Reducing Drivers of Deforestation

Fire: the main culprit

MCDI analysed the different drivers of deforestation in Kilwa District to estimate annual carbon losses (report). After evaluating the different drivers – e.g. shifting agriculture, logging and fires – it was determined that fire is the most significant driver in the still sparsely populated and highly forested project areas. We estimate that roughly 60% of our project landscape burns each year. This happens mostly during the mid-to-late dry season when new farms are cleared – fire is used as the tool to do this, and often burns out of control beyond the areas selected for farming, spreading across large areas when steady breezes blow.

A model developed by MCDI’s partner, University of Edinburgh, suggests that 0.5 - 1 tonne of carbon can be lost from dry forests every year as a result of regular hot fires from a combination of two mechanisms:

  • Hot fires substantially increase tree mortality rates - the premature death of just one or two large trees in a year can amount to considerable decreases in carbon stocks.
  • Regular hot fires retard regeneration, slowing biomass recovery following large tree deaths.

Reducing Fire as a Driver of Forest Degradation

In order to generate carbon offsets, we aim to reduce both fire intensity and frequency in the Village Land Forest Reserves (VLFR) where our REDD projects are being implemented.  We will achieve this by launching a community-based fire management programme that focuses on early burning (burning early in the dry season when fuel loads are lower to reduce risk of out of control fires).   Specifically, we will introduce four management techniques:

  1. Early burning of a buffer strip or fire break of 50 – 100m wide around the entire VLFR.
  2. Patchwork early burning inside each VLFR to ensure that every part of the VLFR that is most vulnerable to fire will burn every few years to reduce the risk of these areas burning during an uncontrolled and unmanaged fire.
  3. Prioritizing fire management, early burning and prevention at the areas closest to farming areas or where prevailing winds are most likely to bring wild fires.
  4. Carry out additional burns later on where grass fuel loads have unexpectedly returned and need to be managed.

This fire management approach is appropriate for this type of semi-arid miombo ecosystem, which is fire-adapted and some species are dependent upon fire as part of their natural life-cycle for seed germination (e.g. Pterocarpus spp.). Further, community members are experienced with fire, as this is a normal tool used for farming, so extensive training will not likely be required.

Community-based Fire Management

At the beginning of this programme, MCDI field office staff will lead on fire management at the project sites.  However, over time we hope to train community teams to manage the process themselves, thus controlling costs and allowing for early burning across a wider area.  We will raise awareness with communities early on about the importance of controlled fires and what this means for maintaining a healthy forest.