Structure of the GPDD database

 

GPDD was reindexed when the database and web site was migrated to new servers.  The data is still the same, but the ID numbers have changed. More information is available here.

 

Taxon Biotope Location Datasource Main Data

The Main table

The Main table is the core of the GPDD database. Each row in the Main table refers to a separate dataset, and contains details pertinent to that dataset only. As well as listing the relevant ID numbers from the taxon, biotope, location and datasource tables, it contains information on the sampling units, the units used for data presentation, the years when the study started and finished and some simple statistics.  The notes field also contains any information thought to be relevant in the context of sampling. A representative example is shown below.

Top

The Taxon table

The Taxon table provides information about the organism sampled in each data set. There is one row for each organism, which might be referenced by more than one dataset. The TaxonID in the Main table contains the appropriate organism's ID number in the Taxon table.  A representative example is given below.

Taxonomic information and classification is contuinually being updated and reviewed for many species. If you have any corrections or updates to the information in this table, we'd be pleased to hear from you.

Conservation Status has been checked using the searchable databases that are part of the World Conservation Monitoring Centre web site (http://www.unep-wcmc.org/).
The animal data is taken from the WCMC Animals Database (http://www.wcmc.org.uk/species/animals/animal_redlist.html) and represents the principle contents of the 1996 IUCN Red List of Threatened Animals.

Film, photographs and audio recordings of endangered species are being amassed and preserved digitally and will soon be available on the ARKive web site. http://www.arkive.org/

Top

The BIOTOPE table

The Biotope table describes the habitat of the organism studied in each dataset.  This enables you to restrict database searches to certain habitats or environments.  There are 181 different habitats listed in GPDD, divided into 21 broad biotope types.  Like the taxon, each dataset's habitat is recorded via the BiotopeID in the Main table, which contains the ID number of the appropriate entry in the Biotope table. A representative example is given below.

Top

The Location table

The Location table provides geographical details of the monitoring site or area where the population sampling was carried out. This enables you to restrict database searches to certain areas, or latitudes or countries. Many of the sampling sites have no particular names, or have local names which are unlikely to appear on most maps. Where possible, therefore, we have included the name of the nearest town, to assist you in finding the sites if you need to. We provide latitude and longitude to the nearest minute for named sites on which we have accurate information. In some cases, the sampling sites cover very large areas of land or sea. In these cases, where no named site exists, we have provided the coordinates of the approximate centre of the area. Caution should therefore be exercised when making use of this information.

The following table lists t he fields in the Location table, together with an example. Note th at fields are empty where information i s not availabl e, or where the field is not appropriate for that record.

Top

The Datasource table

The Datasource table provides a reference to the original source of the data which appear in the other tables.

Population data are presented in published material in many ways. Obviously, it is easiest to extract data from a table of numbers. However, population trends and variations are best depicted in a line plot or histogram and it is in these graphical forms that most data are published. In some cases it has been possible to retrieve the raw data from the author(s) in question, but in others it has been necessary to reconstruct the data from the graphics. This has its limitations, and the accuracy of the derived numbers may be compromised where logarithmic scales are used or where printed copy is of poor quality or where diagrams and graphs are very small. We are using the latest scanning and data retrieval software for this purpose and, generally speaking, we believe that errors due to data extraction are within acceptable limits. The Datasource table includes a field which states the format in which the data originally appeared so users will be able to draw their own conclusions about data accuracy.

In the following table we show the field names for the Datasource table, together with a typical example. Note that Elton himself drew on a variety of data sources for this particular book. We have provided these sub-sources where they are mentioned in our source document. Note also, from the ID numbers in the Notes field, that nearly thirty time series were extracted from this one source.

Top

The Data table

The Data table houses the time series data.  It is the simplest table of all, comprising just three fields: Main ID, Time of sample and Population size. The majority of populations in the database are sampled annually, and it is therefore a simple matter to list the year of sample and the population in each year. There are, however, many variations on this basic sampling regime which we have also taken into account. For example, there is a collection of insect time series in which  sampling took place on a four-weekly basis during the summer months only. There are thus thirteen (52/4) equal sampling periods per year, even though no actual sampling took place in the winter periods. For the sake of time series continuity, we have incorporated the unsampled months into the data table.

In a more extreme case (tropical insects), each of the twenty years' data contains 52 weekly samples. However, for any one species, the great majority of samples are zero counts. To avoid filling the database with hundreds of thousands of zero values we have chosen to sum the data to get monthly totals. A zero in the Population field indicates that sampling took place but no individuals were found, and a blank indicates that no sampling took place in that time period. Unless flagged, all data sets conform to this format.

Not all of the data in GPDD is in the public domain; some are provided under licence from the BTO.  We require permission before we can distribute some of the data sets.  The data in these datasets have been removed from the DATA table in this online version of GPDD, but the metadata in the other tables has been left in place. 

Top