Poking their tender tips into the hostile world, matumi seedlings choose the unlikeliest spots to germinate, unfurling their leaves happily while rooted in cracks in bare rock. The tenacious nature of Breonadia salicina (mingerhout) made it an ideal subject for scientists trying to use the environment to measure how sedimentation was affecting the Sabie River in the Kruger National Park (KNP). The decreasing amount of water and increasing amount of sediment entering the park through the Sabie River is of concern to park management, as it can have implications for the biodiversity of the plants that grow in and alongside the river.
Some, like the matumi tree, need rocks to thrive while others prefer more sediment. Increased silt could harm the bedrock-reliant vegetation. To predict when it was time to be concerned about the sediment load in the Sabie, a computer model using the matumi tree as an indicator was developed in the 1990s to forecast when a threshold of potential concern would be reached. However, the floods of 2000 may have swept more than debris downstream – the raging waters may also have serious implications for their electronic watchdog.
The floods have swept away innumerable matumi trees of all shapes and sizes, leaving only a tiny fraction of the previous population in some places. Overall, in 1990 there were 914 matumi trees per hectare. Immediately after the flood, the density was reduced to 37 trees per hectare. Lisa Dowson from the Centre for Water in the Environment at Wits University has visited 16 carefully selected sites all along the 120km of the Sabie in the park, recording the matumi trees in each, from the tiniest seedlings to the tallest mature trees.
Her measurements of each individual will be used in her Master’s thesis to determine if the “Breonadia model” is still accurate. Having left four years for the matumi seedling population to re-establish itself, Lisa came to the park armed with tape measure and measuring stick in September last year. The river was categorised into four main channel types. The rapid system describes areas that have bodies of clear, calm water that end in the water spilling over rocks.
Bedrock anastomosing (interlinking) sections have the river flowing through rocks that direct the water into certain channels before leading them back into a single river, while the mixed anastomosing sections are characterised by more vegetation, with both rocks and sandbanks directing the water flow. The last river type is known as braided. In these sections the river is divided into several channels within a large, flat, sandy riverbed.
Four sites in each river type were chosen, and Lisa spent days tracking down matumis of all shapes and sizes. As well as measuring the diameter and height of each tree, she made note of any flood damage, and if trees had been uprooted or tilted by the flood. She checked to see if the trees had responded to this damage by producing many new shoots, or coppicing. Earlier this year she returned to the park to see if the trees in the chosen sites were flowering, or if the flood was still influencing the trees four years later.
Once surveyed, trees of similar sizes are grouped together in categories, and the initial model showed that there were many very small saplings. As the trees grew larger, the number of individuals in each size group decreased dramatically. Very few large trees are found in the population. This produces a graph that looks like the mirror image of the letter J. As the matumi seedlings germinate and survive best in rocky outcrops, the tree grows in both the pool-rapid and anastomosing sections of the river.
However, with the most exposed rock, the pool-rapid parts of the river were chosen as being the most sensitive to changes in sediment load. Since the model was initially developed by James Mackenzie, these areas have been used to monitor sedimentation in the river. Now Lisa has found that after the flood most of the matumis are gone and seedlings have failed to germinate in the expected numbers in the pool-rapid regions.
However, to the surprise of some of the scientists, matumi seedlings have popped up in the mixed and bedrock anastomosing sections of the river in the same pattern that they had expected to find in the pool-rapid areas. Although her work is still in progress, Lisa says that this might mean that for future modelling of river health, the pool-rapid areas may not be the best place to identify matumi seedlings for input into the computer model. As with all good science, answering one question inevitably leads to several more unanswered questions.
Park management may now start taking a closer look at the matumi trees in the mixed and bedrock anastomosing parts of the Sabie river to determine if increasing sediment from upstream is upsetting the downstream vegetation. At the same time, the question of when to return to monitoring the pool-rapid systems which are more sensitive to siltation will be come a priority. Only time will tell when enough water has flowed under the bridge to return the Sabie River to an approximation of its pre-flood state. Meanwhile, the tiny matumi seeds will continue to take their chances and germinate in the most improbable locations, just in time to be counted by the next researcher.