How did grassland respond to fire treatments in the Cathedral Peak catchments?

By Paul Gordijn, Field Technician, SAEON Grassland Forest Wetland Node

An understanding of vegetation dynamics is key to unlocking the bigger picture of hydrological and carbon cycles.

Vegetation cover intercepts precipitation and plant roots hold the soil together, helping to prevent the loss of soil, or erosion. Through the process of photosynthesis atmospheric carbon-dioxide (CO₂) is absorbed by plants. The carbon absorbed is assimilated into the soil, creating carbon deposits that other processes and organisms depend on.

The Cathedral Peak area forms part of the Maloti-Drakensberg mountain range, which is known as South Africa’s “water tower”. As the name implies, the mountain range supplies drinking water to many South Africans, being the source of some major rivers such as the Orange (longest river in South Africa) and Tugela. Without vegetation, the Drakensberg would not function as we know it.

The majestic Cathedral Peak mountain range viewed from Organ Pipes Pass. By afternoon, the clouds visible on the right will have developed into large storm clouds. These bring the frequent afternoon summer showers experienced in the Maloti-Drakensberg Mountains. (Photo: Paul Gordijn)

The importance of the Maloti-Drakensberg range was recognised in the early-to-mid 20^th century. At Cathedral Peak fifteen research catchments were established to assess water yield under a number of treatments. The recording of stream flow measurements and precipitation began in these catchments during the late 1940s.

The principal comparison examined was between afforested (grassland catchments were planted with Pinus patula, a commercial pine tree) and non-afforested catchments. The results revealed that afforestation reduced catchment water yield by up to approximately 68% (the average loss was ~40%), opposed to catchments with indigenous grasslands (Nänni 1970). The indigenous grassland catchments had a significantly greater water yield than catchments planted with pine trees. In the non-afforested catchments different fire regimes were applied to test their effects on grassland botanical composition.

Effects of fire

In the early 20^th century, the effects of fire in grasslands were poorly understood. Initially, fire was perceived as undesirable and for a short period the only burning that was done was for fire breaks. Research has since shown that fire is necessary in the grasslands of the Drakensberg. Without fire, grasslands transform into shrublands or forests.

Killikia grandiflora, a rare botanical jewel encountered during the sampling at Cathedral Peak. This species' distribution is limited to a few of the catchments at Cathedral Peak and is found nowhere else. The exclusion of fire has restricted the habitat of this species, pushing it towards the brink of extinction. A beetle visiting the flower. Very little is known about this species' autecology. Photos: Paul Gordijn

The frequency and timing of a burn has been shown to affect grass species composition at Cathedral Peak (Everson 1985). There are a suite of grass species, each preferring a specific fire frequency, from annual burns through to octennial burns. At Cathedral Peak, different fire regimes were imposed from the early 1970s, so that their effect on the grassland vegetation and interaction with the hydrology of the catchments could be examined.

Vegetation monitoring programme

Derrick Tomlinson (former Department of Forestry employee) set up a monitoring programme for the vegetation in the research catchments in mid-1970. Subsequently, his vegetation monitoring plots were resampled by Dr Terry Everson in mid-1980.

Falling within an area of phenomenal diversity, Cathedral Peak boasts a species list of approximately one thousand. Many of the species are not described in botanical field guides. Detailed photos (e.g. the grass Digitaria flaccida) showing some distinct characteristics, such as the ligule, conspicuous venation and sparse hairs (trichomes) on the upper leaf blade of the grass are useful for in field identification.

A SAEON-University of KwaZulu-Natal (UKZN) collaboration team began resampling Tomlinson’s plots in December 2013. The collaboration began with Sue Janse van Rensburg, Coordinator of the SAEON Grassland Forest Wetland Node and Dr Terry Everson of UKZN. The node’s intern, Sinethemba Ntshangase and field technician, Paul Gordijn joined the team, with Dr Everson guiding the work. Professor Tim O’Connor, SAEON’s Observation Science Specialist, also provided valuable advice to the programme.

To ensure that the new data collected would be comparable to Tomlinson’s data, old instrumentation was recreated. Tomlinson used a “levy bridge” for his sampling - essentially a stand, or bridge, that drops sharpened pins perpendicular to the ground. The plant species base closest to the pin base is recorded. Levy bridges typically have ten pins, with equal spacing between the pins (Tomlinson used a bridge with 15 cm spacing between the pins). By placing the bridge systematically within each 18 x 8 m plot, 20 times, a measure of species composition is obtained.

The levy bridge used by Tomlinson was heavy and cumbersome. Known for his ingenious craftsmanship, SAEON Field Technician Matthew Bekker embarked on producing a light-weight, elegant alternative that would do exactly the same job as Tomlinson’s levy bridge. To make sampling efficient, a novel CyberTracker application was made for collecting data directly into a digital format.

Another novel initiative was a specially created digital herbarium. Intricate photos of key grass and forb characteristics were taken to facilitate in field identification. This initiative is set to grow into a valuable resource for vegetation research within the area.

Analysing the data

With the majority of the field work complete, we are looking forward to analysing the data collected. Some changes in species composition are predicted. It remains unknown exactly how the variable fire regimes applied will affect grassland species composition. Particularly, some catchments have experienced a convergent fire regime after the initial different fire treatments were applied. How would this affect the species composition of these grasslands? Do they have the resilience to revert to original species composition? Have certain species been lost? Perhaps some plots have been “taken over” by the grass species (e.g. Festuca or Merxmuella species) that are able to take advantage of the recent increase in atmospheric CO₂ concentration associated with climate change?

Dr Terry Everson of the University of KwaZulu-Natal uses the new light-weight levy bridge created by SAEON’s Matthew Bekker. Photo: Paul Gordijn

The analysis of the vegetation change over the years from Tomlinson’s first study will be valued by all scientists with vested interests at Cathedral Peak. The work will not only provide an indication of the interaction of vegetation composition with fire, but is required as a baseline for the analysis of hydrological dynamics over time.

Moving into the future, it is likely that the vegetation sampling methods will be updated. The outputs will be used to establish a more detailed programme aimed at detecting if any directional shifts in vegetation in relation to climate change are occurring.

This study looks forward to placing an important puzzle piece in the understanding of past and present vegetation dynamics and helping predict future developments in an area known as the “water tower” of South Africa.

Sinethemba Ntashangase, an intern at SAEON, has been able to expand her botanical knowledge of grasslands while assisting with the sampling. Photo: Sue Janse van Rensburg

References

• Everson, C.S., 1985. Ecological effects of fire in the montane grasslands of Natal. University of Natal.
• Nänni, U.W., 1970. Trees, Water and Perspective. South African Forestry Journal, 75: 9-17.