The Science of Instream Flows: A Review of the Texas Instream Flow Program (2005)

The Technical Overview Document (TOD; TPWD, TCEQ, and TWDB, 2003) outlines the methodological aspects of conducting instream flow studies in Texas rivers. The act of drafting this document is acknowledged as formidable because it must simultaneously provide (1) methods that are specific enough to guide technical evaluations, and (2) guidance that is broad enough to be applicable in individual subbasins across the different river systems in Texas. The Texas agencies faced a dilemma in writing the TOD because uniform approaches towards technical methods will be of little value to Texas with its wide range of riverine conditions, but the TOD cannot possibly make methodological prescriptions for every river system in the state. Therefore, the difficult task of preparing the TOD involves finding middle ground between these two options.

This chapter reviews and comments on the technical sections of the TOD and provides recommendations for its improvement. The TOD was evaluated for technical accuracy in the context of the instream flow program. The review of the TOD begins with a brief summary and description of the document’ s contents. Subsequently, a section on the overall findings of the TOD is presented, followed by individual evaluations of the subsections of hydrology and hydraulics; biology; physical processes; water quality; and integration and interpretation in the order they are presented in the original TOD. Implementation aspects are discussed in Chapter 6.

The TOD describes the methods to be used to collect, analyze, and integrate technical information among hydrologic, biologic, physical processes, and water quality aspects of instream flow study. The TOD is a fairly detailed document (74 pages) with more than 2,000 pages of supplemental, highly detailed, technical appendices. The appendices contain information mostly about the water quality programs in Texas, although other topics are

also covered. Appendices, background, and introductory material aside, the TOD has 8 major sections that correspond to the flowchart in the Programmatic Work Plan (PWP; see Figure 4-1): (1) study design; technical evaluations for (2) hydrology and hydraulics, (3) biology, (4) physical processes, and (5) water quality; (6) integration and interpretation; (7) study report production; and (8) monitoring and validation.

The TOD opens with two sections, the Introduction and Ecological Setting, that present important background material that introduces the motivation of the Texas instream flow program and necessary components of an instream flow study. In the context of the state mandate to maintain a “sound ecological environment,” the ecological setting of rivers is described. Biology, hydrology and hydraulics, geomorphology, water quality, and connectivity are defined and introduced as the components of an instream flow study.

Study Design (Section 3) is a short section that identifies the major steps necessary to begin an instream flow study. Basic steps for starting an instream flow study include compiling and evaluating existing information; identifying stakeholders; identifying appropriate study areas; conducting field reconnaissance, or initial technical assessments; preliminary biological and physical surveys; and the development of geographically-specific objectives and study plans. Without much detail, this section lays out the general approach to design an instream flow study in Texas.

By far, the Hydrology and Hydraulics section (Section 4) is the most detailed section of the TOD. In it, technical aspects of hydrologic evaluation are discussed, such as historical, naturalized, and environmental flows and flow duration curves. Examples of types of hydrologic models are mentioned. Aspects of hydraulic modeling relevant to instream flow study are major segments, too, including some guidance on how to select a representative reach and methods for data collection. One- and multiple-dimensional modeling options are detailed. Large woody debris is consid-

ered a special challenge in hydraulics, and is discussed separately in this section.

The Biology section of the TOD (Section 5) outlines with specific methods for conducting baseline surveys and understanding instream habitat. Methods for surveys of instream habitat, fish, riparian systems, and macroinvertebrates are discussed in moderate detail. The TOD Biology section describes how to sample assemblages and measure habitat conditions, calculate habitat suitability criteria, integrate calculations with simulations of aquatic physical habitat, and integrate these calculations with simulated patterns of physical habitat dynamics. Instream and riparian habitat heterogeneity are also discussed in this section.

Physical processes in the TOD (Section 6) refer to hydrogeomorphic riverine processes. Compared to the previous sections, physical processes is notably brief. This section of the TOD presents compact discussions of river classification, assessment of the current status of a river in terms of its geomorphology, and sediment transport processes. Flushing flows and valley, riparian, and channel maintenance physical processes are explained. For this section, the TOD focuses primarily on describing these processes, and only scantly mentions some general methods that can be employed to assess and measure physical processes in an instream flow study.

Water quality is unlike the other technical aspects of instream flow study in Texas because it is regulated at the federal and state levels. There are several well established water quality programs in Texas. The TOD section on water quality (Section 7) describes the state programs and provides relevant background and administrative history of these programs. The section on water quality for instream flow studies (Section 7.3) notes that applying water quality models used in the total maximum daily load (TMDL) and Texas Pollutant Discharge Elimination System (TPDES) programs to the instream flow studies will provide consistency among state programs. Water quality models for instream flow studies, according to

Section 7.3, should take into account spatial and temporal scales; geomorphic and hydraulic conditions of the water body; and the constituents of concern. Sampling or modeling methods for instream flow studies are not presented in the TOD section on water quality.

Findings from the technical evaluations (i.e., biology, physical processes, hydrology and hydraulics, and water quality) will be integrated to develop a flow recommendation. The integration section of the TOD (Section 8) describes the integration process in a framework (Section 8.1; see Figure 5-1) described simply as “the steps needed to develop flow regimes.” It also specifies that a quantitative analysis will be performed to identify critical relationships among the various technical aspects of an instream flow study. Instream habitat is defined as the integration of biology and hydraulics (Section 8.3) and will be predicted using a geographical information system (GIS)-based physical habitat model. The TOD presents ways in which such a model can be used. Habitat time series and habitat duration curves are described as tools for determining flow recommendations (Austin and Wentzel, 2001). The TOD stresses that many combinations of these spatial and temporal analyses can be used to identify target flow regimes. This section very briefly mentions how hydrology, physical processes, and water quality also need to be integrated into a flow regime recommendation. Quantitative Analysis (Section 8.7) includes a combination of statistical, time series, and optimization analyses. The TOD acknowledges that the “precise formulation of the instream flow optimization exercise has yet to be defined or tested,” but presents examples and scenarios in which such analyses could be useful in an instream flow study. Finally, this section of the TOD briefly discusses implementation issues (Section 8.8), but does not mention how flow recommendations will be implemented administratively, scientifically, or in combination with existing Texas water statutes and regulations.

The TOD ends with a very short section that states a study report (Section 9) will be produced and submitted for peer review and the final section, Monitoring and Validation (Section 10), that mentions the importance of monitoring the effectiveness of the implemented flow regime(s). The

FIGURE 5-1 TOD Integration of instream flow study elements.

SOURCE: Adapted from the TOD (TPWD, TCEQ, and TWDB, 2003).

Monitoring and Validation section refers to Texas Commission on Environmental Quality’ s (TCEQ) surface water quality monitoring procedures and lists elements deemed important for a comprehensive monitoring program.

The TOD sets out to prescribe the technical aspects, including methodologies, for conducting the detailed technical evaluations in the Texas instream flow program. This is a difficult charge to meet in a single document that is intended for diverse river subbasins across a large state. The TOD is evaluated in the following sections. The overall strengths of the document are listed first, some overarching opportunities to improve the TOD are presented next, and, finally, opportunities to improve the individual technical sections of hydrology and hydraulics; physical processes; biology; water quality; and integration and interpretation are discussed.

The Texas agencies are commended for drafting a document that has several strengths. The main strength of the TOD is that it encompasses the primary elements of separate technical evaluations relevant to a larger instream flow study. Technical areas of hydrology and hydraulics, physical processes, biology, water quality and connectivity are recognized as important elements and described in the TOD. The TOD also includes initial approaches for integrating results into a flow recommendation. It cannot be overstated how complicated inter-disciplinary instream flow studies can be, and Texas has made a commendable effort in designing its instream flow program to be comprehensive. The biology and hydrology and hydraulics sections reflect a commanding understanding of the relevant issues for instream flow work in Texas rivers. Finally, the TOD represents cooperation among three state agencies with separate missions. Presentations in the TOD reveal the relative expertise of each agency and hint at the promise of these three agencies working together successfully to design and implement a benchmark instream flow program in Texas.

The TOD is composed of several individual pieces that comprise the technical aspects of the Texas instream flow program. The TOD presents each technical piece and the process by which the pieces will be integrated (Figure 5-1) into a flow recommendation. The sharp focus of the TOD is on the distinct, technical pieces of the instream flow study; however the real challenge in instream flow science, and the weakness of the TOD, is the connections among these pieces. Landscape ecology metrics and connectivity can help strengthen these connections.

Chapter 3 outlines seven principles of a state-of-the-art program. The top three principles are to (1) preserve whole functioning ecosystems; (2) mimic, to the extent possible, a natural flow regime; and (3) expand the spatial scope of instream flow studies beyond the river channel to include the riparian corridor and floodplain systems. The whole ecosystems, natural flow regime, and the expanded spatial scale can be viewed as landscape ecology metrics of instream flow science. Used as the focus of the technical evaluations, these metrics can guide the development of instream flow recommendations. Methods by which landscape ecology elements guide an instream flow study are also listed in Chapter 3 as the last four principles: conducting studies using an interdisciplinary approach; using a variety of tools and approaches tailored to the subbasin characteristics; using adaptive

management; and involving stakeholders in the process. Together, these seven principles, viewed as landscape metrics and methods, can be used to set instream flow requirements in a state-of-the-art instream flow program.

Connectivity is defined in the TOD as the “movement and exchange of water, nutrients, sediments, organic matter, and organisms within the riverine ecosystem” (TPWD, TCEQ, and TWDB, 2003). It is discussed in two paragraphs in Section 2, Ecological Setting, but not in the subsequent technical evaluation sections of the TOD. Whereas the TOD defines the concept of connectivity well, it never addresses how connectivity is a part of the Texas instream flow program. The brief, early section on connectivity lays out a nice structure in which technical evaluations could be designed or integrated; unfortunately, connectivity is not revisited in subsequent sections of the TOD. Connectivity reflects important aspects of instream flow science and the TOD should address how connectivity will be used in the detailed technical evaluations.

The dimensions of connectivity occur laterally, longitudinally, vertically, and temporally. These dimensions could be used as organizing axes for developing the conceptual models (see Table 3-2) and designing technical evaluations. Connectivity dimensions also can be used to calibrate the spatial scale of the technical evaluations to ensure compatibility and smooth integration of results. For example, the lateral dimension across a stream channel and associated floodplains could establish the spatial scale for technical evaluations of sediment erosion and deposition, flooding frequency and magnitude, aquatic and riparian species, and variation in water quality. Reconnaissance data could be collected on these assays and entered into a matrix (i.e., Table 3-2) to create a conceptual model that informs the detailed technical evaluations.

Three other overarching findings come from evaluating the TOD. First, for each technical evaluation (i.e., hydrology/hydraulics, physical processes, biology, and water quality), the TOD makes little distinction among individual basins and presents one or a very few approaches that may not be appropriate in all basins and subbasins. Of course, the TOD cannot possibly list all approaches for all possible scenarios in Texas rivers, but it should identify that a range of models, approaches, and tools may be necessary to address the highly variable characteristics of each study subbasin.

Second, considerable inconsistency is found in the level of detail among the technical sections. Some sections have highly detailed methodological processes (hydrology and hydraulics, biologic sampling), but other sections outline only very general sampling methods or none at all (integration, physical processes, and long term monitoring and validation).

Finally, many of the methods presented in the TOD lack context because measurable instream flow goals are not clearly articulated. It is under-standable that the goals for the individual basin studies will vary from basin to basin and all goals cannot be identified in the TOD. Still, the TOD very briefly mentions goals (i.e., biological diversity and biological integrity) and does not discuss the methods in the context of a state-wide program or goals. Without a clearly defined goal statement or process to identify it in the PWP or TOD, the context for these technical studies is unclear.

Overall, the section on hydrology and hydraulics demonstrates a solid understanding of the state-of-the-art in hydrologic and hydraulic methods used in the scientific and engineering community. Compared to other chapters in the TOD, Section 4 (Hydrology and Hydraulics) is quite specific about what tasks will be performed and the tools that will be used. The Texas agencies are commended for the high level of sophistication and detail presented in the hydrologic/hydraulic section of the TOD.

The weakness with the hydrologic/hydraulic TOD material is that the detailed and sophisticated methods presuppose that all techniques are applicable in all Texas basins. The hydrologic/hydraulic TOD section describes specific approaches that may not be necessary in or appropriate for all instream flow studies. A better approach to hydrology and hydraulics is to outline specific methods that could be applied in different circumstances to assure a consistent approach across the state that has enough flexibility to accommodate the variety of river systems within Texas. This improved approach will strengthen the study design phase and reduce the cost of hydrologic/hydraulic sampling and modeling by eliminating unnecessary analyses. Furthermore, connections are not explicit between the hydrologic/hydraulic techniques presented and a sound ecological environment, instream flow study goals, or the other technical disciplines of instream flow study, such as biology or water quality.

The purpose of hydraulic modeling is to define the streamflow characteristics (e.g., depths and velocities) as a function of discharge. As presented in the TOD, results from hydraulic modeling subsequently will be used to assess biological, water quality and physical processes in instream flow systems. The problem in the TOD is that these models and the results from these models are related loosely, if at all, to the other technical elements and studies. It is unclear in the TOD whether the spatial scale of the hydrologic/hydraulic studies coincides with the spatial scales of the biology, physical processes, or water quality empirical studies. A stronger connec-

tion among “the master variable,” hydrology, and the other instream flow technical elements will be very important to ensure that the sophistication of the hydrologic/hydraulic tools, models, and methods is appropriate and efficient for achieving instream flow study goals. Therefore, the TOD should be revised to include explicit connections to the other technical studies to ensure that hydrologic/hydraulic technical assessments are relevant to achieving instream flow study goals, including a sound ecological environment.

The TOD and the PWP mention how hydrology is affected by human uses in the watershed. Several of the supporting documents indicate the profound effect of reservoirs on Texas rivers. Reservoirs are important considerations in Texas, as all major rivers in Texas are dammed for hydropower, municipal, or irrigation purposes. One important element missing from the hydrologic/hydraulic TOD section is a method to relate reservoir operations to instream flows. For some distance below a dam, a river’s hydrology, water quality, substrate, and biota will be greatly affected by the dam’s operation. The TOD also does not discuss how instream flow characteristics may change due to watershed and land use changes, such as increases in urbanization, irrigation, and impervious surface area in the watershed. Water managers have many options to affect instream flows, including dam operation, as well as issuance of water permits to withdraw water from or discharge water to a river. The TOD needs revision to consider approaches for predicting instream flow levels that take into account reservoir operations, permitting, and other watershed land uses.

Some of the two- and three-dimensional models presented in the TOD are highly sophisticated. These models require high quality input data to produce high quality, sensitive and very detailed output data about streamflow characteristics. The hydrologic/hydraulic models presented in the TOD appear too detailed for and therefore misaligned with, some of the other technical studies. For example, aquatic habitat in Texas is classified in the Aquatic Life Use scale as “Exceptional,” “High,” “Intermediate,” or “Limited.” The TOD suggests that results from hydraulic analyses can be used to broadly classify aquatic habitat in these categories. If aquatic habitat is classified in such qualitative terms, then highly quantitative outputs of hydraulic modeling may not be needed for such classification.

Another example of misalignment is with spatial scale. The accuracy of two- and three-dimensional hydraulic models is dependent on the spatial density of the data. These time- and resource-intensive models require suitably accurate input data. Some biological or geomorphic empirical studies will take place over larger spatial areas, such as over the floodplain of a segment or the home range of a key fish species. In these cases, the limited spatial scale of a hydraulic model is too fine to be of use to the geomorphic

or biologic assessments. The methodologies presented in hydrology/hydraulics section of the TOD need better alignment with the other technical aspects of the instream flow studies in terms of model sophistication, sensitivity of model output, and spatial scale.

The authors of TOD Section 4 clearly have a good understanding of hydraulics, methods for gathering hydraulic data, and one- and two-dimensional flow modeling. The remaining challenge is the development of quantitative relationships between hydraulics and specific elements of a sound ecological environment. Given that rather short duration streamflow phenomena can be critical for many aquatic and riparian biota, the TOD appropriately proposes to develop naturalized flow series using daily time steps by disaggregating naturalized monthly flows from the water availability models (WAM) used for water rights permitting in Texas. Nevertheless, the TOD needs revision to make stronger connections among the naturalized flow series and biologic or other aspects of instream flow.

Hydrology is often referred to as the “master variable” in an instream flow context because all the other aspects relate to it. Biology, physical processes, and water quality aspects of instream flow work all can be tied to components of the hydrologic regime. Indeed, the TOD and PWP suggest strongly that the intention of the Texas program is to capitalize on these naturally occurring connections to develop a strong, comprehensive instream flow program for the state. Quantifying streamflow characteristics requires highly technical methods and models, and the hydrology/hydraulics section of the TOD reflects an impressive knowledge of such approaches. Despite its level of detail and sophistication, the hydrology/hydraulics section of the TOD needs significant revision to:

• Include explicit connections to the other technical studies to ensure that hydrologic/hydraulic technical assessments are relevant to achieving instream flow study goals, including a sound ecological environment

• Consider approaches for predicting instream flow levels that take into account reservoir operations, permitting, and other watershed land uses

• Align more closely with the other technical aspects of the instream flow studies in terms of model sophistication, sensitivity of model output, and spatial scale

• Make stronger connections among the naturalized flow series and biologic or other aspects of instream flow

The general strengths of the Texas TOD section on biology include a strong introductory section that provides an excellent overview of the literature and major issues associated with choice of biological response variables and methods of data collection and analysis for instream flow recommendations. The TOD also provides an outstanding general discussion of the important issues of habitat scale, ecological processes, and species life histories. However, the TOD Biology section gives a limited description of the program’s rationale and plans for implementing alternative methods for field sampling, data analysis, and derivation of flow recommendations. The connection between the biological surveys and goals of the instream flow program or of individual studies is not discussed in this section. Program elements related to biology are discussed here in the order that they appear in the TOD.

The TOD baseline information section correctly identifies the starting point for a biology survey that is part of an instream flow study. The TOD highlights steps to take at the beginning of a biology sampling effort: compiling existing information, soliciting stakeholder involvement, and inventorying the types of information that will likely be needed in the biological survey (i.e., life history traits, environmental requirements, species distribution, community composition, and connectivity considerations). The TOD sets forth four types of surveys, or field reconnaissance, to be done in the process of gathering baseline information: instream habitat, fish, macroinvertebrate, and riparian surveys. All except the riparian survey section are described at a decent level of detail; the riparian survey section is brief and lists only the types of information to be considered or collected.

The field reconnaissance measures, as outlined, appear logical and sufficient, and the need for gathering and evaluating baseline information is well defended in the TOD. However, the TOD does not adequately illustrate the availability of specific data sources, the manner in which data will be gathered and analyzed, and how these analyses will influence the design and implementation of specific studies.

The TOD states that “ecological integrity” will be assessed at the reach scale, but the specific metrics for estimating ecological integrity are not identified, with the exception of using the TCEQ standard protocol for determining the appropriate Aquatic Life Use designations of surface waters. The metric developed by Texas Natural Resources Conservation

Commission (TNRCC; now TCEQ) is for statewide application and has limited ability to be tailored for specific conditions of different subbasins.

Indices of biotic integrity (IBIs; Linam et al., 2002) hold the promise of providing fast, cheap, yet comprehensive ecological indicators for long-term monitoring in instream flow studies. Given that Texas is a large state with diverse geography, climate and cultures, IBIs and other aggregate metrics must be customized for different biotic regions, river basins, and, in some cases, river segments. Regionalized IBIs that are applicable in streams of different sizes and biogeographic sub-regions would be immensely useful in the Texas instream flow program. TCEQ has recently adopted regionalized IBIs in its state-wide metric system (D. Mosier, TCEQ, personal communication, 2004) and developed these IBIs to assess water quality goals in wadeable streams. As yet, the regionalized IBIs have not been tested as a means of evaluating modified flow regimes, particularly in large, nonwadeable rivers.

Despite the strengths and promise of regionalized IBIs, they are not appropriate in all settings. Regionalized IBIs require further research and revisions to adapt the metric to apply to a range of stream sizes and orders within each region. Before being used to monitor the effectiveness of the Texas instream flow program, the Texas regionalized IBIs should be evaluated for application to instream flow studies and larger rivers. These evaluations should be published in the open, peer-reviewed scientific literature as a means to validate the Texas approach.

The choice of biotic response indicators and assessment methods should be as standardized (repeatable) as possible at all relevant spatial scales. Explicit data quality assurance/quality control and precise sampling methodologies, taxonomic identification, and quantitative methods are needed such that separate, independent groups of researchers could repeat sampling with comparable results across the state. The Biology section of the TOD should be revised to clarify biotic response indicators and assessment methods of the sampling protocol to be reliable, precise, and related to program objectives.

The TOD refers to the mesohabitat (pools, riffles, runs, rapids, and chutes) as the spatial scale for most of the biological studies and surveys. The method for designating mesohabitats is essentially visual (Vadas and Orth, 1998). Unfortunately, visual criteria and mesohabitat designations may be subjective and applied differently in different river basins. Given the TOD emphasis on mesohabitats and habitat guilds (Leonard and Orth,

1988; Vadas and Orth, 2001), this apparent subjectivity could influence the success of the Texas instream flow program. Objective criteria need to be developed to designate mesohabitats in Texas’ diverse river systems. This fundamental issue needs to be addressed more thoroughly in the TOD.

The TOD discusses fish sampling with specificity, citing seines and electrofishing as primary empirical methods. These methods can be effective in some situations, but the TOD accurately mentions that they have limitations, too. For example, the TOD states that microhabitat utilization data will be collected quantitatively, but seining and electrofishing are unlikely to provide information at this fine spatial scale in any systems except the smallest streams. The TOD does not outline how these specific methods can be standardized across different size streams and various geographic regions across the state. The TOD places too much emphasis on correlational habitat suitability criteria (HSC) approaches involving fishes to the exclusion of other viable approaches and biotic components, and focuses on fish almost exclusively as aquatic fauna.

Some ecological categorizations of fish species cited by the TOD (see Linam et al., 2002; Linam and Kleinsasser, 1998) require further study and revision (e.g., the spiny-cheek sleeper is not an “omnivore,” the Mexican tetra is an “omnivore”, etc.). The TOD states that surveys of mesohabitats are to be conducted when flows are at or below median conditions, but the document leaves unclear how organism—habitat associations will be determined for high flow events. The problem is that a model that projects fish habitat use at median and low-flow conditions likely will not predict habitat use under high flow conditions accurately. Conversely, ecological interactions that would not occur during low-flow conditions (when there is greater habitat segregation) may occur during high flow events. Fish habitat use should be considered under base flow, subsistence flow, high flow pulse, and overbank flow conditions.

The TOD describes three methods for macroinvertebrate surveys in detail: kick nets, woody debris (snag), and hand-picked sampling methods. The rationale for the selection of these methods is not presented. It is implied in the text that these three methods are appropriate for all Texas riv-

ers, when, given the range of river conditions across the states, these methods alone may not be suitable for all Texas rivers.

The TOD provides limited information about selection of biological variables and survey and analysis methods for riparian habitats. The focus on connectivity of off-channel (floodplain) aquatic habitats is appropriate for some, but not all of Texas’ river systems. In comparison to the fish, macroinvertebrate, and instream habitat surveys sections, the riparian survey section and presented methods are very brief. Riparian ecosystems are important components of the river ecosystem and the TOD needs to balance the treatment of riparian surveys with those of aquatic fauna. The TOD should be revised to more strongly emphasize riparian habitats as elements of a sound ecological environment, augment the methods presented for riparian surveys, and present ways to relate riparian sampling results to flow needs necessary to maintain a sound ecological environment in Texas rivers.

The TOD presents instream habitat models (Section 5.3) in two parts, the Quantity and Quality of Microhabitat and Habitat Heterogeneity, both of which model aquatic habitat availability in response to discharge. The TOD section on Quantity and Quality of Microhabitat (Section 5.3.1) details four steps for quantifying and qualifying these habitats: (1) sample assemblages and measure habitat conditions; (2) calculate habitat suitability criteria; (3) integrate criteria with simulations of instream habitat over a range of flows; and (4) develop habitat time series. TOD Section 5.3.2 outlines the model used to determine habitat heterogeneity.

Section 5.3 presents material that is inconsistent with other sections of the TOD. For example, macroinvertebrate sampling presented in this section is different from methods presented in the macroinvertebrate surveys section (Section 5.2.3). Section 5.3 also seems to confuse meso- and microhabitat spatial scales, sampling methods for each spatial scale, and ways to use data collected at each scale. The TOD needs careful revision to clarify inconsistencies in instream flow habitat models.

Proposed methods for calculation of habitat suitability criteria for indicator species and mesohabitat guilds seem to follow currently accepted approaches. The TOD mentions multivariate criteria for combining multiple

variables (e.g., depth, velocity) simultaneously, but these criteria are not discussed. The use of habitat guilds is logical for many of the stream and river systems in Texas. However, it is likely that criteria for identification of habitat guild members will vary from basin to basin, and perhaps even from stream reach to stream reach. The degree to which habitat guild designations will be transferable between spatial units of study is unknown, and the team should consider statistical methods to estimate transferability. A reference system for habitat guilds would be useful to define mesohabitats based on biological criteria, and the derived units could be used for examining species associations in a different study system. The TOD could be revised to include such a reference system or some other means to designate mesohabitat based on biological criteria.

Modeling approaches, variables, and survey methods should be limited to the most relevant parameters consistent with available time and resources. In some cases, the potential severity of ecological risks or the complexity of the ecological setting may demand multiple approaches, some perhaps involving considerable investment of time and resources, to provide sound recommendations. Rather than attempt to apply a diverse set of methodologies to project responses by a diverse set of biological response variables at a diverse set of spatial and temporal scales, the instream flow program should develop consistent study plans using the fewest possible biological response indicators to derive defensible flow recommendations.

The Biology Section of the TOD (Section 5) gives a solid overview of the main biological considerations in an instream flow study. The strengths of the section include a strong general discussion of the important issues of habitat scale, ecological processes, and species life histories. Opportunities for improvement include strengthening connections between the detailed sampling methods and study goals; increasing consistency within the document with respect to spatial scale and sampling methods; and sharpening approaches for conducting biological surveys in dissimilar river systems across Texas. The Biology section of the TOD provides highly detailed accounts of how to conduct some sampling or modeling methods, but gives scant attention to how modeled and empirical data will be communicated, related to program goals, or integrated with other aspects of an instream flow study to derive a flow recommendation. Recommendations for addressing biological issues in the TOD include the following:

• Texas regionalized IBIs should be evaluated for application to instream flow studies and larger rivers; these evaluations should be published in the open, peer-reviewed scientific literature as a means to validate the Texas approach.

• The Biology section of the TOD should be revised to clarify biotic response indicators and assessment methods of the sampling protocol to be reliable, precise, and related to program objectives.

• Objective criteria need to be developed to designate mesohabitats in Texas’ diverse river systems.

• Fish habitat use should be explored under base flow, subsistence flow, high flow pulse, and overbank flow conditions.

• The TOD should be revised to more strongly emphasize riparian habitats as elements of a sound ecological environment, augment the methods presented for riparian surveys, and present ways to relate riparian sampling results to flow needs necessary to maintain a sound ecological environment in Texas rivers.

• The TOD needs careful revision to clarify inconsistencies in instream flow habitat models.

• The degree to which habitat guild designations will be transferable between spatial units of study is unknown, and the Texas instream flow team should consider statistical methods to estimate transferability.

• The instream flow program should develop consistent study plans using the fewest possible biological response indicators to derive defensible flow recommendations.

The physical processes section of the TOD (Section 6) presents riverine physical processes in four main sections: the introduction, classifying a river segment, assessing current conditions in the river, and sediment transport. Strengths of the physical processes section include: recognition of the natural and anthropogenic variability in river system status and processes, presentation of reasonable techniques for specific components of the physical processes evaluations; and acknowledgment for the need of a variety of techniques. Nevertheless, the physical processes section needs significant revision and expansion. After the hydrology component, the channel geometry may be the most important component of an instream flow study, but the TOD gives physical processes very cursory treatment. This section is noticeably shorter and less comprehensive than the hydrology/hydraulics, biology, and water quality technical segments, and it needs

to be augmented to address the important issues associated with physical processes.

One major omission from the TOD is the mention of flood-dominated river regimes in Texas. Texas has a hydrological regime with high frequency of flash floods (Beard 1975). There is a spatial gradient in flash flood potential, high in west Texas and decreasing toward east Texas. Rivers with high flash flood potential may rarely or never achieve equilibrium, since channel morphology, physical habitat and flood features may be substantially rearranged in each flash flood. The geographic and geomorphic variations in flood variability are important considerations in the development of an instream flow program for the state. Different types of hydrological, hydraulic, biological, and physical assessments may be needed for river systems in different portions of the state, and criteria and expectations for instream flow management will need to accommodate dramatic difference in river hydrology across Texas. This is potentially a key factor in physical processes in Texas rivers, but it is never mentioned in the TOD.

Since physical processes vary spatially within a river system, understanding of physical processes is best constructed in a geographic context. The initial assessments should determine whether the channel and its watershed are in a state of dynamic equilibrium, and the TOD should describe processes to determine equilibrium, such as sediment budgets, models, aerial photographs and GIS, as applicable. GIS is a tool that allows efficient storage and viewing of environmental data, helps identify linkages, and improves stakeholder access to and understanding of scientific data collected in the instream flow study. While GIS is mentioned in the TOD, a general framework for GIS and its analytical role is not described. Compilation of data in a geospatial data base structure and use of a GIS to store, display, and analyze data for all parts of the study will improve the quality of the study and also provide documentation for subsequent review, reassessment and adaptive management. A GIS database should be used for data storage and analysis in instream flow studies and the instream flow program at the state level.

Physical processes vary through time. Population growth and land use change are ongoing processes that affect river systems. Changes in climate patterns over the next fifty years may also have significant effects on river discharge patterns and therefore on physical processes (U. S. National Assessment Synthesis Team, 2001). These concepts are not mentioned in the TOD. Instream flow recommendations should take into account trends in watershed and river conditions, probable future human demands on the river system, and probable future climatic change.

Geomorphic classification of river segments and reaches is an important component of the study that will be useful for documenting and analyzing physical processes, for selecting representative reaches and study reaches for instream habitat analysis, and for water quality analyses. Geomorphic classification provides a spatially explicit framework for analyzing physical processes and instream habitat, and a framework for selecting representative reaches for detailed analysis and modeling. Channel morphology includes a number of distinct components, including cross-section form and size, planform, slope and bed morphology (Knighton, 1998), which need to be represented in a geomorphic classification. A variety of geomorphic river classification methods have been developed since the 1980s and have been ably reviewed by Thorne (1997), Montgomery and Buffington (1998), and Kondolf et al. (2003).

The geomorphology of flood-dominated rivers presents enormous challenges for geomorphological classification, assessing the dynamic status of a river based on morphological indicators, and “maintenance flows” for sediment transport. Flood-dominated rivers can radically restructure themselves physically during individual hydrological events. This restructuring can lead to major changes in river form that serve as the basis for classification and dynamic assessment. Moreover, a flood-dominated flow regime can overwhelm attempts to maintain specific substrate and physical habitats through “maintenance” flows for sediment transport. Flash floods have real implications for changing the spatial and temporal structure and connectivity of physical habitat, both instream and in the riparian zone. Physi-ographic and hydrologic setting also relate to river classification. Whether reaches are gaining (water supplied by groundwater sources) or losing (supply water to groundwater sources) water can be critical in determining whether sufficient flows are provided for physical processes as well as aquatic and riparian biological needs. An oversight of the TOD is its silence on these aspects of classifying a river.

The TOD briefly describes and promotes the Rosgen (1996) method to classify streams in Texas. While the Rosgen system is widely used by land management agencies, there is considerable disagreement as to its efficacy (for example, Federal Interagency Stream Restoration Working Group, 1998; Juracek and Fitzpatrick, 2003; Kondolf et al., 2003; Miller and Ritter, 1996). A channel classification system that is hierarchical (in the sense of Bisson and Montgomery, 1996; Frissell et al., 1986) and physically-based (see Kondolf et al., 2003) may be more appropriate for instream flow studies. In addition to classification systems focusing solely on the river channel, geomorphic classification of floodplains (Nanson and Croke, 1992) and

riparian zone classification (NRC, 2002c) may be useful for understanding floodplain equilibrium, channel-floodplain connectivity, and linkages between physical processes and ecological conditions.

Assessing the status of a river reach or segment can be done qualitatively or quantitatively. Qualitative assessment can be based on morphological indicators (see Table 5-1). Morphological indicators appropriate for the study area can be developed and tested with field observations. The initial equilibrium status assessment can be verified by examining historical maps and aerial photos, or with quantitative methods, such as using channel evolution models, calculating bed level changes from gaging station records, analyzing channel width changes from historical maps and aerial photography, and assessing deposition and erosional features (McDowell, 2001; Phillips, 2003; Simon and Castro, 2003; Smelser and Schmidt, 1998). Assessing equilibrium conditions must also take into account the fundamental mode of river dynamics associated with the prevailing flood regime.

The TOD proposes identifying “any recent changes (perhaps within the last 30 to 50 years) that may have occurred to the watershed or channel” as a means to assess the current status of a river reach or segment. Trends in

watershed land use and river conditions are relevant elements to understanding a river’s dynamics, but this guidance is too general to guide consistent, repeatable technical evaluations.

Furthermore, the TOD needs to recognize that such trends can be deceptive in flood-dominated systems where individual events can drastically restructure a river system in addition to changes from on-going trends. The guidance in this section should be made more specific, and identification of the river’s geomorphic equilibrium status should be included as part of the geomorphic classification.

The basic form of a river channel is a direct result of interactions among eight variables: discharge, sediment supply, sediment size, channel width, depth, velocity, slope, and roughness of channel materials (Heede, 1992; Leopold, 1994; Leopold et al., 1964). In an undisturbed watershed, there exists a dynamic equilibrium between sediment loading and the stream’s capacity for sediment transport.

Sediment transport and related hydrogeomorphic processes are discussed in TOD Section 6.4, with some emphasis on valley maintenance, riparian maintenance, channel maintenance and flushing flows. The TOD appropriately underscores the importance of sediment transport and deposition among the physical processes necessary for maintaining a sound ecological environment. The TOD provides a good general discussion of sediment transport processes in Section 6.4, but fails to state how sediment transport analysis might be used in the physical process evaluations or how it relates to a sound ecological environment.

Methods for establishing a general context for sediment and its potential influences on habitat are not described. As part of the physical process evaluations, it may be necessary to quantify the sources, sinks and throughflow of sediment of different sizes—in other words, to do a reconnaissance, semi-quantitative or qualitative sediment budget (see Campbell and Church, 2003; Reid and Trustum, 2002). As land use changes within the contributing watershed, discharge levels needed for channel maintenance flows and flushing flows may also change. In such cases, a sediment budget may be needed to define current levels of sediment flux and predict future levels. Depending on the goals of the study, it may be important to develop sediment budget estimates for appropriate representative time periods, such as historic, pre-dam, agricultural, and post-dam conditions.

The TOD refers to several models that make quantitative predictions about flow and sediment transport (e.g., HEC-6, SED-2D), but model pre-

dictions about how much sediment will be yielded are subject to large errors (Simon and Senturk, 1992). This potential for error should be acknowledged in the TOD to help justify making adjustments to flushing or channel maintenance flow recommendations, if deemed necessary by monitoring and validation efforts.

The TOD recognizes three key discharge levels linked to physical processes that should be evaluated in the physical processes evaluations: floodplain maintenance, flushing flows, and channel maintenance. These terms are broad, perhaps too broad to be useful, and a more detailed breakdown of ecological and management objectives may be more helpful in instream flow studies (Kondolf and Wilcock, 1996). The four-part flow regime (i.e., subsistence flows, base flows, high pulse flows, and overbank flows) is recommended as a structure to link ecological and management objectives, physical processes, and discharge (see Table 3-2). The specific ecological or management objective that a key discharge level is intended to satisfy must be specified prior to analyzing flow requirements. The specific objectives listed in Section 6.4.4 are to restore/enhance riffle habitat; remove superficial fine sediment deposits; and remove interstitial fine sediment from gravel (Kondolf and Wilcock, 1996), but no criteria are presented in the TOD to measure progress towards achieving these objectives. In addition, the biological objectives of flushing flows should be clarified in the TOD. These biological objectives are not discussed in Section 5 (Biology) as was stated in Section 6.4.4.

For determining channel maintenance flows, sediment transport alone may not be adequate. The TOD correctly notes that flow duration, not simply instantaneous peak flow, is important in defining the channel maintenance flow (IFC, 2002). The IFC (2002) describes an empirical approach using suspended sediment rating curves, bedload rating curves, and daily discharge records to compute the channel-maintaining effective discharge (Knighton, 1998). As the TOD points out, establishing suspended and bed load rating curves requires field measurement of sediment transport at a wide range of flows, a labor intensive process that typically takes several years to complete.

An alternative approach to determine maintenance flows suggested in the TOD is to assume that bankfull stage is the minimum channel maintenance flow and occurs once every 1.5 years. This assumption is not safe. There has been a great debate on whether the 1.5-yr flow, bankfull discharge, effective discharge, and channel forming discharge are equal (Knighton, 1998). The 1.5-yr flow is bankfull for many streams, but bankfull discharge frequency can vary from less than 1 year to several tens of years, depending on the river system. Some additional alternative proce-

dures for defining channel maintenance flows are suggested by Kondolf and Wilcock (1996).

Understanding physical processes for instream flow involves consideration of hydrologic regime; channel morphology; processes that form floodplains, channels and physical habitat; sediment transport; historical alteration of the channel and floodplain; and future changes in the watershed. The TOD physical processes section identifies important problems related to geomorphology and some techniques for addressing those problems, but it does not consider the hydrologic regime in geomorphic assessments. Several important elements involved in conducting physical process evaluations are not discussed, such as the import and relevance of hydrologic regime, generally, and flood-dominated systems, specifically; GIS applications; sediment budget considerations; and impacts of changes in land use and population in the watershed over time. The strong spatial gradient in flash-flood potential and, by extension, physical processes, make necessary a wide range of assessments and tools for physical processes. This section needs significant augmentation to address the physical processes along this gradient. Currently, the TOD sets forth a thin, single set of analytical approaches for physical processes in Texas rivers that are unlikely to address the range or complexity of physical processes that exist.

Recommendations for addressing physical process concerns include:

• Augmenting this section to equal in detail the hydrology and hydraulics and biology sections and to discuss Texas hydrologic regimes, GIS applications, sediment budget methods, and impacts of changes in land use, population, and climate in the watershed over time.

• Basing instream flow recommendations on prevailing flood regimes, as well as trends in watershed and river conditions, probable future human demands on the river system, and probable future climatic change.

• Including an assessment of geomorphic equilibrium status and study of historical alterations of the channel and floodplain of the river area under study.

Water quality in the Texas instream flow program is treated differently than the other technical sections of hydrology and hydraulics, biology, and physical processes. Unlike the other technical aspects of instream flow, water quality is regulated by the Clean Water Act at the federal level and a number of well-established water quality programs at the state level. As specified in the statement of task, this report reviews aspects of the instream flow program relevant to the TMDL water quality program and its associated water quality models. Therefore, what follows are two sections: the evaluation of the TOD and an evaluation of the TMDL program and its associated water quality models, specifically QUAL-TX.

The TOD section on water quality for instream flow studies (Section 7) notes that applying water quality models used in TMDL and Texas Pollutant Discharge Elimination System (TPDES) to the instream flow studies will provide consistency among state programs. While using the same models for multiple state programs would indeed provide consistency among them, this approach will work best if the models fulfill the needs of the instream flow goals as well as those of the water quality programs. Current water quality models can be used for discrete aspects of instream flow studies, but no current model exists that can model all water quality elements needed in an instream flow evaluation.

The TOD section on water quality contains a summary of each of the TCEQ water quality programs, and also of the Texas water availability modeling program. Summaries are presented of the Water Quality Standard and Assessment, Surface Water Quality Standards, Texas Water Quality Inventory, TMDL, TPDES, and the Water Rights Permitting and Availability programs. The primary water quality model that Texas uses, QUALTX, is also described. Further detail on the water quality programs is contained in three appendices to the TOD that collectively contain 26 documents and nearly 2,000 pages of material. The strength of this material is that it describes the very comprehensive structure for water quality management that the state has progressively built up over many years. A significant limitation of the TOD water quality section is that it presents a mass of documents without delineating how (1) those documents relate to the instream flow program, (2) the water quality component of an instream flow assessment should be conducted, or (3) instream flow and water quality considerations can be integrated with each other.

The appendix material clearly describes Texas’ existing administrative ways to combine water quality with biology; it is done in through the Aquatic Life Use standards. All of the classified rivers and streams in Texas have a designated level for Aquatic Life Use defined in the Texas Administrative Code. Designations of Aquatic Life Use are “Exceptional,” “High,” “Intermediate,” and “Limited.” Aquatic Life Use designations rely on measurable criteria that establish levels of ecological integrity in classified water segments, including IBIs and some aspects of water quality. As coarse as these classifications are, Texas currently uses them as a systematic method of measuring aspects of a sound ecological environment in streams and rivers. Table 5-2 helps to define the attributes of these aquatic life classifications.

However, there are limitations to using Aquatic Life Use standards in instream flow studies. First, it is unclear whether the IBIs used to determine Aquatic Life Use standards are sensitive to flow variation. That is, if the flow regime changed, it is unclear whether the IBI would respond appreciably. An instream flow program may be better served by more simply defined ecological indicators that are directly related to the flow regime. Second, there are some aspects of a sound river environment that are not covered by the Aquatic Life Use component of water quality management system, such as riparian vegetation and water quality in oxbow lakes.

Finally, the Texas water quality standards account for Aquatic Life Use levels and provide a method for assessing whether these levels are being attained, but these standards do not relate aquatic life to instream flow. The Texas Water Quality Standards empower the TCEQ to protect aquatic life from degradation by pollutants but not from degradation by lack of stream flow.

If “a sound ecological environment” for instream flows differs from “ecological integrity” used in water quality assessment, the TCEQ Commissioners could be faced with two separate sets of requirements for assessing the ecological conditions of Texas streams and rivers. At a minimum, the existing Aquatic Life Use goals should be considered in implementing instream flow recommendations to avoid conflict or even establish support between the water quality and instream flow programs. Integrating the instream flow and water quality programs will provide clearer direction for all parties involved. Streamlining related programs will also reduce the potential for inconsistent or conflicting recommendations among the programs, reduce costs, and eliminate redundant analyses. Therefore, the instream flow program should be integrated with water quality, water permitting and other water-related programs in Texas.

The TMDL program in Texas¹ is the primary mechanism to remedy impairments to water quality. A TMDL is “the total amount of a pollutant a water body can assimilate and still meet state water quality standards” (TNRCC, 1999). TMDL development has come to prominence in recent years because many water bodies are not “swimmable and fishable,” despite significant water quality improvements due to controls on end-of-pipe wastewater discharges. TMDLs include point- and non-point sources, such as pollution from watershed runoff, atmospheric deposition, and contaminated sediments.

The Texas TMDL program relies on water quality models that estimate nutrient, bacterial, and other pollutants in surface waters. QUAL-TX and seven other models are currently used in TMDL studies in Texas. The applicability of these models in an instream flow context, however, is untested.

Of the models used in the TMDL program (QUAL-TX; Mass Balance or CSTR; HSPF; WASP; QUAL2E; SWAT; EPIC; and EFDC), QUAL-TX has been relied upon most heavily and is therefore the focus of this discussion. QUAL-TX is a modification to the federal QUAL2E² model. It has been tailored for Texas river conditions, such as a site specific equation for stream reaeration. QUAL-TX is a steady state model for which the discharge is set at a small value, such as the 7 day 2 year low flow. It is most often used to estimate effects of wastewater discharge on dissolved oxygen (DO) during very low flow conditions. In some river basins, such as the San Antonio Basin, the TCEQ has developed and maintained a suite of QUAL-TX models for segments of the San Antonio River and its principal tributaries to help assess wastewater discharge permit applications.

QUAL-TX is a mainstay of the Texas wastewater discharge permitting process, and it has also been applied in about one third of the TMDL studies undertaken to date by the agency (Table 5-3). However, this model has several limitations when considering instream flow specification. Principal among them is that the model is a static or steady state model, which means

it operates only for a single streamflow discharge, but an instream flow assessment has to consider a whole range of flows that may occur and their time patterns of occurrence. QUAL-TX operates on river or stream segments ¼ mile to 1 mile long. The model yields an average dissolved oxygen value for a river reach that contains many mesohabitat zones and associated aquatic communities. QUAL-TX also accounts for spatial variations in water quality between water in the center of a stream and that along the banks, and the vertical variations in dissolved oxygen content with depth.

Another limitation of QUAL-TX is that it assumes a flat-bed stream, i.e., the bottom area of the stream does not change as the flow approaches zero. A real stream has spatially varying bed topography; therefore higher areas of the stream bed are exposed and become dry as flow diminishes, and lower parts remain submerged longer than would be if the bed were flat. Although some of the other water quality models are dynamic and offer a greater range of possible bed geometries than QUAL-TX, the other models used in the TMDL program still employ spatial computational units of the same order of size as QUAL-TX. It might be possible through field scale research to quantify the spatial and temporal variations of dissolved oxygen within a reach so that with a daily and reach-averaged dissolved oxygen concentration available, some type of “down scaling” process could be applied to infer the spatial and temporal patterns of dissolved oxygen at the meso- and microhabitat scale.

In the four part instream flow regime (subsistence flows, base flow, high pulse flows, and overbank flows) (see Figure 3-2, Table 3-2), QUALTX may be useful in establishing the subsistence flow. In other words, QUAL-TX could estimate the flow needed to maintain minimum water quality standards. Since depressed DO impairs Aquatic Life Use in many streams in Texas, QUAL-TX may be a useful means of examining what

flows are needed to maintain adequate levels of dissolve oxygen during low flow conditions.

However, the water quality component of an instream flow technical evaluation should involve other aspects, as well, including suspended sediment, temperature, and other water borne nutrients and pollutants. The flow regime and the various constituents of water quality act together to produce a sound ecological environment and these aspects need to be considered when defining instream flow requirements for a particular river.

Ideally, a water quality and temperature simulation model for instream flow assessment needs would allow for:

• Time varying hydrology across the full range of flow variation from floods to drought low flows

• The effect of management variations such as alternative strategies for releasing water from reservoirs

• watershed processes for sediment production and nonpoint source pollution generation

• point sources of pollution from wastewater discharges instream processes of chemical transformation and sediment transport

• local scale variations in flow and water quality characteristics within stream mesohabitats and microhabitats

There is no single simulation model currently available which can perform all of these functions. A mechanism is needed to combine hydrologic, water quality and hydrodynamic models across spatial scales to achieve this range of capabilities. The emerging technology of Hydrologic Information Systems is providing some capabilities that could contribute to this goal (Maidment, 2002).

While just as important as hydrology and hydraulics, biology, and physical processes in a Texas instream flow program, water quality is treated differently than its sibling components. Water quality is a regulated entity in Texas and has a well established set of state and federal programs. The TOD ably describes these programs. These programs, administered by TCEQ, meet their purposes of ensuring that surface waters in Texas comply with regulatory standards. Instream flow considerations are not the focus of the state’s water quality programs. Therefore, the instream flow

program’s elements that contend with water quality must be aligned with the existing water quality programs, so as to avoid conflicting requirements for maintaining sound ecological environments in Texas rivers. The TOD presents more than 2,000 pages of water quality material. A significant limitation of the TOD water quality section is that it does not refer to this material or discuss how the water quality component of an instream flow assessment should be conducted, or how instream flow and water quality considerations can be integrated with each other.

The Texas TMDL program’s aim is to improve water quality in Texas surface waters. In total, eight water quality models are used in the TMDL program, with QUAL-TX used more than the seven others. QUAL-TX is a steady state model that models DO. QUAL-TX is applicable to instream flow studies in that (1) DO is an important constituent of water quality that strongly influences aquatic biology; (2) QUAL-TX has an established record of use in Texas, including use with Aquatic Life Use designations; and (3) it operates on the same spatial scale as many biological sampling efforts. The primary limitations of the model include (1) it models DO for only one discharge at a time and instream flow studies require water quality measurements over a range of flows; and (2) DO is only one constituent of water quality, when others, such as suspended sediment, also influence a sound ecological environment. The eight models used in the TMDL program can address discrete pieces of water quality as it relates to instream flow studies, but none can simulate all of the aspects that need to be included in a comprehensive instream flow technical evaluation.

The water quality component of the Texas instream flow program reflects the existing strengths of the Texas water quality management program. These strengths include a comprehensive water quality database for Texas streams and rivers, an established set of water quality standards, and procedures for assessing compliance with them, including standards for aquatic life. The TOD does not define how these existing procedures will need to be adapted or refined for use in an instream flow assessment. A significant limitation of the bioassessment component of the existing water quality program is the lack of a comprehensive database of empirical biological information compared to extent and history of the data maintained by the TCEQ on water quality.

The major findings and recommendations for the water quality component of the TOD are:

• The TOD, with appendices, presents thorough documentation of the Texas water quality programs, but does not outline how this program can be integrated with or used in an instream flow program.

• QUAL-TX is a steady state model that can accurately model DO for a single rate of flow, a limitation for a comprehensive instream flow technical evaluation. However, there is no single simulation model currently available which can model all instream flow functions, and a mechanism is needed to combine hydrologic, water quality and hydrodynamic models across spatial scales.

• A more comprehensive method is needed for storing all the biological and physical data acquired during Aquatic Life Use assessments, and a more complete digital inventory of biological data on the past condition of Texas streams and rivers needs to be compiled.

• The instream flow program should be integrated with water quality, water permitting and other water-related programs in Texas to avoid conflict or establish support between the water quality and instream flow programs.

Integration of the results of the hydrology, biology, water quality and physical process investigations into flow recommendations is critical to the success of any instream flow study. This is a very difficult task because the methods for integration are not well documented (see Chapter 3), and too often, the individual investigations are not designed to be integrated with each other. The Integration and Interpretation section of the TOD (Section 8) presents a process (Figure 5-1) to derive a flow recommendation that uses instream habitat models to integrate hydraulics and biology. The Integration section of the TOD has an in-depth review of quantitative analyses and a very brief section about hydrology, water quality, and physical processes integration.

The TOD describes a vague integration process that is based on several assumptions. It is assumed that relevant aspects of aquatic habitat can be modeled by habitat models. It is also assumed that a standard set of techniques and models will be applied in all river basins. The TOD states that integration is to be accomplished mainly through quantitative analyses, but these analyses are not described in enough detail or in context to guide consistent, repeatable studies in river basins across the state. These assumptions need to be better explained and defended in the TOD to provide much needed support for the integration process presented.

The main weaknesses of the integration section are that (1) it is represented by a complicated Integration Framework (Figure 5-1) that is never thoroughly explained; (2) it does not mention the goals of the study as part

of the integration process; (3) the process of how technical evaluations (Sections 4, 5, 6, and 7) are used to derive flow recommendation is not well described; and (4) the integration of biology and hydraulics is given far more attention than the other technical aspects of water quality, physical processes, and hydrology. While the purpose of integration is to pull all of the elements together, this section of the TOD ultimately stands alone. With few exceptions, the material in this section makes no reference to the technical evaluations of the previous sections of the TOD, and the previous TOD sections do not mention that results from sampling efforts will ultimately be used in the Integration Framework.

That said, the integration phase of any instream flow study is decidedly the most difficult, and methods to integrate several interdisciplinary studies into a single flow recommendation are not well documented in the current literature (IFC, 2002; Postel and Richter, 2003). Examples of possible approaches to Integration are the Building Block or Percent-of-Flow (Flannery et al., 2002) approach (see Chapter 3). The Building Block approach essentially builds a recommended instream flow hydrograph, or set of hydrographs, using key pieces of information developed during technical studies. The percent-of-flow approach uses results from the technical evaluation to determine appropriate levels of allowable flow depletion (typically expressed as percentages of the natural flow) during different times of the year, or during different water year types. These are just two approaches to integrating the various aspects of instream flow studies. Other approaches are being used and developed, but very few are well documented.

Integration should be conceived early in the study design phase to ensure that studies fit together conceptually and are aligned with each other and program and subbasin goals. This type of integration between disciplines may require different models than those models routinely used within disciplines, different sampling methodologies, or sampling at different spatial scales. In these and maybe other ways, the state agencies may have to adopt new approaches to data collection and analyses, since the agencies normal intra-disciplinary practices may not lead to an integrated approach.

Arguably, the Integration Framework (Figure 5-1) which is intended to illustrate “the steps needed to develop flow regimes” is the most critical element of the TOD’s integration process. However, the framework in the TOD is presented very briefly in one short paragraph; the framework figure does not indicate any order of sequence; and the boxes of the framework

contain general topics, not “steps,” towards integrating across several disciplines and technical evaluations. The Integration Framework is very difficult to navigate. Furthermore, the framework figure omits program or study goals, without which, the purpose of and connections among the integration efforts and goals is obscure. In order to be useful, the Integration Framework must be described more thoroughly in the text and/ or revised to articulate the specific steps to be taken or specific points of consideration in the process of developing an instream flow recommendation.

The Instream Habitat sub-section (Section 8.2) describes how GIS-based physical habitat models, hydraulic models, and habitat time series can be used to integrate hydraulics and biology for instream flow purposes. The TOD presents these models as a menu of options that can be used separately or in combination to identify flow regimes in all river basins, but it does not give guidance as to under which circumstances each model is most appropriate. The models are adequately described in terms of what function each model fulfills; however their position(s) in the flowchart and methods used to derive a flow recommendation are not explained. This section states that instream habitat models will be used with output from the hydraulic and biological technical evaluations, but the earlier technical sections of the TOD do not indicate that their results or output will be compatible with these instream habitat models. The instream habitat models focus exclusively on integrating hydraulics with biology, and leaves unclear whether any models can be used to integrate hydrology, water quality, and physical processes.

Aside from a passing reference to statistical and time series analyses, the section on quantitative analyses (Section 8.7) focuses almost exclusively on optimization analyses. Optimization analysis is proposed as a technique to “identify and evaluate alternative flow conditions that maximize, or at least preserve, ecological health” (TPWD, TCEQ, and TWDB, 2003). The goal of optimization is to make a “best” or optimal decision. Optimization has the benefits of being quantitative and leading to a single alternative; however optimization has significant shortcomings as a primary method for reaching an instream flow recommendation. A main shortcoming is that optimization is a mathematical function that cannot easily include broad

ecological, legislative or social goals in its syntax. Furthermore, the type of optimization presented in the TOD “has yet to be defined or tested.”

These four sections, plus a fifth, Other Integration Considerations, are very brief recaps of the important elements that should be included in an integration exercise. These sections are too brief to be useful in guiding integration processes and need significant augmentation.

With little doubt, the integration phase is the most difficult and least documented phase in instream flow science. Most often, the purpose of this phase is to pull together results from different technical evaluations into a single flow recommendation, but the process could be more efficient if integration is conceived in the early study design phase and focuses on common goals or objectives. Various models (GIS-based physical habitat models, hydraulic models, habitat duration curves) and quantitative tools can be helpful to derive a flow recommendation, and some of those tools are introduced in this section. The Integration and Interpretation section of the TOD needs significant revision to:

• correspond more strongly to the methods presented in the biology, hydrology and hydraulics, water quality and physical processes sections of the TOD;

• revise the Integration Framework to include sequential steps and clearer direction of how to combine results from the technical evaluations with appropriate models to derive flow recommendations; and

• augment sections on integrating Hydrology (Section 8.3), Water Quality (Section 8.4), Physical Processes (Section 8.5) and Other Integration Considerations (Section 8.6) to equal in detail and application those presented in Instream Habitat (Section 8.2).

The TOD sets out methods for the technical evaluations of hydrology and hydraulics, biology, physical processes, and water quality in the Texas

instream flow program. The Texas TOD (1) makes little distinction among individual basins and sets forth a standardized set of tools for use in river basins that are highly variable across the state; (2) is inconsistent in the level of detail among the four technical sections; (3) encompasses the primary elements of separate evaluations relevant to a larger, instream flow study, those of hydrology and hydraulics, biology, physical processes and water quality, with tenuous connections among them and vague associations to an instream flow recommendation; and (4) presents methods that lack context because measurable instream flow goals are not clearly articulated.

Therefore, the TOD is recommended to be revised to:

1. strengthen linkages among individual studies on instream biology, hydrology and hydraulics, physical processes, and water quality, and stronger connections between studies and components of flow regime;

2. include greater capacity for and reference to site-specificity at the (sub) basin-scale;

3. design the biological, physical processes water quality, and hydrology and hydraulics instream flow studies at commensurate spatial and temporal scales to improve the ability to integrate findings from the various technical evaluations into a single flow recommendation;

4. strengthen the physical processes section to align more closely with the hydrology and hydraulics and biology sections;

5. clarify methods and the flowchart in the Integration and Interpretation section;

6. describe how connectivity will be used in the Texas instream flow studies;

7. augment the monitoring and validation (i.e., adaptive management) section to monitor progress towards meeting the stated goals; and

8. establish means to set goals for the individual studies that relate to the state-wide definition of a sound ecological environment.

In general, the major findings and recommendations for each technical section are as follows:

1. Hydrologic and hydraulic technical studies reflect a significant understanding of hydrology, hydrologic measurements, and analyses commonly required for performing instream flow studies. The TOD presents highly sophisticated yet standardized hydrologic and hydraulic analyses. Not all models, however, will fit all streams and the analytical approaches should be more closely tailored to the specific objectives of the instream flow study.

2. The physical processes section is notably brief, especially in comparison to the hydrology and hydraulics and biology sections. It omits discussions about Texas hydrologic regimes as they relate to physical processes, GIS applications, sediment budget estimates, and impacts of changes in land use, population, and climate in the watershed over time. The physical processes section needs to be expanded to be comparable to the hydrology and biology sections and include discussions on Texas hydrologic regime, GIS application, sediment budget analyses, and impacts of land use, populations and climate changes in the watershed.

3. Texas regionalized IBIs should be evaluated for application to instream flow studies and larger rivers; these evaluations should be published in the open, peer-reviewed scientific literature as a means to validate the Texas approach.

4. The instream flow program should be integrated with water quality, water permitting and other water-related programs in Texas to avoid conflict between the water quality and instream flow programs.

5. The Integration Framework (TOD Figure 8.1) needs be revised to include sequential steps and clearer description of the proposed process to derive flow recommendations from combining results from the technical evaluations with appropriate models.