000 m and the deepest commercial production well is located at 1853 m in the Roncador field off Brazil (Offshore Engineer 1999). Although dumping of wastes is at present not allowed due to the London Dumping Convention studies are being persued to test the potential for controlled sequestration of CO₂ in the deep sea (Brewer et al., 1999, in press) or for the deposition of municipal and other wastes (Angel and Rice, 1996). Very intense ship traffic adds further anthropogenic impacts to the deep sea floor due to accidental loss of ships and cargo. At present no adequate data base and no tool exists to identify particularly sensitive areas in the deep sea and, thereby, aid political, economical and legal decisions. The deep sea floor has been generally recognized as a key global environment and improvement of the knowledge about this environment has been recommended as one European Grand Challenge in marine research (Le Pichon, 1995, Lochte, 1995).

Many data sets have been collected in the deep sea, particularly in the Atlantic, by different projects and have never been compiled beyond the individual project data management. Therefore, no common data base exists to achieve large scale analyses and assess deep sea processes in the Atlantic as a whole, connected system. The biological and geochemical key parameters describing standing stocks and rates of turnover can for practical reasons only be obtained at a few selected stations, hence, extrapolation of such data from individual points/stations to a larger spatial scale is a difficult task. It requires determination of empirical correlations or modelling of processes which link these limited data to ”master” variables for which a large data base is already available. In this way, the global ocean flux of particulate organic carbon was assessed in a first attempt from a limited set of benthic data (Jahnke 1996). Flux, turnover and burial of organic carbon in deep sea sediments was also assessed globally based on surface water primary productivity and sedimentological data (Romankevich et al. 1999). These two approaches were the first comprehensive attempts to link surface water and deep sea fluxes of organic carbon.

2 Aim of the project ADEPD

Aim of the project was to build up a joint data base for deep sea benthic data from a variety of sources and conduct preliminary geographical analysis of these data. The emphasis was placed on the North Atlantic, since it is this area for which the most comprehensive data sets are available from British, Dutch, French, German, Russian and American studies. Furthermore, it is the most perturbed region in the Atlantic due to intensive human activities. Equatorial, South and northern North Atlantic were included to arrive at a complete description of the whole Atlantic Ocean.

Specific objectives:

• to compile biogeochemical data from Atlantic deep sea sediments (benthic boundary layer) from various projects and from the literature.
• to convert data to common units and uniform variables.
• to link biological (biomass) to geochemical (fluxes of chemical species, etc.) data.
• to extrapolate data of biogeochemical processes at the sea floor obtained at individual stations to a basin wide scale using empirically established correlations to widely measured ”master” variables.
• to develop for well studied regions areal descriptions of ”benthic biogeochemical provinces” in the deep sea of the Atlantic.
• to compare the estimated flux rates at the sea floor with data on surface water productivity and sedimentation.
• to identify gaps in regional coverage as well as in scientific approach to be considered in future projects.

3 Approach

The project organized two workshops:

• Workshop 1: number of data sets, variables and format of data to be compiled were determined
• Workshop 2: progress of data compilation and their geographical coverage was assessed, the analysis of the data was discussed.

All further contacts were via e-mail and the home page of the project (http://www.pangaea.de/Projects/ADEPD).

A list of participating institutions with all scientists involved is given under participants.

The data were delivered to the coordinator of the project and read into the data information system PANGAEA by the data curator and the partner at the AWI, who is managing and servicing PANGAEA. Via the project home page all participants had access to the data bank and retrieval of larger data sets was supported by the data curator. The new developments and the utility of the PANGAEA data informations system are described under “Developments and progress in the information system PANGAEA”. Under “Data collection” an overview of the data collection and the geographical coverage of data is given.

One of the major problems when comparing data from different projects is the use of different methods. It was a major task of this project to convert different types of measurements to common units. This applies in particular to biological measurements where we have attempted for example to obtain common biomass data from as diverse measurements as ATP, phospholipids and microscopic counts. This problem also applies to other types of variables, but is perhaps less obvious. Under “Total microbial activity and biomass” conversions to microbial biomass are described, as these represent the most difficult manipulations.

Linking of biological and geochemical data was achieved by comparing data from the same geographical location irrespective of sampling time. Therefore, seasonal variations had to be ignored in this analysis. If samples were not from the same station, which was only possible in some cases, data averages from geographical grids (1°x1° or 3°x3°) were compared. Since on most cruises geochemical and biological research is not carried out jointly, the number of data pairs which can be used for statistical analysis was surprisingly low. The results of these analyses are shown under "Total microbial activity and biomass”, “Meiofauna”, “Empirical relationships between pigment concentrations and oxygen flux”.

Since the main task of this project was to build up a joint data bank of deep sea biogeochemical data, only first steps of statistical and geographical analysis of these data could be achieved within this project. Regional analyses of benthic fluxes are presented under “Benthic fluxes of nutrients and oxygen”, “Overview of regional analysis of benthic fluxes in the South Atlantic”, “Global benthic fluxes of oxygen and carbon”. It has to be pointed out that due to the complexity of the task only some partners were involved in the analysis stage, but that this work would not have been possible without the joint effort of all partners in bringing the data together.

Recommendations for future data collection and analysis as a result of our discussions and experiences within the project ADEPD are given under “Recommendations for future research”.

All cited literature for all these sections a given under “References”