Predicting moisture dynamics of fine understory fuels in a moist tropical rainforest system: results of a pilot study undertaken to identify proxy variables useful for rating fire danger

Authors

  • David Ray,

    1. Tall Timbers Research Station, 13093 Henry Beadel Drive, Tallahassee, FL 32312-0918, USA
    2. Woods Hole Research Center, 149 Woods Hole Road, Falmouth, MA 02540-1644, USA
    Search for more papers by this author
  • Dan Nepstad,

    1. Woods Hole Research Center, 149 Woods Hole Road, Falmouth, MA 02540-1644, USA
    2. Instituto de Pesquisa Ambiental da Amazonia, Av. Nazare, 669, Nazare, Belem, PA, 66.035-170, Brazil
    3. Department of Botany, University of Florida, Gainesville, FL, USA
    Search for more papers by this author
  • Paulo Brando

    1. Woods Hole Research Center, 149 Woods Hole Road, Falmouth, MA 02540-1644, USA
    2. Instituto de Pesquisa Ambiental da Amazonia, Av. Nazare, 669, Nazare, Belem, PA, 66.035-170, Brazil
    3. Department of Botany, University of Florida, Gainesville, FL, USA
    Search for more papers by this author

Author for correspondence:
David Ray
Tel: +1 850 893 9605
Email: dray@ttrs.org

Summary

  • The use of fire as a land management tool in the moist tropics often has the unintended consequence of degrading adjacent forest, particularly during severe droughts. Reliable models of fire danger are needed to help mitigate these impacts.
  • Here, we studied the moisture dynamics of fine understory fuels in the east-central Brazilian Amazon during the 2003 dry season. Drying stations established under varying amounts of canopy cover (leaf area index (LAI) = 0 – 5.3) were subjected to a range of water inputs (5–15 mm) and models were developed to forecast litter moisture content (LMC). Predictions were then compared with independent field data.
  • A multiple linear regression relating litter moisture content to forest structure (LAI), ambient vapor pressure deficit (VPDM) and an index of elapsed time since a precipitation event (d−1) was identified as the best-fit model (adjusted R2 = 0.89). Relative to the independent observations, model predictions were relatively unbiased when the LMC was ≤ 50%, but consistently underestimated the LMC when the observed values were higher.
  • The approach to predicting fire danger based on forest structure and meteorological variables is promising; however, additional information to the LAI, for example forest biomass, may be required to accurately capture the influence of forest structure on understory microclimate.

Ancillary