The link between the terrestrial and the aquatic realms in riparian zones is the most significant factor in their definition... The riparian landscape is unique among environments because it is a terrestrial habitat strongly affecting and affected by aquatic environments; it has a particular spatial configuration; it has use values derived from these features; and like mountain or desert habitats, is diverse in its structure and function among regions while responding to the same primary factors [Malanson].

The Forest Practices Code [Western North America] defines riparian zones as the land adjacent to the normal high water line in a stream, river, lake or pond and extending to the portion of land influenced by the presence of the adjacent ponded or channeled water. The riparian zone is of critical importance to stream ecosystems. The riparian vegetation contributes nutrients and fish food by providing plant material and insects to the stream, regulates stream water temperatures (tree canopy shading) and delivers large woody debris to the stream. The large woody debris provides much of the fish habitat and also contributes to stream channel stability. The roots of streamside vegetation tend to resist stream erosion by helping to hold the bank materials together. Streamside vegetation promotes overbank sediment deposition and also provides hiding cover or refuge for fish.

What are Buffers?

In their essence, buffers are areas designated around a region or object of significance as to shield or cushion it from outside effects. In our context, we are interested in buffers that designate an area to secure the naturally necessary and important interaction and relationship between riparian zones and aquatic ecosytems. In the past, buffers were generally designated, with good intention, simply as static width areas along rivers, many times disregarding the actual riparian zone. Yet, riparian zones are by no means static in width. It is my belief that with the growing realization of our threatened watersheds, proper buffering of riparian zones and rivers will begin gaining the respect and attention it needs.

The noted Tom Schueler of the Center for Watershed Protection pre-emptied a discussion of urban stream buffers this way:

Up to now, buffer requirements have been relatively simplistic - the "design" of a stream buffer often consists of no more than drawing a line of uniform width on a site plan. As Heraty notes, buffers designed in this manner often become invisible to contractors, property owners, and even local governments. As a result, many stream buffers fail to perform their intended function, and are subject to disturbance and encroachment.

Efforts have been underway to create "spatially dynamic riparian buffer models." One such model, of particular interest to me, was created by Kyle et al. and is targeted at resource management, primarily in the western United States and Canada, although aspects of it are surely applicable elsewhere. An abstract from their paper introducing the model probably introduces it best:

The procedural simplicity of constant width buffering in geographic information systems has compelled many resource manager to apply an un-ecologically sound conservation tool. Constant width buffers do not approximate the dynamic character of riparian environments. In this paper, we [Kyle et al] develop a spatially dynamic riparian buffer model (SpaDRBuM) with an edge mitigation component (EMC). SpaDRBuM_EMC is primarily intended as a simple tool for riparian habitat conservation. The SpaDRBum delineates riparian boundaries and the EMC protects those boundaries against microclimatic changes associated with edges. To test the usefulness of the model, we compared the amound of habitat types between a traditional static width corridor and dynamic corridor created with the SpaDRBuM_EMC. We found that the dynamic buffer model increased the proportion of other habitat. Since a spatially dynamic riparian buffer model with an edge mitigation component more efficiently targets the zone of interest than traditional static width buffers, it reduces the opportunity cost of losing land to "no touch" areas while maintaining a strong conservation agenda for riverine environments.

Specifically, to make their model, they had four requirements for each river: maximum extent of the floodplain, maximum river width, topological data (preferably a digital elevation model), and for the EMC, average adjacent vegetation height. Their model varies a corridor radius between two bounds. "The lower bound is the maximun width of the river and the upper bound is the maximum extent of the floodplain. A lower bound ensures that a river vector aquires sufficient width to buffer the watercourse's actual diameter. An upper bound aids in excluding areas which are intuitively non-riparian (i.e., beyond maximum limits of flooding). All areas within the lower bound are coded as riparian. Additionally, all contiguous areas between the lower and upper bounds that are less than 5% gradient (derived from DEM) are coded as riparian" [Kyle et al.].