Management of the Kruger National Park, in South Africa, acknowledges that savannas are dynamic ecosystems. Complex interactions between edaphic conditions, herbivory, competition, climate, fires and human management coringinfluence ecosystem pattern and process at a range of spatial and temporal scales. Strategic Adaptive Management of the Kruger National Park is based on a heterogeneity paradigm. This approach raises important questions for management and conservation: What determines spatial and temporal variation in savanna tree abundance? Can habitat degradation be distinguished from background variability? What are the critical climatic thresholds and disturbances at which the ecosystems of Kruger park undergoes rapid change? At what point should managers intervene?
Our project utilises palaeoecological techniques like fossil pollen, charcoal and stable isotope analysis to study vegetation change over time-scales of hundreds to thousands of years. We are studying patterns of variability in savanna vegetation, and are investigating the role of climate, fire and disturbance in ecosystem dynamics.
We are working closely with Scientific Services in the Kruger National Park to ensure that our research is relevant to management objectives. Information on long-term variability in tree cover will help ecosystem managers to decide when management intervention is necessary. Our studies of the relationship between fire and savanna vegetation structure, ecotone dynamics, and the influence of historic and recent human influences on the savanna landscape are also relevant to management practices.
The Kruger Environments project has practical applications to ecosystem management, and engages with theoretical debates over equilibrium and non-equilibrium processes in ecology, ecosystem dynamics, disturbance, resilience and complexity. It contributes to a more holistic, integrated understanding of the Kruger Environment, in which ecological knowledge is contextualised according to landscape history, conservation objectives, and societal values.
Fires are a main agent in shaping the heterogeneity of the savanna landscape and understanding the dynamics of fire ecology is important for the ecosystem management of Kruger Park.
The study of fossil pollen grains, spores, microscopic charcoal particles, and 14C dates, from a series of lake or pan sediments in the north-eastern part of Kruger park will contribute to a better understanding of the different mechanisms that shape the savanna landscape over different spatial and temporal scales. A focus will be given particularly to fire management practices. The dynamics between climate and fire over the last 2000 years will be discussed using the palaeoecological data. Research activities will be expanded to the Mozambique side of the park border and this will enable a comparison of vegetation patterns and its linkages to the differences in management on either side of the Mozambique border during the last century.
The project will facilitate the understanding of the role of fires, natural and manmade, and its interaction with climate, herbivory and plant competition in shaping the savanna ecosystem over long term, a knowledge which will be of importance for ecosystem management and will contribute to shape policies of fire management in the future.
Interpretation of fossil pollen records requires an understanding of the relationship between the percentages of pollen found in the fossil record and the spatial distribution and abundance of vegetation in the landscape. The project aims to simulate past landscapes based on fossil pollen counts, by comparing simulated pollen assemblages with actual pollen counts.
Recent major advances in pollen / landscape modelling require accurate weighting of pollen data in order to account for differences in pollen productivity and dispersal between taxa. This study presents figures of pollen productivity and pollen fall speeds for key arboreal and indicator southern African savanna taxa. Pollen assemblages are simulated based on vegetation survey data and pollen productivity and fall speeds. These simulated assemblages are then compared with actual pollen counts in order to refine model parameters. The parameters are used to simulate past landscapes based on fossil pollen counts.
The sensitivity of the palaeoecological record to known disturbance events is tested by comparison of high resolution pollen and charcoal records from the past 80 years with records of fires, floods and rainfall variability, as well as satellite images from the past 30 years. These data also allow the effects of changing management practices on the landscape to be assessed, and will complement vegetation survey data.
By linking present day, decadal and palaeo time-scales in a nested hierarchy of temporal scales, this project, investigates the relationship between fire, climate and landscape change, and improves and refines the interpretation of palaeoecological data, thus facilitating its application in ecosystem management.
The savanna-grassland ecotone consists of different vegetation assemblages at the limits of their environmental and competitive tolerance; it is therefore particularly sensitive to changes in climate and disturbances. The project aims to study the spatial and temporal changes in the savanna-grassland boundary and identify the position of climatic and disturbance thresholds. The project will contribute to the development of conceptual models which explore how feedbacks between climatic and anthropogenic factors can cause phase transitions in ecosystems.
This project uses past vegetation change from analysis of fossil pollen in sediments, and fire history from the analysis of charcoal abundance to study the spatial fluctuations in the savanna-grassland boundary over a timescale of approximately 1000 years. Analysis of stable isotopes from soil profiles give insights into changes in the abundance of C3 and C4 plants over time and will provide complimentary data on the boundary transitions.
The understanding of the mechanisms which causes phase transitions facilitates the need of the ecosystem managers of Kruger park to establish thresholds of potential concern and enable managers to intervene if ecological system or process is close to an intrinsic limit.
We are grateful to the Trapnell Fund, the Higgins Trapnell Family Foundation and the Andrew W. Mellon Foundation for funding this research. The project is hosted by the African Environments Programme and the Long-Term Ecology Laboratory. We thank staff of Scientific Services, South Africa National Parks for their guidance and for logistical support.