Atlantic coast joint venture waterfowl implementation plan
|
|  Yüklə 7.28 Mb.
|
3. HABITAT STATUS AND TRENDS 3.1 Status and Trends of Wetlands Since the settlement of the United States by the Europeans, wetlands have been viewed as impediments to progress, harbingers of insect pests and disease, and wastelands to be drained for farming (Koneff and Royle 2004). For nearly 200 years, this view of wetlands predominated and millions of acres were lost for agriculture, silviculture, development, and insect and disease abatement. Approximately 89.5 million hectares (220 million acres) of wetlands originally covered the conterminous United States at the time of European settlement (Dahl 1990, U.S. Environ. Prot. Agency 1998, Dahl 2000). By 1997 only an estimated 42.7 million ha (105.5 million acres) of wetlands remained in the conterminous U.S. (Dahl 2000). The greatest rate of loss occurred from the 1950s to the 1970s with an average annual net loss of 185,400 hectares (458,000 acres) (Frayer et al. 1983). Between the mid 1970s and mid 1980s that rate had slowed to 117,400 ha (290,000 acres) per year and was further reduced between 1986 and 1997 to 23,700 ha (58,500 acres) per year (Dahl 2000). Forested wetlands have shown the heaviest losses in recent years with two-thirds of the loss due to agriculture, development, and other practices and the remaining one-third lost to conversion to other types of wetlands (Hefner et al. 1994). Most states in the ACJV have lost on average 25% to 50% of their original wetlands with Connecticut losing an estimated 74% while New Hampshire has lost only 9% of their wetlands (U.S. Environ. Prot. Agency 1998). Although the status assessments of wetland change (e.g., Dahl 1990 and Dahl 2000) in the U.S. are informative, they are based on a stratified, randomly selected sample of plots (1,040 ha or 2,560 acres) such that the results provide estimates of change for the U.S. as a whole. One cannot use these data to depict spatial patterns at a scale useful to developing explicit waterfowl management objectives. Towards that end, the ACJV commissioned a study that sought to create a spatially-explicit model of wetland loss (Koneff and Royle 2004). They developed zero-inflated logistic regression models to predict the area of six wetland classes within 1,036 ha (2,560 acres) grid cells throughout the ACJV. The models predict the areal extent of each wetland class in each of four decades: 1950s, 1970s, 1980s, and 1990s. Overall estimates of wetland change for the ACJV between the 1970s and 1990s, showed similar trends to those estimated by Dahl (2000) (Table 3.1). Overall, approximately 8.5% of wetlands were lost or converted to other wetland types between the 1950s and 1970s with more palustrine emergent wetlands being lost than other types. The amount of wetland loss declined between the 1970s and 1990s due to most states enacting wetland protection laws, but the ACJV still lost approximately 5.6% of the remaining wetlands during those two decades. Palustrine forested wetlands suffered the greatest loss during this period, declining by almost 6%. Only two wetland classes, lacustrine and palustrine unconsolidated bottom (i.e., ponds), showed a general increase across the 4 decades, 0.9% and 1.7% respectively. The previous summary ignores the spatially explicit aspect of Koneff and Royle’s (2004) approach. That is, their results allow one to map wetland change with a minimum resolution of 10.36 km2 pixels, thus allowing one to depict where the greatest change for each wetland class has occurred (Fig 3.1). Examination of this map shows that forested wetlands have decreased across a large proportion of the ACJV. It is important to note that there are several important Table 3.1. Predicted area (ha) and percent change of predicted wetland extent based on zero-inflated logistic regression models in the ACJV for the 1970s and 1990s. Standard error estimates for state level predictions currently not available but will be added as they become available. Total wetland area and total percent change are shown at bottom of table. Predicted percent change reflects a combination of conversion both to and from upland as well as to and from other wetland classes. Taken from Koneff and Royle (2004).
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||