Vegetation Mapping

   One of the future goals of this research is a system of mapping, not only vegetation by species, but also vegetation communities, so that the habitat for each endangered species may be determined and protected. Another long-term research priority is to develop a hydrological and habitat model of the delta, so the effects of water releases on the riparian and wetland habitats and the wildlife that use those areas, can be predicted. Land cover and land use information in the delta is also needed to target the most productive area for restoration projects. The need for land cover classes is necessary to determine the extent and magnitude of land cover change and to solve a range of environmental problems. Land cover change, especially the growing extent of agriculture and human settlements, impacts the habitability of Earth and is fundamental to changes in climate, carbon and biogeochemical fluxes, hydrologic cycling, surface energy balance, and the overall functioning of the ecosystem (Townshend et al., 1991). They showed that traditional classifications are deficient and conventional methods for collecting ground data are time-consuming, not easily repeatable, and expensive.

Current mapping systems used for vegetation on the Colorado River

   In the case of the Colorado River, there are two vegetation classifications used to assess land cover and land use. The first is the Anderson-Ohmart system and the other is the BoR's Lower Colorado River Accounting System (LCRAS) (Figure 1, APPENDIX A). The Anderson-Ohmart system (Ohmart et al., 1988) is based on a series of studies that described the "community structure" of the riparian corridor from Davis Dam, below Grand Canyon, to the Southerly International Boundary at San Luis. Mapping units are approximately 2 ha in area. Using aerial photography and ground-truthing methods, each map unit is classified into one of several vegetation types based on the dominant species. Then, it is classified into one of six vertical complexity classes based on the proportion of understory, midstory and overstory vegetation. This sytem purportedly captures the major habitat types in terms (primarilly) of bird usage. However, it is only semi-quantitative and cannot be used to actually calculate how many acres of riparian zone are covered with particular species, such as cottonwoods and willows.

   The LCRAS system, by contrast, relies on satellite imagery and attempts to be quantitative at least in terms of water use by crops and phreatophytes along the river (Congalton et al., 1998). However, the resolution of the TM images on which LCRAS relies is too low to capture different community types in the riparian corridor - river vegetation outside the agricualtural areas is simply lumped together as "phreatophytes". Furthermore, the ET estimations for phreatophyte vegetation is based on an assumed (untested) crop coefficient that lumps all plant types together and has only a single step between summer and winter ET. A new mapping system, more quantitative than the Anderson-Ohmart system yet containing biologically significant mapping units, is needed not only for the delta of the river but for the entire lower Colorado River.