Emerging Issue Summary
Overview and Management Relevance:
Significant efforts have been undertaken to produce baseline maps of North Slope vegetation. These efforts have included work by the USFWS, USGS, and BLM with Ducks Unlimited, the LandFire Program, Alaska Geobotany Center, BP, ConocoPhillips, and others. In addition, Torre Jorgensen of ABR Inc. has produced an ecotype map of the North Slope that can be used to infer vegetation community composition based on landscape parameters. These map-based products, combined with a completed NSSI land cover map (which is being coordinated with many of these same entities) and perhaps some additional baseline mapping on State and private lands, will provide an excellent starting point for developing a change detection program at the North Slope regional scale.
In some cases, differences in scale, vegetation classes, and other aspects of methods will make it difficult to combine maps from different sources. Nevertheless, the logical next step for the NSSI is to integrate the various baseline maps into a regional product. This would consolidate existing data, lead to suggestions for a standardized mapping approach, and identify areas that may need further mapping. Part of this effort would include an assessment of the methods used, the degree to which ground truthing has been done, and some attempt to deal with the problem of heterogeneity of vegetation within pixels (typically 30 m by 30 m) that make up the smallest units in many of the map products.
One recommended approach for addressing vegetation heterogeneity within pixels is to develop fractional vegetation cover products for key vegetation types (e.g., percent shrub cover, percent bare ground, percent herbaceous cover). Fractional vegetation cover products have the advantage of increased flexibility (for instance, land managers could choose their own threshold values to create maps identifying vegetation of interest) and are potentially more sensitive to subtle changes in vegetation cover. Development of fractional vegetation cover products for the entire North Slope will likely require additional field data. Using other remote sensing datasets in addition to Landsat can enhance vegetation mapping efforts. Finally, acquisition of remote sensing data with matching vegetation phenology can enhance North Slope-wide vegetation mapping.
However, compilation of map data is not the only important step forward. It is also important right now to ascertain whether we have captured enough detail in the existing map inventories to allow for the sort of change detection needed to address management questions. Unfortunately, actual use of the expensive and extensive map-based vegetation data for change detection has to date been largely absent. As a result, many issues related to the scale, level of detail, and accuracy of the map products have yet to be fully resolved. North Slope lands include tundra, barrens, and shrublands. Trajectories of change within each of these ecotypes differ markedly, but through a regional approach, broad trends are likely to be delineated.
Most of our current understanding of broad trends in changing North Slope vegetation has been the result of academic research programs. These have been based on satellite remote sensing (specifically temporal changes in NDVI) and comparative photography; or have resulted from limited studies in or near Barrow, Prudhoe Bay, and Toolik Lake that were undertaken for project-specific purposes, rather than historic inventory maps.
Long-term monitoring sites may provide a level of detail that cannot be seen through mapping programs or other methods. Information about existing long-term vegetation sites has not been compiled. While the number of sites and their locations are not known, it is reasonable to assume that more permanent sites are needed and that, ideally, a standard method should be established. Therefore, we recommend as the next step that NSSI tackle directly the problem of compiling existing long-term vegetation site locations and data as well as work toward the development of a minimum standard methodology for data collection. One advantage of tackling this problem at the NSSI level is that it will allow for regional detection of change.
The recommended detection of change alone could prove to be extremely useful in management decision-making where North Slope vegetation is involved, but of even greater utility would be mechanistic models of vegetation response to climate change, disturbance (human or natural), and fire that could be used to predict future rates of change. Two approaches to mechanistic modeling that might be worth exploring rely on a) retrospective studies, and b) climate-based models of vegetation distribution (often known as the Köppen Climate Classification, these climate-based models are already being investigated by a group from Woods Hole).
The issue of changing vegetation has a direct impact on a number of high-profile management issues and is therefore important to pursue. These impacts include:
No formal change detection plan currently appears to be in place. We therefore recommend moving as soon as possible beyond the inventory stage in this area and beginning to attempt to delineate change through formal programs. This will, at worst show the shortcomings of data and methodology; and at best provide change trajectories.
Another group of management concerns revolves around the ramifications of changing vegetation on permafrost, herbivores, food-webs, birds and so on. Sadly, we are still struggling to figure out what the “new” vegetation community composition is likely to be, and until we can master that, determining ramifications of future changes remains elusive. What might be possible is to assemble a group of experts and have them work out an “intuitive” system model wherein ramifications of changes in vegetation on other systems can be suggested, similar to an effort funded by BP and undertaken by The Nature Conservancy about 5 years ago. This would be an exercise in scenario modeling, largely based on analogies, but it could prove quite informative.