Aim Our main goals were to develop a map of the life zones for the conterminous United States, based on the Holdridge Life Zone system, as a tool for ecosystem mapping, and to compare the map of Holdridge life zones with other global vegetation classification and mapping efforts.
Location The area of interest is the forty-eight contiguous states of the United States.
Methods We wrote a PERL program for determining life zones from climatic data and linked it to the image processing workbench (IPW). The inputs were annual precipitation (Pann), biotemperature (Tbio), sea-level biotemperature (T0bio), and the frost line. The spatial resolution chosen for this study (2.5 arc-minute for classification, 4-km for mapping) was driven by the availability of current state-of-the-art, accurate and reliable precipitation data. We used the Precipitation-elevation Regressions on Independent Slopes Model, or PRISM, output for the contiguous United States downloaded from the Internet. The accepted standard data for air temperature surfaces were obtained from the Vegetation/Ecosystem Modelling and Analysis Project (VEMAP). This data set along with station data obtained from the National Climatic Data Center for the US, were used to develop all temperature surfaces at the same resolution as the Pann.
Results The US contains thirty-eight life zones (34% of the world's life zones and 85% of the temperate ones) including one boreal, twelve cool temperate, twenty warm temperate, four subtropical, and one tropical. Seventy-four percent of the US falls in the ‘basal belt’, 18% is montane, 8% is subalpine, 1% is alpine, and < 0.1% is nival. The US ranges from superarid to superhumid, and the humid province is the largest (45% of the US). The most extensive life zone is the warm temperate moist forest, which covers 23% of the country. We compared the Holdridge life zone map with output from the BIOME model, Bailey's ecoregions, Küchler potential vegetation, and land cover, all aggregated to four cover classes. Despite differences in the goals and methods for all these classification systems, there was a very good to excellent agreement among them for forests but poor for grasslands, shrublands, and nonvegetated lands.
Main conclusions We consider the life zone approach to have many strengths for ecosystem mapping because it is based on climatic driving factors of ecosystem processes and recognizes ecophysiological responses of plants; it is hierarchical and allows for the use of other mapping criteria at the association and successional levels of analysis; it can be expanded or contracted without losing functional continuity among levels of ecological complexity; it is a relatively simple system based on few empirical data; and it uses objective mapping criteria.