Our review of the potential sources of information available to resource managers on the composition and structure of undisturbed riparian forests of Ohio suggests that these characteristics can be quite variable, depending largely on the influence of surficial geology, stream valley geomorphology, soils, and predominate disturbance patterns. Such relationships are similar to those documented by many others, including Goebel et al. (1996), Hupp and Osterkamp (1996), Pabst and Spies (1998), and Bendix and Hupp (2000).
The result is that riparian vegetation is often arrayed in predictable patterns across stream valleys and, with an understanding of the factors that regulate their development, reference vegetative states can be determined.
By and large, while statewide pre-European settlement maps and regional county-wide surveys provide generalized information on the broad forest types that once occurred across the state, they lack the specific information on forest stand structure and vegetation-environment relationships needed to adequately predict reference vegetative states. As such, these sources of information are too coarse for developing restoration templates for on-the-ground riparian forest restoration.
However, the studies of the few remaining old-growth forests in the state elucidate many of the finer-scale vegetation-environment relationships, providing considerable data on the overstory composition of these reference riparian systems. Less information is available on the composition of the herbaceous ground-flora, or the structure of either the overstory or ground-flora vegetation.
Consequently, the available information on reference conditions for riparian forests in the state is limited, making it difficult to identify specific reference vegetation states for riparian areas. Additionally, without some idea of the reference conditions, it is nearly impossible to measure the success of riparian forest restoration efforts.
It should be noted, however, that various small old-growth remnants exist across the state with small streams or rivers (Figure 2). In many of these remnants, some information is available on the riparian vegetation-environment relationships. For example, there are several remnant forests distributed across the state, and information on these forests has been published (e.g., McCarthy et al., 2001); however, no published information was found detailing the composition or structure of the riparian forests, including those at Dysart Woods, Fowler Woods, Davey Woods, and Lawrence Woods (Figure 2). Additionally, there are likely other areas in the state that can be used to help develop riparian reference conditions, including forests owned by The Nature Conservancy and other private organizations.
Based on our review, it is clear that a framework is needed to help identify reference conditions for riparian forests across the state. Hierarchy theory (O'Neill et al., 1986) provides the conceptual basis for an approach we suggest to understanding interrelationships among riparian forests and multi-scale environmental factors that can be used to develop reference or benchmark conditions for riparian areas across Ohio.
When applied to riparian landscapes, hierarchy theory predicts that the upper levels of the hierarchy (e.g., ecoregions, physiographic systems) constrain a complex array of hydrogeomorphic processes that in turn mediate the dynamics of lower hierarchical levels, including stream valley shape, fluvial landforms, and plant communities (Allen and Starr, 1982; Baker and Barnes, 1998; Bendix and Hupp, 2000). Thus, we suggest that through detailed characterizations of the remaining old-growth and least-disturbed mature second-growth riparian forests, reference vegetative conditions for these systems can be developed. Specifically, by quantifying the landscape-scale and local-scale geomorphic features that shape the composition and structure of riparian areas, we can better understand the natural variation in riparian forests and develop suites of reference conditions within and among Ohio's different ecoregions.