Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Afrique du Sud → 2007 → Long-term effects of fire on nitrogen cycling in a broad-leaf savanna, Kruger National Park, South Africa

University of Cape Town (2007)

Long-term effects of fire on nitrogen cycling in a broad-leaf savanna, Kruger National Park, South Africa

Coetsee, Corli

Titre : Long-term effects of fire on nitrogen cycling in a broad-leaf savanna, Kruger National Park, South Africa

Auteur : Coetsee, Corli

Université de soutenance : University of Cape Town.

Grade : Doctor of Philosophy 2007

Résumé partiel
Fire with herbivory, climate, and soil properties including nutrients are said to be important in regulating the structure and function of savanna ecosystems. Frequent fire is often held responsible for a decrease in nitrogen pools and availability and the maintenance of low fertility conditions. However, previous research in the Kruger National Park (KNP) and elsewhere found conflicting results for the effects of fire on nitrogen pools and transformation rates. The main aim of this study was to gain a better understanding ofthe long-term effects of fire on nitrogen cycling in the KNP. The KNP provided an ideal opportunity for this study because of the initiation of a fire experiment in 1954. The Experimental Bum Plot (EBP) experiment was initiated in four representative landscapes of the KNP to determine the effects of fire on vegetation structure. I tested the effect of burning on nitrogen cycling and productivity in four fire treatments situated in Pretoriuskop Sour Bushveld (broad-leaf savanna). The fire treatments included a late winter, annual bum (August), late winter and summer triennial bums (August and February) and a fire exclusion treatment. Total soil nitrogen, available nitrogen, woody biomass and herbaceous production were measured. 1 hypothesized that vegetation adapted to low N conditions with low N foliage would be expected to dominate in frequently burnt areas if fire was to decrease N pools and fluxes. To test this, I measured grass species composition and accompanying traits across the fire gradient. I also used ols.N of soil and foliage as a proxy of nitrogen cycling and measured ols.N in wood to obtain a longer term perspective of fire treatment on nitrogen cycling. The results show that fire did not affect total soil nitrogen and carbon regardless of soil depth. The high intensity triennial bum decreased the cumulative amount of available N more than the annual bum. As herbaceous production compensate for decreased woody production, frequently burnt savanna managed to stay as productive as fire excluded areas. Less frequent burning, especially in the dry season, was more detrimental to productivity than annual burning. However, the grass composition, daily rates of nitrogen mineralization and ecosystem l)1s.N values were not affected by fire treatment. Longer term signals of nitrogen cycling (l)l values in wood) indicate that the annual burns have been moving persistently towards a more open nitrogen cycle, whereas the fire exclusion plots have had more variable olsN values through time. The effect of fire on the nitrogen cycle in this study is ameliorated by an interaction between low losses of nitrogen and limited species change-over with frequent fire. The persistence of trees, even in the annual bums and the presence of herbivores may contribute to relatively low fire intensities and the conservation of nitrogen. Furthermore, the exclusion of fire has not lead to closed canopy forest. Soil organic matter o\3e values showed that this area has been in a savanna state for a considerable length oftime and were not forest previously.


Version intégrale

Page publiée le 5 février 2019