Irrigation plays an important role in the agricultural production of South Africa, using 52% of available water resources in South Africa during 1986. Water losses from irrigation schemes are therefore of great concern. It is estimated that over 25% of water released into the irrigation schemes is lost during peak demand periods due to blockages caused by aquatic weeds. The weed of greatest concern is the green alga Cladophora glomerata ("draadalg").

At present, this alga is controlled by mechanical removal from gratings and other structures, and the combined use of copper-sulphate and sulphuric acid

Biological control of weeds is the use of living organisms to control unwanted plants. A number of biological control programmes directed at various aquatic weeds have been successful worldwide, however few have investigated the use of pathogens against an alga.

The overall goal of this investigation was to determine what pathogenic organisms occur in Cladophora glomerata in South African irrigation canals, and whether any of these should be further investigated as biological control agents for this problematic alga.

The aims of the project were:

1. Determine which natural enemies of Cladophora glomerata occur in natural populations of the alga in South Africa.
2. Identify these natural enemies.
3. Screen these natural enemies for their potential as biological control agents of the alga in laboratory and glasshouse tests.
4. Determine the host range of these natural enemies.

In retrospect these aims were too ambitious. Aims 1) and 2) were successfully completed, and 3) was partially completed. Aim 4) was not completed.

A number of field trips were conducted at a number of government irrigation schemes where Cladophora glomerata has been reported as being problematic. Algal samples were collected at each of these schemes, and observed by means of dissecting and light microscopes to determine what potentially pathogenic organisms occurred in the algae. Fungi were isolated from small pieces of C. glomerata that showed signs of disease.

A wide diversity of organisms were observed. Nematodes were often present, these were identified as a species of Prodorylaimus. These may feed on algae. Insect larvae, possibly a species of midge, on occasion appeared to feed on the algae and make shelters out of the frass and chewed pieces.

Bacteria were extremely common, but did not appear to cause any disease.

A number of fungi belonging to the groups Chytridiomycetes and Oomycetes were observed. A number were observed on only one occasion. However a chytrid, possibly a species of Entophlyctis, was commonly observed, as was a species of Olpidium. Although these fungi can not be grown on artificial growth media, they may be important pathogens of C. glomerata and therefore a useful component of an integrated control programme.

Fungal hyphae were also commonly observed in C. glomerata filaments, including in still alive cells. The chloroplasts in these cells appeared to be in the process of degenerating. The most commonly isolated fungi from the algae proved to be various species of Pythium. It is therefore assumed that these hyphae observed in the diseased algal filaments belonged to Pythium species, and that they are pathogenic.

Pythium isolates were divided into various groups depending on type of zoosporangia produced, production of oospores or hyphal bodies, and growth rate at 15 and 25°C. Eleven such groups were identified. Only one isolate produced oospores, as taxonomic detennination depends on such structures, none of the groups were identified to species level.

Twenty three isolates were tested in initial pathogenicity tests, most proving to be able to penetrate slightly damaged cells of C. glomerata within a 24 hour period, confinning their pathogenic status. However after 48 hours, the longest period of observation, no adjoining healthy cells had been penetrated.

Copies of this report are available from the Foundation, price £15.00, less 20% to FWR members.