5

Overarching Issues

The preceding chapters focus on assessing whether the biological objectives in the Habitat Conservation Plan (HCP) will meet the biological goals and on the effectiveness of minimization and mitigation (M&M) measures to meet the biological objectives. In reality, both the biological goals and objectives are imperfect targets that were defined many years ago during the development of the HCP—that is, without the benefit of the more recent information that has been collected on the system. In addition, there are potential stressors to the system that may be more severe than the Drought of Record on which the HCP was based. Finally, it must be recognized that the ultimate success of the HCP is based upon the protection of the listed species and not the surrogates for that protection, which are the biological goals. As the Edwards Aquifer Authority (EAA) plans for implementation of Phase 2 of the HCP and ultimately a renewal of the incidental take permit, it should begin to consider several overarching issues and concerns that may ultimately suggest improvements to the biological goals and objectives to better protect the listed species.

FOUNTAIN DARTER

Although the habitat-based biological goals for the fountain darter are reasonable because they are easy to measure and quantify (see Chapter 2), the ultimate goal is to ensure that the fountain darter population is sufficiently large to provide a buffer against environmental variation and other possible factors that can affect population abundances. This requires estimates of the total numbers of fountain darters in each system. Vari-

ous reports present fountain darter abundances obtained by multiplying fountain darter densities times the acreage of submerged aquatic vegetation (SAV), and then summing these products over the long-term biological goal (LTBG) and restoration reaches. The Committee is not aware of analyses that support the idea that the target numbers of fountain darters in the LTBG and restoration reaches calculated by this method would reduce the risk of jeopardy or how much this calculated abundance would contribute to recovery. Various estimates of fountain darter population abundance are available from field data (Schenck and Whiteside, 1976; Linam et al., 1993) and used with fountain darter population modeling (Grant et al., 2017). Further exploration of the population abundance of fountain darters, especially with the monitoring data available and commonly used modeling tools, could help determine the viable population abundance.

An approach to examining how well population abundances offer a buffer to variation and can lead to recovery of the species is population viability analysis (PVA) modeling. Modeling analyses to determine viable population estimates and project recovery trajectories in response to restoration are commonly used for well-studied species in biological opinions and are often part of the broader conservation strategy, of which the HCP is part. PVA methods are well documented and relatively easy to apply (Brook et al., 2000; Beissinger and McCullough, 2002; Possingham et al., 1993). The analyses done to date, including the fountain darter ecological modeling (Grant et al., 2017), are considered deterministic approaches that do not explicitly deal with possible underachievement of goals (contingency planning) or stochastic events. In contrast, PVA can directly address the question, How many fountain darter adults are needed for a viable population that is resistant to take, drought, and massive loss of habitat? It would be nice to know that as the EAA fine-tunes the biological goals (as was done recently via the nonroutine adaptive management action), there is some confirmation of the total numbers of fountain darters that are dictated by the current habitat-based goals. The information needed to develop a PVA model for fountain darters is available. Such an effort could use the same approach of teaming local experts with population ecologists as used with the fountain darter ecological model.

For the fountain darter in each system, a challenge for PVA modeling is deriving how flow and habitat influence stage duration (growth), mortality, and reproduction. These three vital rates determine the population growth rate in models commonly used for PVA analyses. Much has been learned about fountain darters through the Applied Research Program, continued monitoring, and responses to extreme events and restoration, and much of the needed information is now available as part of the development of the existing fountain darter ecological model. Whether the information is sufficient to develop the relationships is worth exploring. However, one can

make significant progress by determining how expected habitat and flows would affect these process rates without actually requiring explicit functions in the model; a range of changes in growth, mortality, and reproduction expected from habitat and flow conditions would be considered (see the “implicit approach” in Rose et al., 2015).

The results of a PVA would complement the habitat-based goals of the HCP by providing information on the effectiveness of the habitat-based goals in achieving healthy and sustainable fountain darter populations. Such analyses would allow for determination of which life stages and processes (growth, mortality, reproduction) provide the largest boost to the population, as well as the largest risks of decline. These results can be used to influence the specifics of how habitat and flow objectives can be formulated to be more effective ecologically and more efficient economically.

PVAs for organisms other than fountain darters could be similarly useful in better understanding the success of the species and in identifying data gaps. There are PVA models that can be used for data-limited situations (typical of most of the listed species in the Edwards Aquifer) (see Beissinger and McCullough, 2002) and plant species (Zeigler et al., 2013). Note, however, that PVA is not a common approach to evaluating management of SAV populations because all aquatic plants are clonal in nature, such that recruitment cannot be simply characterized in terms of sexual reproduction. This is not to say that PVA could not be useful in this context, but there is much additional data collection that would be required, such as documentation of seed banks and dispersal and characterization of the role of clonal reproduction versus ramets from seeds. These aspects of recruitment and dispersal were difficult to capture in the SAV modeling that was attempted for these systems because of the lack of data.

SUBMERGED AQUATIC VEGETATION

While the existing M&M measures for SAV were found to be effective (see Chapter 4), there are some issues worth considering as work proceeds and certainly in planning for the next phase of the HCP. The Committee suggests a relaxation of the targets for species-specific areal SAV coverages and a stronger attempt to identify which factors control SAV success. Both of these suggestions might lead to lower overall effort without sacrificing the ultimate goal for fountain darters.

The first issue is the continual maintenance required to reach or hold ground on the specific SAV coverage targets—targets that have been adaptively modified already with clear and substantiated justification. There are two main lines of argument suggesting specific areal targets may not be necessary. First, and most directly relevant to the fountain darter, is the relatively small difference in fountain darter densities across the species of

SAV subject to active management (Figure 5-1). This is also apparent in the statistical analysis described in Appendix K3 of the 2017 HCP Annual Report (Blanton and Associates, 2018). The big differences are between bryophytes, filamentous algae, and the rest of the SAV species. In particular, the species most commonly managed (Ludwigia, Sagittaria, Cabomba, and Potamogeton) are not substantially different from one another, with high variability about the mean. This brings into question the fine-scale, precise management of areal targets that is currently being implemented; indeed, there is even mention (in Appendix L of 2017 HCP Annual Report—Blanton and Associates, 2018) of removal of Ludwigia and Cabomba if the target area is exceeded. Such strict interpretation of the species-specific areal targets may come at the expense of fountain darter habitat maintenance, generally. It might be worth running some modeling scenarios to see

what kind of changes in relative SAV cover are necessary to shift estimated fountain darter abundances. However, continual monitoring will remain necessary because either directional changes in coverage or establishment of new invasive species may lead to undesirable conditions.

The second line of argument to rethinking the SAV targets is to better understand controls on SAV in general and relative species contributions in particular. For example, there are likely only so many potential acres of Ludwigia habitat, and the target for Ludwigia should reflect that potential, but one cannot compute that potential without more information on what the drivers are for Ludwigia. There are many environmental factors and biotic interactions affecting SAV, and at present it is not clear that their relative importance has been worked out. The SAV submodel of the fountain darter ecological model (Grant et al., 2017) represents one effort to synthesize mechanistic understanding of how these factors affect SAV. However, as the Committee previously outlined, there are significant omissions and drawbacks to that modeling effort (NASEM, 2016). For instance, there is repeated mention of light limitation, but this shallow, clear-water system seems unlikely to be light limited in the usual sense. The SAV ecological submodel lays out environmental parameters (PAR, extinction by the water column) showing that about 75 percent of incident light would reach 2-m depth, which even on a not very sunny day (1,000 uE per m²/s) yields light ~ 20 times the half-saturation value used in the model (14 uE per m²/s). The model does include a significant plant-shading effect such that competition for light may significantly shift species growth rates, but this is a normal “sorting” of species rather than some environmental control on initial establishment. Given the fairly small differences in fountain darter abundance across the vegetation types under active management, some shifting in SAV species coverage due to competition for light seems unlikely to greatly harm the fountain darter population.

Flow velocity is another factor often implicated in SAV species habitat selection. Across a reasonable range of modeled discharge conditions there was no evidence of response of SAV cover to variation in discharge (Appendix K4, HCP Annual Report 2017—Blanton and Associates, 2018), but this is likely due to the omission of flow dependency in the model formulations (NASEM, 2016). On the other hand, the SAV report (BIO-WEST and Watershed Systems Group, 2016) suggests that planting success at least partially depends on flow conditions, and so the potential influence of flow seems worthy of further investigation. Nutrient limitation is dismissed as a potential control, even though these are low-phosphorus systems. There is likely to be substantial local knowledge or experience gained by the contractors about what environmental factors control which species of SAV. Should new efforts to improve mechanistic understanding proceed, an opportunity may exist to evaluate SAV species targets with the benefit

of species-specific habitat requirements that would better inform restoration efforts.

Overall, given the effort and cost of maintaining specific coverage targets for SAV, it seems justified to consider a lower level of management with the exception of monitoring for invasive species. If a slightly different blend of coverages leads to an indistinguishable darter population size, then some relaxation seems warranted. In addition, better understanding of controlling factors would lend more confidence and effectiveness to SAV management.

MACROINVERTEBRATE DATA ANALYSIS

Although the macroinvertebrate monitoring program is not formally part of the HCP, a long-term monitoring program has been in place since 2003 (Beaver Creek Hydrology, 2018; Perkin, 2018; Perkin et al., 2018). The Committee applauds the EAA for the addition, continuation, and refinement of the macroinvertebrate monitoring program, as it responds to a recommendation for developing a more holistic ecological understanding of the two ecosystems (NRC, 2015). Multiple aspects of the macroinvertebrate monitoring program provide great potential to the HCP, and the Committee urges the EAA to continue to tap into that potential. First, macroinvertebrate monitoring could serve as a general proxy for the overall ecosystem health of the two spring systems, like that routinely done throughout the United States for wadeable streams (e.g., Barbour et al., 1999; Bonada et al., 2006). Second, the general monitoring of aquatic invertebrates can provide substantial understanding of, and a powerful database on, the complex natural history of the aquifer. Third, comparisons of the general invertebrate community composition and dynamics could be paired statistically with Comal Springs riffle beetle (CSRB) population estimates to provide an evaluation of the cotton-lure sampling approach. Finally, standard ecological community analyses for macroinvertebrates could ultimately serve as a useful surrogate metric for evaluating the overall HCP, and specifically, the efficacy of the M&M measures related to protecting all troglobitic invertebrates in the Edwards Aquifer.

While the macroinvertebrate monitoring program is a positive development, there remain limitations to the current program and analyses. It would seem that given the importance of water quality as exemplified by the biological goal of no more than 10 percent variance from historical conditions, water quality evaluation of associated reaches should accompany macroinvertebrate sampling. Not only would such data be useful in exploring relationships between community composition and structure and water quality, the data could be used to provide quantitative support for the assumption that ≤10 percent deviation in water quality provides

sufficient protection to the covered invertebrates and other troglobitic species. The latter issue may be addressed with the new Texas Commission on Environmental Quality (TCEQ)-based sampling protocol (2014; adopted in 2017), but might also be addressed with the collections associated with the refugia program, since collections were made relatively frequently throughout 2017. Ultimately, a solid temporal and spatial monitoring program, including water quantity and quality measures, will be required. Whether through the refugia collections or via independent monitoring, such a program would prove invaluable in providing a solid link between the LTBGs and the macroinvertebrate communities, including the covered species.

To maximize the utility of the macroinvertebrate monitoring, it also will be important to simultaneously maintain the old and new sampling regimes to provide enough overlap so that the two datasets can be compared for complementarity, and ultimately be combined potentially for a longer-term analysis of population and community dynamics. There will be limitations to combining the two datasets, but the exercise could prove valuable and render the older data useful in assessing the efficacy of the HCP. Finally, we encourage the EAA to bring both the data generated through Applied Research projects and the data being collected under the auspices of the refugia program to bear on better comprehension of the interrelationships between invertebrate population estimates and water quality and other biological variables in the two systems.

INVASIVE SPECIES, EXOTICS, AND DISEASE

The species covered by the HCP are adapted to a stable physical environment and a biotic community with which they have co-evolved. The HCP articulates a broad array of detailed measures to protect covered species (e.g., adequate spring flows and water quality, native vegetation restoration, recreation management), largely by maintaining this stability. While most of the threats that these measures address are generally straightforward to identify, some, such as those posed by potential introductions of nonnative species or diseases, are more nebulous.

Both the Recovery Plan (FWS, 1996) and the HCP note that nonnative species can pose a significant threat to the listed species via competition, habitat modification, or predation, or as vectors for diseases or parasites. Decreased spring flows may exacerbate the problems posed by nonnative species. Thus, the HCP addresses control of nonnative species “to minimize and mitigate the impacts of low flows.”

Current control efforts focused on nonnative species are already present in the Comal and San Marcos systems (Blanton and Associates, 2018). In addition to extensive efforts to remove nonnative vegetation, both the City of New Braunfels and the City of San Marcos devote substantial effort to

monitoring and removal of nonnative fishes—primarily suckermouth catfish (Hypostomus plecostomus), sailfin catfish (Pterygoplichthys disjunctivus), and blue tilapia (Oreochromis aureus)—via spearfishing, bow fishing, gill netting, seining, and other means. Effort is also devoted to monitoring and removal of giant ramshorn snail (Marisa cornuarietis), a nonnative herbivore, and the red-rimmed melania (Melanoides tuberculatus), a nonnative snail that is the first intermediate host of Centrocestus formosanus, the gill parasite that infects fountain darters. There are, however, instances where management of existing, established invasive species may have short-term negative or inconclusive effects on attainment of goals. For instance, large areas of invasive SAV species have been removed from the two systems despite knowledge that these species were habitat for fountain darter (NASEM, 2017). Additionally, removal of exotics from riparian areas may have left exposed soil susceptible to erosion. These examples suggest that management of existing invasive species requires individual consideration rather than blanket actions.

Unidentified future invaders may pose even greater risks than those already present in these systems. The literature is replete with accounts of well-meaning, intentional introductions of nonnative species having serious, and sometimes catastrophic, unanticipated consequences. Such impacts may be even more likely for unintentional introductions. While some species known to prey on fish or other aquatic organisms might be immediately recognized as a potential threat to fountain darters, salamanders, or invertebrates, it would be unwise to consider even apparently benign species as safe. For example, red shiners (Cyprinella lutrensis) have been introduced into many streams in the western United States and often occur at high densities. Although not typically thought of as a piscivore, this small cyprinid will feed on larval fishes, and its predation has been implicated as a major constraint on recruitment of the Colorado pikeminnow (Ptychocheilus lucius), an endangered species endemic to the Colorado River basin (Bestgen et al., 2006).

Species introductions can also introduce diseases or parasites. Although background levels of pathogens are common components of natural ecosystems, some outbreaks, particularly of introduced pathogens (e.g., fish diseases such as largemouth bass virus, whirling disease) can cause major mortality and alterations to community composition. A fungal rust species, Ustilago esculenta, is endemic to Zizania latifolia, a congener of Texas wild rice native to China, and known to infect other plant species where it is introduced. Watson (1991) reported a significant outbreak of Ustilago esculenta on wild rice crops in California. The disease attacks plants at flowering, and it is easily spread on seeds by wind (Terrell, 2007). Inadvertent introduction of this fungal species to the San Marcos system could be catastrophic for Texas wild rice. The slime mold Labyrinthula

zosterae infected populations of the SAV Zostera marina as “wasting disease” and led to regional declines in this species in the 1930s (Muehlstein et al., 1991). Labyrinthula spreads via rhizomes and root structures, making clonal plants particularly susceptible. This vulnerability may extend to other SAV species and fungal infections. The chytrid fungus, Batracho chytrium salamandrivorans, a deadly pathogen that can precipitate severe declines and extinctions of salamander species, has emerged as a major conservation concern in Europe. Though it has not yet been detected in the United States, it would pose a severe threat to San Marcos salamanders and Texas blind salamanders if it were to reach the Edwards Aquifer system. Other environmental stressors (e.g., low flows and declines in water quality) can make organisms more vulnerable to disease.

The HCP is heavily focused on maintaining homeostasis of the unique environments inhabited by the covered species. Refugia populations provide redundancy and genetic representation to reestablish populations in the wild once habitat has been restored after a catastrophic event. But some events, such as introduction and establishment of a high-impact nonnative species, could make these systems permanently uninhabitable for one or more covered species, even if all these suitable habitat conditions are maintained. The opportunity to eradicate an introduced species is often limited to a short period before it becomes abundant or widely distributed. Once a population is well established, it can be difficult or impossible to eliminate, rendering reintroduction of covered species from refugia populations infeasible or ineffective.

It is understandable that HCP efforts have been largely focused on dealing with threats that are clear and present rather than potential threats that may or may not be realized in the future. Nonetheless, the threat posed by potential introduction of a nonnative species needs additional attention, since such an event may pose the greatest risk to long-term viability of covered species in the wild. There is an urgent need to develop and implement a plan for early detection of nonnative species and for rapid response to eradicate them before they become established. Given the intensive sampling and monitoring that occur in both spring systems, formalizing an early detection strategy should not be difficult. The plan for responding to a new invader will need to have contingencies for different types of species. Risk analysis is an established approach for responding to existing or potential invasive events and has formed the basis for many management and policy decisions (e.g., Lodge et al., 2016).

Given that humans are the likely vector for nearly all species introductions (and that species introductions can be vectors for disease and parasite introductions), efforts to educate the public about the potential catastrophic effects of species introductions are critical. Both the City of New Braunfels and the City of San Marcos conduct multipronged educational campaigns

designed to inform the public about the negative impacts of introducing nonnative aquarium and bait species. These efforts include distribution of educational materials on social media websites and via fliers posted in Texas State University dormitories and local pet stores (though not all are willing to participate), outreach at public events, signage at river access points, presentations to school groups and local organizations, and establishment of a pet fish drop-off location in San Marcos to deter aquarium dumps into the river system (Figure 5-2). The City of New Braunfels has established an ordinance prohibiting fishing with live bait, and the City of San Marcos has also prohibited release of fish, plants, or other organisms into waters in its city parks. These efforts are important and should be expanded to frequently and consistently reach all members of the greater community. Given the potentially irreversible nature of most introductions, multiple layers of deterrence are warranted.

CATASTROPHIC EVENTS

The HCP represents a detailed and comprehensive planning process that is focused on meeting the recognized challenges to the listed species and the Comal and San Marcos spring and river systems. Increasingly, however, there is the potential for catastrophic events that are far outside the historical record and could pose unrecognized challenges to listed species and the systems. Although not part of the HCP, these should begin to receive evaluation for possible inclusion in future take permits and HCP planning.

The type of events that might affect the function of the system is illustrated by the behavior of Hurricane Harvey. On August 25, 2017, Hurricane Harvey came ashore between Houston and Corpus Christi and stalled, ultimately turning toward the north and settling over east Texas. More than 40 inches of rain fell in areas around Houston and Beaumont and caused catastrophic flooding and substantial erosion in waterbodies such as the San Jacinto River. What if Harvey had stalled closer to San Antonio and over the Comal and San Marcos rivers? What if climate change increased the frequency and intensity of such events in central Texas? Loss of habitat in these systems due to storm-related erosion was noted in a storm event during October 2015 (Blanton and Associates, 2016). The stormwater control measures being put in place (Chapter 4) would likely reduce the impacts of smaller storms but would be completely overwhelmed by an event such as Hurricane Harvey. Although the volume of water available to fall as rain would likely be reduced because of the area’s inland location, such an event could completely destroy much of the restored SAV in the Comal and San Marcos rivers, directly affecting Texas wild rice and fountain darter habitat. Hutchinson and Foote (2017) show no effects on SAV coverage for either system across a range of discharge up to 450 cfs, although they suggest that minor losses would be experienced at 1,000 cfs, and flows of 4,000 cfs could scour much of the SAV. The October 2015 event exhibited far higher instantaneous flows: 20,900 cfs in the San Marcos and 14,100 cfs in the Comal (Blanton and Associates, 2016), suggesting that the river systems and SAV may be more resilient than Hutchinson and Foote (2017) suggest. A high-rainfall event could also lead to substantial erosion and sedimentation in areas of the rivers, affecting silt-sensitive species. Particularly significant would be riparian bank failure along the western edge of Landa Lake, which could lead to siltation of the spring runs housing the CSRB.

It may not be possible to completely plan for such an event, but evaluating the potential impacts may be useful. Organisms with a substantial portion of their life cycle in the aquifer, such as the Texas blind salamander, may be largely unaffected, whereas fountain darter, SAV, and Texas wild rice populations may be severely affected. The refugia may be the best and perhaps only means of restoring the populations of these species once the

habitat has been restored, although an event that decimates both the river systems and the primary refugia located in the same area cannot be eliminated from consideration.

The MODFLOW model is a potential tool to partially address some scenarios that could occur in the future. Results of the model parameter estimation and uncertainty analysis may inform refinement or bracketing parameters in future scenarios. For example, climate variations may lead to observations that such events are more frequent, last longer, and exhibit greater intensity than is found in the historical record. As mentioned in Chapter 4, asymmetrical droughts across the model domain and variations in recharge and interformational flow could lead to precipitous declines in spring flow. These scenarios could impact whether the triggers for flow protection measures predict the timing of declines. The MODFLOW model could inform adaptive management specifically by evaluating how the four flow protection measures operate in extreme scenarios such as these. Other modeling tools would be needed to evaluate other processes, for example, the effect of extreme events on overland flow, surface water hydrology, sediment transport, and habitat loss.

This discussion of future scenarios is not to suggest that EAA should undertake formal contingency planning or expenditures to build resilience in the face of events such as these (at least without a better understanding of the potential likelihood of their occurrence). However, an examination of how the system might respond to such events may make them easier to address should such events occur.

FINAL THOUGHTS

In keeping with its statement of task, the Committee has largely focused on the biological goals and objectives as identified in the current HCP. Some of the challenges identified in this chapter, however, go beyond the scope of the HCP and may ultimately limit the ability of the HCP to protect the listed species. In the future, as the HCP is revised and renegotiated, the Committee hopes that the new information being collected will allow a more holistic look at the spring and river systems in order to address challenges that the current HCP could not. Each iteration of the HCP offers an opportunity for improvements and to apply lessons learned. Sources of new information include the results of the Applied Research Program, the routine monitoring data and those data collected to support and evaluate the M&M measures, as well as the recommendations of the three previous National Academies reports (NRC, 2015; NASEM, 2016, 2017). It is typical in this type of review process involving several years, three committees, and four reports for recommendations to get lost and for some that the EAA thought they addressed (see implementation reports – EAA,

2015, 2017) to warrant revisiting. As part of the review process for this report, the Committee recommends that the EAA do an end-of-the-review synthesis and review all of the recommendations again. Over time, results and recommendations that may not appear useful or relevant may become so as knowledge is gained about the system.

REFERENCES

Linam, G. W., K. B. Mayes, and K. S. Saunders. 1993. Habitat utilization and population-size estimate of fountain darters, Etheostoma fonticola, in the Comal River, Texas. Texas Journal of Science 45(4):341-348.

Lodge, D. M., P. W. Simonin, S. W. Burgiel, R. P. Keller, J. M. Bossenbroek, C. L. Jerde, A. M. Kramer, E. S. Rutherford, M. A. Barnes, M. E. Wittmann, W. L. Chadderton, J. L. Apriesnig, D. Beletsky, R. M. Cooke, J. M. Drake, S. P. Egan, D. C. Finnoff, C. A. Gantz, E. K. Grey, M. H. Hoff, J. G. Howeth, R. A. Jensen, E. R. Larson, N. E. Mandrak, D. M. Mason, F. A. Martinez, T. J. Newcomb, J. D. Rothlisberger, A. J. Tucker, T. W. Warziniack, and H. Zhang. 2016. Risk analysis and bioeconomics of invasive species to inform policy and management. Annual Review of Environment and Resources 41:453-488.

Muehlstein, L. K., D. Porter, and F. T. Short. 1991. Labyrinthula zosterae sp. nov., the causative agent of wasting disease of eelgrass, Zostera marina. Mycologia 83(2):180-191.

NASEM (National Academies of Sciences, Engineering, and Medicine). 2016. Evaluation of the Predictive Ecological Model for the Edwards Aquifer Habitat Conservation Plan: An Interim Report as Part of Phase 2. Washington, DC: The National Academies Press.

NASEM. 2017. Review of the Edwards Aquifer Habitat Conservation Plan: Report 2. Washington, DC: The National Academies Press.

NRC (National Research Council). 2015. Review of the Edwards Aquifer Habitat Conservation Plan: Report 1. Washington, DC: The National Academies Press.

Perkin, J. 2018. Preliminary Evaluation of Long-Term EAHCP Biomonitoring Data. Presentation to the National Academies Committee to Review the Edwards Aquifer Habitat Conservation Plan. January 4.

Perkin, J. S., E. Kosnicki, and J. Jackson. 2018. Analysis of the Comal Springs and San Marcos Springs Long-Term Monitoring Dataset. Edwards Aquifer Authority.

Possingham, H. P., D. B. Lindenmayer, and T. W. Norton. 1993. A framework for the improved Possingham, H. P., D. B. Lindenmayer, and T. W. Norton. 1993. A framework for the improved management of threatened species based on population viability analysis (PVA). Pacific Conservation Biology 1(1):39-45.

Rose, K. A., S. Sable, D. L. DeAngelis, S. Yurek, J. C. Trexler, W. Graf, and D. J. Reed. 2015. Proposed best modeling practices for assessing the effects of ecosystem restoration on fish. Ecological Modelling 300:12-29.

Schenck, J. R., and B. G. Whiteside. 1976. Distribution, habitat preference and population size estimate of Etheostoma fonticola. Copeia (4):697-703.

TCEQ (Texas Commission on Environmental Quality). 2014. Surface Water Quality Monitoring Procedures, Volume 2: Methods for Collection and Analyzing Biological Assemblage and Habitat Data. RG-416. Water Quality and Planning Division, TCEQ.

Terrell, E. E. 2007. Zizania L. Pp. 47-51 in Flora of North America, vol. 24, M. E. Barkworth, K. M. Capels, S. Long, L. K. Anderton, and M. B. Piep (eds.). New York: Oxford University Press.

Watson, T. 1991. Smut of Manchurian wild rice caused by Ustilago esculenta in California. Plant Disease 75(10):1075D.

Zeigler, S. L., J. P. Che-Castaldo, and M. C. Neel. 2013. Actual and potential use of population viability analyses in recovery of plant species listed under the U.S. Endangered Species Act. Conservation Biology 27:1265-1278.