Computerisation of herbaria in southern Africa has been a core activity within the SABONET Project since 1997 and we have published regular updates on the progress of this important output. Maps showing the geographical distribution of specimens computerised by several southern African herbaria were included in both the August 1998 and December 1999 editions, and bar graphs were included in the August 1998, December 1999, and December 2000 editions of SABONET News (Arnold & Willis 1998, Willis & Arnold 1999, Siebert & Willis 2000).
The importance of computerisation was emphasised during the Project’s Midterm Review in 2001 (Timberlake & Paton 2001). The herbarium specimens computerised by the participating herbaria of the region are of great importance to southern Africa, as this is the first attempt ever to compile a comprehensive, organised database of all known plant taxa housed within selected herbaria of the sub-continent. As the SABONET Project is coming to a close, we will have to look at improving our computerisation rate during the final months. Indeed, Decision 2 of the Tripartite Review Recommendations of August 2000 and Recommendation G of the Midterm Review of February 2001, requested that IT issues be given priority and sped up and that the level of training be addressed where possible. This activity has become a major priority for the 2001–2002 period.
We define computerisation of a southern African herbarium as the process whereby the information on the data labels of herbarium specimens is entered into the PRECIS Specimen Database. A Data Entry Clerk is responsible for removing specimens from the herbarium cabinets and typing the relevant information on the specimen label into the database. The encoded information is then stored and managed in computer databases throughout the region. However, the computerisation process to capture, encode, or digitise herbarium specimen information is obviously prone to problems as a consequence of human error.
The following report is an evaluation of the status of data capturing in the participating herbaria of southern Africa. It investigates some of the problems that can occur, and that negatively affect the quality of the data or hamper the computerisation process.
Considerable progress has been made with the total number of specimens computerised since the start of the Project. To date, the participating herbaria have computerised a total of 375,000 specimens as part of SABONET. Up to May 2002, the best progress was made by SRGH (National Herbarium of Zimbabwe)—they had computerised 104,000 specimens. WIND (National Herbarium of Namibia) have finished computerising all their specimens, approximately 75,000. NBG (Compton Herbarium, South Africa), MAL (National Herbarium of Malawi), and NH (Natal Herbarium, South Africa) have also made excellent progress and have computerised 55,000, 28,000, and 27,000 specimens respectively. The National Herbarium of South Africa (PRE) started computerising their specimens long before SABONET was initiated and currently has the largest database of computerised specimens. They have computerised 65% of a total of ca 1,200,000 specimens. The National Herbarium of Swaziland will probably become the second herbarium to complete the computerisation of their collections, as they have already entered the information of approximately 6,800 of a total of 7,400 specimens. We believe that these statistics show that the computerisation initiative is successfully developing into a major taxonomical capacity-building programme.
A regional decision was taken at the Fourth SABONET Steering Committee meeting held during September 1997 in Zomba, Malawi, to focus initially on computerising the grass specimens (Poaceae) in the participating southern African herbaria. Although many problems are hampering the computerisation process, approximately 160,000 grass specimens (94% of the total number) housed at participating SABONET herbaria have so far been encoded. Of the 16 participating herbaria, 11 have already computerised all the grass specimens of collections made in their own countries; three have computerised more than 90%. The remaining two herbaria have computerised 25% and 35% respectively—the reason for the low percentage of computerised Poaceae specimens by these two herbaria is that LUAI (Angola) prioritised another group for computerisation and PSUB (Botswana) did not regard computerisation as a priority.
As the computerisation of the Poaceae is a policy decision and of regional interest as self-acquired knowledge, the following proposal was made by the Midterm Review and endorsed by the SABONET Steering Committee at the 10th meeting in Windhoek, Namibia (February 2001): countries should also produce Poaceae checklists for their countries before the end of the Project. As a result, three National Poaceae Checklists are being developed and published by Lesotho (Mr Khotso Kobisi), Namibia (Ms Esmerialda Klaassen), and Zimbabwe (Mr Christopher Chapano).
—by Trevor Arnold & Stefan Siebert
SABONET News 7.2: 92