Fire is the most ubiquitous disturbance affecting
terrestrial ecosystems, most prominent in savannas, Mediterranean
woodlands, and boreal forests.
It is an important component of the
Earth system affecting the vegetation, the carbon cycle and the radiative
forcing through fire emissions.
It has the potential therefore to
amplify or reduce initial climatic changes through its different feedbacks
on climate. However, tackling its exact influence on climate still
remains challenging as fire varied in the
past both at temporal and spatial scales, responding to different climatic
variability over different boundary conditions.
To go beyond these issues, my main aim is at
identifying and quantifying the different controls of biomass burning
(fire) through time.
Figure illustrating the different
feedbacks of fire on climate (designed by A-L Daniau)
Figure illustrating natural controls of biomass burning (designed by
Climatic conditions are the primary control of the
incidence of fire, but fire is also influenced by the nature of the
vegetation. Vegetation on the other hand depends itself of climate changes
on timescales from interannual (vegetation
productivity) to multi-millennial (vegetation dynamics and distribution).
Another superimposed control on fire is natural (by lightning storms) or
anthropogenic ignition. Humans have been put forward to explain both
increased and reduced periods of fire through fire use intensification for
ecosystem management. This can lead to an increase of fire, as well as fire
suppression by fragmenting landscapes and reducing fuel load.
I examine how key climatic and vegetation
variables governed biomass burning in regions that are today sensitive to
fire, in particular the Mediterranean region and south-western Africa. I analyse microcharcoal
particles preserved in long and continuous deep-sea
sedimentary sequences, focusing on orbital and millennial time
scales. This approach is worth to draw on changes in
biomass burning directly in relation with vegetation (pollen grains) and
climate at regional scale.