• We examined the relationship between vegetation and landform characteristics and caribou distributions during the calving and post-calving period.:
• We obtained vegetation maps of summer habitat of the PCH and the CAH and incorporated them in a Geographic Information System (GIS) for caribou modeling and scenario analysis.
• We procured annual AVHRR NDVI maps and set them up on the GIS database for the PCH. The NDVI data allow us to correlate vegetation changes on a landscape scale with calving and summer distribution of caribou.
• We think that habitat ruggedness with to calving distributions of caribou of the CAH (Nellerman et. al. 1996).
• NDVI, measures of both plant green-up and biomass, correlate with caribou calving distributions and early post-natal habitat use.
• We have applied NSF Land-Atmosphere-Ice-Interaction (LAII) experimental results on climate-induced vegetation changes from Toolik Lake to a model of vegetation community changes.
• We developed a complete list of major regional vegetation types for five climate zones (boreal forest, treeline, low Arctic, high Arctic, polar desert), four moisture regimes (dry, moist, wet, riparian), and two soil pH regimes (acid and non-acid). Each of these regional types is accompanied by a species list, with associated abundances for each species, and at least a total biomass, partitioned by either species or growth form depending on what data were available in the literature.
• We have modeled a change in vegetation under a climate change scenario. The modeling begins with the assumption that plant community composition and biomass are primarily a function of summer temperature, moisture, and nutrient supply. Given a long enough time, if only one variable is changed, the community composition will eventually shift to the composition currently present in communities with those same conditions. For example, a moist, non-acid low Arctic community will eventually shift to a moist, non-acid treeline community if the climate shifts to summer temperatures associated with treeline. Figure 1 on the next page shows a theoretical landscape with six regional community types present. The maps show percentage biomass of two major growth forms, graminoids and shrubs, for an initial climate and following a complete transition to a warmer climate.

1. Normal – this scenario approximates average conditions given the current climate.
2. Lowland - this scenario is based on the expectation that, over the next decade or so, spring snowmelt will occur earlier, snows will be deeper in the winter, and that insect conditions will be bad. As a result, calving and post-calving would occur closer to the coast.
3. Shrub – this scenario takes us further into the future and represents a shift in vegetation community types to those dominated by shrubs. There is a preponderance of birch rather than willow. The data for this run comes from the George River herd where they are subsisting on just such a mix of vegetation. This scenario assumes deep winter snows and a summer that is - too hot and not humid enough for mosquitoes.

The results of our initial simulations indicate that an increase in the frequency of early snowmelt coupled with deeper winter snows and high insect harassment would result in the Porcupine Caribou Herd declining at an average rate of three percent per year. Under current common conditions, the herd increases at an average rate of three percent per year. Thus, a climate change induced shift in weather patterns could lead to a reversal from population growth to population decline.

• Our initial simulation of the effects of a longer term shift to a more shrub dominated vegetation indicate a reversal from Porcupine Caribou Herd population growth at an average of three percent per year to population decline at an average of one percent per year.
• In year 3 under our joint NSF/US MAB grants, we will incorporate the results of new research results in the caribou ENERGY and POPULATION models. These changes will improve our ability to simulate the effects of climate change and oil development activities on the herd. The changes will affect the model’s treatment of protein dynamics, reproductive strategies for adult female caribou, the ability to simulate variation among years in snow and temperature, and the relationship of human harvests on the herd. We will also produce simplified versions of the models to operate on the internet.