1995). A Vermont study in a mountainous landscape advocated widths of 75–175 m to optimize bird diversity (Spackman and Hughes, 1995), while a South Carolina study in bottomland hardwoods described the need for riparian areas 500 m wide (Kilgo et al., 1998). And these examples are drawn from only one taxon: birds! Brinson et al. (1981) graphically summarized the spatial distribution of selected riparian vertebrates in relation to streams, with distances ranging from a few meters in the case of salamanders to several kilometers in the case of herons (Figure 5-13). Far-ranging species such as cougar or bear might require widths that are measured in kilometers (Smith, 1993). Several tables of documented habitat widths for a variety of taxa are found in Verry et al. (2000). Increasingly, such information on widths, combined with data on reference sites, can be used by resource managers in their design of leave areas or engineered buffer strips, although the wide disparity in desired riparian widths associated with various taxa presents an ongoing challenge.
Any question of width must also include concern about negative effects associated with fragmentation of habitats, often loosely grouped under the term “edge effects.” Edge effects can result from reduction in habitat area as well from increases in isolation of habitat patches, predation and parasitism, and disturbance from adjacent lands (Noss, 1983; Robbins et al., 1989; Robinson, 1992). The linear nature of riparian areas, and the abrupt transitions between some uplands and riparian lands, make them particularly susceptible to such effects. Riparian areas severely affected by fragmentation and edge effects may cease to function as breeding habitat—particularly for birds (e.g., Robinson, 1992; Trine, 1998). Work in Maine conservatively estimated that negative edge effects adjacent to a clearcut extended 25–35 m into the adjacent forest (Demaynadier and Hunter, 1998). A bird community will persist, even in the narrowest riparian fringe, but often it is not the community that would have typified the riparian area in a less degraded condition. Clearly, a landscape approach is needed to address edge-effect issues when managing riparian areas for optimal biodiversity.
Management decisions in riparian areas must consider the potential simplification of habitat that can characterize degraded riparian areas. A loss of plant species, a reduction in understory diversity, the elimination of flooding or other disturbances such as fire, a reduction in amount of snags and woody debris, and disruptions of ecosystem continuity by roads, trails, or recreational facilities can lead to a parallel decline in animal diversity. In disturbed riparian areas, often only the most tolerant species remain, as was shown in a study from an Iowa agricultural landscape (Stauffer and Best, 1980). Likewise, in a comparative study of grazed and ungrazed riparian areas in Colorado, mammal and bird species needing more complex vertical structure and lush herbaceous understory were displaced in grazed areas by more tolerant species such as American robin and deer mouse (Schulz and Leininger, 1991). Similar findings of lower bird diversity and species richness, coupled with an increase in a few common spe-