these restoration measures to take place. Many of these pothole sites could be described as falling in the lower right portion of Figure 6.3.
Bottomland hardwood forests of the southeastern United States have undergone rapid reductions in area and changes in composition (Box 6.4). The lower Mississippi River floodplain is an example of a large-scale disturbance in which the physical condition of the wetland area has been altered and cumulative impacts have occurred. Large areas of mature forested wetlands have been removed or totally disrupted through diking, draining, and clearing for conversion to agriculture, and through urban encroachment. Extensive federal water management and flood control programs in the Mississippi watershed have altered the basin hydrology on a large scale and have allowed major changes in land use to occur. Thus, within this landscape, only 23 percent of the area of floodplain forest remains (Tiner, 1984). The converted agricultural lands typically retain the natural alluvial soils even though the hydrology of these floodplain and backwater areas has been substantially altered. If crop cultivation ceased, wetland vegetation (including forests) would be reestablished eventually in much of the area because the poorly drained alluvial soils hold sufficient moisture. Blockage of small drainage ditches and canals and breaching of levees or dikes would enhance recovery of riverine overflow hydrologic conditions. These former forests could be characterized as falling somewhere in the middle of the model shown in Figure 6.3 and would require a longer time for regrowth than would marshes.
Reforestation with bottomland species is being undertaken in several large-scale efforts on federal lands, with wildlife habitat and increased timber values as the goal. Much of the converted wetland area in the lower Mississippi valley is privately owned, however. Furthermore, removal of large water-control structures is not politically or economically realistic. Thus, the major constraints on wetland restoration in the lower Mississippi valley are not so much physical or technical as institutional and economic.
If the model shown in Figure 6.3 is correct, then the two disturbance variables (for the site and the surrounding landscape) can be extremely useful in predicting the restoration potential of various wetland systems. Understanding the factors that limit restoration potential can lead to setting realistic goals for systems that will be most difficult to restore and to making better decisions when restoration is proposed as mitigation for further destruction of wetland habitat.
Because some landscapes have lost the majority of their wetland area to irreversible uses, restoration opportunities may be few in these areas. The remaining degraded systems may be highly stressed, yet