My recent post, The Whale Trail in the Pandemic, has attracted enormous traffic, by the standards of my site at least. In it I opine that biological control of Acacia cyclops (rooikrans) in De Hoop Nature Reserve is failing because my – admittedly unscientific and subjective – observations of its spread over 19 years of walking The Whale Trail indicate “exponential” infestation. A windsurfing acquaintance of many years ago, who now lives in Switzerland, picked the story up via our LinkedIn connection, and sent it to his friends with whom he had hiked the trail three years ago. One of them knows Prof John Hoffmann of the Plant Conservation Unit at UCT, who is an expert in biological control of aliens, and forwarded it to him. The resultant email response was shared with me again. So I contacted Prof Hoffmann and asked if I could quote his responses in this blog post, because I think it is important that readers of my original post read the expert, scientific perspective as well.

In summary, Prof Hoffmann’s message is optimistic and encouraging: rooikrans is on the backfoot in De Hoop and elsewhere and biological control is working, based on scientists’ data analysis. He points out that although rooikrans seeds prolifically, seed mortality is high in the soil. Fire destroys 90% of the seeds and predation by mice reduces the number of seeds that germinate to a minimum. With the introduction of the gall midge and seed weevil, the numbers of seeds being produced has declined enormously which has tipped the scales against rooikrans.

He notes that my deduction that biological control takes a long time to become effective, is largely correct. The gall midge and seed weevil do not kill the trees. Some external intervention such as fire or mechanical removal is necessary to do that. Based on the fact that rooikrans thickets burn every 15 to 20 years, it will take two fire events, thus 30 to 40 years, before any change becomes obvious. By the time the second fire occurs, there are few seeds left in the soil and more are destroyed by the second fire so that there are insufficient seeds left to replenish the burnt trees. Monitoring progress thus requires sustained, longitudinal studies by scientists over a long period – a period equivalent to the whole research career of an invasion biologist I would think.

Obviously, my opportunistic observations are unscientific and subjective. Scientists have proposed four parameters to categorise biological control outcomes1:

  1. Density – the number of trees in a defined area;
  2. Biomass – the amount of plant material in a defined area;
  3. Area – the extent of coverage by the trees;
  4. Rate of spread – changes in the amount of time taken for the trees to expand into uninvaded patches or beyond the boundaries of their current distribution.

The four parameters are assessed in three outcome categories following sustained observations by experts, by comparing rooikrans’ status according to the parameters before and after the establishment of the biological agent. It sounds a little complicated and it’s a long-term exercise. There appears to be little published data as present because the studies are ongoing. So we have to trust the experts, who have access to the data, when they tell us that the data is showing significant effects on rooikrans.

I speculate that my observations might roughly speak to Area and, at a stretch, Rate of Spread: the area that I observed is larger than before and the spread seems to have accelerated in recent years. However, I don’t have any empirical data to back this sense up with – and so I am left with some cognitive dissonance.


  1. V. Cliff Moran, Costas Zachariades, John H. Hoffmann,
    Implementing a system in South Africa for categorizing the outcomes of weed biological control,
    Biological Control,
    Volume 153,
    ISSN 1049-9644,