Evidence: This aerial view shows the removal of renosterveld vegetation that has been replaced by farmland in the area between Albertinia and Riversdale. Photo: Homebrew Films Company
The South African Environmental Observation Network (SAEON) has come up with an app to protect the renosterveld vegetation, a critically endangered and the fastest disappearing habitat in South Africa.
Known for its biodiversity, particularly geophytes, the renosterveld is in the fynbos biome in the Western Cape. It is generally good land for crop farming such as vineyards and orchards, which has led to only 5% of its original extent remaining.
“Of its former extent, only relatively small fragments remain, often in areas that are too steep or rocky [to access],” says Cape Town botanist Zoë Poulsen.
Other threats are heavy grazing, mismanagement, invasive alien plants and urbanisation (for example in Cape Town where remnants of renosterveld vegetation occur on Signal Hill), according to the article Renosterveld Shrublands in One Earth.
The Renosterveld Ecosystem Management Plan lists the various types of renosterveld as the Matjiesfontein Shale Renosterveld (De Doorns to Gamka Poort, north of the Groot Swartberg and south of the Great Karoo Basin), Montagu Shale Renosterveld (Little Karoo between Montagu and Barrydale), Swartland Shale Renosterveld (Swartland and Boland areas), Western Rûens Shale Renosterveld (Western Overberg: Villiersdorp to Napier), Eastern Rûens Shale Renosterveld (Eastern Overberg towards Riversdale), Swartberg Shale Renosterveld (northern slopes of the Groot Swartberg), Uniondale Shale Renosterveld Sebrasfontein to Uniondale), Swartland Granite Renosterveld (Swartland and Boland), Breede Alluvium Renosterveld (Breede River from Worcester to Ashton) and the Rûens Silcrete Renosterveld (Riviersonderend to Riversdale).
Data from the observation network showed that between 2016 and 2020, 478 hectares in the Overberg area was lost to illegal ploughing. This is nearly 1% of the surviving renosterveld found there.
Based on rough estimates provided by the director of the Overberg Renosterveld Conservation Trust, Odette Curtis-Scott, fewer than 30 000 hectares of land — which is made up of strips along water courses — out of more than 360 000 hectares of land that the vegetation would have covered, remain.
“30 000 sounds like it could be one massive reserve but it is these little speckles of veld, like one hectare here, five hectares there and in a few places, a few hundreds of hectares together and that is all that is left,” said Curtis-Scott, who is working with farmers on a conservation model.
Glen Moncrieff, a data scientist at the observation network network, said: “Through analysing satellite data collected over the entire region every week, this new app uses machine learning to automatically scan the landscape for any signs that renosterveld is being converted to agriculture. If habitat loss is detected, this information is immediately passed on to law enforcement for further investigation. There is now no way for farmers to destroy renosterveld without being seen.”
It is hoped that this app will create awareness among farmers and the public of the loss of this ecosystem.
Developed over a period of two years, the app has gained international recognition that will allow it to be used in other vulnerable habitats.
“We have recently been awarded a R10-million grant from Nasa to roll out this technology across the entire Western Cape together with collaborators from the University at Buffalo in the US,” said Moncrieff.
“We hope to complete this by 2024. We really believe that this will be a game-changer for conservation in the region, providing a real-time view of the health of all the natural vegetation in the province.”
He added that a collaborative effort was needed to protect South Africa’s landscapes and vast biodiversity of plants and animals.
“Climate change, habitat loss and other threats are leading to the rapid decline in this natural heritage. Our national and provincial environmental protection agencies do not have the resources to address these issues alone,” said Moncrieff.
“Rapid advances in artificial intelligence and machine learning, together with new satellite images, promise to greatly improve our ability to monitor biodiversity.”