Patch Dynamics and Clonal Expansion of Reeds in the Letaba River, Kruger National Park

A multiscale approach was used to study the dynamics of reed patches in a section of the
Letaba River flowing through the Kruger National Park (KNP), South Africa's largest conservation
area. The valley floor consists of a mosaic of patches of reeds, water, sand and
rock. Reed vegetation change in space and time is closely linked to geomorphic change in
the rivers of the KNP, because of reciprocal interactions between reeds, flow and sediment.
Reeds are thus considered a potentially important species around which to formulate management
goals for these rivers.
A study of clonal growth characteristics at the scale of individual reedbeds was combined
with a study of reed patch dynamics at the river reach scale. Mechanisms affecting clonal
expansion across reedbed boundaries were combined with short-term patterns of reed
persistence, loss and gain at the reach scale, and longer-term trends in reed cover, to
produce a conceptual model of proposed relationships between flow conditions and reed
cover in the Letaba River.
Reed areal cover increased between 1988 and 1999. Despite this, reed persistence was
low, and reeds were a dynamic element of the river landscape in the short term. Turnover
of water and reed patches was high, within a matrix of sand. Complex patterns of reed
gain and loss occurred at the boundaries of reed patches, with gain exceeding loss throughout
the period of study.
During a drier period (1988-1996), clonal expansion of patches and establishment of new
patches contributed equally (50% each) to the increase in reed cover. During a wetter
period (1996-1999), the relative contribution of clonal expansion increased to 90%. Reed
persistence was higher in the wetter period due to less reed loss, and reed cover increased
more rapidly.
The relative proportions of reed, water and sand patches did not remain constant over 3
years or over 8 years in the alluvial section of the river studied. At the time and space scales
of this study, therefore, the Letaba River did not display the kind of compositional stability
described in other river systems. This points to the difficulty of applying equilibrium
concepts of either geomorphology or vegetation dynamics to semi-arid rivers.
A clear spatial patchiness was present in the distribution of sediment and nutrient resources
on the macro-channel floor. Reed patches were associated with `current shadow'
features, the characteristic raised elevation of the central, denser parts of reedbeds suggesting accumulation of sediment within reedbeds. Silt and clay particles accumulated
preferentially within reedbeds. The results suggest that this accumulation was achieved
primarily by selective retention of nes within reedbeds, and to a lesser extent by selective
deposition of stones within reedbeds.
Total carbon, nitrogen and phosphorus were all higher beneath reedbeds than in uncolonised
sediment. The distributions of total N and total P were not correlated with
that of total C, suggesting that nutrient accumulation does not simply reflect breakdown
of reed organic matter. It was proposed that the nutrient accumulation recorded in this
study was due primarily to interactions between reedbeds and flow, and to a lesser extent
to the input of reed organic matter. The fact that surface resource distribution patterns did
not always correspond to subsurface patterns illustrates the importance of viewing reed
patches as three-dimensional entities.
A high stem density and limited plasticity in spacer characteristics of Phragmites mauritianus
in the Letaba River were consistent with a consolidation strategy of clonal growth.
Physiological integration within within the clone was, however, limited—a notable departure
from the consolidation strategy.
While Phragmites mauritianus showed little evidence of plasticity in the horizontal spacer
characteristics traditionally measured in studies of clonal growth patterns (internode length,
branching pattern and bud activity), it was proposed that plasticity in rhizome characteristics
related to regenerative capacity is important for reed persistence and clonal expansion
in this habitat. The added vertical dimension of the rhizome system in species such as
P. mauritianus limits the usefulness of horizontal spacer characteristics in explaining patterns
of clonal expansion. Progressive accumulation of sediment above the rhizomes and
upgrowth of reeds to the surface was proposed as a mechanism whereby the depth of the
rhizome layer is increased, thus increasing the density of reed stems and enhancing the
persistence of the reedbed.
Rhizomes at reedbed boundaries did not show particular adaptations for rapid expansion
into available space. Clonal expansion and consolidation of the space aroud established
reedbeds was limited both by features of the rhizomes themselves (high proportion of
dormant buds, low branching probability and shallow depth), and by external factors
(flow disturbance).
Management concerns over reed `encroachment' in the Letaba River appear unfounded.
While rapid reed expansion may occur during favourable periods, conditions suitable for
large increases in reed cover occur relatively infrequently and persist for short periods of
time. It was proposed that wet periods with a high frequency of medium flows and no
large floods are most likely to be associated with large increases in reed cover.