Making a market in environmental credits II: Watershed moments

Abstract

This pair of papers examines and describes the state action necessary to make markets function as environmental policy instruments and as strategies of governance. We do this through a detailed look at the mechanics of environmental credit compliance markets in the US states of Oregon, Ohio, and North Carolina in which stream credits are privately created and sold to developers who have impacted protected stream systems. In this paper, we observe that the governance of streams as water resources requires the state to create a scalar hierarchy that fixes certain characteristics of streams at certain scales of state action. These fixes attempt to resolve, bracket, or ignore the temporal and spatial variability of streams that can confound governance; however, these variabilities are essential to the scientific study of streams. At each of the four distinct scales, four different operations crucial to market function were observed; at each scale, elements of natural variability were fixed or confined to be expressed only within the given scale. These observations reveal principles of how scale functions within environmental governance, as well as failures where gaps and resistances appear that create unforeseen outcomes in market-led policy. In three different state settings, the establishment of a fixed scale of governance is made in different ways that depend on the local institutional and social context. However, they all act to render an unruly set of temporal and spatial flows as instead occurring within certain fixed scalar boundaries, and thus amenable to governance with markets.

Keywords

Neoliberal natures water governance water pollution trading multiscalar governance environmental governance

Introduction

This is the second of a pair of papers that examines the detailed operation of environmental governance with markets through the example of stream credits: commercially produced environmental credits created by restoring stream and river systems. Whereas the previous paper focused on the tools and techniques used to define environmental flows as stable objects, in this paper, we examine the variety of scales at which environmental governance acts on ecosystems. As with the previous paper, we find a reduction, fixing, and simplification of ecological flows into something static and discrete for the operation of governance.

One of the many difficulties in manufacturing a river for sale is that rivers themselves are represented and understood by scientists in ways that transgress most of the scalar hierarchies normally used in governance. In the previous paper, we explored the process of constituting value in environmental credits, focusing on the creation and use of metrics for defining and measuring the stream credit as an object. In this paper, we take a different (though intersecting) path through the creation of markets in environmental credits. We examine and describe how governments and bureaucrats commodify ecosystems through establishing scalar and temporal fixes. These fixes resolve—or at least bracket—the temporal and scalar transience and transgressiveness of river systems.

Since natural systems are spatially and temporally dynamic, this bracketing of ecological transience and transgression may be observed in the commodification of all ecosystems, not just river systems. Ecologists define ecosystems as sprawling spatial networks of matter and energy which are dynamic on many different temporal scales. This is easily seen in the way riparian science describes rivers, and as Heraclitus famously observed, a river's primary characteristic is transience. “False as water” was Othello's charge that Desdemona lacked constancy, in the way that water takes the form of its container and has no form itself, and from Catullus comes the expression that worthless promises are “written in water.” If the purchase of a commodity is an expression of faith in the constancy of its properties over time and space, the prospect of rivers as commodities presents a challenge. The production of stream credits means not only defining the credit as a stable object that can bear value (as we described in the previous paper) but aligning the temporal and spatial scales of market governance with the scales through which we understand the natural world.

It can be done, however. In this paper, we describe how governments in the United States have attempted to implement scaled environmental market governance. The issuing of national regulations undergirding a market in stream credits in 2008 (Corps & EPA, 2008) allows us to see this at work in three different US states: Oregon, Ohio, and North Carolina. “Stream mitigation banking” is active in each of these states, a business in which private firms restore degraded river systems and sell the environmental credits they produce by doing so (Lave et al., 2008). Stream credits are bought by developers to compensate for a project which will adversely impact stream ecosystems, under the requirements of the US Clean Water Act (CWA). Therefore, stream banking is a regulatory market rather than a free market: buyers are compelled to purchase stream credits by regulation, in order to compensate for permitted impacts to streams. It is part of the same regulatory system in which wetland banking operates (see Robertson 2006, 2007), a comparison to which we will return below.

Fixes at scales

By exploring how a variety of public and private-sector actors—regulators, entrepreneurs, and scientists—interact through the stream credit market, we can trace how they attempt to overcome the transience and dynamism of environmental phenomena like rivers. To do so, they must fix or frame rivers at distinct temporal and spatial scales—and here we use “fixing” in the sense of reducing its variability in order to treat it as if it has constant properties over an extent of space and/or time. This is a “spatial and temporal fix” in the sense specifically used by Bob Jessop (2000; 2013; also Bakker 2009), which is somewhat distinct from its usage by David Harvey (e.g., 1996), whose work on spatial fixes has stimulated a voluminous critical literature. Jessop's “fix” is an achievement of temporary ontological stability which serves to support an accumulation strategy. It is similar to what Noel Castree called the “cerning” of nature, in which nature is “‘fixed’ in specific ways from particular perspectives and with particular implications for how we might behave toward ‘it’ and each other” Castree (1995, 15; see also Hannah, 1999). It is also similar to the Marxian notion of “moments” which crystallize a particular form out of an organic flow, as described by Harvey (1996) and Hartsock (1998). Any given fix is strategic and likely transient.

Fixing creates necessary and useful fictions: every apparently fixed object is in some kind of ontological flux which we find ways to ignore in affirming its stable identity over space and time. We treat the objects and services we buy as though they have ontological stability over space and time, often ignoring evidence to the contrary: you consider your car to be the same object, even after it has aged for a year or been driven to a distant city. For most objects we encounter daily, we do not renegotiate or investigate their basic properties as objects as they move through time and space.

The case of markets for restored streams shows what Castree (2008a, b), and Bakker (2009) refer to as ecological or environmental fixes: the use of ecological concepts (or information from the natural sciences) to create the spatial and temporal stability required for markets to function. Fixes are one way to overcome the problem that nature's materiality (however conceived; see Robertson, 2012) is seen to pose for capitalist accumulation:

An “ecological fix,” simply put, refers to strategies of externalization and internalization of socio-environmental conditions, in search of profit, both by states and by capitalists. … Castree follows this line of argument in defining “environmental fixes” as a set of strategies adopted by fractions of capital (or the state) in order to combat barriers to accumulation, and foster continued economic growth (Bakker, 2009, 1782).

In this paper, we are observing one of the central tasks of environmental governance: to either contain or ignore the dynamism of ecological and biophysical processes—their failure to sit still and be well-behaved within semantic categories and geographic boundaries (Bakker, 2005; Lave, 2012; Prudham, 2004; Robertson, 2012; Sayre, 2017). To achieve an ecological fix for stream credits, we find that the state addresses different stream properties at different scales: some attributes of streams are defined at a broad and national scale, others at a very narrow or local scale. The creation of durable static abstractions from dynamic natural ecosystems is achieved by establishing a scalar hierarchy that fixes various aspects of rivers into static elements at various scales across the hierarchy. Using this hierarchy, the sprawling flows that constitute river systems can be considered simply as a set of discrete objects that are adjacent in space or time.

Achieving a spatial fix can be more difficult with rivers than with other resources. For example, when we look at the case of forests, we can see that it is permissible and common for forest ecologists to simply stop at the edge of the trees. While they may point to edge effects and soil catenas that could extend the forest-object well beyond the physical extent of the trees, these considerations do not usually trouble the exercise of forest governance. Conversely, to define a river only by the extent of its water at any given moment is obviously unsatisfactory for those who draw the line. Stream scientists define streams, instead, by their watersheds (Doyle & Bernhardt, 2010): they insist on a dependency between object and context in a way that forest ecologists or wetland scientists typically do not when mapping the spatial extent of the resource.

Below, we observe that four scales emerge as sites of the ecological fixing of rivers: (a) the federal scale, (b) the subordinate state scale, (c) the watershed scale, and (d) the site or parcel of real estate. As related in the previous paper, by 2010 it was common to see state and regional environmental managers across the country revising or developing criteria to define stream credits (Doyle et al., 2013). But while the USEPA's regulation of water is national in scale, typically the definition of stream credits was handled by regional or state offices. Thus, the credit commodity definition was constant only across roughly US state-sized chunks of territory, rather than the entire nation. Meanwhile, the stream scientists who create the assessment metrics refer mainly to the scale of watersheds. And the stream bankers who produce credits acquire and manage specific parcels of real estate. The key scalar referent varies by actor and positionality within governance.

Before describing these fixes at four scales, we should clarify what we mean by scale. The turn away from the nation-state as the default scale of analysis in an ontologically prior, static, nested hierarchy is now decades-old in Geography. It is now common to recognize that state power (over nature or anything else to be governed) is not somehow at a scale “above” the scale of daily life but rather constituted by it and through it (Larner, 2003; Ong, 2006; Scott, 1998; Whitehead et al., 2007; among many others). Likewise, we can think of the nested hierarchy of scales in which states organize environmental governance as an achievement rather than something that is structural or natural necessity (Brenner, 2001; Bridge, 2002; Brown & Purcell, 2005; Ferguson & Gupta, 2002; McCarthy, 2005; Norman & Bakker, 2009; Sayre, 2005, 2017; Swyngedouw, 2000). Neoliberalism is said to involve a reallocation of the roles played by different levels of government—Jessop's (1990) famous “hollowing out of the nation-state”—and we should investigate rather than assume the necessity of the spatial scales in play (Bridge, 2002; Jessop, 2013; Jessop & Sum, 2006; Peck, 2004). This shuffling of tasks among scales of governance can take the form of a spatial or temporal fix, about which Jessop (2013, 9) says

[Spatio-temporal fixes] establish spatial and temporal boundaries within which the always relative, incomplete, provisional, and institutionally-mediated structural coherence of a given order are secured—to the extent that this occurs.

Jessop's large body of work shows that the stabilization of the global capitalist economy is inherently improbable due to its many disequilibria. However, stability can be provisionally achieved by corralling enough temporally and spatially variable phenomena into a framework where they can be regarded as homogenous or stable, for a while. Race, occupation, age, and gender are all census categories that the state uses to “fix” into exclusionary rigidity what are actually daily variable and non-discrete states of being on the part of state subjects (Moore, 2005). Fixes are the policies or practices that achieve this, and here we bring Jessop's principle down to the empirical case of stream credits, focusing our view on fixes to streams’ disequilibrating tendencies.

The stream credit market is just one case within a broad and global set of policy practices aiming to make markets in carbon, wetlands, and the habitat of endangered species. All of these are well-developed in the United States, but are also features of environmental policy in Europe (Regnery et al., 2013), Australia (Maron et al. 2015), and a number of regional and global governance arrangements (ten Kate et al., 2018; Peterson et al., 2018). In all of these, capital and governments face the problem of establishing the spatial and temporal boundaries within which order is secured.

In this paper, we draw on more than 60 interviews conducted between 2010 and 2015 in the US states of Oregon, Ohio, and North Carolina with regulators, scientists, and the stream restoration business community. We begin by discussing watersheds—the way stream systems are most commonly organized hierarchically by the US government—and then show what kinds of fixes appear at the national level, the subordinate state level, the landscape level, and finally the level of the project site. We will conclude by discussing what the case of streams means for scholarship on the scalar politics of environmental governance.

What is a watershed?

Any body of water is connected, through surface or groundwater, to nearly every other water body on Earth. One can speak of individual lakes, or wetlands, or river reaches, but the boundaries between them are literally fluid. To discipline the boundless landscape of water resources into fixed objects, the watershed is often called into service. The watershed provides a naturalized spatial extent or scale at which to consider problems of water governance (Cohen, 2012; Linton, 2010; Norman & Bakker, 2009).

The US Geological Survey created a set of watersheds known as the Hydrologic Unit Code (HUC) system (Seaber et al., 1987). At the highest scale, this system divides the United States into 21 hydrologic basins, each of which is given a two-digit code in the National Hydrography Dataset¹. Each of these regions is subdivided into further designations with larger numbers of digits. Thus, the 2-digit HUC designating the watershed of the Ohio River is “05,” and has a spatial extent covering the land that delivers surface water to the point where the Ohio discharges into the Mississippi River. But the Ohio River basin is further subdivided into four-digit HUCs; hence, the Kentucky River, which discharges into the Ohio, is given the four-digit HUC designation “0510.” Further subdivisions describe the nested 6, 8, 10, 12, and 14-digit HUCs. The word “watershed” can apply to any or all of these subdivisions. With HUC-14s, the government's spatial subdivision stops, describing units that usually measure between one and 30 square miles. The United States comprises tens of thousands of HUC-14s (Figure 1).

Figure 1.

The 8-digit, 10-digit, and 12-digit Hydrologic Unit Codes (HUCs) in the State of Massachusetts.⁹ 14-digit HUCs have been defined but are not shown here.

The authors of the HUC system offer up an array of seven naturalized scales from which to implement governance and manage different kinds of water resource problems. It appears to emerge without politics but serves a crucial function in governance. As Norman & Bakker (2009) note about the watershed concept generally, the HUC system does not argue for any particular scale of environmental governance—rather, it organizes space in a nested hierarchical way that can serve to distribute decision-making at several scales. Stream ecologists and geomorphologists would say that any stream site is connected to much of the continental United States through more or less attenuated chemical, physical and biotic flows; the HUC system provides a way to bracket and fix, to exclude data from outside a given geographic area and isolate spatial units of a variety of sizes.

Cohen's analysis of the watershed as a boundary object demonstrates clearly how a naturalized spatial unit can serve a political role by providing the scale at which stakeholders are defined and problems conceptualized:

… water governance is an exercise not only in making decisions about how water is to be used and allocated, but also about defining the geographic areas bounding a particular initiative. … The move away from jurisdictional boundaries and toward hydrologic ones is reflected in the rescaling of water governance in Canada and elsewhere in the uptake of the ‘watershed approach’, which typically includes a shift from political to hydrologic boundaries, increased extragovernmental participation in decision making, and some degree of delegation to watershed-scale organizations (2012: 2210).

The USGS HUC system is thus part of a larger family of landscape classifications that aim to parcel the natural world into discrete units, achieving a spatial fix that brackets or erases the flows that cross unit boundaries. Within the US government, the US Fish and Wildlife Service uses two systems to divide the United States into hierarchically nested ecoregions for regulatory purposes, Bailey (1980) and Omernik (1987), both of which are founded on the concept of an ecoregion as a set of consistent ecological relationships covering a discrete area of space. Wetland ecosystems are likewise classified and mapped according to Cowardin's (1978) taxonomic system. No project of environmental governance can begin without a landscape classification that achieves this fix. The HUC system is an example of a scalar taxonomy that makes environmental governance possible, the activity to which we now turn.

The scale of the federal state

Fixing the spatial and temporal variability of streams at four scales allows the state to ignore stream data that transgress each scale and define the stream as homogenous and static within that scale. At the scale of the national state, this was done using the 2008 US Wetland Compensation Rule, which set regulatory standards for the stream and wetland credit market. This Rule was the result of eleven years of intense lobbying by entrepreneurial wetland and stream credit producers, who wanted demand for their product bolstered by the firm requirements of national regulation (Lave 2012). Too often, they complained, regulatory agency staff decisions at small regional offices had resulted in inconsistent policies concerning what constituted a satisfactory stream or wetland credit, how and when they could produce credits, and to whom they could sell them (Albrecht & Wenzel, 1996; Mogenson, 2006). The 2008 rule did, in fact, create a certain amount of national consistency concerning the market in stream credits. However, it was silent on a large number of issues crucial to the market, leaving these to be resolved by jurisdictions at lower scales of governance.²

The 2008 Rule says very little about ecological standards and criteria that would apply to streams and wetlands across the country. The EPA's position was that there were no ecological criteria that could be used equally to evaluate wetlands and streams in places as ecologically diverse as Alaska, Maine, Florida, and Hawaii. During the rule-writing process, any suggestion of nationwide standards for ecological performance elicited strenuous objections from federal staff at EPA's 10 regional offices. In the section on “Ecological performance standards”, the rule has only this to say:

Performance standards must be based on attributes that are objective and verifiable. Ecological performance standards must be based on the best available science that can be measured or assessed in a practicable manner” (40 CFR 230.95(b)).

Decisions about what is “best”, “available,” and “practicable” are left to the 38 District offices of the agency issuing the permits, the US Army Corps of Engineers. As the Rule's preamble says:

Neither the proposal nor today's rule prescribe the individual variables or metrics that should be used to evaluate each aquatic resource type potentially restored, enhanced, established, or preserved in compensatory mitigation projects. Given the extremely large variation among the aquatic resource types found across the country, and the constant advances in the science of aquatic ecosystem restoration, overly prescriptive requirements would be impractical. (Corps & EPA, 2008, 19597)

Thus, in the face of ecological variability, the federal government refused to offer a national spatial fix to some of the most difficult questions about what counts as a stream credit. This task was instead given to the many regions (see below). The 2008 rule did, however, set nationwide administrative standards, requiring any plan to create stream credits must contain 18 different elements, including monitoring requirements, financial assurances, a legal site protection instrument, and default and closure provisions³. These elements create a national administrative architecture for the production of stream credits. The federal Rule thus creates a baseline assurance that the stream credit exists in a legally coherent form, and provides for the financial and liability arrangements assuring that the stream credit will continue to exist as a durable object of property.

Non-ecological qualities of the stream credit were fixed at the national scale, but what happened when regulators attempted to treat ecological features as static and homogenous across the same scale? One key element of any environmental credit market is its service area. This is the area within which a customer can seek to buy credit to compensate for their impact. If service areas are small, customers may have to purchase credits very close to their impact. Larger service areas cover more customers and therefore tend to provide a thicker market for credit producers (BenDor et al., 2009; Womble & Doyle, 2012). Thus, credit producers tend to push for larger service areas, while ecologists—concerned that compensation sites will stray too far from impact sites—tend to push for smaller service areas. Before the federal rule was issued, the size of service areas was set differently by each Corps District; often each bank would have its own bespoke service area. They frequently involved agglomerations of adjacent HUC-8 s; some were as large as a HUC-6.

It was assumed among observers that the federal government would use the Rule to determine an appropriate size for service areas (Mogenson, 2006)—that is, there would be the national-scale answer to the ecological question of “how far away can a compensation site be from the impact site?” As far back as 1990, the US Congress had called for the Corps to subdivide the entire nation into spatial markets for credits and determine “the appropriate geographic scope for which wetlands [and stream] loss may be offset by restoration.”⁴ However, the idea of specifying service areas at the federal level was rejected in the rulemaking process, and the final rule decrees that:

The service area must be appropriately sized to ensure that the aquatic resources provided will effectively compensate for adverse environmental impacts across the entire service area. For example, in urban areas, a U.S. Geological Survey 8-digit hydrologic unit code (HUC) watershed or a smaller watershed may be an appropriate service area. In rural areas, several contiguous 8-digit HUCs or a 6-digit HUC watershed may be an appropriate service area. [40 CFR 230.98(d)(6)(ii)(A)]

A consistent area across which stream credit commodities could be considered equivalent was not set. Instead, the determination of what is “appropriate” is recognized as a question that must be answered in a case-by-case or regional matter:

The service areas suggested in the text of this section may not be appropriate for some mitigation banks …. For these sponsors, it may be infeasible to have relatively small service areas for their mitigation banks, such as those based on 8-digit hydrologic unit codes, because they incur a relatively small amount of debits per year. Also, having relatively small service areas for some single user mitigation banks may discourage the establishment of large mitigation banks that provide substantial amounts of habitat and other aquatic resource functions and services. (Corps & EPA, 2006, 15529)

The federal state's concern for the “feasibility” of exchange across variable landscapes prevents it from acting. Because ecological factors must be balanced with market thickness in setting service areas, this is best done at lower scales of governance. Currently, the geography of credit equivalence continues to be a relatively uncoordinated patchwork (Womble & Doyle, 2012).

The scale of subordinate jurisdictions

Most important elements of the stream credit market were not treated as static at the national scale. Thus, the creation of stream credit markets is strongly inflected by the specific social and ecological context of lower jurisdictions, and we find the methods for defining the stream credit and its value are held constant across areas roughly the size of US states. In the CWA, regional environmental governance is exercised by the 38 Districts of the US Army Corps of Engineers, the 10 Regional Offices of the US EPA, and the various state and tribal governments of the United States. This regionally differentiated exercise of federal power is common in the United States, in which regional offices of national agencies implement national laws and regulations with considerable variance and interpretative discretion (Owen, 2017, Doyle et al., 2013). The technical measurements and criteria that allow a regulator to assess a stream were developed by regional resource agencies and by Corps Districts, applying only to their territories. Regional staff developed the architecture of the market: performance standards, credit release schedules, and mitigation ratios (some of which are described in the previous paper). They did so in ways that reflected regional experience, resulting in a considerable diversity in how stream credit markets work. We will look here at examples from the states of Oregon and Ohio.

Oregon: The green economy drives a statewide standard for measuring streams

Oregon's economy has a very vigorous environmental sector, and its environmental governance has been strongly shaped by its recent history. The political controversy which emerged in Oregon in the late 1980s around the protection of the endangered Northern spotted owl eventually became central to the 1992 presidential campaign and the first year of the Clinton Administration. Through the Northwest Forest Plan (FEMAT, 1993), President Clinton made an enduring commitment of federal funds to plan and manage environmental change in the region, and to support a forestry industry in decline. As more species, including river species such as salmon, were listed as endangered, both the environmental consulting economy and the environmental bureaucracy have expanded. The State of Oregon's environmental agencies, together with the regional offices of several federal agencies in Portland and Seattle, coordinate closely on all proposed impacts. A massive private-sector environmental consulting industry has arisen around these requirements, and it is estimated that $1 billion was spent annually on stream restoration in the Pacific Northwest states from 2000−2012 (Li & Boxall, 2012). In Oregon alone, the Watershed Enhancement Board (a lottery-funded agency) has spent over $370 million on watershed-focused site acquisition and restoration since 1999.⁵

In short, a lot of money is ready to flow into the production and consumption of stream credits in Oregon, if only the market can be established. Because of the vigor of the sector, a 2007 Oregon business leaders’ summit met to ask for an Oregon-specific ecosystem credit commodity definition that would translate well across many different state and federal regulatory requirements, with the goal of reducing regulatory complexity for Oregon developers. They proposed that state law require all regulated ecosystems be compensated for in units of ecosystem service: “Individual markets are already active in carbon, wetlands, habitat, open space and hazard reduction, but the proposed initiative recognized the enormous potential value of an integrated ecosystem services marketplace.” (INR 2008, III) With a coalition of Oregon environmental non-profits, these business leaders proposed bills to the Oregon Legislature in 2009, 2011, and 2013, requiring that the state “Address the need for a consistent methodology to describe and quantify ecological values” (SB 513, 2009), and “adopt and implement programs that use market-based approaches to conserve or enhance ecosystem services” that “have clear and predictable rules and provide for transparent, accountable and credible transactions” (HB 3337, 2013). Note that this was precisely what the federal Rule had refused to do in 2008.

After this law was passed, an Oregon nonprofit group received $1.3 million from the US government for the purpose of “building a transparent and trusted accounting system” for environmental credits. This system was called Counting on the Environment (COTE⁶; see Parametrix, 2009), and was piloted between 2009 and 2011. It consisted of simple, easy-to-use procedures for measuring salmon habitat or streambank forest functions. The nonprofit was initially focused only on proving COTE's use within a fairly small area of the Willamette River watershed near Portland. However the nonprofit quickly realized that, to respond to state legislation, an accounting system would have to extend across the state:

We began kind of naively thinking that we would create this infrastructure for the Willamette Basin and learned over the process that it was silly, that we were working at the wrong scale. That is was silly to have a separate infrastructure created in every little region of the country because it's too expensive, it's too complex, it's not a good use of money to keep reinventing the wheel over and over and over again. (Oregon environmental non-profit director, February 2012)

So while the federal scale was too large, the Willamette River basin was too small: a statewide stream credit methodology was needed to define the commodity. This methodology was produced in response to the 2012 RFIP discussed in the previous paper. But our return to Oregon here demonstrates something important about environmental governance. On the continuum of water resources from the site to the continent, the territory contained by a US state was considered homogenous enough to define consistent ecological criteria across it. The definition and measurement of stream function were fixed—through interactions among scientists, regulators, and other actors—at the US state scale. As with all such fixes, this outcome was not ordained by physical features, but depended heavily on the relationships of social and political power between actors in the stream credit policy realm.

Ohio: budgetary conflict drives a statewide standard for measuring streams

The same scalar fix to an ecological continuum is seen in the state of Ohio, although the state's role in defining stream credits arose from a more conflictual relationship between scales of government. The 1972 CWA authorized that millions of dollars of federal money be given to state governments for the purpose of cleaning up urban sewage pollution by building new treatment plants. President Nixon opposed this massive expense, and vetoed the Act, only to be overridden by Congress. However, he successfully blocked the transfer of funds to states that would help states implement the Act's requirements. Ohio's Governor Jim Rhodes was thus concerned that the CWA forced his state to develop stringent new water quality standards without federal funding.

In response, Governor Rhodes authorized a tiered approach to water quality standards that would focus the state's effort only on the most polluted streams. In order to categorize streams into tiers of impairment, Ohio's Environmental Protection Agency developed a stream assessment and classification system by conducting biological surveys of streams across the state. This created an exhaustive dataset on Ohio's stream ecology that could be used to locate new sewage treatment plants where they would do the most good. “So by 1990 you have basically the largest, longest fish data set in the world. In fact today I think it still stands as the largest, longest monitoring program of fish anywhere in the world,” recalled one former staffer; “No other state had this windfall”, said another. With these biological data on streams spanning the spectrum of conditions from polluted to pristine as benchmarks, Ohio state government ecologists in the 1990s were able to develop standards and goals for stream systems based on criteria that were drawn from and calibrated to actual Ohio streams. These were formalized as the Qualified Habitat Evaluation Index (QHEI); a stream's QHEI score (Figure 2) is a number indicating the stream's quality.

Figure 2.

One of six metric scoresheets from the Qualified Habitat Evaluation Index (QHEI) (Ohio EPA 2006).

Although the QHEI's six metrics mainly involve measures of physical geomorphology, their calibration is made against Ohio's massive benchmark dataset consisting of vegetation, fish, and macroinvertebrate surveys. This is extremely unusual. In most stream assessment methods, the physical form of the stream is treated as though it determines the stream's water quality and biology; therefore, in most states, stream assessment focuses on its physical form (Somerville & Pruitt, 2004). In Ohio, by contrast, physical morphology is merely a proxy for the real goal: measuring biological quality.

When Ohio developers and regulators urged the creation of markets in stream credits, the QHEI became the foundation of the definition of stream credit. By doing so, the state defined stream credits using actual field data on biota, rather than depending on the general mathematics of hydrological science. This use of regional field data sets Ohio distinctly apart from stream assessment practices in most of the rest of the nation. In Ohio, as one regulator said, “biology drives the bus,”⁷ and the richness of Ohio's biological dataset is a legacy that state regulators take great pride in. By 2012, Ohio regulators had established a state-wide system for defining a stream credit that was uniquely grounded in regional stream ecology. It depends on the work done by teams of biologists in the 1970s, profiting from the strategic successes of their state government working to solve a federally-imposed fiscal problem. It has produced a stream credit commodity measured in a way unlike any other in the United States—but still, emphatically and only, at the scale of the US state.

As these examples from Oregon and Ohio show, the state's historical relationship to environmental issues and to the federal government strongly conditioned the way the stream credit commodity was defined. In each case, it produced a unique commodity measure inflected by that history. There is no sense in which the “ecology of Ohio” or the “hydrogeology of Oregon” is meaningfully consistent in a way that determines this scalar fix. Rather, it is the play of power within the structure of US federalism and its sub-national governments which in turn imparts unique forms to the environmental credit markets they govern.

The scale of the watershed

We have shown some elements of the stream credit market put in place at the federal and subordinate state scale, but more work remains before a stream credit can be traded. Credit producers create stream credits on individual property parcels, but the value of those credits depends on factors beyond the parcel boundary. How far beyond? How big is the landscape of forces and factors that are allowed to influence investment and design decisions at a stream site? That landscape is larger than the site but smaller than the state. And here we come to a stalemate where the spatial fix has proven very difficult: stream scientists strongly assert that the wider landscape has important effects on a stream credit production site, while stream credit producers strongly assert that they can only be responsible for what happens within parcels of real property that they own.

The 2008 federal Rule says that choices about creating and locating stream compensation sites should be guided by “watershed planning” that identifies preferred locations for stream restoration within “watersheds” (which the Rule does not define). Thus, regulators often use off-site data to assess individual stream credit sites. For example, in the Wilmington (North Carolina) Corps District, how you sample for macroinvertebrates at a stream site varies depending on the size of a site's watershed. In Ohio, the size of a site's upstream catchment determines whether you use QHEI or a different method to assess the site.

Credit producers, for their part, strongly resist being held responsible for conditions outside their credit production site. One Oregon engineer was frustrated, trying to get credit for doing work at stream sites located in polluted watersheds:

… all the stuff upstream is still junk and you’ve got all this crap coming down, how do you come up with a functional lift in a [stream] reach here? … you aren't gonna have much luck restoring even adjacent [stream reaches] to pristine wilderness settings. It isn't gonna happen. Or it isn't gonna happen for just less than outrageous amounts of money that involve lots of features that are way off-site. (Stream restoration consultant, Oregon)

A counterfactual may help to illustrate the challenge of this spatial fix. One stream restoration professional, asked how he would approach the project of stream restoration if he had access to unlimited resources and power, said simply:

You start buying watersheds. You buy from ridgetop to offshore. Not just stop at the estuary, but literally out 3 miles. Because those are the units that control what goes on. … That's what a [stream credit] bank would look like. It would look like an entire watershed. (Stream restorationist)

His point is that the condition of any stream restoration site depends on the condition of a large area beyond it—a watershed—so the only way to ensure the quality of a site is to control all other sites that are connected to it. This extreme position is obviously incompatible with a real production strategy: instead, it is a refusal to bracket spatial and temporal complexity, a refusal to fix. It nods, through absurdism, to a biophysical understanding of a stream credit by which the system must be viewed as a whole, and dictates that any definition which does not incorporate the entire web of ecological connections around it is illegitimate.

The power of this vision shows that the connectivity between a stream credit site and its watershed remains disruptive. How to fix or ignore these flows between stream site and landscape is an ongoing debate in most jurisdictions, and has proven exceptionally hard to resolve. In the eyes of stream scientists, the limits of property boundaries are obviously irrelevant to stream conditions. For them, the study of the flow of energy and materials through stream systems confirms the importance of spatial and temporal scales profoundly different from that of daily human experience. For many, it is what draws them to love their work:

It's very streamlined to put wood in streams is because if it moves, great, it's providing function someplace else. In the estuary, it's providing habitat as it moves through the system. Just as the gravel moves through the system it's providing habitat in different places. (Federal environmental regulator, Oregon)

I love things like these metabolism measurements that integrate across the whole system. You don't have to measure a bunch of rocks and scale up. You just measure the whole thing, and it's elegant, awesome stuff. (Stream scientist, Oregon)

For stream scientists, the very concept of a stream depends on connectivity. As one stream scientist observes, in determining what objects are included in the category of “stream”:

… really what we needed to show was connectivity, show that the site was actually connected to surface waters downstream—because otherwise it wouldn't be a stream. And we had to show that there was a net offsite flow of water, otherwise it wouldn't really be a stream. (Stream hydrogeologist, North Carolina)

Thus paying attention only to one spot in a stream network makes little sense because any material additions such as gravel or wood at one site will move downstream, and not be replaced. One is not actually, according to the geologist above, even restoring a stream unless one attends to landscape and watershed flows. Adding woody debris (an important element of habitat and structure) without creating the conditions upstream for the natural addition of more wood in the future is therefore pointless:

The reason this site is degraded might be because that site is degraded upstream. And fixing this site without fixing that site doesn't actually fix your problem. And I think that's a common failing, right? It's like [saying], “We need to add some wood [and] this is the place we can put it in. Let's add our wood here.” And you haven't really addressed the fact that you’re not recruiting any more wood, and eventually this stuff's gonna get blown out. (Stream scientist, Oregon)

These scientists’ views are crucial because, as discussed in the first paper, stream scientists play an essential role in developing and certifying the measurements used in stream credit markets. But stream credit bankers take an equally strong position that they can only make improvements to a stream site within a real estate parcel controlled by the stream banker. For the state to require them to be responsible for flows and influences beyond the parcel boundary, in their view, threatens the entire basis of their industry—and here the struggle over scale is visibly the struggle over a capitalist accumulation strategy. A consultant reflects on trying to meet a water quality criterion by restoring a short parcel of stream length:

Now how on earth are you going to change water chemistry? It's a nice idea but you can't do it, I mean it's a watershed issue. You got legacy pesticides in an area where you got ag[ricultural] activities. Your 50 feet of stream isn't gonna change anything. But you may get dinged for [the pesticides] and you may have to assess it. And for what reason? (Stream restoration consultant, Oregon)

This credit producer objects that he cannot control land-use in the watershed above their site and thus cannot control the quality of his credits if the state insists they must be defined and assessed with reference to water coming from offsite (unless, of course, he buys the entire watershed). Another team of environmental consultants expressed this tension vividly:

Consultant #1: First of all, our goal was to come up with a mitigation bank that actually had a credit and debit system. But when you think about that—nutrient dynamics, etc.—from a biological standpoint, how are you going to quantify those functions within a stream that is just an isolated section of a much larger system? We just tried to wrap our heads around that and just, and couldn't come up with it.

Consultant #2: Basically we just failed at the boundaries. I mean it's just impossible to set a boundary for an isolated piece of stream when you have the whole watershed.

“Failing at the boundaries” is a way of saying that the spatial fix is not working to either ignore or stabilize the dynamic flows across the parcel boundary. The producers worry that this lack of fix undermines the stability required for capital investment in commodity production. And they’re right. Short of “buying watersheds”, this lack of fix continues to make credit producers responsible and liable for conditions outside the parcel of real estate they control. This threatens the stability of the governance strategy.

The scale of the project site

The scale of the property parcel presents a final set of challenges for fixing ecological flows into a static form. At the scale of individual property parcels, credit producers must acquire the rights to real estate on which the stream credit is assessed. This may sound like a straightforward transaction, but the process of finding appropriate land on which to develop credits falls into the profession of real estate (or, as one credit producer put it, “chasing dirt”). Stream bankers almost never acquire land fee-simple for credit production—they would be left holding the land after the last credit is sold, which is not in their interest. Rather, they typically seek landholders on whose land they develop stream credits, arranging for profits to be split. No matter how convincing the stream banker is, most US rural landowners will be reluctant to commit to an in-perpetuity land use subject to a tangle of hard-to-interpret federal and state rules. The more so when these rules frequently change and are subject to reinterpretation by the US Supreme Court and successive Administrations. For a typical stream banker seeking a land-owning partner, getting access to the parcel is a process of finding appropriate sites in a landscape that has, for centuries, been subdivided and arrayed for the production agricultural commodities. This is not so much an ecological fix, then, as a spatial fix of Harvey's variety: stream bankers encounter a landscape delineated and arrayed for one kind of production, and seek to shake loose existing spatial fixes in order to impose their own for a new kind of production.

To do this well requires that they know the territory and the population of landowners intimately. This excerpt from an interview with a stream bank entrepreneur provides some of the flavors of that task, and highlights the high degree of competition for scarce land between different conservation interests trying to convince farmers to convert their land to producing environmental credits:

So look at the whole landscape, to find the best place, and you quickly can rule out a whole bunch of [sites]. It just ain't going to work. You know, various people own it. We sit in the conference room and take fly-arounds where we just fly around the watershed and discuss all the different things, and what happens is you develop a mosaic of people you know out there. Over 15 years, even in a 34 million acre state, we got stories from every single county, and that's invaluable, you know? So it becomes a real kind of personal thing: “Oh yeah, down on Moccasin Creek down in Johnston County we had a request there three years ago for something. We talked to that guy back in November.” So it's just this kind of institutional knowledge that's really important. (Stream mitigation banker, North Carolina)

The character of local real estate markets exerts a powerful influence on the conditions of possibility for stream credit banking. Stream credits must be generated from parcels containing streams, of course, but most property parcels are laid out in blocky shapes that do not follow the linear, narrow path of stream corridors. Stream bankers don't want to acquire a lot of “excess” dry land that isn't usable for stream credits, so the ability to acquire an appropriate parcel often depends on a landowner's willingness to subdivide their parcel or overlay it with easements. Landscapes dominated by a few large landowners are preferred because it is expensive to agglomerate many small parcels and satisfy many partners:

Some areas you don't have these big property owners… usually those folks that have a lot of land are the ones that you usually can negotiate to take some of that land, because it really isn't going to hurt them. They’ve got so much. It's the little guy that has a little small parcel that you’re going to cut in half, that's a hard sell. And the negotiation part of this process should not be understated, because the more landowners you add—[effort] doesn't grow linear, it grows exponential. Everybody else is doing the same thing. They’re looking at GIS, looking at aerials. They’re out there beating the pavement and invariably we’re all going to stumble on that big parcel that has a big cow pasture creek through it, and so sometimes you end up three, four, even five different companies on the same door. (Stream restoration engineer, North Carolina)

As the end of this comment makes clear that, to the extent that there is an active stream banking industry in an area, there will be an expert and competitive winnowing of possible sites down to those parcels that present the opportunity for profit. All the achievements of environmental governance at higher spatial scales—watershed planning and credit definition and administrative assurances—do not, in the end, assure the production of stream credits. These will come to naught without someone at the other end of a real estate transaction who successfully re-arranges the geography of production. Such a transaction is often an intensely personal event and is expressed as the bottleneck moment for the entire market by one North Carolina regulator:

The real logjam of this whole thing—you can do all the planning you want. You can do all the talking to people you want, but if the dude ain't going to sell you his property, he ain't doing it. End of the game.

This testimony from regulators and bankers on site acquisition suggests that successful stream credit banking is an exercise in re-arranging economic interests and boundaries in real estate, a last necessary step in unfixing the settlements that had arranged the landscape for agricultural production and reinscribing them in ways that allow the activation of the national, state and watershed-level decisions about how stream credit markets will function.

Stream credit banking turns out to depend as much on personal familiarity with individual sites and networks of trust as did the small-scale productive industries analyzed as Marshallian nodes since the economic geography of the 1980s. In that tradition, geographers and other economic scholars found that some of the transformations occurring under the rubric of “globalization” or the emergence of some elements of the “post-Fordist” global economic order were observable only at the level of personal relationships and manufacturing traditions at the scale of small municipalities (Amin & Thrift, 1992; Markusen, 1996). Success in such economic strategies was found, in the end, to be dependent on acts of learning and experience that are “not fully codifiable” (Storper, 1997, 19), occurring at the finest grain of a scalar hierarchy.

Conclusion

Governing water in streams requires constraining and simplifying its spatial and temporal complexity. This is a task that is general to not only environmental governance but daily life. As citizens, we are reductively coded racially and geographically in ways that must remain constant from day to day for the government to effectively govern us (Ong, 2006). Of course, we all overflow the limits of such categories and descriptors, a phenomenon sometimes called catachresis (Best, 1999): no term can ever fully capture the complete meaning of the object named. Natural objects like forests and streams tend to be fixed and encoded as spatially bounded objects with static measures quantifying their nutrient cycling or biodiversity. There are obvious ways in which any instance of nature overflows such categories, in which nature is catachretic and tends to resist fixing. The conception of a river as fully integrated with its watershed has long been a tenet of riparian science (Leopold & Maddock, 1953; Hynes, 1970), complicating the simple blue line on the map for over half a century. In this paper we have shown the ways that river systems as flows of energy and materials are fixed at various scales of governance, but this project is never complete—rivers, like our own selves, are catachretic and overflow with meaning that is not always captured or bracketed by attempts to fix their, and our, dynamism. The integrated concept of river systems and their watersheds that underlies the riparian science informing state regulation is in fundamental conflict with the state's concept of resources as property parcels, river reach segments, or fungible commodities. Watching the emergence of a market in stream credits, as we have done across the previous paper and this one, is an exercise in observing spatio-temporal fixes succeed and fail. Let us reiterate what we said above: the struggle over scale is visibly the struggle over a capitalist accumulation strategy, and fixes will be strategic and transient. This is far from a novel concept in geography, but it is an under-appreciated element of environmental governance.

It is common to understand social processes, as well as ecological processes, as “flows”—to assert that everything nameable is at the same time a discrete object and an effect of fluid social relations (Hartsock, 1998). The conception of social space or globalization as a “space of flows” in Castells (1996) sense is familiar now and is a legacy of systems theory that has found a durable home in the theorization of social systems (Jessop et al., 2008). The resonance of the concept of flows is due in part to the sense of catachresis we experience in any attempt to bracket complex systems in order to isolate one of its functions or features. It is precisely the veiled connections adhering to the commodity, our ability to notice its ongoing entanglement with flows of labor and value, that initiates Marx's Capital. The contradiction between a commodity's discrete object-ness and its existence as a flow of value, and the need to fix it in time and space, is at the root of a contradiction in capitalism (Jessop, 2013). Rivers also flow and draw our attention in an unusually physical way to the object/flow distinction. Their occasional literal overbank flows are analogous to their catachretic resistance to being bound within a spatially-bracketed strategy that is meant to produce governable landscapes and fungible commodities.

Where spatial and temporal fixes met resistance at one scale, they moved down-scale, and we have shown that these fixes don't always succeed. In fact, the struggle to create the ecological fixes necessary to commodify streams is visible in the market data: the economy of stream credits in the United States is small and shows signs of inadequate supply. In 2015 there were 5.2 million linear feet of stream credits available, and this had grown to 7.18 million by 2019. This may seem like a large amount (the latter number is roughly the straight-line distance from Kansas City to Los Angeles), but even assuming a generous 200-foot buffer on a channel centerline, it represents only 33,000 acres of real estate. Nationwide, the number of stream bank sites with credits available has grown from 298 to 442, but in only 25 of the 37 Districts of the US Army Corps of Engineers (up from 21 in 2015); 97% of the stream credits, and 77% of the stream banks, are clustered in just 10 districts in a coastal arc between Delaware and Texas. By comparison, in the relatively stable and mature market for wetland credits, there were about 742,000 acres worth of wetland credits available in the United States for sale in 2015. Although this had decreased to 561,000 acres by 2019, the number of wetland bank sites grew from 966 to 1075, and are present in every District.⁸

It is hard to assess whether this supply of stream credits is adequate to demand or not, but the abundance of wetland credits and a relatively small number of stream credits is in sharp contrast with Corps data showing robust permitting for stream impacts: there were permitted impacts to over 4 million linear feet of streams in 2013 alone, compensation for which would have consumed more credits than were available in that year (IWR, 2015). By comparison, there were permitted impacts to 7000 acres of wetlands in 2013, a potential demand for about 1% of available banked wetland credits. The production of stream credits, on the scant data available from the US government, appears to struggle to keep pace with demand, whereas the wetland market experiences oversupply. One may at least infer that there are difficulties in producing stream credits which, we argue, are in part constituted by the complexity of the spatial and temporal fixes necessary to bring stream credits to market, some of which we have documented here.

The success or failure of ecological fixes is a matter of empirical and descriptive research. Our work suggests that this story will be different for every attempt to precipitate service commodities out of nature's flows, whether in carbon, wetlands, species, or anything else. Spatial fixes must generally happen in order to structure a world of flows as a stable field of action on which the state can govern and the market can trade. However, these spatial settlements exist in the realm of unresolvable tensions and second-best solutions (Jessop, 2000) that characterize all situations in which apparently discrete objects depend on flows we are constantly invited to ignore. The individual laborer is sustained by unrecognized flows of care and nourishment stretching great distances; the value of a single coin is sustained by digital capital markets a world away; and a length of stream channel is sustained by flows of wood, fish, and nutrients from the mountains to the sea.

Footnotes

Acknowledgments

The authors are deeply grateful for the insightful comments of Eric Nost, Jen Rose Smith, and a host of anonymous reviewers. The authors are profoundly indebted to the community of stream regulators and restorationists who shared their wisdom and expertise. All errors are the responsibility of the authors.

Highlights

Successful environmental governance must either contain or ignore the spatial and temporal dynamism of biophysical processes.

Regarding markets in stream credits, we observe this dynamism is fixed or constrained at four different scales of governance.

The watershed is a scale-less concept that is useful in naturalizing the spatial scale of governance.

Where spatial and temporal fixes met resistance at a scale, governance measures tended to move to less extensive spatial scales.

Not all governance attempts to simplify the spatial and temporal complexity of streams were successful.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Science Foundation Division of Behavioral and Cognitive Sciences, (grant number BCS-0961551).

Notes

References

Albrecht

Wenzel

(1996) A view from the private sector. In: Marsh

Porter

Salvesen

(eds.) Mitigation Banking: Theory and Practice. Washington, DC: Island Press, 77–87.

Amin

Thrift

(1992) Neo-Marshallian nodes in global networks. International Journal of Urban and Regional Research 16: 571–587.

Bailey

(1980) Description of the ecoregions of the United States. No. 1391. Washington DC: US Department of Agriculture, Forest Service.

Bakker

(2005) Neoliberalizing nature? Market environmentalism in water supply in England and Wales. Annals of the American Association of Geographers 95(3): 542–565.

Bakker

(2009) Neoliberal nature, ecological fixes, and the pitfalls of comparative research. Environment and Planning A 41: 1781–1787.

BenDor

Sholtes

Doyle

(2009) Landscape characteristics of a stream and wetland mitigation banking program. Ecological Applications 19(8): 2078–2092.

Best

(1999) Postcolonialism and the deconstructive scenario: Representing gayatri spivak. Environment and Planning D: Society and Space 17: 475–494.

Brenner

(2001) The limits to scale? Methodological reflections on scalar structuration. Progress in Human Geography 25(4): 591–614.

Bridge

(2002) Grounding globalization: The prospects and perils of linking economic processes of globalization to environmental outcomes. Economic Geography 78(3): 361–386.

10.

Brown

Purcell

(2005) There’s nothing inherent about scale: Political ecology, the local trap, and the politics of development in the Brazilian Amazon. Geoforum; Journal of Physical, Human, and Regional Geosciences 36: 607–624.

11.

Castells

(1996) The information age: economy, society and culture, Volume 1: The rise of the network society. Oxford: Blackwell.

12.

Castree

(1995) The nature of produced nature: Materiality and knowledge construction in marxism. Antipode 27(1): 12–48.

13.

Castree

(2008a) Neoliberalising nature: The logics of deregulation and reregulation. Environment and Planning A 40(1): 131–152.

14.

Castree

(2008b) Neoliberalising nature: Processes, effects, and evaluations. Environment and Planning A 40(1): 153–173.

15.

Cohen

(2012) Rescaling environmental governance: Watersheds as boundary objects at the intersection of science, neoliberalism, and participation. Environment and Planning A 44(9): 2207–2224.

16.

Corps [US Army Corps of Engineers] & EPA [US Environmental Protection Agency] (2006) Compensatory mitigation for losses of aquatic resources; proposed rule. Federal Register 71:15520–56.

17.

Corps [US Army Corps of Engineers] & EPA [US Environmental Protection Agency] (2008) Compensatory mitigation for losses of aquatic resources; final rule. Federal Register 73(70): 19594–19705.

18.

Doyle

Bernhardt

(2010) What is a stream? Environmental Science & Technology 45(2): 354–359.

19.

Doyle

Lave

Robertson

, et al. (2013) River federalism. Annals of the Association of American Geographers 103(2): 290–298.

20.

FEMAT [Forest Ecosystem Management Assessment Team] (1993) Forest Ecosystem Management: An Ecological, Economic and Social Assessment. Washington DC: Forest Service, US Department of Agriculture. July 1993.

21.

Ferguson

Gupta

(2002) Spatializing states: Toward an ethnography of neoliberal governmentality. American Ethnologist 29(4): 981–1002.

22.

Hannah

(1999) Skeptical realism: From either/or to both-and. Environment and Planning D: Society and Space 17(1): 17–34.

23.

Hartsock

NCM

(1998) Moments, margins, and agency. Annals of the Association of American Geographers 88(4): 707–712.

24.

Harvey

(1996) Justice, Nature and the Geography of Difference. Cambridge, MA: Blackwell.

25.

HB 3337 [Oregon Legislative Assembly B-Engrossed House Bill 3337] (2013) An Act relating to ecosystem services .

26.

Hynes

H.B.N.

1970. The Ecology of Running Waters. Toronto: Univ of Toronto Press.

27.

IWR [Institute for Water Resources] (2015) The Mitigation Rule Retrospective: A Review of the 2008 Regulations Governing Compensatory Mitigation for Losses of Aquatic Resources. IWR Report 2015-R-03. Alexandria, VA, 146pp.

28.

Jessop

(1990) State Theory: Putting the Capitalist State in its Place. University Park, PA: The Pennsylvania State University Press.

29.

Jessop

(2000) The crisis of the national spatio-temporal fix and the tendential ecological dominance of globalizing capitalism. International Journal of Urban and Regional Research 24(2): 323–360.

30.

Jessop

(2013) Revisiting the regulation approach: Critical reflections on the contradictions, dilemmas, fixes and crisis dynamics of growth regimes. Capital & Class 37(1): 5–24.

31.

Jessop

Brenner

Jones

(2008) Theorizing sociospatial relations. Environment and Planning D: Society and Space 26(3): 389–401.

32.

Jessop

Sum

N.-L

(2006) Beyond the Regulation Approach: Putting Capitalist Economies in their Place. Northampton, MA: Edward Elgar.

33.

Larner

(2003) Neoliberalism? Environment and Planning D: Society and Space 21: 509–512.

34.

Lave

Robertson

Doyle

(2008) Why you should pay attention to stream mitigation banking. Ecological Restoration 26(4): 287–289.

35.

Lave

(2012) Fields and streams: Stream restoration, neoliberalism, and the future of environmental science. Athens, GA: University of Georgia Press.

36.

Leopold

Maddock

(1953) The Hydraulic Geometry of Stream Channels and Some Physiographic Implications. USGS Professional Paper 252. US Government Printing Office, Washington, DC.

37.

Boxall

(2012) Application of the Riverscape Perspective in Developing Restoration and Monitoring Strategies. Short course at River Restoration Northwest Conference. Stevenson, WA. 30 January 2012.

38.

Linton

(2010) What is water: The history of a modern abstraction. Toronto: UBC Press.

39.

Markusen

(1996) Sticky places in slippery space: A typology of industrial districts. Economic Geography 72(3): 293–313.

40.

Maron

Bull

Evans

, et al. (2015) Locking in loss: Baselines of decline in Australian biodiversity offset policies. Biological Conservation 192: 504–512.

41.

McCarthy

(2005) Scale, sovereignty, and strategy in environmental governance. Antipode 37(4): 731–753.

42.

Mogensen

(2006) Mitigation banking: It’s no myth. National Wetlands Newsletter 28(2): 15–16. 21.

43.

Moore

(2005) Suffering for Territory: Race, Place, and Power in Zimbabwe. Durham NC: Duke University Press.

44.

Norman

Bakker

(2009) Transgressing scales: Water governance across the Canada–US borderland. Annals of the Association of American Geographers 99(1): 99–117.

45.

Ohio EPA [Ohio Environmental Protection Agency] (2006) Methods for Assessing Habitat in Flowing Waters: Using the Qualitative Habitat Evaluation Index (QHEI). OHIO EPA Technical Bulletin EAS/2006-06-1. Columbus, OH, 71pp.

46.

Omernik

(1987) Ecoregions of the conterminous United States. Annals of the Association of American Geographers 77: 118–125.

47.

Ong

(2006) Neoliberalism as exception: Mutations in citizenship and sovereignty. NC: Duke University Press: Durham.

48.

Owen

(2017) Little streams and legal transformations. Utah Law Review 1: 1–55.

49.

Parametrix (2009) Salmon Credit Calculation Rules and Equations. Portland, OR: Parametrix, Inc.

50.

Peck

(2004) Geography and public policy: Constructions of neoliberalism. Progress in Human Geography 28(3): 392–405.

51.

Peterson

Maron

Moillanen

, et al. (2018) A quantitative framework for evaluating the impact of biodiversity offset policies. Biological Conservation 224: 162–169.

52.

Prudham

(2004) Knock on wood: Nature as a commodity inDouglas-fir country. New York: Routledge.

53.

Regnery

Couvet

Kerbiriou

(2013) Offsets and conservation of the Species of the EU habitats and birds directives. Conservation Biology 27(6): 1335–1343.

54.

Robertson

(2006) Emerging markets in ecosystem services: Trends in a decade of entrepreneurial wetland banking. Frontiers in Ecology and Environmental Science 6(4): 297–302.

55.

Robertson

(2007) Discovering price in all the wrong places: commodity definition and price under neoliberal environmental policy. Antipode 39(3): 500–526.

56.

Robertson

(2012) Measurement and alienation: Making a world of ecosystem services. Transactions of the Institute of British Geographers 37(3): 386–401.

57.

Sayre

(2005) Ecological and geographical scale: Parallels and potential for integration. Progress in Human Geography 29(3): 276–290.

58.

Sayre

(2017) The politics of scale: a history of rangeland science. Chicago: University of Chicago Press.

59.

Scott

(1998) Seeing like a state: How certain schemes to improve the human condition have failed. New Haven: Yale University Press.

60.

SB 513 [Oregon Legislative Assembly Enrolled Senate Bill 513] (2009) An act relating to ecosystems, and declaring an emergency .

61.

Seaber

Kapinos

Knapp

(1987) Hydrologic Unit Maps . U.S. Geological Survey Water-Supply Paper 2294: 63. p.

62.

Somerville

Pruitt

(2004) Physical Stream Assessment: A Review of Selected Protocols for Use in the Clean Water Act Section 404 Program. September 2004, Prepared for the U.S. Environmental Protection Agency, Office of Wetlands, Oceans, and Watersheds, Wetlands Division (Order No. 3W-0503- NATX). Washington, D.C. 213 pp. Document No. EPA 843-S-12-002.

63.

Storper

(1997) The Regional World: Territorial Development in a Global Economy. New York: Guilford.

64.

Swyngedouw

(2000) Authoritarian governance, power, and the politics of rescaling. Environment and Planning D: Society and Space 18(1): 63–76.

65.

ten Kate

von Hase

. & Maguire

(2018) Principles of the Business and Biodiversity Offsets Programme. In: Wende

Tucker

Quétier

Rayment

Darbi

(eds), Biodiversity Offsets: European Perspectives on No Net Loss of Biodiversity and Ecosystem Services. Cham, Switzerland: Springer.

66.

Whitehead

Jones

(2007) The nature of the state: excavating the political ecologies of the modern state. Oxford: Oxford University Press, 17–26.

67.

Womble

Doyle

(2012) The geography of trading ecosystem services: A case study of wetland and stream compensatory mitigation markets. Harvard Environmental Law Review 36: 229–2296.