Vegetation monitoring

Back to Conserving terrestrial ecosystems

We like to think about vegetation as the ‘capital’ required to fund an ecosystem, with its degradation leading to ‘ecosystem bankruptcy’. For this reason, we understand it’s important to keep track of changes in plant species composition and structure by periodically monitoring the vegetation of a reserve.

When Malilangwe’s vegetation was first surveyed in 1994 (Stalmans 1994), 18 plant communities were identified, and 32 permanently marked sites were established for monitoring. Range condition and trend was comprehensively assessed in 1997 (O’Connor 1997). Comparison with historic datasets showed that there had been distinct changes in the herbaceous layer over the past 50 years.

 

The sward had been converted from a diverse community of long-lived perennial grasses (predominantly Cenchrus ciliaris) to one dominated by Urochola mossambicensis, with an associated decline in carrying capacity. The unprecedented drought of 1992 was largely responsible for the changes, but past heavy grazing by cattle and the exclusion of fire were also responsible.

The botanical composition recorded for the 1970’s (abundant Cenchrus ciliaris) was suggested as a medium-term target for management. A more comprehensive vegetation survey, identifying 38 vegetation types, was carried out between 2000 and 2003 (Clegg and O’Connor 2012). The herbaceous, shrub and tree layers were sampled at 201 sites to create a benchmark against which future changes could be measured.

We know that elephants are the primary cause of vegetation change at Malilangwe. Bulls can be very destructive, and their impact on trees is a cause for concern. To measure their effect, sampling plots located in the most heavily impacted vegetation types were resampled in 2004 (Kaschula 2004) and 2014 (Ferguson 2014). The results of these studies revealed unsustainable levels of damage to trees. If left unchecked, elephants were likely to convert Malilangwe’s woodlands to shrublands. To prevent this, the elephant population was controlled by translocating 53 animals to another conservation area in 2008.

Collection of detailed information on the impact of elephants is time consuming and the associated expense necessitates a long-time interval between sampling.

To provide information at more frequent intervals, a ‘quick look’ method of assessing damage to large trees was developed in 2008 (Clegg 2010b). Kaschula (2004) identified four tree species that could be used as indicators, namely:

  • Marula (Sclerocarya birrea)
  • False marula (Lannea schweinfurthii var. stuhlmannii)
  • Wing pod (Xerroderis stuhlmannii), and
  • Tall firethorn corkwood (Commiphora glandulosa).

At Malilangwe, these species are particularly sought after by elephants, and their rate of damage is considered the worst-case scenario, with other species being impacted to a lesser extent.

50 trees of each indicator species were located and permanently marked with a numbered tag. Between 2008 and 2017, the following stats were recorded annually for each tree:

  • Status (alive, dying or dead)
  • Percentage canopy biomass removed
  • Percentage trunk circumference debarked, and
  • Whether roots had been excavated or not.

Time series plots of the monitored variables revealed interesting results. Essentially, corkwood is in the process of being eliminated from large parts of the reserve. These trees are pushed over by elephants to access palatable shoots at the canopy crown. Being poor re-sprouters, the trees don’t recover and soon die. The rate of mortality is also exacerbated by fire. Unlike many tree species, corkwoods don’t have a thick layer of protective bark, which makes them vulnerable to fire.

Consequently, where fires are common, corkwoods are restricted to rocky or bare sites where grass production (which is fuel for fires) is limited. It is likely that corkwoods at Malilangwe expanded their range beyond natural limits in response to a 50-year absence of both fire and elephants (both were excluded prior to the takeover by Malilangwe). With the reinstatement of these factors in 1994, the distribution is once again being restricted to historic refugia.