Reducing Drivers of Deforestation

Fire: the main culprit

MCDI assessed the annual carbon losses from different drivers of deforestation – e.g. shifting agriculture, logging and fires – in Kilwa District (report). We found that fire is the most significant driver in the still sparsely populated and highly forested project sites, and 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 our 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. This is because:

  • 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 suppress 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 the frequency and intensity of forest fires in the Village Land Forest Reserves (VLFR) where our REDD project is being implemented.  We achieve this using a community-based fire management programme that focuses on early burning (controlled fires that are lit early in the dry season when fuel loads are lower to reduce risk of out of control fires later on). The programme combines four management techniques:

    1. Early burning a buffer strip, or fire break, of 50 – 100m wide around the entire VLFR.
    2. Patchwork early burning inside each VLFR to ensure that areas that are most vulnerable to wildfires are early-burned every few years; this reduces the risk of these areas burning during an uncontrolled and unmanaged fire later on.
    3. Prioritizing fire management, early burning and prevention at areas closest to farms, 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 our REDD project area because the semi-arid local Miombo Woodlands are fire-adapted and some trees are dependent upon fire for seed germination, as part of their natural life-cycle (e.g. Pterocarpus spp.). Moreover, local communities are experienced with using fire, which is a normal tool used for farming.


    Community-based Fire Management



    We introduced wide-scale early burning by drip torches to Tanzania, an idea that may now be taken up in the National Fire Strategy. To begin with, MCDI field staff will took the 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.