A landscape indicator approach to the identification and articulation of the consequences of land-cover change in the Mid-Atlantic Region, 1973-2001

Abstract

Landscape indicators, derived from land-use and land-cover data, hydrology, nitrate deposition, and elevation data, were used by Jones and others (2001a) to calculate the ecological consequences of land-cover change. Nitrate loading and physical bird habitat were modeled from 1973 and 1992 land-cover and other spatial data for the Mid-Atlantic region. Utilizing the same methods, this study extends the analysis another decade with the use of the 2001 National Land Cover Dataset. Land-cover statistics and trends are calculated for three time periods: 1973-1992, 1992-2001 and 1973-2001. In addition, high-resolution aerial photographs (1 meter or better ground-sample distance) were acquired and analyzed for thirteen pairs of adjacent USGS 7.5 minute quadrangle maps in areas where distinct positive or negative changes to nitrogen loading and bird habitat were previously calculated. During the entire 30 year period, the data show that there was extensive loss of agriculture and forest area and a major increase in urban land-cover classes. However, the majority of the conversion of other classes to urban occurred during the 1992-2001 period. During the 1973-1992 period, there was only moderate increase in urban area, while there was an inverse relationship between agricultural change and forest change. In general, forest gain and agricultural loss was found in areas of improving landscape indicators, and forest loss and agricultural gain was found to occur in areas of declining indicators related to habitat and nitrogen loadings, which was generally confirmed by the aerial photographic analysis. In terms of the specific model results, bird habitat, which is mainly related to the extent of forest cover, declined overall with forest extent, but was also affected more in the decline of habitat quality. Nitrate loading, which is mainly related to agricultural land cover actually improved from 1992-2001, and in the overall study, mainly due to the conversion of agriculture to forests and urban. The high-resolution imagery analysis was significant in that it confirmed, at a very local level, the specific land-cover changes that were driving the landscape metrics and model results that were calculated from moderate resolution land-cover data and models. These were generally subtle changes in patch size of agriculture, forest, and urban areas, but had substantial effects on bird habitat and nitrogen loadings. This analysis of high-resolution imagery demonstrates and confirms the important ability of moderate-resolution land-cover data to capture significant landscape-level activity that is directly related to specific metrics of ecological significance. It also demonstrates consistent landscape-scale relationships between data derived from high-resolution, moderate-resolution and landscape-model sources. Finally, many of the areas of improvement and decline in bird habitat and nitrogen loadings appear to be potentially regional in nature and likely reflect some local trend in landscape activity. Although the use of ecoregions as sampling units has been criticized in recent years, these results show that basic changes in Level 1 land-cover categories, such as forest and agriculture, may still reflect ecoregional patterns and considerations at some scale of mapping and analysis. This is a potentially important area for future landscape-indicator research. This and other follow-on research opportunities are discussed.