Abstract
Characterizing forest structure and composition at the landscape scale can assist in understanding the impact of global phenomena such as climate change on complex ecosystems. In this study, remote sensing data and output from a computer forest succession gap model are compared with field results in a vast and biologically diverse area in Far Eastern Russia. Projected climate change is applied within the model to determine the effect on the trophic network of the endangered Amur, or Siberian, tiger.
The forest succession gap model FAREAST has been validated at 31 sites in Far Eastern Russia, where simulated results matched forest inventory data at 23 sites in terms of the four dominant tree genera by biomass. FAREAST produces old growth forests using inputs of temperature, precipitation, site quality and species characteristics.
To calibrate and explore the model’s capabilities, output at 1,000 randomly selected points was compared with land covers developed from a synthesis of the Moderate Resolution Imaging Spectrometer (MODIS) land cover product, the Global Land Cover 2000 (GLC 2000) map, and a Russian forest map. Comparing remotely sensed with modeled land covers suggested that disturbance occurred at 461 of the 608 points (76%) where forest types differed.
FAREAST estimates of canopy height (based on the tallest tree) and biomass were compared with the ratio of canopy height to biomass derived from the Geoscience Laser Altimeter System (GLAS) on ICESAT. The LiDAR-derived ratio was continuous with the FAREAST results (y = 6.3178e 0.1004x, R2 = 0.9468). This reaffirms the confidence with which both approaches can be used to estimate forest biomass from canopy height.
FAREAST results were compared with field data at eleven sites in three strictly protected reserves. Simulated total basal area approximated observed at seven of eight sites with primary old growth forests, and the model was able to correctly identify forest type at all old growth sites. FAREAST results most closely approximated observed at the southernmost sites, close to Changbai Mountain in northeastern China, where the model was developed and verified. The presence or absence of large Korean pine trees at field sites accounted for much of the difference between modeled and observed results at other sites with old growth forests.
Model parameters were adjusted for climate change to represent the approximate average temperature increase for the research area in 2070 – 2099 compared with 1961 - 1990 under two Intergovernmental Panel on Climate Change (2007) scenarios, B1 (+3.5 oC) and A1F1 (+6.0 oC). Testing was conducted to determine the temperature increase that would cause the disappearance of Korean pine, a critical food species for Amur tiger prey. Climate change of 1.5 – 1.8 oC at the southernmost field sites in Ussurisky Reserve would cause Korean pine to decline, and warming of 2.4 – 2.8 oC would cause it to disappear. Climate change up to 3.9 oC would allow Korean pine to expand range in Sikhote-Alin Biosphere Reserve in Central Sikhote-Alin, but the species would disappear with climate warming of 4.2 oC. At the northernmost sites, warming of 2.1 – 2.5 oC caused Korean pine to decline and 2.5 – 3.2 oC warming resulted in its disappearance. At 1,000 random points representing sites across Amur tiger range, warming of 3.5 oC caused dark conifer forests to convert to mixed deciduous broadleaf/Korean pine/conifer within 60 years of the new temperature regime (160 years from present) and southern mixed forests to change to deciduous broadleaf. With 6 oC warming, mixed and dark conifer forests convert to deciduous broadleaf, and forests disappear at about 10% of sites. These results are consistent with simulations for the field sites. Atmospheric temperature is expected to warm more quickly in East Asia than globally, and global average temperature increases projected for the end of this century (~1.8 oC for B1; ~4.0 oC for A1F1) would cause these levels of warming to occur in the study area.
