The Everglades of south Florida once encompassed about 4,600 mi2 (three million acres) of slow-moving water and associated biota that stretched from the Lake Okeechobee drainage basin in the north to Florida Bay in the south (Figure ES-1).1 Today, human settlements and associated flood-control structures have reduced the Everglades to about half its original size.
To remedy the degradation of the Everglades, the Comprehensive Everglades Restoration Plan (“Restoration Plan”), was unveiled in 1999 with the goal of restoring the original hydrologic conditions to what remains of the natural ecosystem. Also in 1999, the National Academies established the Committee on Restoration of the Greater Everglades Ecosystem in response to a request from the Department of the Interior on behalf of the South Florida Ecosystem Restoration Task Force. The committee’s task (see Box ES-1) was to provide the Task Force with scientific advice in respect to the restoration activities and plans. This report evaluates the many storage options considered by Everglades restoration planners, including some options that are not in the Restoration Plan. Storage is a critical aspect of the Everglades ecosystem and of the Restoration Plan, but other critical factors, such as timing of land acquisition, intermediate states of restoration, and evaluating tradeoffs among competing goals or ecosystem components, provide the context for choosing and implementing storage options. Therefore, this report considers them as well.
WHY IS STORAGE IMPORTANT?
A basic premise of the Restoration Plan is that if the water is “right,” then the ecosystem will become “right” as well. The amount of water in the Greater Everglades Ecosystem today, and its spatial and temporal distributions, are very different from conditions in the natural system, which included the Kissimmee River drainage north of Lake Okeechobee, the lake, and the Everglades system south of the lake. Before drainage and other human modifications to the landscape that began in the late 1800s, seasonal variations in the amount and distribution of water in the system were strongly damped and the system was not as prone as it is today to rapid water-level changes that cause flooding and drying. In addition, the human demand for water in south Florida is much greater than it was 100 years ago, and there often are competing goals for the use of stored water.
This NRC activity (CROGEE) provides scientific guidance to multiple agencies (the South Florida Ecosystem Restoration Task Force, or SFERTF) charged with restoration and preservation of the Central and South Florida aquatic ecosystem, i.e., the greater Everglades. The activity provides a scientific overview and technical assessment of the many complicated, interrelated activities and plans that are occurring at the federal, state, and nongovernmental levels. In addition to strategic assessments and guidance, the NRC provides more focused advice on technical topics of importance to the restoration efforts when appropriate.
A major feature of the restoration plan is providing enough water storage capacity to meet human needs while also providing the needs of the greater Everglades ecosystem. One of the primary assumptions of the restoration effort has been that “getting the water right” is the most important single factor leading to sustainable ecologic restoration. Given the importance of storage to the restoration effort the CROGEE, with the SFERTF endorsement and cooperation, undertook a review of hydrologic and ecological analysis and other considerations with respect to analysis of size and location of water storage components proposed in the Restudy.
Early modifications to the landscape drained many areas and increased peak flows in others. Overall, they reduced the amount of water stored within the Everglades Ecosystem and thus increased the risk of desiccation of wetlands in the southern part of the system during droughts. However, at the same time, these modifications increased the risk of flooding in many areas. For all those reasons, many control structures such as levees and canals were built, and the Water Conservation Areas (WCAs) were created. The result is that parts of the Everglades are water-starved at times, other parts are submerged, and the natural timing and amplitudes of high-water and drying events have been severely disrupted. Large pulses of fresh water diverted to sea have also had detrimental effects on estuaries. As a result, the Restoration Plan includes large amounts of new, constructed storage to replace lost natural storage and supply the water that is needed for both people and the ecosystem when and where it is currently in shortest supply.
It is not clear exactly what ecological conditions will accompany hydrologic change, but there is merit in concluding that more natural hydrologic conditions will lead to improved ecosystem functioning. Thus attempting to “get the water right” (or at least better) is a reasonable approach to restoration.
MAJOR STORAGE AND WATER-CONSERVATION COMPONENTS IN THE RESTORATION PLAN
The major aspects of the Restoration Plan involve currently available and planned storage facilities. The largest existing storage components are Lake Okeechobee and the WCAs. Additional components are in place or planned for the completed Restoration Plan.
Lake Okeechobee historically was the key hydrologic link between the mainly upland ecosystems to the north and the wetlands ecosystems to the south. The lake receives an annual average of 1.6 million acre-feet of water through the Kissimmee River and discharges 416,000 acre-feet to the sea through the Caloosahatchee River and the St. Lucie Canal. Additional water is discharged to the WCAs and to adjacent agricultural areas.
Despite many hydrologic changes to the system, the lake still provides substantial water storage. While current operating rules are not designed primarily to maximize storage, they do permit up to 470,000 acre-feet of storage capacity for each foot of drawdown. Planned modifications to the operating rules, which are intended to reduce fluctuations in lake level to protect the littoral zone, water supply, and levee integrity, will decrease available storage capacity in the lake.
Lake Okeechobee has higher nutrient concentrations, especially phosphorus, than would be ideal as a source of water for the Everglades, despite extensive efforts to limit nutrient inputs. These concentrations are substantially above stated goals for the lake. It is a goal of the Restoration Plan to reduce nutrient concentrations in the lake, largely by reducing nutrient inflows.
Water Conservation Areas
The central Everglades was converted into surface reservoirs called WCAs when levees were completed in 1963. They are currently used to detain excess surface water. WCAs serve many competing uses, including controlling floods, storing water to augment supplies along the east coast and in Everglades National Park, recharging groundwater, reducing seepage of water to the coast, and providing habitat for wildlife. Their combined storage capacity is 1,882,000 acre-feet. The WCAs still contain substantial remnants of original Everglades landscapes and thus offer a major opportunity for restoration. Restoration Plan projects are planned to “decompartmentalize” the WCAs and enhance sheetflow.
Conventional Surface Reservoirs
The Restoration Plan includes large conventional reservoirs in the Kissimmee basin north of Lake Okeechobee, the Everglades Agricultural Area, and the Upper East Coast plus additional smaller reservoirs and stormwater treatment areas (STAs) to remove nutrients, especially phosphorus. Together these features will provide new storage capacity of about 1,120,000 acre-feet.
Land acquisition costs for new reservoirs would be significant, especially for the Upper East Coast reservoirs. The long-term effectiveness of STAs is still untested, but their lifespans are finite. Some water-quality issues remain to be resolved.
Aquifer Storage and Recovery (ASR)
ASR involves pumping water into subsurface geologic formations, then recovering it as needed. It has a planned annual average capacity of over 500,000 acre-feet of storage and a cu-
mulative capacity of more than 4 million acre-feet. Even with 30 percent loss of water during injection (as assumed in simulations by the SFWMD), the ASR systems account for about three-quarters of the new storage capacity of the Restoration Plan. ASR would not require large amounts of land; in addition, water stored underground would not experience evaporative losses. However, ASR—especially on as large a scale as envisioned in the Restoration Plan–is an untested technology. It also will require large amounts of energy for injecting and recovering water, and the water might need treatment to meet quality standards.
This component is planned to consist of reservoirs constructed in former quarries up to 80 feet deep with a storage capacity of about 330,000 acre-feet. Two of these west of Miami are anticipated to cover 9,700 acres; the area likely will be mined whether or not the quarries are converted to reservoirs. The conversion will require seepage barriers. As is true for ASR, the technology for creating such barriers at this scale has not been developed or tested, and so costs and feasibility of this option are uncertain. Therefore, pilot studies are planned, but these are not yet under way. Estimated construction costs are higher than for conventional surface reservoirs, and the seepage barriers will likely incur maintenance and repair costs over the long term. Water quality issues also are unresolved at present.
Seepage across levees that bound the WCAs and Everglades National Park can exceed one million acre-feet per year. Seepage management reduces this loss or recovers it and returns it to the interior. It is not a storage component, but as a water-conservation component it would have the same net effect as storage.
Water Reuse and Conservation
This component envisions two wastewater-reuse facilities in Miami-Dade County, ultimately slated to produce 220 million gallons per day or about 250,000 acre-feet per year. It requires advanced waste treatment with high capital and maintenance costs. This option involves conservation rather than storage, to be implemented in the likely event that more economical sources of water are not available.
Costs and Effectiveness
The storage options can be compared in terms of their costs and effectiveness in several ways. Of the new storage components that will be created by the Restoration Plan, conventional storage reservoirs have the advantages of using proven technology and of needing less input of energy and money than water reuse or ASR. ASR systems are the most expensive to site and build when compared on the basis of average annual outflow, but they are the least expensive
when compared on the basis of the maximum storage they can provide in a single year. Other factors, such as reliability, environmental consequences, and social and political acceptability also are important.
The creation of new water-storage capacity through implementation of the Restoration Plan involves large-scale re-engineering of much of the Greater Everglades Ecosystem and it consists of many individual projects. With so many components and constraints on this ambitious project, how the components are ordered in space, and especially in time, can profoundly affect the outcomes of the project.
The project’s overall plan does impose some constraints on sequencing of its components, as is true for any construction project. The committee judged two criteria to be most important in deciding how to sequence components of such a project.
Protect Against Additional Habitat Loss
The first criterion is that the sequencing should protect the system against any damaging changes in external or environmental conditions—especially for habitat that is or has the potential to be ecologically valuable—that would adversely affect the project’s success and that could not be reversed if such changes occurred. In the case of the Restoration Plan, the most striking such environmental change would be the loss or irreversible2 alteration of land-surface required to implement the plan. The most urgent and overriding sequencing criterion should be to protect from irreversible development all land that is or potentially could be included in the Restoration Plan. This kind of protection can be achieved by acquisition of the land, by obtaining easements, by zoning restrictions, or other methods. The Restoration Plan specifies that the method to be used is acquisition. Despite annual expenditures for land acquisition of between $100 million and $200 million, the plan for acquisition of the needed lands that have not yet been acquired extends over more than two decades. Irreversible development of land not yet acquired and increases in the price of land are almost certain. Therefore, delays in acquiring or protecting critical lands risk compromising the outcome of the Restoration Plan.
Provide Ecological Benefits as Early as Possible
The second criterion is that the sequencing should provide ecological benefits as early as possible. As the restoration of the Everglades begins, there is continuing reduction in species’ distributions and loss of habitats distinctive of the Everglades. There is high potential for these losses to be irreversible. In addition, invasive species continue to increase in number and distribution in the Everglades, despite efforts to eliminate some of them. As the ridge-and-slough and tree-island landscapes continue to erode, there is increased homogeneity of Everglades land-
scapes. Communities of marl prairies and periphyton mats continue to diminish in areal coverage, and nutrient loading continues to be above historic levels. These continuing losses and degradation of habitat and ecological functioning and the great uncertainty associated with implementation of the Restoration Plan and ecological restoration goals all argue for increased emphasis on achieving near-term ecological results. Those uncertainties are discussed below.
The ability of the various storage technologies in the Restoration Plan to provide the quantity and quality of water required to achieve the Plan’s goals is surrounded by a variety of uncertainties. Some uncertainties are inherent in measurement and interpretation of both hydrologic and ecological data. In addition, there are uncertainties related to natural variability and unpredictability of ecological systems (process uncertainty) and model applications. Natural system uncertainties include processes such as climate change—certain to occur, but uncertain in magnitude, rate, and direction—and ecological system responses to such changes. Model uncertainty arises from the use of simplified, abstract representations of complex systems and from model error, i.e., misunderstanding of variables and the functional form of the model. There also is uncertainty about historical information on the system, which is used for the Natural System Model to identify hydrologic goals for the Restoration Plan. Further, model projections are based on a fully implemented Restoration Plan, but there is uncertainty about what hydrologic and ecological conditions will occur during the transition from current conditions to the final restored conditions. Finally, there is uncertainty about what the restored hydrologic and especially ecological conditions will be, when they will be attained, and how variable they will be. These unavoidable uncertainties underscore the importance of procedures that can accommodate them.
Large uncertainties surround the future population size and its distribution in the region, and human activities, both in and outside the region. Population projections for south Florida have a history of being too low. At some unknown future time, however, population growth will slow and stop, and the slowing likely will not be well predicted either. Other uncertainties are associated with changes in societal values and restoration policy, including uncertainty about the location, size and timing of future stressors to the system. Funding for the Restoration Plan will be influenced by changes in values and policies. Events outside the region also are likely to affect what happens in the region.
Specific issues that introduce uncertainties include the Endangered Species Act (ESA) and its effect on implementation of water management, even though recovery of endangered species is an explicit objective of the Restoration Plan. For example, even management actions that could have beneficial effects on the Everglades in terms of Restoration Plan goals could be prevented, by Fish and Wildlife Service regulation or litigation by others if those actions adversely affected an endangered species, even temporarily. As discussed in Chapter 3, lawsuits based on the Clean Water Act also have introduced uncertainty to the implementation of the Restoration Plan.
Another issue is the effect of invasive and irruptive species. The Everglades has many nonnative invasive species, notably the Australian bottle brush tree, Brazilian pepper, and Australian pine. The native cattail seems to have dramatically increased its presence in the Everglades as a result of higher phosphorus concentrations, deeper water, and longer periods of inundation. Several tropical fish species have become established in the Everglades as well.
SUSTAINABILITY OF THE RESTORATION PLAN
The Restoration Plan relies very heavily on engineered solutions such as ASR and the Lake Belt storage system. Although there is a clear need for additional storage to implement the Restoration Plan, experience suggests that natural restoration processes usually produce more satisfactory restoration outcomes than engineered ones. However, opportunities to restore a system in which flows are controlled only by natural processes in natural areas are severely constrained in south Florida. This is the result of the restricted footprint of the remaining natural areas in the Everglades, the proximity of urban and agricultural lands that cannot be subjected to flooding without significant loss of property values, and the current and future demands for urban and agricultural water supply. Many of the natural storage features of the system, which provided essential damping of seasonal and storm-driven flows, have been lost permanently as a result of agricultural and urban development. Simply routing excess water from Lake Okeechobee to the southern Everglades through pipes or other structures that bypass the agricultural area would reduce the detrimental pulses of freshwater discharged to estuaries, but it would generate unnatural timing and magnitudes of flows and water levels, as well as high nutrient concentrations, in the terrestrial ecosystem and in Florida Bay.
Some of the natural storage and damping could be restored if agricultural land south of Lake Okeechobee were converted into a restored corridor connecting the lake to the southern Everglades. However, subsidence due to peat loss in the agricultural area south of Lake Okeechobee has caused the land surface to be lower than in areas to the south. This means that even if the Herbert Hoover Dike were breached, slow sheet flow to the south would not be restored in the area that was historically a sawgrass plain. Instead, the subsided region would become an extension of the lake itself. An expanded lake of this type would provide significant storage and damping of southward flows, but it would also inundate established communities and agricultural lands surrounding the current perimeter of the lake and increase the flooding hazard in other areas to the south and southeast. This type of restoration, therefore, would require additional engineering measures for flood control.
Clearly, some degree of engineering control will be necessary in any plan to restore more natural water levels and flows in the southern Everglades. The framework developed by an earlier NRC committee to consider options for interventions to enhance wild salmon runs in the Pacific Northwest is applicable to the Everglades restoration as well. The earlier committee recognized, as we do, that engineering techniques would be needed, at least in the short term, but recommended that they be used with the ultimate goal of rehabilitating ecosystems to the point where human inputs can be substantially reduced, if not eliminated.
There is a considerable range in the degree to which various proposed storage components involve complex design and construction measures, rely on active controls and frequent equipment maintenance, and require fossil fuels or other energy sources for operation. Storage components that have fewer of those requirements are likely to be less vulnerable to failure and hence are likely to be more sustainable in the long term.
A SECOND LOOK AT CONSTRAINTS, BOUNDARIES, AND ADAPTATION
The planning framework that led to the Restoration Plan resulted from a process of adaptation and compromise among interests and concerns of myriad stakeholders in south Florida,
including governments. As new information becomes available and as the effectiveness and feasibility of various restoration components become clearer, some of the earlier adaptation and compromises probably will need to be revisited if the restoration is to meet its goals. Unanticipated changes that occur will likely require rapid responses. Therefore, it is even more important to deal with changes that can be anticipated in a timely and proactive way. The progressive loss of soil in the Everglades Agricultural Area is an example of a change that can be anticipated in advance. In addition, it is likely to become ever clearer that not all current interests and conditions can be protected while preserving restoration options. We discuss two examples here, the Everglades Agricultural Area and Lake Okeechobee.
Everglades Agricultural Area
The Everglades Agricultural Area (EAA) immediately south of Lake Okeechobee was an important conduit for sheetflow in the unaltered Everglades. Today, this area of rich peat soils is devoted mainly to sugarcane production. The total agricultural value of its produce is more than $640 million annually. However, agricultural drainage has led to oxidation of the peat soils and subsequent subsidence. It is certain that unchecked, subsidence will eliminate the topsoil, making agriculture at best extremely expensive, although it is not certain just when that will occur. The EAA has a variety of potential fates. The worst from the point of view of Everglades restoration would be commercial, residential, and industrial development of the area. It is not clear what continued agricultural production would require, but it is likely that eventually the required treatments would make the area less amenable to restoration. Another possibility would be to consider uses of the EAA more aligned with restoration needs. Those might include turning parts or all of it into a wetland with a cattail-sawgrass gradient. Perhaps it could simply be flooded and the water used for storage and to enhance sheetflow.
As discussed in Chapter 4, many factors unrelated to the Restoration Plan, such as import restrictions on sugar, the number and distribution of people living in south Florida, energy costs, and so on are likely to change, and those changes will affect calculations related to potential uses of the EAA. For these reasons, this committee recommends a re-evaluation of the EAA’s future role in Everglades restoration. This is a complex analysis, requiring estimates of the costs of land acquisition, the feasibility and likely costs of various options, and other matters. Such analysis should begin as soon as possible.
Lake Okeechobee is a major component of the Everglades ecosystem. It was a key natural hydrologic link between upland ecosystems to its north and the marshes and prairies of the Everglades to the south, and, especially before the hydrologic modifications made in the twentieth century, it moderated the effects of variations in rainfall. It also provides drinking water to nearby communities and recreational opportunities. The lake has the capacity to provide much more storage than it does under its current operating rules. Several issues, including water quality, flood control, and the extent and functioning of the littoral zone, need careful consideration if the lake is to serve the latter purpose.
Several options available for increasing the storage capacity of the lake have been considered in the development of the Restoration Plan; they would have extreme effects on lake levels and would diminish the lake’s ecological value and its value for fishing. Other more moderate options or combinations might have a better array of costs and benefits. Given the possibility that some of the components of the Restoration will be more costly or less effective than envisioned, the committee judges that the use of Lake Okeechobee for storage should be revisited. Other storage options have their own environmental and financial costs, and the analysis could lead to a beneficial change in the overall plan. For both the EAA and Lake Okeechobee, any actions taken after the re-evaluations should be done using adaptive management. An added incentive for the re-evaluation is the potential to provide ecological benefits earlier in the restoration.
A Conceptual Restoration System Performance Measure
In previous reports of this committee, the importance of evaluating the restoration effort was discussed. This has been a major focus of the Restoration Plan scientists as well. Major difficulties are associated with such evaluations. One is translating the general and societal goals of the Restoration Plan into realistic targets and performance measures. Restoration of the ecosystem to its pristine state, however that might be defined, is not possible, because so much has changed in south Florida. Another difficulty is that restoration of one aspect of ecosystem functioning or of biological diversity might have to come at the expense of another. And not all aspects of ecosystem structure and functioning are equally valued by all sectors of the public or even by all agencies in the region. Thus, any overall restoration goal will require tradeoffs among subsets of ecosystem goals and among desired endpoints. For these reasons, the committee proposes a system performance measure based on multi-attribute decision making that could be used to help evaluate restoration progress and alternatives. The measure is akin to a utility function in economics, is based on hydrologic performance measures, and can be expressed mathematically as the weighted sum of individual performance measures.
The performance measure is intended to complement rather than replace other evaluation tools already in place. Its main value would be in the context of an inclusive process involving stakeholders to evaluate policy and management tradeoffs and alternatives. Its properties are described in Chapter 5, and its use as an analytic tool is recommended. In addition to the numerical performance measure, and based in part on it, there is a need to embed learning into the processes of project planning, evaluation, implementation, and operation (adaptive management).
MAJOR FINDINGS AND RECOMMENDATIONS
Finding 1. The historic resilience of the ecosystem was a direct consequence of the continuity and the diverse mosaic of natural system communities found over a wide range of spatial scales. As the spatial extent of the ecosystem is reduced, the resiliency of the system is reduced and susceptibility to unexpected and irreversible change is increased. Although a considerable amount of money ($100-200 million annually) is allocated to land
acquisition, it seems certain that some land not soon acquired will be developed or become significantly more expensive before the two-decade-long acquisition program can be completed. Protecting the potential for restoration, i.e., protecting the land, is essential for successful restoration.
Recommendation 1. Preservation of the remaining areal extent of the potential natural system should be a priority. Land should be purchased or conservation easements should be obtained now to prevent additional loss of land to development and to provide a buffer between the built and natural environments. (Chapter 3.)
Finding 2. A restoration as ambitious and complex as the Everglades Restoration Plan has the potential to allow—and perhaps even cause—irreversible changes to the Everglades ecosystem as it proceeds. Some processes of deterioration might continue to an undesirable endpoint before the restoration is complete, and in some cases, it is possible that an intermediate stage between current conditions and the restoration goal could result in additional damage.
Recommendation 2. Efforts should be made to prevent irreparable damage to the ecosystem during the restoration. The focus should be on interim changes in the system as well as the end point of the restoration to avoid losses in the short-term that will prevent ecosystem restoration in the long term. (Chapter 3.)
Finding 3. Some aspects of the restoration are likely to benefit the target ecosystem components while adversely affecting others, at least until the restoration is completed. In other cases, finite resources and other factors are likely to lead to differing restoration goals for different parts of the ecosystem and among different stakeholders.
Recommendation 3. Methods should be developed to allow tradeoffs to be assessed over broad spatial and long temporal scales, especially for the entire ecosystem. Development of methods now, such as the overall performance indicator described in Chapter 5, will allow alternatives to be tested quickly and modifications to the restoration to be developed when surprises do occur. (Chapters 3, 4, and 5.)
Finding 4. It is likely that some components of the Restoration Plan will be more costly or less effective than envisioned. The high degree of uncertainty associated with all phases (economic, social, political, engineering, and ecological) of the Restoration Plan necessitates the allocation of significant effort to establish alternative approaches to restoration (contingency planning). Even if the Restoration Plan “gets the water right,” there are circumstances that might prevent restoration of the Everglades to the conditions envisioned by the plan. The multi-species recovery plan, efforts to eradicate invasive species, changes in water-quality legislation, and many other factors may have major influences on the restoration effort.
Recommendation 4. In addition to the contingency planning that already is being undertaken, more intensive and extensive planning should be pursued. In particular, options such as those discussed in Chapter 4 should be considered for using the Everglades Agricultural Area and Lake Okeechobee as elements of the Restoration Plan in ways that are not now part of the plan. Any such change in the use of EAA and Lake Okeechobee
should be undertaken using adaptive management, and it has the potential to bring ecological benefits earlier. (Chapter 4.)
Finding 5. A variety of economic, political, financial, engineering, and other factors and constraints have resulted in a restoration plan that provides most of its ecological benefits towards the end of the process. Some of the delay is unavoidable, because some engineering structures must be in place before other elements of the plan can be implemented. However, the longer the provision of such benefits is delayed, the more likely that continued degradation will occur, that loss of species and habitats will continue, and that at least some political support will be lost as well. These factors argue for increased emphasis on ecological results earlier in the plan.
Recommendation 5. Restoration projects should be implemented in a way that provides benefits to the natural system sooner rather than later by accelerating storage projects that are not as reliant on technology or use short-term storage solutions to achieve benefits to the natural system until more technologically advanced methods are proven. An example of such a benefit to the natural system would be providing more natural flows (in terms of seasonal timing, volume, and flow velocity) to Everglades National Park. Doing so might not require large-scale changes in sequencing; instead, incremental changes could add up to be significant. (Chapter 3.)
Finding 6. Many projects that will contribute to or otherwise affect the restoration of the Everglades are not part of the Restoration Plan. To the degree that there is coordination or at least communication among those projects, benefits of economy and of effectiveness are likely.
Recommendation 6. Coordination and communication among the various restoration efforts should continue to receive high priority. (Chapter 3.)
Finding 7. Considering the 40-year time frame of the Restoration Plan and perhaps a century of system response, a regional information synthesis center would enable the systematic provision of evolving, reliable knowledge in support of the policy process and the interested public who affect and are affected by the program. Such a center also would help implement adaptive management on a system-wide basis.
Recommendation 7. Incorporation of integrated assessment models, long-range-development scenarios, and a regional information-synthesis center into an adaptive-management and assessment program in the Restoration Plan should be considered. Monitoring is an essential part of adaptive management, and models have the potential to help design, assess, and evaluate the results of monitoring programs. (Chapter 3.)