CHAPTER SEVEN ELEMENTS OF AN ADAPTIVE MANAGEMENT STRATEGY FOR THE LAKE TAHOE BASIN Patricia N. Manley, John C. Tracy, Dennis D. Murphy, Barry R. Noon, Mark A. Nechodom, and Christopher M. Knopp

©1999 J.T. Ravizé. All rights reserved.

CHAPTER SEVEN

ELEMENTS OF AN ADAPTIVE MANAGEMENT STRATEGY FOR THE LAKE TAHOE BASIN

Patricia N. Manley, John C. Tracy, Dennis D. Murphy, Barry R. Noon, Mark A. Nechodom, and Christopher M. Knopp

Introduction poses a threat not only to the of natural Adaptive management is manage- ecosystem processes within the Lake Tahoe basin ment informed by research and monitoring. This but to the sustainability of the basin’s social, cultural, chapter presents some key elements of an adaptive and economic systems. Lake Tahoe’s recreation- management strategy that can help managers and based economy depends to a large degree on the policy-makers in the Lake Tahoe basin proceed with health of its , the availability of scenic alpine restoration efforts in the face of limited information. vistas, and the quality of the lake’s . The sim- With constant feedback and revision, management ple recognition that the health of the society and can become more effective, efficient, and account- economy of the basin is related to the health of the able. Because adaptive management essentially en- environment underscores the need to assess the tails “learning by doing,” as well as “action based on conditions and trends of the basin as a whole sys- learning,” management actions, data gathering, and tem. decision-making must interact and keep pace with Although a complete understanding of the each other. Ideally, management and research are integrated nature of in the Lake Tahoe designed to maximize information gain, the course basin has not been achieved, restoration activities of management is readily evaluated in light of new need to proceed. Approximately $200 million already information as it becomes available, and manage- has been spent since the early 1980s on improve- ment direction is efficiently revised in response. To ments to ecosystem health within the basin. Another do so, coordination of and management in $187 million has been invested in state and federal the Lake Tahoe basin will be paramount—new lines acquisitions of ecologically significant in urban of communication and inter-organizational links will intermix areas. The Tahoe Regional Planning be necessary. Agency’s Environmental Improvement Program Establishing an effective adaptive manage- (EIP) in 1998 identified $900 million in future pro- ment process will take time and investment. The jects to restore the Lake Tahoe basin to a more de- Lake Tahoe basin has already made substantial in- sirable condition. However, no specific process has vestments in data acquisition, information manage- been developed for integrating the role of science ment systems, and formal and informal mechanisms into the implementation of the EIP. Scientific re- of communication. However, the rate at which man- search will continue within the basin, but, without agement and restoration activities must be applied in focused effort, it cannot be well coordinated with the basin in order to meet current conservation and management nor can it efficiently contribute to restoration goals suggests that time is of the essence meeting management goals. and that development and implementation of an The concept of adaptive management was adaptive management strategy is critical. developed more than three decades ago, based on Previous chapters of this assessment have the observation that science and management were documented degraded elements of the Lake Tahoe engaged in an inefficient partnership (Walters 1986). basin ecosystem. If left unchecked, this degradation It is often described as a cycle analogous to cycles of

Lake Tahoe Watershed Assessment 691 Chapter 7 birth, death, and rebirth. New information and mation acquisition and assessment phase of the cycle changing perceptions, needs, and desires force the and a brief reference to how information can best be death of old ideas, structures, and processes. A pe- transferred to the evaluation and decision-making riod of disorganization is followed by revitalization, phase. The other two stages are critical to developing as the birth of new approaches, paradigms, and di- a fully functioning adaptive approach to the man- rections emerge from the synthesis and evaluation of agement of resources in the basin; however, they new information and context. When new directions largely pertain to public policy development and and approaches become solidified, a period of stabil- participatory evaluation processes rather than to the ity follows in which management direction, proce- direct relationships between scientific research and dures, and protocols are made institutional and rou- management. tine. This stable phase abides until new information The information acquisition and assessment and changing social preferences once again precipi- phase of the adaptive management cycle includes tate a period of reorganization. research, monitoring, and modeling activities, which In a much more applied sense, the cycle of provide new information for making management adaptive management can be described in four decisions (Figure 7-1). Each of these three activities phases: information needs, information acquisition is an essential and complementary component of the and assessment, evaluation and decision-making, and information gathering stage of adaptive manage- management actions. This chapter focuses on two of ment. Research, monitoring, and modeling are iden- the four phases: an in-depth discussion of the infor- tified in Figure 7-1 as three distinct activities in the

Identification Information Needs Acquiring and Assessing Information

Monitoring

Planning Prioritization and Resources Allocation of Resources Modeling Information

Research

Information Needs Identifcation

Evaluation and Decision-making Management Action

Sustainability Stakeholders Goals Agency Projects

Project Project Resources Prioritization New Regulations

Figure 7-1—A schematic diagram of an adaptive management planning cycle.

692 Lake Tahoe Watershed Assessment Chapter 7

information gathering and assessment phase of an the basin closer to the ecosystem health objectives adaptive management cycle; however, in practice articulated during the Presidential Forum in 1997. they are highly interconnected. Research provides new knowledge about system interactions and The Role of Science and Research in Adaptive dynamics, including basic information about Management resource interactions and validation of assumptions The role of science in the adaptive man- used in developing management direction. Scientific agement of ecosystems broadens scientific research research can be used to understand processes related as it is usually defined. While science in adaptive to environmental conditions, which in turn can be management includes the acquisition of new knowl- used in making decisions about how to monitor a edge through basic research, scientists must also given activity. Monitoring helps us to understand the assist managers in the interpretation and incorpora- status and trends of resource conditions, to assess tion of new knowledge as it applies specifically to progress toward management goals, and to develop a management problems. Scientists have become in- better understanding of management effectiveness. creasingly involved in management planning Monitoring data can either be used directly as and analysis as concerns for the sustainability of bio- decision-making information (to assess compliance physical and sociocultural systems have become with existing regulations) or it can be synthesized to more pervasive. Examples are many. Regional ex- evaluate the effectiveness of decisions (whether a amples include the science-driven planning restoration activity is having the desired effect). efforts in the Pacific Northwest, in which two dec- Modeling can be used to extract more information ades of research focussed on imperiled species, the and to learn from data collected through research structure and function of old-growth forest stands, and monitoring efforts. Synthesized monitoring and and the physical processes that sustain both have research data can also be used to improve process redefined management practices and definitions of descriptions or to adjust parameters in order to sustainability. Closer to Lake Tahoe, the CalFed Bay- refine models that describe watershed behavior. Delta Ecosystem Restoration Program uses focussed Modeling tools can be used to support management research, coupled with large-scale management in- decisions, to help decision-makers understand trade- tervention and pilot studies, all subject to scientific offs among various management options, and to review, to provide conservation planning guidance in facilitate broader public involvement in decision- the face of political and economic . making. Together, research, monitoring, and model- Even on private lands, basic research and effective- ing increase our scientific understanding of ecosys- ness monitoring are paired to inform a more than tems and their responses to management actions. five-million-acre plan in re- This chapter outlines the key elements of an sponse to endangered species concerns in Clark adaptive management strategy and discusses how County, Nevada. scientific information can be efficiently generated The newly expanded role of research en- and effectively applied to management in the Tahoe compasses the following activities: (1) developing basin. To do so the chapter departs from the rest of new information of relevance to the document by offering direct recommendations planning, (2) approaching research in a more inte- for developing and implementing an adaptive man- grated manner, working across disciplines and at agement strategy for the basin. The status of re- larger scales, (3) packaging information so that its search, monitoring, and modeling activities in the meaning and application are readily accessible to Lake Tahoe basin are described in this chapter in nonscientists, (4) working directly with managers to relation to future acquisition and application of as- develop and implement adaptive management ex- sessment tools and efforts. We present some ap- periments and monitoring, and (5) participating in proaches to adaptive management that have shown the planning process by providing scientific validity success in other geographic areas and then close with assessments of current information and its applica- a description of some next steps that would bring bility to individual planning and implementation

Lake Tahoe Watershed Assessment 693 Chapter 7

efforts (Committee of Scientists 1999). These activi- • Identifying explicitly key areas of uncer- ties warrant the more detailed discussion below. tainty; • Formulating multiple competing models or Development of New Information hypotheses regarding system function; The acquisition of new information through • Assigning probabilities to alternative out- research can take the form of passive or active adap- comes generated by competing models; tive management (Walters 1986). Passive adaptive • Testing iteratively competing models at a management refers to the linear sequential process scale commensurate with management; of gathering information on system conditions and • Whenever possible, designing management responses using a single model of system dynamics. actions in an experimental framework (ver- The validity of the system model (i.e., the current sus experimentation being carried out inde- understanding of resource links and interactions) and pendently from management); the management approach are tested with data. • Working collaboratively across disciplines Modeling and management approaches are revised if and at large geographic scales; and necessary, then new data are gathered to evaluate the • revised system model and management regimes. Ac- Taking advantage of environmental “sur- tive adaptive management, in contrast, is supported prises” (large-scale stochastic events) by by the formulation and testing of multiple models studying them to learn more about system simultaneously in an effort to speed the process of dynamics. information acquisition and to improve management These data are intentionally pursued to help more rapidly. Actively probing areas of answer specific questions related to key uncertainties on an experimental basis is a hallmark of active about system function and trajectory in an efficient adaptive management. and cost-effective manner. However, the sometimes In their recent report, which reviewed and narrow focus of active adaptive management re- recommended amendments to the National Forest search can miss broader, perhaps unexpected, Management Act, the Committee of Scientists high- changes in system conditions that may convey a lighted the complementary nature of passive and great deal of information. It is passive adaptive man- active adaptive management and suggested that a agement that can provide this service. In addition, combination of passive and adaptive management discoveries from basic research can and will provide provides the strongest approach to information ac- unexpected keystones to understanding system dy- quisition. Data acquisition that contributes to passive namics. adaptive management consists of (1) monitoring Who decides which research questions are system conditions to assess changes over time, (2) most relevant, what information to be obtained has evaluating the effectiveness of specific management highest priority, and who will be most successful in directions already in practice (effectiveness of ripar- conducting the research in a timely manner are im- ian protection zones in meeting intended goals), (3) portant open questions. Funds and time are always validating assumptions adopted in the formulation of limited. At a broader scale, several institutions, such existing management direction, and (4) generating as the National Science Foundation (NSF) and the new basic information on system form and function National Institute of Health (NIH), fund hundreds (Committee of Scientists 1999). The combination of of millions of dollars of research each year. A peer these data provides useful information but may not review process has proved to be an effective and reduce key uncertainties about system function, in- scientifically defensible means to evaluate and priori- tegrity, and trajectory. It is active adaptive manage- tize research proposals for funding. NSF, for exam- ment that provides this service. ple, establishes panels of scientists each year to allo- Active adaptive management reduces uncer- cate funds to scientists for research; in turn those tainty more efficiently than passive adaptive man- panels call upon networks of peer reviewers to agement through a number of approaches as follows: evaluate proposals in diverse topic areas. In the Lake

694 Lake Tahoe Watershed Assessment Chapter 7

Tahoe basin, the process of information prioritiza- ment agencies and stakeholders to define a research tion and resource allocation is not well developed agenda that most appropriately serves the breadth of and is often conducted ad hoc. Restoration goals contributions from science. Progress on these fronts within the basin would be better served by the estab- will require -taking, motivation, institutional sup- lishment of a formal prioritization and allocation port, and funding. process that involved scientific review. The coordination of scientific activities with Packaging Scientific Information management actions is at the core of an adaptive Generating new information is only one of management approach. Scientific research used to several important steps in adaptive management. increase knowledge of the Lake Tahoe basin must be McLain and Lee (1996) posit that effective manage- closely coordinated with management activities in ment requires societies not only to acquire knowl- order to achieve the restoration goals set for the ba- edge but also to change their behavior in response to sin. For example, experiments can be designed new information about the systems in which they around restoration activities. Active restoration ef- live. An organization’s effectiveness and responsive- forts can constitute significant environmental distur- ness can be measured by its ability to translate in- bances, and therefore can be used as impact vari- formation into appropriate action (Westley 1995). ables in research design. In this manner, a wider set Receptivity to new information depends in part on of research experiments can be undertaken, poten- the form of scientific information. Research has a tially leading to a greater and quicker understanding role not only in generating new information but also of ecosystem behavior. in facilitating access to information and in packaging information in a manner that can be readily under- Integrated Research stood and assimilated. Based on the successes and The broad range of contributions de- failures of previous land management planning ef- manded of research in adaptive management pre- forts, Westley (1995) identified characteristics that sents new challenges both to individual scientists and promote a rapid incorporation of information into to the agencies that employ them. Among those decision-making processes: results are unambiguous challenges is the growing need for research institu- and presented and explained simply, interpretation tions to work collaboratively with one another— of the results is placed in a management context (i.e., acknowledging the normally competitive nature of low probability of multiple interpretations), and re- funding and hiring practices. In addition, the rewards sults and potential implications are packaged in the for participation of scientists in environmental plan- context of the problem at hand (place-based assess- ning commonly do not conform to conventional ments with predefined issues and applications). measures of productivity (peer-reviewed publica- It is important not to underestimate the ca- tions). Finally, adaptive management calls for new pacities of the management and policy communities roles for science that lie outside the formal academic to understand and assimilate scientific information. training of most scientists. However, patterns of knowledge acquisition and use Full engagement of research in adaptive can differ widely among the end users of that knowl- management will require a change in how scientists edge (Weeks and Packard 1997). It is incumbent are viewed in the workforce. Both research institu- upon scientists actively involved in adaptive man- tions and management agencies need to invest in agement to present results in a manner that is not staff scientist positions that reflect the expanding unnecessarily laden with technical jargon, to assist in spectrum of contributions required to approach land translating new knowledge into an applied language and in an adaptive manner. As of management and policy, and to highlight links is suggested below, a coalition needs to be formed between new information and current land manage- among the research institutions in the Tahoe basin, ment objectives. and, in turn, they need to join forces with manage-

Lake Tahoe Watershed Assessment 695 Chapter 7

objectives related to needed goods and services. In The Role of Monitoring in Adaptive the Lake Tahoe basin, the main objective of moni- Management toring is to provide information on the condition of Monitoring will be an integral part of adap- biological, physical, and socioeconomic resources tive management of the Lake Tahoe basin. Also re- and how management is affecting those resources ferred to as environmental surveillance, monitoring relative to desired effects. A monitoring program for is the “measurement of environmental characteristics the Lake Tahoe basin must be able to describe the over an extended period of time to determine status status and trends of resource conditions and to dif- or trends in some aspect of environmental quality” ferentiate the effects of environmental factors that (Suter 1993). In the context of the watershed as- are outside the control of managers (intrinsic varia- sessment, an expanded definition that encompasses tion) from the effects of management activities (hu- three different forms of monitoring is appropriate: man-induced patterns of change) on resource condi- monitoring of management activities in relation to tions. planned activities (implementation monitoring), Specific management objectives and desired monitoring of the status and trend of resource con- conditions have been developed for the Tahoe basin ditions and their change agents (status and trend in the form of agency direction, primarily repre- monitoring), and monitoring of the effectiveness of sented by TRPA’s thresholds and the Forest Ser- current management practices in achieving desired vice’s standards and guidelines, with additional direc- conditions or trends (effectiveness monitoring). tion in the form of specific management direction It is helpful to further differentiate status for state lands and municipal land holdings. These data from trend data. The most common reason to sources of direction need to be melded into a cohe- monitor specific environmental indicators is to de- sive set of information objectives to be addressed tect differences in values among locations at a given through monitoring. Once these monitoring objec- time (status) or differences in value across time at a tives are established, they need to be refined by con- given location (trend). For example, changes in ob- sidering the purpose that information will serve and served Secchi disc depth as an indicator of lake clar- how it will be applied to decision-making. The Na- ity are useful in that they indicate adverse changes in tional Research Council (NRC 1995) has identified the ecosystem of the basin. Such trend data are par- two general approaches, “retrospective” and “predic- ticularly valuable information because they can signal tive” monitoring, to designing monitoring programs potential future conditions associated with system that serve to circumscribe the primary monitoring degradation. Nonetheless, the timeframe for status purposes. As the NRC suggests, “retrospective or and trend monitoring is frequently left unspecified effects-oriented monitoring seeks to find effects by because of uncertainties inherent in funding and the detecting changes in status or condition of some impacts of human behavior and population growth. organism, population, or community.” Retrospective But while timeframes may not always be readily monitoring identifies resource attributes of interest, specified, three key features of monitoring efforts the primary environmental factors that could influ- always serve the adaptive management process: iden- ence their condition, and the management actions tification of the goals, objectives, and questions to be that are likely to also affect their condition (inten- addressed, selection of indicators and their interpre- tionally or otherwise). Retrospective monitoring does tation, and application of monitoring data to man- not require knowing cause-and-effect relationships. agement decisions. These three areas are addressed In contrast, “predictive or stress-oriented monitoring below. seeks to detect the known or suspected cause of an undesirable effect (a stressor) before the effect has Monitoring Goals, Objectives, and Questions had a chance to occur or to become serious.” Predic- tive monitoring is more narrowly focused on ex- The overarching goal of monitoring is to pected changes and requires the identification of determine whether current management practices are cause-effect relationships between stressors and re- maintaining the ecological integrity of the target eco- source conditions. logical systems, as well as achieving socioeconomic

696 Lake Tahoe Watershed Assessment Chapter 7

Both retrospective and predictive effects Monitoring is most informative if it includes monitoring have value for the Lake Tahoe basin. monitoring of stressors as well as resource condi- They provide complementary information about tions. Environmental stressors are anticipated extrin- resource conditions, management activities, and the sic factors that may compromise the integrity of the environment, improving the ability to differentiate ecosystem and its component resources. Stressors, as the effects of management from other environ- defined here, can be both human-induced and mental influences. While retrospective monitoring “natural.” provides a broad spectrum of information about To return to the example of old-growth resource conditions and potential influential factors, forest monitoring, stressors that might be included predictive monitoring provides more detailed infor- in a monitoring scheme that targets old-growth for- mation on a more limited set of conditions that are ests include the following (also see Barber 1994): suspected to be at greater risk of detrimental change. • Prescribed and natural fire as a link to loss Once the balance of emphasis on retrospec- of late seral habitat; tive and predictive monitoring is determined, moni- • Dams and diversions as links to alterations toring questions can be developed. For example, of hydrologic cycles; monitoring questions to address the amount and • Altered climatic regimes as links to in- distribution of old-growth forest at Lake Tahoe creased sediment loads to streams from might include the following: storms; • Is management direction being followed in • Urbanization as a link to the reduction, loss, management actions? (implementation or fragmentation of habitat; monitoring); • • Road construction as a link to changes in What is the amount and distribution of for- the horizontal transport of and nu- est age classes, including old-growth forests, trients; and at the scale and how is it chang- • as a link to reduction in lake ing over time? (status and trend monitor- clarity from atmospheric deposits of nutri- ing); • ents. What are the distributions of patch size, Cause-and-effect relationships may be more patch interior area, and interpatch distances credibly derived from a monitoring program when for old-growth forests at the landscape scale the status of stressors is accurately documented. and how are they changing over time? Again, the choice of indicators and stressors is nec- (status and trend monitoring); essarily guided by the questions one seeks to address. • What is the biological diversity in old- growth forests and how is it changing over The Use of Conceptual Models for Indicator time? (status and trend monitoring); Selection • What is the correlative relationship between A well-constructed and well-implemented tree mortality and biological diversity in old- monitoring program should explicitly link scientific growth forests and key stressors (air quality, knowledge of ecosystem conditions to the selection fire, recreational uses)? (effectiveness moni- and interpretation of indicators. The use of concep- toring); • tual models of system dynamics to inform and What changes have been produced by man- document indicator selection is recommended by the agement actions in the amount and distribu- NRC (1995) and is becoming an increasingly com- tion of forest stand structure? (effectiveness mon practice (Noon et al. 1999; Manley et al. in monitoring); and press). The likelihood of choosing appropriate indi- • Is management effective at leading to an in- cators is greatly improved if a monitoring scheme’s crease in the amount and distribution of conceptual model thoroughly characterizes system old-growth forest? (effectiveness monitor- dynamics and accurately reflects the effect of stress- ing). ors on system conditions. Furthermore, the use of

Lake Tahoe Watershed Assessment 697 Chapter 7

well-designed conceptual models can enable a moni- worksheet to characterize stressors and their antici- toring program to investigate relationships between pated effects on ecosystems and their component stressors and environmental consequences, and it elements, where scale was considered by allocating can provide the foundation for developing detailed the effects of specific stressors to various levels in predictive models. the ecological hierarchy—landscape, community/ Well-developed conceptual models outline ecosystem, population/species, or genetic levels (see the interconnections among ecosystem resources also Noss 1990). (key system components) and between ecosystem resources and environmental stressors, the strength Selecting and Interpreting Indicators for and direction of those links, and attributes that char- Monitoring acterize the state of the resources and stressors. Once monitoring questions have been ar- Models should demonstrate how systems work, with ticulated and a conceptual model has been devel- particular emphasis on anticipated system responses oped, appropriate indicators can be selected. By to stressors. Conceptual models also should indicate convention, measured environmental attributes are how systems respond to natural disturbances referred to as indicators, under the assumption that (changes in successional pathways) and how they their values in some way indicate the quality, health, develop resilience to disturbance. In most cases it or integrity of the larger system to which they belong will be sufficient to model restricted, but relevant, (Hunsaker and Carpenter 1990; Olsen 1992). The components of systems to identify appropriate indi- ultimate success or failure of an adaptive manage- cators and to provide a foundation for more detailed ment program may be determined by the selection of modeling. In other words, complete descriptive indicators. Even if a monitoring program is fully models of ecosystems are seldom necessary in order funded and implemented for many years, it will fail to proceed with a reliable monitoring program. to be effective if the wrong indicators are selected. As a general goal, management associated For purposes of this watershed assessment, envi- with most monitoring programs will strive to main- ronmental attributes at Lake Tahoe can be broadly tain ecological and associated sociocultural proc- defined to include biological, physical, and socioeco- esses. Many ecological processes, however, are diffi- nomic features that can be measured or estimated. cult or impossible to measure directly. Conceptual The task of detecting and recognizing models can identify structural and compositional meaningful change in ecosystems is complex because components of the resources affected by underlying those systems are inherently dynamic and spatially processes. A conceptual model should clearly iden- heterogeneous. Moreover, many changes are not tify the processes and pathways by which stressors human-induced and in many cases are not amenable are linked to changes in ecosystem composition, to management intervention. At least three kinds of structure, and process and how particular indicators changes are inherent in natural systems: stochastic are suited to represent these stressors, conditions, variation, cyclic variation, and successional trends and processes. following disturbance. These changes, or sources of The ability to measure and draw inferences variation, need to be recognized and accounted for from ecosystems is affected by the scale of observa- in a monitoring program in order to differentiate tion. The temporal and spatial scales at which proc- their effects from management effects on resource esses operate and resources respond must be esti- conditions. mated and identified in the conceptual model in or- Indicator selection processes should closely der to determine the appropriate scale of measure- follow the goals, objectives, and questions estab- ment for a given indicator. Conceptual models with lished for the monitoring program. The same re- hierarchical structures are helpful in addressing mul- source may be monitored using very different indica- tiple scales. Ideally, the model will reflect processes tors, depending on whether the objective is to de- that operate at a range of temporal and spatial scales scribe the status and trend of an ecosystem condi- and that accommodate the constraints operating at tion (retrospective monitoring) or to obtain an early each scale (Allen and Starr 1982; Allen and Hoekstra warning of detrimental change in a condition (pre- 1992). For example, Noon et al. (1999) developed a dictive monitoring). For example, on a parcel of

698 Lake Tahoe Watershed Assessment Chapter 7

public land, the Endangered Species Act may require • The costs of indicator measurement should monitoring of the status and trend in the population not be prohibitive. of a high trophic level vertebrate predator, such as Additional evaluation criteria for screening the bald eagle. The history of this species (long- candidate indicators are presented by the National lived, high survival rate, low fecundity, and high site Research Council (NRC 1990) and in Barber (1994). fidelity) exhibits lags in its response to environ- mental change. Status and trend monitoring would Considerations in Data Collection require measuring attributes that tend to change in In general, determining the status of an in- longer cycles. However, if the goal were to seek early dicator is a challenge in estimating the value of an warning signals for declines in bald eagle popula- unknown parameter within some specified bounds tions, then related attributes that change more of precision. Estimates of trend address the pattern quickly, such as sizes of prey populations, condition of change over time in the status of the indicator. of roost sites, or survival of young, would be more How to efficiently acquire these estimates lies in the appropriate to monitor. realm of survey and sample design (Cochran 1977). A variety of evaluation criteria are helpful to Proper design requires substantial statistical exper- consider in selecting indicators. Once a conceptual tise; fortunately, there exists a large body of statistical model is developed and consulted, candidate indica- literature on parameter estimation, hypothesis test- tors can be proposed for monitoring and subsequent ing, and trend estimation that is relevant to monitor- field-testing. Indicators should meet the following ing (Sauer and Droege 1990). Some debate exists criteria: over whether parameter estimation or hypothesis • They should reflect underlying ecological testing is the correct statistical framework for moni- processes and changes in stressor levels; toring (Stewart-Oaten 1996). For purposes of a Lake • They should represent the larger resource Tahoe basin adaptive management strategy, imple- of which they are a structural or composi- mentation and status and trend monitoring are best tional component; and served by parameter estimation, whereas effective- • They should be measurable. ness monitoring is best approached through hy- Before field or simulation testing, the list of pothesis testing. candidate indicators can be narrowed to identify final Determining effect size and statistical indicators by focusing on those with the following power is an important element of monitoring design. properties: Effect size (the magnitude of change to be detected), • They should exhibit dynamics that parallel the precision of estimation (Type I error rate or al- those of the larger environmental compo- pha), and sample size are intradependent. A nent or system of ultimate interest; monitoring program should be able to detect the • They should show a short-term but persis- magnitude of change in the value of an indicator or, tent response to change in the status of the in statistical terms, the effect size. Acceptable levels environment; for a type I error (concluding a change or difference • They should be accurately and precisely es- when none exists), type II error (concluding no timated; change or difference when in fact one exists), natural variability of the indicator, and the sensitivity of the • The likelihood of detecting a change in their test determine the effectiveness of a sampling effort magnitude should be high, given changes in for a given effect size. the status of the system being monitored; • Statistical power is a function of the prob- Each should demonstrate low natural vari- ability that a difference of a given size will be de- ability or additive variation, and changes in tected (i.e., power = 1 - type II error rate). Managers their values should be readily distinguish- must implement monitoring programs with suffi- able from background variation (i.e., a high cient statistical power to detect meaningful changes signal-to-noise ratio); and in the values of the indicators. For monitoring de-

Lake Tahoe Watershed Assessment 699 Chapter 7

signs and analyses to be meaningful, a desired statis- change or a particular value of an indicator (condi- tical power should be selected before it is imple- tion or stressor) that may herald declines in the lar- mented to determine sample size needs. Also it ger ecosystem. At Lake Tahoe, the term threshold should be calculated after monitoring data is gath- also applies to management thresholds, broad indica- ered (post hoc) in order to interpret the true power tors of ecosystem health that were established to of tests that failed to detect a change or to reject the alert policy-makers to conditions incompatible with null hypothesis differences (see Skalski 1995 and desired conditions for the basin. To avoid confusion, Zielinski and Stauffer 1996). we use the term “trigger point” to refer to ecosystem In practice, addressing questions of statisti- thresholds associated with individual indicators. cal power requires determining the minimal magni- Trigger points serve as red flags intended to tude of change in the indicator variable that is envi- raise awareness and prompt response by managers or ronmentally significant (this value must be estimated policy-makers. Trigger points can be designed to by a scientifically defensible process). Initial esti- provide an early warning of undesirable changes, or mates of an appropriate effect size applied to the they can be poised at an estimated juncture of irre- indicator can be based on spatial or temporal varia- versible environmental degradation. Responses to tions under baseline or reference conditions (Skalski such trigger points might range from simply revisit- 1995). Given this information, practical sampling ing conceptual models and management effective- issues, such as numbers of samples and resampling ness to actions that include a moratorium on devel- intervals, can be addressed. A comprehensive discus- opment (such as what occurred during the 1980s in sion of statistical power and its relevance to deci- the Tahoe basin), direct ecosystem intervention and sion-making in the context of responsible manage- restoration activities, and changes in land manage- ment of natural resources is found in Peterman ment policies and practices. The identification of a (1990). Emphasis should be placed on minimizing trigger point depends on the intended response; the of type II errors as opposed to type I er- therefore, it is important to clarify the function of a rors, particularly when declines in resource condi- given trigger point, as well as the appropriate re- tions are irreversible (see Shrader-Frechette and sponse, during the design of the monitoring scheme. McCoy 1993). Importantly, effect sizes and trigger points should be considered in concert with one another. If trigger Interpreting the Ecological and Management points are designed to serve an early warning func- Significance of Indicator Values tion, then effect size can be calibrated to detect Ecosystems are complex systems subject to changes similar in size to differences between cur- stochastic variation and unpredictable behaviors. It rent conditions and the trigger point. However, trig- should not be surprising that the task of monitoring ger points that identify irrevocable degradation whole ecosystems and drawing reliable inferences to should be accompanied by effect sizes small enough system integrity has historically proven to be such a to detect a trend toward that point. daunting task. Interpretation of the significance of Defining the trigger point of an environ- changes in the value of an indicator is complicated mental indicator that can determine a management by nonlinear cause-and-effect relationships between response is difficult and complex. But existing man- indicators and stressors. However, indicator values agement direction in the Lake Tahoe basin actually must be interpreted in reference to “trigger points” provides a solid basis by which monitoring thresh- or “thresholds” in order for monitoring results to be olds may be determined. As a general rule, as the risk applied to management decisions and actions. of environmental loss increases, trigger points The term “thresholds” is used in this dis- should be made more sensitive. Threshold values for cussion in a different manner than the legally binding environmental indicators may be established by ref- “environmental carrying capacity thresholds” that erence to documented historical values or prelimi- underpin the regional plan in the Lake Tahoe basin. nary baseline monitoring of a nonaffected or “pris- In ecosystem theory, a threshold is a magnitude of tine” system may be conducted. In the absence of

700 Lake Tahoe Watershed Assessment Chapter 7

reference systems or adequate historical data, it is phase in the adaptive management cycle, the evalua- difficult to establish expected values or require a tion and decision-making phase (Figure 7-1). An given future trajectory of indicator variables. None- effective information transfer strategy involves and theless, trigger points must be identified as thought- informs decision-makers early and often in the proc- fully as possible. Incomplete historical data com- esses of data collection and analysis. This is so that bined with some notion of a “desired future condi- the data and results are familiar and decisions can be tion” must serve as the basis for identifying trigger made. It provides frequent and readily understand- points (see discussion in Bisson et al. 1997). Models able reports of progress in monitoring so that the can help in this pursuit by providing predictive capa- information is available to interested parties. And bilities and appropriate trigger point values. information transfer strategy packages assessment All evaluations of monitoring data, includ- and evaluation reports in a manner that is accessible ing trigger points, require that the appropriate data to nonscientists and that directly addresses the ques- analysis also be determined before monitoring data tions facing management. are collected. For each indicator, will one summary value be estimated for the entire basin, or will values The Role of Modeling in Adaptive Management be estimated for multiple analysis units? Is the spatial As chapter five suggests, it may take up to distribution of conditions of interest? Each of these thirty years to see changes in clarity that result from options likely calls for different sampling considera- immediate reductions of nutrients going into Lake tions, analysis approaches, trigger points, and man- Tahoe. Some scientists have concluded that if the agement responses. Sampling and analysis ap- buildup of nutrients in the lake is not reversed within proaches need to be fully developed before data col- the next ten years, the costs of solving the problem lection begins, if data collection is to result in useful will be so great and the impacts so extreme that they information. will exceed the currently available capacity for reso- Monitoring has limitations. Care must be lution. This situation creates a unique dilemma. How taken to understand what can and cannot be inferred do we ensure that the actions taken during the next legitimately from monitoring data, particularly when ten years will be effective when it may require three the ability of a monitoring program to assess attain- decades to obtain measurable results? The only prac- ment of management objectives is judged. Monitor- tical solution is to understand the ecological system ing programs can neither unambiguously ascertain sufficiently to model accurately the effects of various the cause of a change nor decide on how much management treatments and thereby to predict fu- change is acceptable. Moreover, monitoring pro- ture consequences of today’s actions. This challeng- grams themselves cannot decide on the threshold ing situation illustrates why computer modeling is an values of indicators that will trigger specific man- essential component of an adaptive management agement actions. Monitoring simply provides data as strategy for the Lake Tahoe basin. Modeling pro- designed, much like any research effort. It is the re- vides an opportunity to look ahead to the likely out- sponsibility of scientists and managers to ensure that comes of management approaches and facilitates the monitoring design is scientifically sound and that adjustments to management before thresholds are it meets the information needs of managers and pol- crossed. icy-makers. If monitoring results indicate that condi- Models are increasingly being employed in tions lie outside an acceptable range, specific changes watershed management planning efforts. The South in land management practices or resource policy Florida-Everglades Restudy Project uses simulation should be triggered. Facilitating the transfer of moni- models to illustrate ecosystem responses to toring results to the decision-making phase is a criti- management strategies to direct research efforts in cal consideration in the design and implementation hydrology and . A model developed in a con- of a monitoring program. The reception and applica- sensus process involving agencies and stakeholders tion of monitoring results belongs in the subsequent simulates hydrologic and economic processes in the

Lake Tahoe Watershed Assessment 701 Chapter 7

Devil’s Lake, North Dakota, watershed allows for or their outcomes. Quantitative models display nu- adaptive control and protection. And decision merically defined relationships among ecosystem support systems, simulating key biotic processes, use components, typically consisting of mathematical economic and agricultural inputs to guide mainte- relationships or logical arguments. Statistical models nance of the Upper Snake River Basin in Wyoming are a subset of quantitative models in which quanti- and Idaho. tative relationships are established and derived through statistical analyses. Predictive models are a Types of Models and Their Applications subset of statistical models in which statistically de- Models can serve a variety of purposes. In a rived relationships are used to predict the value of a management context, models can clarify system resource of interest through time or space. Finally, functions, conditions, and trajectories and can facili- decision support models use a variety of models to tate translation of scientific data into information for establish a scientific foundation upon which multiple decision-making. Ecological models can also serve as interests and stakeholders can weigh the conse- links between the natural community and quences of various management options. the public sector, using mathematical and statistical The ideal modeling tools for the Lake Ta- relationships to translate the language of the physi- hoe basin will efficiently and accurately describe and cal, biological, and social sciences into the language predict the effects of background environmental of social preferences (see Figure 7-2). Effective factors or management activities on air, water, and modeling depends on the availability of scientific biotic resources, and socioeconomic conditions. information generated through research and moni- Such modeling tools can be used by agencies and toring. However, models can also be used to cope citizens groups involved in decision-making in the with information gaps in decision-making processes. basin to evaluate management options and to de- A variety of models exist, five of which can velop social consensus on expected conditions of be outlined in terms of their general function and each resource. It is unrealistic to expect immediate potential contribution to management and planning development of such modeling tools, primarily be- in the Lake Tahoe basin: conceptual, quantitative, cause they take many years to develop; however, a statistical, predictive, and decision support models. strategy for model development should be part of an Conceptual models illustrate the components of a integrated adaptive management plan for the Lake system and their links but do not contain quantita- Tahoe basin. tive information regarding component interactions

Modeling Public Data Tool

Social Physical Mathematics Sciences Sciences

Figure 7-2—The role of modeling as a tool to aid decision-making.

702 Lake Tahoe Watershed Assessment Chapter 7

Taking a Systems Approach ing modeled output to actual measurements (empiri- Models are simply manifestations of a view cal data), with the difference between the measured of certain discreet aspects of the world around us. A and modeled output serving as an index of accuracy. “systems approach” to modeling is particularly ap- An alternative measure of model accuracy uses ex- plicable to adaptive management for three reasons. pert opinion to evaluate the accuracy of model pre- First, a systems approach explicitly recognizes the dictions. To accomplish this, model simulations pro- dynamic nature of ecosystems and management di- duce predicted outcomes based on a management rection. Management does not proceed in a linear scenario. Then, topic experts with knowledge of the fashion but rather adjusts in response to predictable, focal system assess the accuracy of predicted out- as well as unexpected, events as do natural systems. comes. If no significant discrepancies are found be- A systems approach builds into models the ability to tween model simulations and expert opinion, the accommodate unexpected changes and to amend model can be considered accurate from a manage- models as scientific knowledge advances and the ment perspective. If some significant discrepancies needs of decision-makers change. Second, a systems are found, further modifications to the model must approach works toward linking the information ac- be undertaken to resolve differences. The use of quisition and assessment phase (research, monitor- empirical data to assess model accuracy is preferable ing, and modeling) to the other phases of the adap- if data are available; however, the use of expert opin- tive management cycle. Foremost in a systems ap- ion is a legitimate and valuable approach to calibrat- proach to modeling is development of a common set ing models in the absence of data. of environmental and socioeconomic variables that The second model criterion, efficiency, re- can serve as a common language shared by all con- fers to ready execution and updating of a model. tributors to the adaptive management cycle. In link- Ideally, a fully developed model has no errors, never ing the elements of the adaptive management cycle fails to execute, and is easily updated. Unfortunately, using a common planning language, the flow of re- the more complex a model, the less efficient it tends sources and information through adaptive manage- to be. Complex models have a greater likelihood of ment can be greatly enhanced. Third, a systems ap- embedded errors because they rarely go through proach can readily accommodate both qualitative formalized and rigorous error checking or debugging and quantitative models ranging from models of procedures. Errors in a model can render its results environmental interactions to models that can assist invalid, halt its execution, or make it difficult to be allocation of financial resources among research, transported from one computer platform to another. monitoring, and modeling efforts. An ability to read- In addition, ideally a model can be updated as im- ily embed detailed quantitative models into broader proved estimates of model parameters are developed conceptual models, such as the conceptual model or a better understanding of system dynamics be- supporting the selection of indicators, or the even comes available. As models become more complex, broader model of the adaptive management cycle, is by necessity, modifications and updating become a highly valuable trait of the systems approach, al- more expensive and tenuous. Models that have few lowing for a strong scientific foundation and useful problems related to these issues are referred to as informative tools for decision-making. robust, with the most robust models considered the easiest to maintain and execute by technical staff. Criteria for Evaluating Model Utility The third model criterion, utility, is an as- Four criteria are useful in helping to guide sessment of the accessibility and user-friendly char- the development and evaluation of models in sup- acter of a model. Specifically, the utility of a model port of management: accuracy, efficiency, utility, and typically relates to how well graphical user interfaces acceptance. The first model criterion, accuracy, de- are designed to allow the user to input proposed scribes the ability of a model to predict the behavior management activities and to view predicted changes of a physical system, given a set of impulses to the to the system. For models that are intended as deci- system. Model accuracy can be assessed by compar- sion-making tools, interfaces need to be developed

Lake Tahoe Watershed Assessment 703 Chapter 7

to produce a balance between ease of use and flexi- within the Lake Tahoe basin must be understood bility of simulation. Providing too detailed a list of all before satisfactory management goals and activities possible model outputs can result in information can be developed. In addition, where strong interac- overload, making the results of the model difficult to tive variables exist between the resources, significant interpret by the user. Alternatively, developing an feedback loops can develop where the conditions of interface that limits the range of management scenar- two or more resources sequentially interact to create ios that can be modeled will prevent the evaluation a downward spiral in resource conditions. Feedback of all possible management alternatives. To attain a loops have been identified in the system dynamics high level of model utility, an equitable compromise literature as being important elements to consider in between flexibility of simulation and ease of use for developing effective management plans for any sys- the model interface must be developed. tem, natural or otherwise. If the feedback loops of a The fourth model criterion, acceptance, re- system are not thoroughly understood, an action flects how well users themselves believe that a model taken to improve the condition of one resource accurately predicts changes in the system submitted within a watershed ultimately could degrade the to modeling. In many respects, this is the most im- condition of that very same resource or another. portant consideration in the development of models Models can provide substantial assistance in identify- to be used as decision-making tools for land man- ing, displaying, and describing such complex re- agement. A model that is not widely accepted can be source interactions. used by an agency to develop management plans, but An example of the challenges that resource the risk is high that a debate among constituencies interactions can present to managers is provided in will focus on the validity of modeling results rather Figure 7-3, which outlines how management activi- than on the effect of the management activities on ties can affect directly and indirectly the condition of the condition of the watershed. Acceptance will de- multiple resources. A prescribed fire regime calling pend on the reliability of the information used to set for more frequent fires may be intended to reduce parameters for the model, on the level of contro- fire risk to life and property and to improve the versy associated with underlying assumptions used health of the forest ecosystem. However, a new pre- by the model, and on the socioeconomic relevance scribed fire regime may also have a number of unin- of the output. tended effects, such as smoke discharged into the Many of these criteria are conflicting and atmosphere, increased nutrients released into the can be difficult to reconcile. For example, the accu- as ash, and increased water yield resulting from a racy of a model may require greater complexity. reduction in cover. Additional unintended ef- Every modeling exercise necessarily involves some fects are likely to stem from direct effects and could level of compromise among each of the four issues, include decreased visibility from airborne particu- the appropriate balance will depend on the intended lates, nitrogen deposition into Lake Tahoe from application of the model, and compromises in model smoke particles, increased sedimentation from lack performance should be determined based on com- of plant cover, and potential declines in habitat con- munication and discussion between agencies and ditions for species of particular concern. Qualitative stakeholders, who ultimately will use the models, and and quantitative models can both be helpful in better the technical personnel involved in developing the understanding, anticipating, and avoiding undesirable models. unintended management effects in the course of pursuing management goals. Integration through Modeling Actions taken to improve the condition of Decision Support Tools one resource can significantly affect the condition of Decision support tools are models designed other watershed resources. Some understanding of to aid decision-making by clearly displaying what is the tradeoffs between the condition of each resource known, what is uncertain, and what is predicted,

704 Lake Tahoe Watershed Assessment Chapter 7

Management Activity: • Change in fire frequency through prescribed fire

Intended Effects: Unintended Direct Effects: • Reduce fire risk • Generate smoke emissions • Improve forest health • Increase nutrient discharge • Increase water yields

Unintended Indirect Effects: • Decrease visibility • Increase atmospheric N deposition • Increase sediment yields • Decrease habitat for focal species

Figure 7-3—Example of intended and unintended effects that can result from a management action.

given various courses of action. A decision support in a manner that can be understood and endorsed tool for the Lake Tahoe basin should accurately pre- and that can make them accessible to all of the con- dict the effect of any environmental factor or man- stituencies who are affected by or who influence agement activity on the condition of air, water, bi- decisions regarding the resource. otic, and socioeconomic resources within the basin. Ideally, a shared-vision modeling tool facili- Decision support tools can be used by agencies and tates an integrated approach to information acquisi- citizens groups involved in evaluating management tion and allows decision-makers to predict and weigh options. the impacts of potential projects on management As discussed above, every model represents goals. A modeling tool that is intended to serve such some balance of four basic attributes: accuracy, effi- a decision support role will have three primary ele- ciency, utility, and acceptance. All of these attributes ments: an input interface that allows decision-makers are critical to the success of decision support mod- to input information related to regulatory and man- els, and model performance must be balanced agement activities, one or more technical modeling among these attributes to achieve the best model for components, often referred to as “black boxes,” and the application. A variety of previous studies have an output interface that allows decision-makers to pointed out the need to compromise on these mod- view the impacts of their proposed actions on re- eling issues (Keyes and Palmer 1995; Tracy 1995), source conditions. The combination of these three especially when the models are to be used in a po- elements provides a powerful integrated manage- tentially contentious political environment. Keyes ment tool that can facilitate decision-making regard- and Palmer used a collaborative process to develop ing difficult resource trade-offs. The elements that “shared vision” models for conflict mediation in support such a tool warrant description. planning and management activities. Shared vision The information required to develop an ef- models attempt to represent resources and their links fective input interface element for an integrated de-

Lake Tahoe Watershed Assessment 705 Chapter 7

cision support model starts with a list of potential socioeconomic). The variables should be quantitative management and regulatory actions. Information on descriptions, which, if predicted accurately, will sig- the spatial location, cost, time-scale of implementa- nificantly help decision-makers understand whether tion, and expected physical impact are recorded for regulatory or management actions improve or de- each action. A graphical user interface is needed to grade the condition of each resource. Without meth- manipulate potential management and regulatory ods to quantify the condition of each resource, no actions. A quantitative interpretation of how each mechanism exists to convey how the condition of management or regulatory action alters management each resource is affected by regulatory and manage- variables in the resource models must be developed. ment actions. Variables that effectively describe the Management agencies will have to collaborate with condition of each resource can be developed the scientific community to arrive at a consistent through a stakeholder process, where the variables method to describe how the management variables are identified based on consensus among interested will change due to a proposed activity. These meth- parties. An example of an integrated research, moni- ods obviously do not provide a perfect interpretation toring, and modeling effort at the watershed scale is of how management activities alter parameters that provided in Table 7-1. It illustrates a design for col- govern the behavior of resources, but they can facili- lecting input data that facilitates useful output data in tate discussion and decision-making under uncer- the form of appropriate descriptive variables. Ideally, tainty. these variables serve the needs to each phase of the Only a small fraction of the processes that adaptive management cycle, thus creating the com- govern the condition of Lake Tahoe’s resources cur- mon planning language discussed above in the con- rently can be modeled quantitatively. The quantita- text of a systems approach. tive models currently under development are state- of-the-science, hence their development was time Information Acquisition and Assessment in the consuming and expensive. It is not realistic to expect Lake Tahoe Basin to be able to engineer quantitative models of similar The preceding chapters provide in-depth in- rigor for every important process and resource inter- formation on the status of our knowledge in the four action. A more tenable goal is to develop a modeling key issues areas addressed in the assessment. That capability for each process that can be realized with information provides a base to answer three primary existing information and that can be improved as questions about the current state of knowledge and better information and process models become how best to improve on that knowledge: available. Existing system dynamics modeling tools, • How well can we describe the processes such as STELLA and VENSIM, are available, but that govern the behavior of air, water and they do not have the capability of incorporating so- biotic resources and socioeconomic condi- phisticated process-based models in their simula- tions in the Tahoe basin? tions. Rather, a modeling platform that can integrate • How well do we understand the relation- sophisticated process-based models with these more ships among air, water, biotic, and socio- simplistic “stock and flow” modeling concepts economic elements and the environmental, should be developed. Such a modeling platform management, and interactive variables that would allow for the rapid development of quantita- link these resources in the Lake Tahoe wa- tive models for all of the watershed resource proc- tershed? esses not currently being modeled within the Lake • Tahoe basin. How can current scientific information be used to evaluate potential future manage- ment investments and aid managers and de- Output Interface Element cision-makers? The first step in developing an output inter- The current understanding of relationships face element for an integrated modeling tool is to among resources is examined here to better define develop a set of variables that best describe the con- data gaps regarding resource interactions (Table 7-2). dition of each of the resources (air, water, biotic, and

706 Lake Tahoe Watershed Assessment Chapter 7

Table 7-1—Examples of information integration activities suggested by the watershed assessment that would im- mediately contribute to building an integrated information strategy.

Resource Area Information Integration Activity All • Develop and implement a multiresource, basin-wide adaptive management strat- egy. • Develop, test, and refine a decision support model. Air Resource • Expand and improve air quality monitoring efforts. Biotic Resource • Develop and implement a prescribed fire implementation, research, and monitor- ing plan that integrates concerns and objectives across resource areas. • Develop and implement an old strategy that includes monitor- ing and research elements. • Develop and implement a conservation strategy for species and com- munities, including terrestrial and aquatic elements. Water Resource • Identify and quantify sources of biologically available nitrogen and phosphorus and adopt effective control strategies. • Integrate the results of the lake clarity model into a broader decision support model. Socioeconomic Resource • Embark on an effort to identify what constitutes a healthy and robust community for full-time residents of the basin. • Identify key links between socioeconomic well-being and environmental health.

Our analysis notes whether a given relationship can known—neither a positive nor a negative impact can be described quantitatively or qualitatively, or if it be predicted. For example, little information exists cannot be described at the current time. Quantitative on how a vegetated area disturbed by fire affects the descriptions consist of equations that can be devel- flux of phosphorous to upland streams and eventu- oped for use in predictive or descriptive models. For ally to Lake Tahoe. It may increase or decrease the example, there is a relatively large body of knowl- flux of phosphorous to streams or the lake, but at edge on the relationship between traffic density and the current time no scientific consensus exists on emissions from vehicle tailpipes; these relationships how the flux will change. This type of uncertainty can be quantified and used to inform policy. Qualita- precludes any predictive capability. tive descriptions consist of statements regarding the The vast majority of resource interactions positive or negative effect of changes in the state of can be understood only in a qualitative sense, with one resource on the state of a linked resource. For only a small set of links being understood quantita- example, it is understood that exposing bare soil by tively (Table 7-2). Some of the qualitatively described removing vegetation leads to a greater potential for resource interactions are so uncertain that the direc- hillslope erosion from precipitation. However, no tion, positive or negative, of the influence of a factor quantitative descriptions have been developed that on a resource condition remains unknown. Oppor- relate the increase in sediment load to streams as a tunities to improve certainty about these interac- function of the fraction of vegetative cover near a tions, including the ability to quantitatively describe given stream. Relationships among resources that and model relationships, vary by resource. In some cannot be described at this time consist of interac- instances, aspects of the relationship could be quan- tions that are known but whose outcome is un- tified if funds were made available to do so, whereas

Lake Tahoe Watershed Assessment 707 Chapter 7

Table 7-2—Current state of uncertainty regarding key links among resources and the factors that shape them.

Quantitative Factors Links Qualitative Links Uncertain Links Interactive Environ. Factors: Exhaust emissions Air, socioeco- nomic. Nutrient flux Air, biotic, water Air, biotic, water Sediment flux Air, biotic, water Smoke emissions Air, socioeconomic Visibility Air, socioeconomic Development/urban landscape Biotic, water, socioeconomic and quantity Biotic, water, socioeconomic Stream morphology Biotic, water Vegetative landscape Biotic, water, socioeconomic

Management Activities: Tailpipe regulations Air Socioeconomic Fire management Air Socioeconomic Vegetation management Biotic, water, socioeconomic Range management Biotic, water, socioeconomic management Biotic, socioeconomic populations Biotic, socioeconomic Woodstove regulations Air Socioeconomic regulations Biotic, water, socioeconomic Road management Biotic, water, socioeconomic

Independent Environ. Factors: Wind Air Biotic, water Temperature Air, socioeco- Biotic, water nomic Humidity Air, water Precipitation Biotic, water Socioeconomic Evapotranspiration Biotic, water Solar radiation Biotic, water

708 Lake Tahoe Watershed Assessment Chapter 7

in other instances other factors are more limiting smoke, small, medium, and large prescribed fires, than funding, such as the difficulty in isolating a par- and wildfires. Forest fires are known to contribute ticular cause-and-effect relationship. large amounts of fine particulate matter to the at- Where additional investments can lead to an mosphere, but the coupling of this source of particu- improved understanding, the appropriate level of lates to deposition and lake nutrification is necessary investment will vary depending on the uncertainty of to make valid predictions of the ecosystem impact of the link. An improvement in the understanding of fires used for forest management. highly uncertain links typically is achieved by invest- 3. Research into transportation-related air ing in scientific experiments to understand the basic quality concerns is needed to better understand the behavior of the interaction more fully. An improve- role of mobile combustion and re-entrained roadway ment in understanding of qualitatively described dust at Lake Tahoe. Studies would best involve par- links is achieved typically by investing in monitoring ticulate and gaseous pollutant collection and analysis, activities that are designed to identify correlative detailed vehicle counts, and distribution of vehicle relationships and to determine trend behavior. An types by roadway, velocity, and acceleration. While it improvement in existing quantitative models of links is unclear what the transportation contribution to is achieved by investing in model development and ecosystem deterioration is, diesel and gasoline com- monitoring activities that are designed to further bustion is known to contribute substantial fine parti- calibrate and verify quantitative relationships. cles to the atmosphere. In addition, potentially large amounts of phosphorous and nitrogenous particu- Research Needs late matter may be re-entrained by automobiles me- The Lake Tahoe Watershed Assessment chanically grinding roadside dust. Evaluating these identified numerous issues of management concern transportation-related pollutant sources could help that require further investigation through research reduce one of the largest uncertainties regarding pol- and monitoring. Those efforts should include, but lutant sources in the Tahoe basin. should not be restricted to, the research targets out- 4. Mitigating air quality impacts depends on lined below. an accurate accounting of sources, both natural and anthropogenic (human-induced), and local and Air Quality transported. Innovative sampling and analysis tech- niques will be necessary to provide information on 1. Increased air quality monitoring at several gaseous composition, particle size and composition, locations throughout the basin is a major priority. and other signatures that link sources of pollutants Desired information includes gaseous and particulate to the basin or downwind receptor sites. Such stud- pollutant data that are important for lake clarity, for- ies will require detailed information about gases de- est health, atmospheric visibility, and human health. rived from the Central Valley, especially those arising Measurements should include large particle phos- from increased traffic on Interstate Highway 80 and phorous, nitrogenous species, nitrogenous gases, State Highway 50. It is equally important to study ozone, sulfate, and fine dust by particle size. At a aerosols on the western slope of the Sierra Nevada, minimum, collocating atmospheric monitoring sta- especially those associated with potential increases in tions with TRG deposition bucket samplers will con- the use of prescribed fire. tribute to reducing uncertainties that will otherwise limit the value of models. Ideally, these sampling sites would be placed in ma- Water Quality and Lake Clarity jor watersheds to be representative of the entire ba- 1. There is a need to determine the portion sin-wide . of total phosphorus loading that is biologically avail- 2. Evaluating the contribution of prescribed able to algae in Lake Tahoe over ecologically rele- fires and wildfires to diminishing lake clarity will be vant time scales. This information, in conjunction necessary to reduce the uncertainty regarding this with increased knowledge of nutrient sources, is important particulate matter source. For maximum likely to provide the most immediate improvement utility this research should address both the amount in restoration efforts. and composition of particulate matter in domestic

Lake Tahoe Watershed Assessment 709 Chapter 7

2. The preliminary nutrient input budget 9. There is a significant need for compre- presented for Lake Tahoe should be further resolved hensively evaluating BMP effectiveness and project to identify and quantify sources of nutrients and longevity with a standardized monitoring framework. sediment within each of the significant categories— 10. A formalized ranking system should be direct runoff, atmospheric deposition, and tributary developed to prioritize restoration projects for use discharge. This effort should be supported by basin- by all agencies implementing such projects. wide monitoring of direct runoff from both urban 11. A comprehensive geographic informa- and nonurban regions, urban runoff to tributaries, tion system, which includes updated land use, demo- streambank erosion, atmospheric deposition, and graphic, geomorphic, and biological layers, is needed runoff from highways and roads. to support water quality-related management efforts. 3. Research and monitoring should identify Such a tool will speed research development and and quantify the contribution of fine sediments and and will provide a common foundation colloidal nutrients from their various sources. Fo- for restoration efforts. cused research is needed to better understand the 12. Limnological studies of Lake Tahoe, in- transport of those materials through and cluding analysis of long-term data, must be contin- ground water and to determine the effects of both ued. In addition, nutrient and sediment loss proc- natural infiltration and BMPs to limit concentrations esses in Lake Tahoe need to be further quantified, of those materials. along with phosphorus cycling in the lake. 4. There is a need for more comprehensive 13. Increased research and monitoring is information on nutrient cycling in watershed vegeta- required on all aspects of prescribed burning as it tion and soils in the Tahoe basin. Such information facilitates nutrient deposition to the lake. The me- will allow a greater understanding of how manage- chanics of smoke deposition and nutrient cycling ment and development affect water quality. needs the most immediate attention. 5. Available LTIMP stream discharge and 14. Research on wetland and riparian zone loading data should be subjected to analysis to de- ecology and nutrient cycling is needed. Wetlands are termine long-term trends, differences among water- believed to provide a critical facility for removing sheds, and relationships among flow, sediment load- nutrients from stream waters. Such information ing, and nutrient input. could provide key information to guide future wet- 6. The tributary sampling design for LTIMP land restoration efforts. should be reevaluated to accommodate current regu- 15. Further development and refining of latory, monitoring, and research needs. water quality and lake clarity models should be en- 7. Relationships between land use and couraged. When complete, these models are ex- sediment-borne phosphorus loading should be elu- pected to serve as guidance tools to assess lake re- cidated at diverse and relevant spatial scales, includ- sponses to restoration activities and to develop ing loading from single-family residences, residential quantitative targets for nutrient and sediment load- neighborhoods and commercially zoned parcels, ing. subwatersheds, and entire watersheds. Such a nested 16. Ecosystem modeling efforts must be approach is needed to expand our understanding coordinated and modeling components should be from site-specific scales to the larger subwatershed designed to be interactive. and watershed scales. 17. Research facilities in the Tahoe basin 8. Available data from existing BMP and need to be upgraded and expanded to meet current environmental restoration projects should be for- and projected needs. mally evaluated. A centralized clearinghouse for 18. A science advisory panel is recom- these data is needed. A listing and descriptions of mended to assist in guiding and coordinating adap- historical, current, and proposed projects should be tive management development, incorporating moni- posted on an Internet website. toring results into management actions, and setting research priorities.

710 Lake Tahoe Watershed Assessment Chapter 7

19. Increased attention should be given to out the Tahoe basin that can serve as nuclei around developing a volunteer monitoring program to assist which vegetation can be managed toward old-growth ongoing and future restoration and monitoring ac- status. Logical places to begin this process are the 38 tivities. old-growth polygons of lower and upper montane 20. There is a need for an “experimental vegetation identified in this report. Those polygons watershed” to better understand the factors that af- should be revisited and the neighborhood vegetation fect terrestrial and aquatic ecosystems. Intensified should be quantified, referencing size and condition data gathering in such a watershed could be used to of the old-growth nucleus, topography of the better understand the complexity of nutrient cycling neighborhood (in terms of ease of movement for and to test directly the results of various manage- managers), substrate stability in the face of foot traf- ment strategies. fic by managers, distance from nearest roads or 21. One of the greatest hurdles to effective structures (which may constrain the use of fire as a integrated ecosystem management in the basin is the management tool), ecological distance of neighbor- lack of interactive data sharing among institutions. hood vegetation from old-growth status, and ease of The best and most current information often proves access for tourists and residents if the neighborhood to be the most elusive. Effective data sharing would is to be used as a management demonstration site. greatly improve the ability to learn from research and 4. Research should contribute to determin- to enhance application of results to restoration. ing a minimum sustainable size for old-growth forest patches. This objective ultimately will be determined Biological Integrity in reference to two factors: the minimum home 1. A quantitative scale of old-growth attrib- ranges of old-growth-dependent animal species and utes by which any forest stand could be numerically the minimum patch size that contains enough het- expressed in terms of its ecological “distance” from erogeneity for regenerating sites for characteristic old-growth status needs to be developed. A tree, shrub, and herb species and enough area to half-dozen or more attributes include the density of permit random disturbances in future years (fire, trees per hectare with dbh greater than some mini- landslide, windthrow, epidemic disease) to occur mum, the amount of canopy cover by trees of such without entire site replacement. dbh, the leaf area index of the overstory, and the 5. More information is required to under- ratio of overstory tree density to understory tree stand differences in effects on ecosystems of fire density. The old-growth status of stands should be a behavior among various mechanical treatments sum of all attribute variables. Future research should (thinning versus biomass removal), burning treat- lead to some minimum scale value, below which a ments (pile burning versus area burning), and no forest stand would be labeled as non-old-growth. treatment. Important effects to analyze include 2. A limited set of monitoring protocols measures of vegetation structure and composition should be developed to signal significant change in that can be used to assess wildlife habitat, old for- the old-growth status of any given forest stand. ests, forest health, and fire hazard. Monitoring indicators that might serve as surrogates 6. Predictive modeling of fire effects on for the attributes above include leaf area index above forests and other plant communities requires models a fixed reference point, amount of litter shed during of mortality. Existing mortality models are based on one portion of a year, soil N status at a certain depth, generalized data from relatively few studies in the rate of litter decay or of N mineralization in the soil, western portion of the continent. Mortality varies annual growth rate at breast height of particular un- greatly with local conditions; therefore, customizing derstory or overstory trees, or the abundance of par- mortality models for the Lake Tahoe basin will ne- ticular shrub or herb species. cessitate a combination of fire effects monitoring 3. Locations should be identified through and research.

Lake Tahoe Watershed Assessment 711 Chapter 7

7. Fire effects on vegetation (especially old- distribution of restoration activities. A coordinated, growth, wildlife habitat components, and fuels), nu- spatially based monitoring system should be put in trient cycling, and air quality are poorly understood place. in the Sierra Nevada in general and in the Tahoe 12. The effects and effectiveness of fire basin in particular. Deriving better information will hazard reduction and vegetation restoration projects necessitate a well-designed monitoring scheme; are largely unknown. Measuring vegetation structure however, some issues will require a more concen- and composition before and after using one of these trated research effort. projects would address this lack of information. 8. Little fire history work has been con- 13. Several of Lake Tahoe’s most immediate ducted in the Lake Tahoe basin or elsewhere in the research needs concern the basin’s aquatic ecosys- Sierra Nevada on vegetation types common in the tems, including an assessment of the effects of live- basin. Restoring fire as a disturbance feature in eco- stock grazing on the integrity of meadows, riparian system management is often based on assumptions zones, and stream communities, the impacts of non- of historic fire patterns, including mean and varia- native trout species on reintroduced mountain yel- tions in fire return intervals. Fire history work is par- low-legged frogs, and the efficacy of available con- ticularly needed in montane pine, mixed-conifer, and trol measures for Eurasian watermilfoil. More gener- white and red fir forests in the basin. Because cli- ally, there is a need to characterize the influences of matic patterns vary dramatically around the basin, various types of disturbances on the biotic integrity research on fire history throughout the basin is of lentic and lotic ecosystem types and to evaluate needed to evaluate spatial variation in historic fire the effectiveness of various management strategies. regimes. Additionally, modeling “potential habitat” for rare 9. Riparian communities are a key concern aquatic community types, such as bogs, fens, and in the basin because of their important role in filter- springs, is advisable, as is modeling habitat for all ing nutrients from upland surface waters on their amphibian species in the Tahoe basin. Indicator spe- migration to the lake. Current restoration efforts in cies for use in monitoring aquatic ecosystems types riparian areas are based on little underlying informa- should be identified and tested. tion on processes that shape these vegetation com- 14. Studies should be undertaken to assess munities. Fire is seen as one potential tool for restor- the influences of different anthropogenic distur- ing riparian vegetation, but the historic role of fire bances on the biotic integrity of ecologically signifi- on riparian vegetation patterns and functions is cant areas. In support of that effort, those areas poorly understood. Research on the effects of pre- should be validated through modeling efforts using scribed burning on riparian vegetation, soils, and remotely sensed vegetation data, and an indicators associated functions, such as nutrient cycling, is monitoring scheme should be developed. A refer- needed to assess the effects of fire as a restoration ence condition measure for each ecologically sensi- tool in riparian areas. tive area type can serve as a standard for directing 10. Fire behavior predictions are based on current and future management planning efforts. models that incorporate fire behavior and weather. Focal activities should include identifying minimum There currently is only one weather station that can patch sizes for ecologically significant stands of as- support fire behavior modeling in the basin. At least pen, modeling and validating in the field the loca- one additional station on the north shore of the ba- tions of cushion plant communities, and evaluating sin would greatly improve information on weather deep-water plant beds as spawning grounds for and would provide inputs to predictive models of nonnative fishes. potential fire behavior. 15. Impacts of anthropogenic actions, such 11. Documenting landscape-scale effects of as direct harassment and use practices, various treatments is necessary to on such focal species as bald eagle, northern gos- better understand changes in forest health, wildlife hawk, spotted owl, willow flycatcher, pine marten, habitat, old forests, and fire hazard. Currently, no and amphibians) should be better understood. Like- comprehensive system exists for tracking the spatial wise a better understanding of the effects of exotic

712 Lake Tahoe Watershed Assessment Chapter 7

fish and bullfrogs on native fish, amphibians, and tality, amenity, and recreation industries; aquatic invertebrates will be necessary for successful and reintroduction efforts targeting extirpated species • Developing social, cultural, and community and those at risk of extirpation in the basin, such as indicators by providing expertise in meth- mountain yellow-legged frog and Lahontan cutthroat odologies and approaches. trout. 3. Key links between socioeconomic well- being and environmental health are perhaps best Socioeconomics and the institutional context of observed in the patterns of land use and develop- adaptive management ment in the basin. Who , works, and plays where 1. A key area of research should focus on and why and with what effect on the environment? visitors and part-time nonworking residents whose Multidisciplinary approaches need to be brought to spending and recreational preferences affect the so- bear on determining the economic and ecological cial and economic dynamics of the basin. More spe- roles of the nearly 8,000 publicly owned parcels in cifically, there is a need to investigate the numbers of the urban interface areas of the basin and the likely tourists and seasonal residents, activities that visitors effects of further acquisitions. Moreover, the eco- participate in and their reasons for choosing to par- nomic and social impacts of BMP implementation ticipate in them, patterns of visitor use of shore and need to be analyzed in light of continued site-specific lake areas by season, and patterns of use of urban- biophysical effectiveness studies. The social, eco- suburban trails and backcountry roads and trails. nomic, and cultural effects and impacts of mitigation Within these research areas there is a growing need and restoration need to be better understood. De- to stratify studies by ethnicity, race, and socioeco- termining the values held by different interests, nomic status in order to inform social policy ques- whether residents, visitors, or those who never visit tions concerning environmental and social justice the basin but consider it an important public asset, issues. should be a high research priority. Failure to connect 2. A “creative tension” is identified in this willingness to pay for or to accept impacts with pro- document that describes the Tahoe community’s posed management strategies to protect the basin attempts to serve visitor recreation needs, while pro- (prescribed burning, restricted access to sensitive viding a healthy and robust community environment ecological assets, and transportation strategies) will for those who live and work in the basin. Significant only exacerbate social conflict. Research should fo- research is needed to focus public discussion and to cus on quantifying held values and the relationships reduce the use of anecdotal and impressionistic data of those values to actual management actions in or- in determining social policy. In particular, research der to guide policy choices. should be directed toward the following: 4. The strength and complexity of public- • Refining the affordable housing needs as- private coalitions and social networks in the basin is sessment (begun by TRPA in 1997) in the a key theme of the institutional assessment in this light of the census 2000 results; document. Given historical emphasis in the Tahoe • Accounting for and analyzing the need for basin on collaboration and partnership building, it is and availability of educational, social, and important to continue researching organizational recreational services focused on full-time response options to scientific and social knowledge. and part-time working residents; Specifically, the institutional capacity for implement- • Analyzing the impacts of economic rede- ing an adaptive management strategy in support of velopment and community reinvestment the EIP should be a focus of policy and social re- strategies to determine whether intended search. The experimental nature of public policy and outcomes are achieved and for whom; social network processes in the basin, and their ca- • pacity to inform other regions encountering similar Tracking and analyzing trends in labor socioenvironmental issues, suggests that a great deal populations, particularly those in the hospi- can be learned by careful research and observation.

Lake Tahoe Watershed Assessment 713 Chapter 7

of reaching restoration goals. The Lake Tahoe Inter- The Status of Monitoring agency Monitoring Program is an example of an ef- The environment in the Lake Tahoe basin fort at coordination and collaboration with regard to is monitored by many university and government monitoring water quality conditions in the basin. programs. The primary universities involved are More collaborative efforts such as this would greatly University of California at Davis (UCD), University benefit monitoring goals and associated information of Nevada at Reno (UNR), and the Desert Research needs in the basin. An integrated monitoring strategy Institute (DRI). These efforts include activities such is an essential component of an adaptive manage- as monitoring elements of lake clarity (UCD), moni- ment approach, and its development should be a toring air quality (UCD), and monitoring ground high priority over the next few years. water (UNR). Monitoring conducted by government The Status of Modeling agencies varies widely in its focus and application. The watershed assessment has unearthed The US Geological Survey primarily monitors water and developed important information about the quantity and quality. The USDA Forest Service con- status, dynamics, and interdependencies of resource ducts a variety of monitoring efforts related to water conditions in the Lake Tahoe basin. Some of the quality, forest health, and recreation use. The Forest data are new, but much of the assessment analyzes Service also conducts surveys of wildlife populations, and synthesizes existing information in new ways. such as marten, northern goshawk, osprey, bald ea- Many key findings in this document highlight chal- gle, and spotted owl. Most wildlife surveys, although lenges in designing management actions that achieve conducted annually, are not designed to function as multiple resource goals. monitoring programs. The state parks departments Each of the four resource areas assessed for in California and Nevada also monitor air quality, the Lake Tahoe basin (air, water, biotic, and socio- plant and wildlife populations, fire effects, and rec- economic) can be thought of as responding to envi- reational use in the parks. The Tahoe Regional Plan- ronmental conditions and human actions. Concep- ning Agency monitors the status of the environment tual models were developed for each resource area, with respect to basin thresholds, such as air quality, dividing each into “interactive” and “independent” water quality, plant and animal populations, and land environmental factors. For example, a change in the coverage. The Lahontan Regional Water Quality condition of the air resource within the basin could Control Board primarily monitors compliance with act as an interactive environmental variable on water regulatory standards and performance requirements. resources by significantly altering the rate of atmos- Typically, monitoring data collected by uni- pheric deposition of nitrogen and phosphorus on the versities and agencies have limited applicability to surface of the lake. This in turn, could alter the con- basin-wide ecosystem issues because they are de- dition of the water resource within the basin. In ad- signed for organization-specific purposes, they are dition, the amount and timing of precipitation, an not formally structured monitoring efforts, and they independent environmental variable, can also affect are not coordinated across agencies or topics. The water resource conditions by affecting rates of sur- benefits of coordinated and collaborative monitoring face run-off. Each interactive environmental factor are many, including the ability to pool funds and can be an output from one resource and an input to expertise across agencies and to address complex one or more of the other resources. issues across jurisdictional boundaries. The liabilities A simple conceptual model for each of the of conducting disparate monitoring efforts is that four resources addressed in this assessment identifies large-scale environmental problems may go unde- the environmental conditions and management ac- tected because we are not measuring attributes that tions that can affect resource conditions (see figures are meaningful from a system perspective or we may 7-4 through 7-7). These conceptual models show fail to comprehend its utility to the broader context several elements: the key factors that affect the con-

714 Lake Tahoe Watershed Assessment Chapter 7

dition of resources as inputs to the resource (arrows quality that operate in Lake Tahoe govern the fate of from factors to the resource); processes that operate nutrients, pollutants, and sediments in the lake. within resources and the parameters that describe Management activities that directly affect the condi- the condition of that resource; and the outputs of tion of the water resource include channel modifica- the resource that may be useful measures of resource tions, sediment traps, land use restrictions, and wa- condition (arrows from the resource to indicator tercraft tailpipe restrictions. The condition of the variables). The factors identified as interactive are water resource also can be altered indirectly by a restricted to those associated with the four key re- mixture of both on-the-ground management activi- sources addressed in this assessment. These concep- ties and regulatory restrictions. The interactive envi- tual models can be used to identify influential factors ronmental variables that affect water quality come that are common to more than one resource, repre- from each of the three other resource areas (nutrient senting links among air, biotic, water, and socioeco- deposition from the air resource, the urban land- nomic conditions within the basin. Emphasis fo- scape from the socioeconomic condition, vegetative cuses on the links that managers have at least some landscape from the biotic resource), as well as from control over—effects from management activities upland watershed processes to Lake Tahoe (sedi- and interactive environmental factors. ment fluxes). A potentially overriding environmental factor affecting the water resource is climate; on a Air Quality Conceptual Model short-term basis, the detection of management- The air resource is represented by two related changes in the condition of water quality components in the conceptual model: the regional could be difficult to detect. airshed and the Lake Tahoe airshed (Figure 7-4). The regional airshed describes input of pollutants into Biological Integrity Conceptual Model the Lake Tahoe basin from the Central Valley of The biological resource is represented by California and the neighboring areas. The Lake Ta- two components in the conceptual model: spe- hoe airshed describes the fate of pollutants generated cies/populations and communities/ecosystems (Fig- within or once they arrive to the Lake Tahoe basin. ure 7-6). The species/population component ad- The activities that directly affect the condition of the dresses the distribution abundance of all plant, ani- air resource are automobile and watercraft tailpipe mal, and fungi species in the Lake Tahoe basin. The regulations, as well as restrictions on stoves communities/ecosystems component addresses the within the Lake Tahoe basin. The interactive envi- distribution, abundance, and integrity of vegetative ronmental factors include a variety of factors, but and aquatic communities and ecosystems in the ba- two resource areas have the greatest interaction with sin. Most management actions have at least some the air resource: biotic resources (for example, influence on biological integrity, with some of the smoke emissions) and sociocultural conditions (ex- greatest influences in the basin coming from vegeta- haust emissions). However, a potentially overriding tion management, fire management, land develop- environmental factor affecting the air resource is ment, and recreation. Interactive environmental fac- climate, which on a short-term basis can obscure the tors originate from all three of the other resources, detection of any changes from either management but the most influential are water quality and interac- activities or interaction variables. tive elements within biological systems.

Water Quality Conceptual Model Socioeconomic Condition Conceptual Model The water resource is represented by two The socioeconomic “resource” is repre- components in the conceptual model: upland water- sented by two components in the conceptual model: sheds and Lake Tahoe itself (Figure 7-5). Upland social well-being and economic performance (Figure watershed processes govern the movement of water 7-7). Social well-being is measurable through a num- and fate of pollutants and nutrients as they move ber of sociocultural processes, including population, toward the lake. Processes associated with water community resilience to change, adaptability of

Lake Tahoe Watershed Assessment 715 Chapter 7

Independent Interactive Management Environmental Environmental Activities Variables Variables

Auto Tailpipe Regs. Pollutant Discharge Sediment Flux Wood Stove Regs. Wind Smoke Emissions Watercraft Tailpipe Regs. Temperature Exhaust Emissions Humidity Nutrient Flux Precipitation

Air Resources Regional Airshed Lake Tahoe Airshed

Atmospheric Discharge Atmospheric Discharge Fate and Transport Fate and Transport Atmospheric Deposition Atmospheric Deposition

Indicator Variables

Ozone Concentration Visibility PM 10 and 2.5 Nox Concentration

Figure 7-4—Schematic diagram of detailed air resource conceptual model.

716 Lake Tahoe Watershed Assessment Chapter 7

Independent Interactive Management Environmental Environmental Activities Variables Variables

Urban Landscape Channel Modifications Precipitation Vegetative Landscape Sediment Traps Evapotranspiration Land Use Restrictions Wind Nutrient Deposition Water Tailpipe Regs. Temperature Sediment Flux Timber Harvest Humidity Stream Morphology Water Quality Fire Management Solar Radiation

Water Resources Lake Tahoe Upland Watersheds Hydrodynamics Precipitation Dynamics Nutrient Processes Water Flow Dynamics Sediment Processes Slack Water Dynamics Biotic Processes Nutrient Dynamics Sediment Dynamics

Indicator Variables

Lake Clarity Algal Growth Water Quality Channel Stability

Figure 7-5—Schematic diagram of detailed water resource conceptual model.

Lake Tahoe Watershed Assessment 717 Chapter 7

Independent Interactive Management Environmental Environmental Activities Variables Variables

Roads Precipitation Smoke Emissions Vegetation Management Evapotranspiration Nutrient Flux Restoration Solar Radiation Sediment Flux Stream Morphology Fire Management Temperature Vegetation Management Wind Soil Productivity Urban Development Water Quality Recreation Biological Conditions

Biological Resources Communities/Ecosystems Species/Populations Conifer Forests Vascular Deciduous Forests Nonvascular Plants Shrublands Vertebrates Meadows Invertebrates Riparian Areas Fungi and lichen Lentic Zones Lotic Zones

Indicator Variables

Biological Diversity Biological Integrity Vegetative Landscape Fire Regime Fire Risk Population Viability Exotic Species Loads

Figure 7-6—Schematic diagram of detailed biotic resource conceptual model.

718 Lake Tahoe Watershed Assessment Chapter 7

Independent Interactive Management Environmental Environmental Activities Variables Variables

Auto Emissions Land Use Restrictions Out-of-basin Growth Water Craft Emissions Transportation Restrictions Precipitation Wood Stove Restrictions Temperature Urban Landscape Recreation Management Wildfire Ignitions Vegetative Landscape Sport Tailpipe Restrictions Biological Diversity Road and Trail Development Water Quality Fire Management Visibility

Socioeconomic Conditions Economy Social Well-being Tourist Economy Population Dynamics In-basin Economy Social Structure Business Development

Indicator Variables

Student Preparedness Scenic Value Affordability Index Property Values Standard of Living Economic Growth

Figure 7-7—Schematic diagram of detailed socioeconomic resource conceptual model.

Lake Tahoe Watershed Assessment 719 Chapter 7

social structures, etc. The economic component affect multiple resources (Table 7-4). In addition, represents the generation and distribution of wealth independent environmental factors also affect multi- within the Lake Tahoe basin. Sociocultural processes ple resources (Table 7-5). The substantial level of directly affecting it are recreation and tourism, non- interaction among resource areas indicates strong tourist economic activities, values feedback mechanisms within resource areas and the (market and non-market), and business develop- potential for a chain of events and feedback loops ment. Management activities that directly affect the precipitated by the alteration of the condition of one condition of the socioeconomic resource include resource. These factors should be key areas of dis- land use regulations, transportation infrastructure cussion and careful management. and investment, tailpipe and wood stove emission Climatological factors contribute the inde- controls, recreation management, and road, trail, and pendent environmental inputs that affect multiple public access construction. Socioeconomic condi- resources, with precipitation and temperature affect- tions are generally more directly affected by regula- ing all four resource areas. Understanding the degree tory actions and investment than on-the-ground to which climatic conditions vary within the Lake management actions. The interactive environmental Tahoe basin and to which they are likely to vary variables directly affecting the socioeconomic re- through time will be a key factor in differentiating source are numerous and come from all three of the among the various contributions of climate and other resource areas (visibility and pollution from management activities to resource conditions. It is the air resource, water quantity and quality from the possible that management plans can be effective but water resource, and fishery production, wildlife still not exhibit improvement in resource conditions populations, and forest health from the biological if those plans are implemented under adverse cli- resource). From this conceptual model, it can be matic conditions. Conversely, it is possible that im- seen that socioeconomic conditions can be affected proved resource conditions could accompany inef- significantly by conditions in all of the other re- fective management activities during periods of sources. In addition, socioeconomic conditions beneficial climatic conditions. An example of the could have significant impacts on each of the other confounding effects of climate became apparent resources in the Lake Tahoe basin. Macroenviron- when the clarity of Lake Tahoe showed improve- mental factors that could affect the socioeconomic ment during drought conditions from 1987 through system are population growth in California and Ne- 1993. This improvement may have been due to a vada and climate. weather condition that reduced sediment and nutri- ent transport to the lake. Summary of Factors Affecting Multiple Resources A significant number of factors are com- Quantitative Models of Key Resource Conditions mon to two or more of the Lake Tahoe watershed and Interactions resources. These factors represent the interactive A number of computerized modeling tools components of each resource and the greater envi- have been developed for a variety of applications, ronment. Understanding these factors, their interac- most intended to simulate the processes that affect tions, and their relative influences can assist manag- air, water, biotic, and socioeconomic resources. An ers in designing management activities that maximize extensive list of these tools is presented in Table 7-6, the intended effects. Approximately 65 percent of all along with a description of each model, the proc- interactive environmental factors affect two or more esses they simulate, and references to the literature resource areas (Table 7-3). More than half of interac- on the theoretical development of the model. While tive environmental factors serve as output and input a catalogue of existing modeling efforts is useful in factors within the same resource. All but one of the establishing an adaptive management strategy, evalu- management activities (transportation management) ating the applicability and utility of those models is critical.

720 Lake Tahoe Watershed Assessment Chapter 7

Table 7-3—Interactive environmental factors identified in the conceptual models of air, biotic, water, and socio- economic resources.

Resource Factor Air Biotic Water Socioeconomic Nutrient flux O I/O Sediment flux I I/O Smoke emissions I I I/O Exhaust emissions I I I/O Dust I O Forest health O I Soil productivity I O Visibility O I Stream morphology I I/O Water quantity and quality I I/O I Urban landscape I I/O Vegetative landscape O I I Fishery production O I Wildlife populations O I

Note: I = Input variable, O = Output variable

Table 7-4—Management activities that directly affect multiple resources within the Lake Tahoe basin.

Resource Activity Air Biotic Water Socioeconomic Tailpipe regulations X X X Wood stove regulations X X Land use regulations X X X Road management X X X X Fire management X X X X Recreation management X X X Range management X X Vegetation management X X Restoration X X Development/urban open space X X X X X

Lake Tahoe Watershed Assessment 721 Chapter 7

Table 7-5—Independent environmental factors that affect multiple resources within the Lake Tahoe basin.

Resource Factor Air Biotic Water Socioeconomic Wind X X Temperature X X X X Humidity X X Precipitation X X X Evapotranspiration X X Solar radiation X X

Development of new models and customi- sources, and discharges from prescribed forest fires. zation of existing models will be necessary as an In addition, out-of-basin pollutant sources from adaptive management strategy is put in place. Model neighboring can be accounted for in the predictions will have greater accuracy as the models simulations. The model predicts the atmospheric are adapted to site-specific processes. Site-specific concentration of gasses (CO, NOx, NMHC, SO2) models also can be better tailored for integration and particles (TSP, PM10, PM2.5). A mechanism to with other models, eliminating the need to develop predict deposition of atmospheric pollutants onto interface routines among existing models. Moreover, the surface of Lake Tahoe has been developed, but site-specific models can be adapted more readily to due to a lack of monitoring information, this mecha- produce information that is more relevant to man- nism has not been calibrated or validated. agement issues in the area of application. Several air, water, biotic, and socioeconomic Water Resource Models models are in use or under development in the basin. A series of models, referred to as the Lake Although these models have been developed with- Tahoe Clarity Model, is being developed by the Ta- out benefit of an integrated information strategy, hoe Research Group at the University of California, they can contribute substantially to a more integrated Davis, to predict the condition of water resource basin-wide management planning effort when it is elements within the Lake Tahoe basin. The first ele- developed. Some of the models with the greatest ment is a watershed hydrology and sediment loading potential to contribute to current and future model- model, referred to as the University of California, ing needs in each of the four resource areas are Davis, Watershed Model (UCDWM). The second briefly described below. element is a water quality model for Lake Tahoe. The UCDWM has been used to predict Air Resource Models stream behavior in the Ward Creek watershed and The only air resource model available for will be designed to predict streamflow from rainfall the Lake Tahoe basin is the Lake Tahoe Airshed into small to medium basins. The UCDWM is a spa- Model (LTAM), developed by Cliff and Cahill at the tially distributed model that uses maps of physical University of California, Davis. The model simulates conditions within a watershed to generate modeling processes within the Lake Tahoe airshed and pre- parameters. Models that predict sediment flux and dicts discharges to the atmosphere, transport and snowpack dynamics are being developed, but they fate of pollutants within the atmosphere, and deposi- are not sufficiently well developed to link directly to tion of nutrients on terrestrial and aquatic surfaces in the UCDWM. the Lake Tahoe basin. LTAM can predict in-basin The Lake Tahoe water quality model will pollutant discharges from traffic sources for the ma- simulate hydrodynamics in the water quality jor roadways within the basin, urban pollution

722 Lake Tahoe Watershed Assessment Chapter 7

Table 7-6—Available models to describe the condition of resources within the Lake Tahoe basin.

Model Name Processes Simulated Variables Simulated Agency or Reference

AGNPS Water Resource Water Quantity USDA Water Flow, Sediment and Quality Young and Nutrient Dynamics Nutrient Flux et al. 1989 Sediment Flux

BEAMS Water Resource Water Quantity USDA Water Flow and Sediment Sediment Flux Langendon Dynamics et al. 1998

CALMET Air Resource Visibility USEPA CALPUFF Atmospheric Discharge CALPOST Fate and Transport

CASC2D-SED Water Resource Water Quantity Johnson Water Flow and Slack et al. 1998 Water Dynamics

CCHE2D Water Resource Water Quantity Jia and Wang Water Flow and Sediment Sediment Flux 1997 Dynamics

CE-QUAL-ICM Water Resource Water Quantity USDA Water Flow, Sediment and Quality Cerco and and Nutrient Dynamics Nutrient Flux Cole 1995 Nutrient Sediment Flux Dynamics

CE-QUAL-W2 Water Resource Water Quantity USDA Water Flow, Sediment and and Quality Cole and Nutrient Dynamics Nutrient Flux Buchok 1995 Sediment Flux

Climatological Air Resource Pollutant USEPA Dispersion Atmospheric Discharge Concentration Model Fate and Transport

COMPLEX1 Air Resource Visibility USEPA Atmospheric Discharge Pollutant Fate and Transport Concentration

CREAMS Water Resource Water Quantity USDA Water Flow, Sediment and Quality Knisel 1993 and Nutrient Dynamics Nutrient Flux Sediment Flux

DWAVNET Water Resource Water Quantity Langendon Water Flow and Slack et al. 1996 Water Dynamics

Lake Tahoe Watershed Assessment 723 Chapter 7

Table 7-6—(continued)

Model Name Processes Simulated Variables Simulated Agency or Reference

Forest-BGC Biotic Resource Vegetative Running and Vegetative Communities Landscape Coughlin 1988

EPIC Water Resource Water Quantity USDA Precipitation, Water and Sediment Dynamics

FRAME Water Resource Water Quantity Vieira Water Flow and Slack Water Sediment Flux et al. 1997 Dynamics

GLEAMS Water Resource Water Quantity USDA Water Flow, Sediment and Quality Knisel 1980 and Nutrient Dynamics Nutrient Flux Sediment Flux

HEC-HMS Water Resource Water Quantity US Army COE Water Flow and Slack HEC 1995 Water Dynamics

HEC-RAS Water Resource Water Quantity US Army COE Water Flow and Slack HEC 1997 Water Dynamics

HFAM Water Resource Water Quantity USGS/EPA Water Flow and Slack Water Marino and Dynamics Crawford 1998

HSPF Water Resource Water Quantity USGS/EPA Water Flow and Slack Bicknell Water Dynamics et al. 1997

LAWF Water and Biotic Resource Vegetative Cheng et al. Water Flow and Slack Water Landscape 1998 Dynamics, Lentic and Lotic Communities

MOBILE5a Air Resource Pollutant USEPA Tailpipe Discharge Concentration

PLUVUE Air Resource Visibility USEPA Fate and Transport

724 Lake Tahoe Watershed Assessment Chapter 7

Table 7-6—(continued)

Model Name Processes Simulated Variables Simulated Agency or Reference PSRM Water Resource Water Quantity Aron 1996 Water Flow Dynamics

QUAL2E Water Resource Water Quantity USEPA Water Flow, Sediment and Quality and Nutrient Dynamics Nutrient Flux Sediment Flux

RAM Air Resource Visibility USEPA Atmospheric Discharge Pollutant Fate and Transport Concentration

REMM Water and Biotic Resources Water Quantity USDA Water Flow, Nutrient and Nutrient and Altier Sediment Dynamics, Sediment Flux et al. 1998 Riparian communities Veg. Landscape

RPM-IV Air Resource Pollutant USEPA Atmospheric Discharge Concentration Fate and Transport

SHAW Water Resource Water Quantity Flerchinger Precipitation Dynamics and Saxton 1989

SHE Water Resource Water Quantity Bathurst Water Flow and 1986 Slack Water Dynamics

SMS Water Resource Water Quantity US Army COE Water Flow Dynamics WES, Holland et al. 1998

SWAT Water Resource Water Quantity USDA Water Flow, Sediment and Quality Arnold and Nutrient Dynamics Nutrient Flux et al. 1993 Sediment Flux

SWMM Water Resource Nutrient and USEPA Water Flow, Nutrient Sediment Flux and Sediment Dynamics Water Quantity and Quality UAM Air Resource Pollutant USEPA Atmospheric Discharge Concentration Fate and Transport

Lake Tahoe Watershed Assessment 725 Chapter 7

Table 7-6—(continued)

Model Name Processes Simulated Variables Simulated Agency or Reference USLE Water Resource Sediment USDA Sediment Dynamics Flux Wischmeier and Smith 1978

WASP5 Water Resource Lake Clarity USEPA Lake Hydrodynamics Algal Growth Nutrient and Biotic Processes Water Quality

using a process-based one-dimensional representa- include a fuels map and weather set, which are used tion of flow dynamics and temperature profiles to display the changes in fire behavior characteristics within Lake Tahoe. The hydrodynamics model is with different fuel loads or weather conditions. The subsequently used as the basis for determining the third model is the Fire Area Simulator (FARSITE) transport and fate of nutrients and sediments in the model, which is a spatially explicit fire behavior vertical column. An optical model predicts the clarity model. FARSITE can be used to simulate wildland of the lake based on the vertical distribution of algae fire growth and behavior under complex conditions and sediment concentrations. The Lake Tahoe water of terrain, fuels, and weather. Fire behavior varies quality model uses data from measurements of nutri- with topography, weather input, and fuels encoun- ents from the atmosphere and watersheds that flow ters as a fire spreads. The outputs of the model in- into the lake. clude fire extent flame lengths, and heat per unit area across the area burned. Fourth, the results from the Biological Resource Models FARSITE model can be combined with existing Several models are being developed to vegetation conditions and analyzed with a First Or- simulate biotic resource processes within the Lake der Fire Effects Model (FOFEM). FOFEM is a fire Tahoe basin. These include models that predict the effects model that displays fire severity, resultant tree risk and consequences of fire on vegetation and the mortality, and changes in vegetation. FOFEM is biodiversity in lentic and lotic zones within the basin. designed to predict the effects of prescribed fire and Five fire and fuels models have been ap- wildfire in forests and . Outputs include plied within the basin. First is a fuels model that was quantitative, site-specific predictions of tree mortal- developed by Fites-Kaufman et al. using a rule-based ity, fuel consumption, and smoke. Finally, the Rare approach to continuously update fuel layers as vege- Event Risk Assessment Process (RERAP) model tation conditions change or when vegetation is dynamically calculates the risk of undesired fire treated through management. Input information movement and smoke dispersion before a fire- includes vegetation cover, dominant vegetation form ending event, such as precipitation, can halt its (i.e., shrub, grass, or tree), dominant tree species, spread. Inputs are historical weather data and recent mortality, elevation, recent management FLAMMAP layers. Outputs include probabilities of treatment (i.e., burning, thinning), and land use clas- the fire or smoke event reaching a predefined target, sification (i.e., urban, undeveloped). The fuel model average, or common rate of spread (ROS), and ROS outputs can serve as inputs to the other fire models. as altered by weather events, such as high winds. The second model is FLAMMAP, which produces a RERAP and FARSITE are complementary models. spatially explicit, continuous surface of fire behavior In addition to fire and fuel models, models characteristics (rate of spread, flame length, heat per of riparian biodiversity are being developed by Man- unit area, and likelihood of crown fire). Input data ley and Schlesinger for the Lake Tahoe basin. The

726 Lake Tahoe Watershed Assessment Chapter 7

models use regression equations to predict indices of of Lake Tahoe are so fundamental that actions based biodiversity related to terrestrial characteristics near on our current knowledge have a substantial likeli- lentic and lotic habitats. Models have been devel- hood of missing the overarching target of reversing oped to predict indices of biodiversity for birds, the current decline in lake clarity. The only practical mammals, and vascular plants. Further development alternative is a deliberate and aggressive adaptive and verification of these models is planned. management program focused on key ecosystem functions, with the explicit goal of restoring lake Socioeconomic Models clarity and forest health. This goal can be accom- A model attempting to simulate the effect plished in a manner that respects and supports the of changes in lake clarity on residential property val- mandates, interests, and authorities of all the basin’s ues is being developed by Bernknopf et al. This stakeholders. However, management and research model portrays relationships between land use re- institutions engaged in activities in the Tahoe basin strictions and property values for privately owned must change how they operate in order to meet this lots in the Upper Truckee watershed, which includes challenge. The need for immediate solutions creates much of South Lake Tahoe, California, and Stateline, an urgency for action. Collaboration is essential to Nevada, the largest urbanized areas within the Lake meeting the information and management needs of Tahoe basin. A regression model has been devel- the basin for two reasons: the complementary nature oped that predicts the property value of a privately of mandates and skills associated with various agen- owned lot as a function of land development restric- cies working in the basin can be leveraged to accom- tions placed on the lot. plish more in a shorter period of time, and, corre- In a related vein, Nechodom et al. are ex- spondingly, a lack of coordination and collaboration ploring the feasibility of modeling the relationships could result in duplication of effort, or worse, in- between public parcels and land and housing values. compatible approaches that result in no net progress The effort is preliminary and attempts solely to test toward restoration goals. Many basin agencies and the usefulness of using existing assessor parcel data stakeholders have contributed to creating the current to measure the effects of public land acquisitions in climate of enhanced collaboration in stewarding discrete neighborhoods in the basin. Further devel- public resources and for refining sustainable ecologi- opment of this model would need to determine the cal, social, and economic practices in the basin. It is additional effects of adjacency of an entire neighbor- our belief that this valuable collaboration needs to hood to larger public holdings, as well as location- continue and grow into a scientifically informed dependent amenity and transportation values, on adaptive management approach to the management housing and land prices. and restoration of the Lake Tahoe basin. The re- In addition, an economic input/output mainder of this chapter addresses specific recom- model constructed at the community-region scale, mendations for the development and structure of an developed by Robison et al., is presented in Chapter adaptive management strategy. Recommendations 6 of this assessment. The purpose of this model is to for a collaborative institutional structure to help de- display the potential effects of specific changes in fine and guide adaptive management is discussed any given industry or economic sector on the rest of first, followed by specific recommendations for next the economy. The ultimate use of this model will be steps for each phase of the adaptive management to track economic impacts of changes in manage- cycle. ment policy. However, most immediately, the model is intended to be used to estimate impacts of strate- Collaborative Structures for Adaptive gies to derive $252 million over ten years to pay for Management the local share of the EIP. We recommend the establishment of formal working groups in order to realize a functioning, Toward the Future multi-agency adaptive management process for the Gaps in our understanding of the ecology Lake Tahoe basin. We recognize that a number of

Lake Tahoe Watershed Assessment 727 Chapter 7

different structures have the possibility of being lished by the recently executed MOU among re- equally successful, but for the purposes of discus- search institutions (signed in June 1999 by University sion, we provide a proposed structure to stimulate of Nevada, Reno, University of California, Davis, discussion and movement toward the establishment USDA Pacific Southwest Research Station, Desert of such working groups. Our proposed structure is Research Institute, and USDI Geological Survey), composed of three main working groups: an Execu- the ECC, AMWG, and SAG should review and tive Coordinating Committee (ECC), an Adaptive make recommendations on research and monitoring Management Working Group (AMWG), and a Sci- activities, particularly those concerning empirical ence Advisory Group (SAG) (Figure 7-8). These relationships among ecosystem elements and the three primary working groups could be established development of sustainable environmental and social and sanctioned through the formation of a memo- policies. The SAG should provide analytical support randum of understanding (MOU). The recom- to project level planning and be responsible for inte- mended roles and responsibilities of each working gration of scales of information appropriate to the group are described below. project. The ECC would be chartered to coordinate The AMWG should convene a task force of policy and communication between management information management coordinators and consult- agencies and the current Lake Tahoe Federal Advi- ants to determine the appropriate protocols for data sory Committee (LTFAC). The AMWG would be storage and analysis. An initial charge of the task chartered to streamline communications, to coordi- force should be to conduct an assessment of infor- nate collaborative implementation of each agency’s mation needs, data formats, data usage and storage, management plans, and to integrate scientific review and hardware and software in use by public agencies. from the SAG. Its ultimate objective is to coordinate The task force should work with the AMWG and adaptive management and restoration activities to the ECC to develop recommendations for an inte- create a fully integrated and interactive terrestrial and grated information management system that is coor- aquatic resources management planning program. dinated with other environmental information sys- The SAG would recommend research, monitoring, tems and information clearinghouse efforts in exis- and restoration activities and priorities. tence throughout the state, such as the Federal Geo- A fully developed adaptive management graphic Data Committee (FGDC), EPA, California’s strategy with collaboration as its hallmark needs to data Framework, California Environmental Re- be brought on line in the next year in order to suc- sources Evaluation System (CERES), and the Cali- cessfully meet restoration goals in the Lake Tahoe fornia Geographic Information Assessment. basin. A five-year plan could focus on the most im- The AMWG should be chartered and sanc- mediate conservation challenges for an initial period tioned by the ECC to facilitate ongoing policy and of approximately five years and could prioritize pro- management dialogues, informed by the scientific ject implementation integrated with research and information in this assessment, as well as future re- monitoring, using whatever criteria established by search and monitoring activities. The SAG should be the AMWG. In 2002, a new 20-year regional plan required to review and prepare background materi- will be developed to guide planning and manage- als, white papers, and problem analyses to inform ment actions. Full implementation of the adaptive the structured dialogues between policy and man- management strategy should coincide with TRPA’s agement. Both the SAG and the AMWG will need new plan in 2002. Information and experience from to seek outside consultation services when necessary a well-developed adaptive management strategy to help distill issues and inform participants in policy should prove invaluable in establishing the goals of and management dialogues. The process should be the new regional plan. guided initially by the key findings and research Under the authorities and guidelines estab- needs identified in Chapter 1 and in this chapter.

728 Lake Tahoe Watershed Assessment Chapter 7

Executive Coordinating Committee (ECC) Has oversight responsibility for the adaptive man- agement process (AMP) and direct responsibility for drafting broad policy changes revealed through the AMP. Is responsive to all agencies and the public. 3 to 5 rotating members

Adaptive Management Working Group (AMWG) Coordinates the AMP among agencies, including setting restoration priorities, restoration and research financing, and data storage and access. Is responsive to the direc- tion provided by the ECC and the scientific informa- tion supplied by the SAG. Proposes policy changes. 1 member per agency

Scientific Advisory Group (SAG) Is responsible for recommending research priorities and programs to the AMWG and for technical quality con- trol of all monitoring, modeling, and research activities. Is responsive to management needs through the AMWG and ECC. Identifies consequences of man- agement actions. 5-9 members represent key scientific disciplines

Consultants Research Institutions

Are responsible for fast Are responsible for long- turnaround research, mod- term research and monitor- eling, and monitoring pro- Agency Working Groups ing that is not time sensitive. jects. Are responsive to Are responsive to the specific contractual re- Are responsible for identifying AMWG and the SAG. quirements issued by the specific restoration needs and for Members—from all interested AMWG. performing restoration actions. institutions Are responsible for agency spe- cific monitoring and are respon- sive to the AMWG and the SAG. Members—agency scientists and mem- bers of SAG

Figure 7-8—Conceptual structure for adaptive management strategic groups in the Lake Tahoe basin. Solid lines = administrative authority, dashed lines = technical exchange.

Lake Tahoe Watershed Assessment 729 Chapter 7

Next Steps for the Adaptive Management Cycle laboration among research institutions and an in- Significant effort will be required for institu- creased emphasis on projects that address the high- tions to absorb the information provided in the wa- est priority information needs. The strategy should tershed assessment and to integrate that information identify opportunities for integrated data collection into policy and management actions. Over the short and analysis efforts and structure avenues of funding term, activities, such as those identified in Table 7-5, to emphasize priority topic areas for investigation, will provide immediate contributions toward adap- favoring projects conducted in a coordinated or col- tive management; however, the simultaneous devel- laborative manner. opment of all phases of the adaptive management Funding of projects and studies at Lake Ta- cycle is essential. Our portrayal of the adaptive man- hoe is currently a piecemeal and ad hoc enterprise. agement cycle consists of four phases: information Much of the funding process is unlikely to change, needs identification and prioritization, information due in large part to institutional arrangements and acquisition and assessment, evaluation and applica- constraints that otherwise generate healthy competi- tion to decision-making, and management action. tion for limited resources among good project op- The wealth of information provided in the water- tions. However, opportunities for fund- shed assessment can be readily applied to effective ing and research can be missed because of differ- action that can be taken in regard to each phase of ences in mandates and priorities among core institu- the adaptive management cycle. tions. There are few institutional homes for the hon- The watershed assessment can be used by est and difficult discussions needed to determine managers, scientists, and policy-makers for future priority basin ecosystem needs—ecologically, eco- investments in data acquisition. Charting such a nomically, and socially—and activities among institu- course will involve evaluating not only what is tions. Opportunities are fewer yet for setting priori- known about the Lake Tahoe basin but also how ties in an open and transparent dialogue and for pro- restoration and management efforts can be informed viding mutually beneficial incentives to reach the by additional data. For example, how much better restoration goals set forth in the Regional Plan, the can the relationship between prescribed fire and 208 Plan, and the EIP. sediment transport to the lake be described by sub- A formal process is needed to identify im- watershed experiments? Would expanding the exist- portant research questions. Then, the feasibility of ing stream monitoring network significantly increase answering those questions should be assessed, priori- our understanding of how BMPs affect nutrient and ties for allocation of funding must be set, and a sediment loading to Lake Tahoe? What information competitive process for the funding of proposals can be gained by developing models that simulate must be established. Research and management pri- the response of Lake Tahoe to new regulatory laws? orities and goals would greatly benefit from the es- Answers to such questions can provide an important tablishment of a formal peer review process. basis for future investments in research, monitoring, and modeling. Information Acquisition and Assessment The watershed assessment highlights many An information acquisition and assessment activities that can be used in prioritizing information strategy would greatly benefit management efforts in acquisition and assessment activities in support of the Lake Tahoe basin. The objective of such a strat- adaptive management. A variety of independent re- egy would be to shorten the time required to fill im- search and modeling efforts are underway in the portant information gaps, to test the effectiveness of Lake Tahoe basin, led by numerous academic and competing management approaches, and to inform research institutions, including the University of potential changes in policy and management. Each California, Davis, University of Nevada, Reno, De- of the other three phases of the adaptive manage- sert Research Institute, and USFS Pacific Southwest ment cycle will require some level of development Research Station. A successful adaptive management and function to maximize the effectiveness of an strategy will demand greater coordination and col- information acquisition and assessment strategy.

730 Lake Tahoe Watershed Assessment Chapter 7

Research. A long list of research needs was the Lake Tahoe basin, resulting in more scientifically identified in the watershed assessment. One of the defensible and effective management actions. first tasks of the SAG should be to identify the time- sensitivity of these research needs and any interde- Evaluation and Application to Decision-making pendence among information needs that requires The most fundamental problem faced by consideration in the prioritization and scheduling of the policy and management community at Lake Ta- data acquisition to fill particular critical information hoe is the fragmentary and unorganized application gaps. Such an assessment could be used to build a of available and technical information; more specifi- research agenda for the next five years. cally, the lack of an institutionalized structure to Monitoring. A critical element of a successful gather, analyze, and disseminate information to sup- information acquisition and assessment strategy is port resource management decisions has impeded the development and implement of a multiresource, achievement of ecosystem integrity goals. Adaptive basin-wide monitoring plan that would serve as a management has been described in this chapter, but link between research and management. Monitoring the application of its precepts to the Tahoe basin should be designed to answer questions regarding remains to be articulated in a way that invites explic- the full range of monitoring questions (implementa- itly defined commitments from the core institutions tion, status and trend, and management effective- to redirect resources or to challenge existing priori- ness). A mechanism for data analysis, interpretation, ties. and reporting should also be developed. Monitoring In addition to highlighting what is known results should be used to inform managers and the about the ecosystems of the Lake Tahoe basin and public about management activities, resource condi- what key information is lacking, the watershed as- tions as they compare to desired future conditions, sessment also identifies information evaluation and and the results of management and restoration ef- application activities that can immediately serve res- forts. A process needs to be developed that links toration goals in the basin. The evaluation and appli- monitoring results to the evaluation and reconsidera- cation phase of the adaptive management cycle is a tion of management and restoration actions. sociocultural interpretation of the relevance and Modeling. Modeling efforts must focus pri- meaning of scientific information generated in the marily on available data and on the immediate needs information acquisition and evaluation stage. With- of restoration planners. This implies that modeling out this phase, the generation of new information tools should be developed in a collaborative envi- has no purpose. Table 7-6 provides some examples ronment that includes scientists and those agency of evaluation and application activities that can be personnel charged with developing and implement- used to integrate information into policy and man- ing restoration activities within the basin. Collabora- agement. These activities lie not in the realm of sci- tion will involve all levels of technical knowledge and ence but at the interface between research and man- modeling capability in conceptual and quantitative agement, requiring the input of scientists and man- formats. For that reason, sophisticated user inter- agers from different disciplines to decipher the most faces will be needed that allow agency personnel a effective and appropriate responses to new scientific level of comfort with model results and will provide information. confidence in model simulations. This may entail Certain agencies are better staffed to ad- some reduction in the complexity of model mathe- dress the stewardship and flow of information in the matics to improve the understandability and utility of Lake Tahoe basin. Data are often collected using the modeling tools. At the same time, modeling tools differing methods, ultimately are subjected to in- must focus on the mechanisms that link the various compatible data storage standards, and commonly resource components of the Lake Tahoe basin. The require different protocols for quality assurance and development and use of decision support tools control. For example, GIS layers have been devel- would greatly aid the evaluation and application to oped in disparate formats, impeding ready exchange decision-making phase of adaptive management in of spatially explicit information among agencies.

Lake Tahoe Watershed Assessment 731 Chapter 7

Large quantities of monitoring data have been col- ments and to protect and restore the ecological and lected, but much available data remains unanalyzed; sociocultural assets of the basin. And they must re- hence, management decisions at Lake Tahoe are ject institutional competition based on positioning, often made without informed guidance from moni- rather than on ideas and innovation. toring. The pathways followed by resource man- References agement information, from data analysis and storage, Allen, T. F. H., and T. W. Hoekstra. 1992. Toward a and then from information transfer to integration Unified Ecology. Columbia University into decision-making, must be better understood and Press, New York. refined in order to correct these key deficiencies. Allen, T. F. H., and T. B. Starr. 1982. Hierarchy: Most management and regulatory agencies know the Perspectives for Ecological Complexity. value of effective information management proc- University of Chicago Press, Chicago. esses. Information management processes should be Alley, W. M., and P. E. Smith. 1982. Distributed mandated, funded, developed, and maintained routing and rainfall-runoff model—version through interagency collaboration and with public II: US Geological Survey Open-File Report involvement and access in a manner that ensures 82-344. pp. 201. availability of information for restoration purposes. Altier, L. S., R. R. Lowrand, R. G. Williams, S. P. Under current conditions, management and Inandar, D. D. Bosch, J. M. Sheridan, R. K. restoration activities are carried out under the aus- Humbbard, and D. L. Thomas. 1998. Ripar- pices of more than a dozen funding and regulatory ian Ecosystem Management Model processes. The core institutions involved in these (REMM), USDA-ARS, Southeast Water- funding and regulatory processes are identified in the shed Research Laboratory, USDA Conser- institutional assessment in Chapter 6 of this docu- vation Research Report. ment. While there are recent examples of networking Arnold, J. G., B. A. Engel, and R. Srinivasan. 1993. and coordination among the core institutions, there Continuous time grid cell watershed model. is no forum with the explicit goal to incorporate new In Application of Advanced Information knowledge into management decision-making and Technologies for Management of Natural management actions in a timely fashion. Existing Resources. examples of successful forums for collaborative fora, Aron, G., T. A. Smith, and D. F. Lakatos. 1996. such as Chesapeake Bay’s collaborative monitoring Penn State Runoff Model PSRM C96, User program, can be used as models for building a col- Manual: Environmental Resources Research laborative adaptive management scheme at Lake Institute, Penn State, Pennsylvania. 54 pp. Tahoe. That scheme will at a minimum prioritize ASCE. 1999. Corps Hopes to Reduce Cost of Ever- research, monitoring, modeling, and restoration pro- glades Plan, Civil Engineering Magazine, jects, will fulfill key scientific information gaps, and 69(1). pg. 16. will monitor the resource conditions and the effec- Barber, M. C. (ed.) 1994. Environmental Monitoring tiveness of restoration activities. and Assessment Program Indicator Devel- This new way of doing business will require opment Strategy. EPA/620/R-94. US Envi- three simple rules of engagement. Participating par- ronmental Protection Agency, Office of Re- ties must commit to open, transparent, and scientifi- search and Development, Environmental cally informed processes that set management and Research Laboratory, Athens, Georgia. policy priorities, that critically evaluate outcomes, Bathurst, J. C. 1986. Physically based distributed and that adapt to new information by changing man- modeling of an upland catchment using the agement strategies when it is clear that current direc- System Hydrologique European model, tions are not effective. The parties must encourage Journal of Hydrology 87, pp. 79:102. robust civil debate, the ultimate purposes of which Bicknell, B. R., J. C. Imhoff, J. L. Kittle, Jr., A. S. are to realize the goals in associated policy docu- Donigian, Jr., and R. C. Johanson. 1997.

732 Lake Tahoe Watershed Assessment Chapter 7

Hydrological Simulation Program—Fortran, Society of Agricultural Engineers, 32, 565- User’s Manual for Version 11, EPA/600/R- 571. 97/080, US EPA, National Exposure Re- Franklin, J. F., and J. Fites-Kaufman. 1996. Assess- search Laboratory, Athens, Georgia. ment of late-successional forests of the Si- Bisson, P. A., G. H. Reevers, R. E. Bilby, and R. J. erra Nevada. In: Sierra Nevada Ecosystem Naiman. 1997. Watershed management and Project, Final Report to Congress, vol. II, desired future conditions. Pages 447-474. Assessments and Scientific Basis for Man- In: Pacific salmon and their ecosystems: agement Options. University of California, status and future options. D. J. Stouder, P. Centers for Water and Wildland Resources, A. Bisson, and R. J. Naiman (eds.) Chap- Davis, California. man and Hall, New York. Garbrecht, J., and L. W. Martz. 1995. An automated Brown, L. C., and T. O. Barnwell. 1987. The en- digital landscape analysis tool for topog- hanced stream water quality model, raphic evaluation and drainage identifica- QUAL2E and QUAL2E-UNCAS: docu- tion, watershed segmentation and sub- mentation and user manual. EPA-600/3- catchment parameterization. Report No. 87/007, US EPA, Athens, Georgia. NAWQL 95-1, National Agricultural Water Cerco, C. F., and T. Cole. 1995. User guide to the Quality Laboratory, USDA-ARS, Durant, CE-QUAL-ICM three-dimensional eutro- Oklahoma. phication model, Release 1.0, Technical Re- HEC. 1995. The HEC Hydrologic Modeling System. port EL-95-15. US Army Corps of Engi- Technical Paper No. 150. Davis, California: neers Waterways Experiment Station, US Army Corps of Engineers. Vicksburg, Mississippi. . 1997. HEC-RAS River Analysis System Cheng, M. S., N. Weinstein, and C. Victoria. 1998. Hydraulic Reference Manual, Version 2. Landscape assessment of wetland functions Davis, California: US Army Corps of Engi- (LAWF): A geographic information system neers. (GIS) model for the evaluation of wetland Holland, J. P., R. C. Berger, and J. H. Schmidt. 1998. functions, in Proceedings of the First Fed- “Finite element analyses in eral Interagency Hydrologic Modeling Con- and : An overview of investiga- ference. USDA, Las Vegas, Nevada. April tions at the US Army Engineer Waterways 19-23, 1998. pp. 7-97:7-99. Experiment Station.” To be published in Cochran, W. G. 1977. Sampling techniques. Wiley, the Journal of Engineering Mechanics, Spe- New York. cial publication for the Third US-Japan Cole, T. M., and E. M. Buchok. 1995. CE-QUAL- Conference on Large Scale Computing and W2: A two-dimensional, laterally averaged, Finite Element Methods, Army High Per- hydrodynamic and water quality model, ver- formance Computing Research Center, sion 2.0. US Army Corps of Engineers In- Minneapolis, Minnesota. struction Report, ER-95-1. Hunsaker, C. T., and D. E. Carpenter (eds.) 1990. Committee of Scientists. 1999. Sustaining the peo- Environmental Monitoring and Assessment ple’s lands: recommendations for steward- Program: Ecological Indicators. ship of the National Forests and Grasslands EPA/600/3-90/060, BP91-14196. EPA, into the next century. USDA Forest Service Research Triangle Park, North Carolina. [online] URL: Jia, Y., and S. S-Y Wang. 1997. “CCHE2D: A two- http://www.fs.fed.us/news/science/. dimensional hydrodynamic and sediment Flerchinger, G. N., and K. E. Saxton. 1989. Simulta- transport model for unsteady open channel neous heat and water model of a freezing flows over loose bed,” CCHE-TR-97-2. snow-residue-soil system I. Theory and de- Johnson, B. E., P. Y. Julien, and C. C. Watson. 1998. velopment. Transactions of the American Development of a storm-event based two-

Lake Tahoe Watershed Assessment 733 Chapter 7

dimensional upland erosion model. In: Pro- McLain, R. J., and R. G. Lee. 1996. Adaptive man- ceedings of the First Federal Interagency agement: promises and pitfalls. Environ- Hydrologic Modeling Conference. USDA, mental Management 29(4):437-448. Las Vegas, Nevada. April 19-23, 1998. pp. Noon, B. R., T. A. Spies, and M. G. Raphael. 1999. 3-39:3-46. Conceptual basis for designing an effective- Keyes, A. M., and R. N. Palmer. 1995. An assess- ness monitoring program. Pages 21-48. In: ment of shared vision model effectiveness B. S. Mulder, B. R. Noon, T. A. Spies, and in planning, In: the Pro- others (technical coordinators). The strategy ceedings of the 22nd Annual Conference of and design of the effectiveness monitoring the Water Resources Planning and Man- program of the Northwest Forest Plan. agement Division, American Society of Gen. Tech. Rep. PNW-GTR-437. USDA Civil Engineers, Cambridge, Massachusetts. Forest Service, Pacific Northwest Station, May 7-11, 1995. pp. 532:535. Portland, Oregon. 138 p. Knisel, W. G. 1980. CREAMS, A field-scale model Noss, R. 1990. Indicators for monitoring biodiver- for chemicals, runoff and erosion from ag- sity—a hierarchical approach. Conservation ricultural management systems, US Dept. of Biology 4: 355-364. Agriculture Conservation Research Report NRC (National Research Council). 1990. Managing No. 26. Troubled Waters: Role of Marine Environ- Knisel, W. G. 1993. GLEAMS (The GLEAMS mental Monitoring. National Academy model plant nutrient component). Univer- Press, Washington, DC. sity of Georgia, Coastal Plain Experiment . 1995. Review of EPS’s Environmental Station, Biological and Ag. Engineering, Monitoring and Assessment Program: Tifton, Georgia. Pub. No. 5, 259 pp. Overall Evaluation. National Academy Langendoen, E. J., R. C. Bingner, and R. A. Kahnle. Press, Washington, DC. 1998. Modeling of long-term changes of Olsen, A. R. 1992. The Indicator Development unstable streams. In: Proceedings of the Strategy for the Environmental Monitoring First Federal Interagency Hydrologic Mod- and Assessment Program. EPA/600/3- eling Conference. USDA, Las Vegas, Ne- 91/023. US EPA, Office of Research and vada. April 19-23, 1998. pp. 3-79:3-86. Development, Environmental Research Langendoen, R. A. 1996. Discretization Diffusion Laboratory, Corvallis, Oregon. Wave Model. Tech. Rep. No. CCHE-TR- Peterman, R. M. 1990. Statistical power analysis can TC-1, Center for Computational Hydro- improve fisheries research and manage- science and Engineering, The University of ment. Canadian J. Fisheries and Aquatic Mississippi. Science 47:2-15. Manley, P. N., W. J. Zielinski, C. M. Stuart, J. J. Running, S. W., and G. Coughlan. 1988. A general Keane, A. J. Lind, C. Brown, B. L. Plymale, model of forest ecosystem processes for and C. O. Napper. In press. Monitoring Regional Applications: 1. Hydrologic bal- ecosystems in the Sierra Nevada: the con- ance, canopy gas exchange and primary ceptual model foundation. J. Environmental production processes, Ecological Modeling, Monitoring and Assessment. v. 42, pp. 125-154. Marino, L., and N. H. Crawford. 1998. Hydrologic Sauer, J. R., and S. Droege. 1990. Survey designs and forecast and analysis modeling (HFAM). In: statistical methods for the estimation of Proceedings of the First Federal Inter- avian population trends. US Fish Wild. agency Hydrologic Modeling Conference. Serv., Biol. Rep. 90(1). 166 p. USDA, Las Vegas, Nevada. April 19-23, Shrader-Frechette, Kristin Sharon, and Earl D. 1998. pp. 8-81:8-88. McCoy. 1993. Method in ecology: Strategies for

734 Lake Tahoe Watershed Assessment Chapter 7

conservation. New York: Cambridge Univer- Fire Science Program. sity Press http:/ffs.psw.fs.fed.us. July 21, 1999. Skalski, J. R. 1995. Statistical considerations in the Weeks, P., and J. M. Packard. 1997. Acceptance of design and analysis of environmental dam- Principles by Natu- age assessment studies. J. Environmental ral Resource-Dependent Communities. Con- Management 43:67-85. servation Biology, 11(1), 236-245. Stewart-Oaten, A. 1996. Problems in the analysis of Westley, F. 1995. Governing design: the manage- environmental monitoring data. Pages 109- ment of social systems and ecosystems 131. In: R. J. Schmitt and C. W. Osenberg management. Pages 391-427. In: L. H. (eds.) Detecting Ecological Impacts: Con- Gunderson, C. S. Holling, and S. S. Light cepts and Applications in Coastal Habitats. (eds.) Barriers and bridges to renewal of Academic Press, New York. ecosystems and institutions. Columbia Uni- Suter, G. W. 1993. Ecological Risk Assessment. versity Press, New York, New York. 593 Lewis Publishers. pp. Swezy, D. M., and J. K. Agee. 1991. Prescribed fire Wischmeier, W. H., and D. D. Smith. 1978. Predict- effects on fine-root and tree mortality in old ing rainfall erosion losses - A guide to con- growth ponderosa pine. Can. J. For. Res. servation planning. Agricultural Handbook 21:626-634. No. 537, USDA, Sci. Educ. Admin. pp. 58. Tracy, J. C. 1995. Selection of modeling tools for a Young, R. A., C. A. Onstad, D. D. Bosch, W. P. conjunctive use decision support system. In: Anderson. 1989. AGNPS: A nonpoint- the Proceedings of the 22nd Annual Con- source pollution model for evaluating agri- ference of the Water Resources Planning cultural watersheds. J. of Soil and Water and Management Division, American Soci- Conservation, 44(2), 168-173. ety of Civil Engineers, Cambridge, Massa- Zielinski, W. J., and H. B. Stauffer. 1996. Monitoring chusetts. May 7-11, 1995. pp. 532:535. Martes populations in California: survey de- Vieira, D. A., E. J. Langendoen, and R. L. Bingner. sign and power analysis. Ecological Applica- 1997. FRAME - A modeling environment tions 6:1254-1267. for watershed processes. In: Proc. Confer- ence on Management of Dis- turbed by Channel Incision, Oxford, Mis- sissippi. pp. 709-715. Walters, C. J. 1986. Adaptive Management of Re- newable Resources. Macmillan, New York. Weatherspoon, P., J. Agee, P. Aune, J. Baldwin, J. Barbour, F. Beall, J. Bentacourt, R. Boerner, M. Busse, C. Edminster, G. Fiddler, C. Fiedler, S. Haase, K. Harksen, M. Harring- ton, R. Hodgson, S. Husari, J. Keeley, M. Landram, B. Laudenslayer, J. McIver, D. Neary, B. Otrosina, R. Ottmar, M. Ritchie, K. Ryan, P. Shea, C. Skinner, S. Stephens, N. Stephenson, E. Kennedy-Sutherland, B. Vihnanck, D. Wade, T. Waldrop, D. Yaussy, A. Youngbolood, and S. Zack. 1999. A National Study of the conse- quences of fire and fire surrogate treat- ments: a proposal submitted to the Joint

Lake Tahoe Watershed Assessment 735