|
Defining Restoration Goals
<br />Table 3. Quantifiable components of selected ecosystem
<br />processes.
<br />Ecosystem Function Quantifiable Component
<br />Energy flow
<br />Food web architecture (number and
<br />kinds of trophic links)
<br />Respiration rates
<br />Ratios of respiration to productivity or to
<br />biomass
<br />Alternate energy flow paths (e.g., detrital
<br />vs. herbivory)
<br />Decomposition rates
<br />Trophic pyramid
<br />Nutrient cycling Mineralization rates
<br />Decomposition rates
<br />Organic matter "quality" indices
<br />Standing stocks of nutrients (total and /or
<br />available)
<br />Losses due to leaching, gas loss, erosion
<br />Different processes for different nutrients
<br />(e.g., immobilization of N, chemical
<br />forms of P, cation exchange, solubility
<br />of ca, anion adsorption of S)
<br />Identity of limiting nutrient(s)
<br />Productivity Rates of growth; biomass change per
<br />unit time
<br />Cover, maximum standing crop biomass
<br />Carbon flow (i.e., respiration rates or 02
<br />uptake rates)
<br />Primary vs. secondary or net ecosystem
<br />production
<br />Primary production: allocation to
<br />different tissue types (e.g., root:shoot,
<br />woody:leaf)
<br />The widely acknowledged association of high plant diver-
<br />sity at low to moderate levels of productivity (Rosenzweig
<br />& Abramsky 1993) also suggests that relationships among
<br />ecosystem functions are complex and not readily predict-
<br />able or generalizable from simple laboratory models or
<br />from studies of a small number of ecosystem types.
<br />with respect to the number of species of organisms in-
<br />volved in the expression of the index.
<br />Furthermore, correlations among the components of
<br />Table 3 (or Table 2) are either unknown or are variable.
<br />For example, in Swedish forests experimentally fertilized
<br />with nitrogen, the amount of carbon in the forest floor in-
<br />creases with added N (= nutrient cycling), but the micro-
<br />bial respiration rate (= energy flow) and biomass de-
<br />crease (Nohrstedt et al. 1989). Experimental studies of the
<br />correlations between biodiversity and various measures of
<br />ecosystem function often show increases in productivity,
<br />plant cover, resilience, and biomass with increasing spe-
<br />cies diversity (Tilman et al. 1996; Naeem & Li 1997;
<br />McGrady -Steed et al. 1997; Tilman et al. 1997). But these
<br />studies are frequently highly simplified microcosms of
<br />protists (McGrady -Steed et al. 1997; Naeem & Li 1997), or
<br />relatively simple experimental grassland or grassland mi-
<br />crocosms. Opposite patterns have been demonstrated in
<br />other ecosystems (e.g., in boreal forests, as studied by
<br />Wardle et al. 1997a, 1997b). No studies have been con-
<br />ducted in systems such as temperate salt marshes, which
<br />are known for both their low plant diversity and their high
<br />productivity, or sedge wetlands of the coastal plain, which
<br />are known for their high diversity and low productivity.
<br />Ecosystem Services
<br />"Ecosystem services" are a third major source of goals for
<br />restoration and conservation. The specification of ecosys-
<br />tem services is as varied as the specification of ecosystem
<br />function. Three recently proposed lists of services are given
<br />in Table 4. The categories of services are overlapping but
<br />not identical; many other such lists in the literature simi-
<br />larly reflect the individuality of their authors. Although the
<br />services themselves are poorly defined, they all share the
<br />characteristic of being driven by considerations of human
<br />valuation. The advantage of using ecosystem services as a
<br />goal for restoration (Table 1) is that the compelling and ob-
<br />vious human interest generates dollars and political sup-
<br />port, just as species of charismatic megafauna generate
<br />support for conservation (Wilcove & Blair 1995). However,
<br />one runs the risk that changes in technology, the economy,
<br />and /or society will undercut or devalue the service. This
<br />risk has been frequently pointed out with respect to efforts
<br />to place a dollar value on ecosystem services, such as that
<br />by Costanza et al. (1997). In addition, it is likely that re-
<br />storing one particular service will preclude the provision of
<br />other services. For example, Marble (1992) details the par-
<br />ticular characteristics of flora, fauna, hydrology, landscape
<br />setting, soils, etc. needed to promote the various ecosystem
<br />services provided by wetlands. Comparison of these lists
<br />reveals that the features needed to produce one service
<br />are sometimes the opposite of those needed for other
<br />services. For example, to promote nutrient removal,
<br />the water source to the wetland should have high nu-
<br />trient concentrations (low quality), but to support bird
<br />habitat, high- quality water (low nutrient concentrations)
<br />is needed. To promote nutrient removal, a restricted outlet,
<br />low flow rates and mineral soils of any texture are recom-
<br />mended, but to support the export of production to down-
<br />stream systems, moderate flow rates, permanently open,
<br />large outlets, and fine- textured soils are required. Al-
<br />though no one expects any given wetland to provide all
<br />possible services, the creation of qualities to maximize one
<br />specified service may well preclude other services which
<br />would otherwise be possible (Table 1).
<br />Several other issues are germane to the discussion of
<br />setting restoration goals. The highly dynamic, ever -
<br />changing nature of ecosystems has been frequently
<br />pointed out (Pickett & Parker 1994), but the implications
<br />of this paradigm for choosing restoration goals are often
<br />conveniently ignored. Efforts to restore a species by cre-
<br />Restoration Ecology MARCH 2000
<br />
|