Technical Appendix:
Compliance of Aguada Regional WWTP
NPDES PR0023736
With Section 301(h) Criteria
Our opposition to EPA’s proposed issuance of a waiver requested under Clean Water Act Section 301(h) by the Puerto Rico Aqueduct and Sewer Authority (PRASA) from the Act’s general requirement for secondary treatment in Section 301(b)(1)(B) is based, in part, on the analysis presented below, as related to the applicable portions of federal law (33 USC 1311(h)) and regulation (40 CFR 125).
This analysis presents the best professional opinions of four Ph.D. marine scientists, with a cumulative total of almost fifty years of experience working on tropical and temperate marine ecosystems, nutrient pollution and other water quality issues, sewage treatment and the ecology of coral reefs, seagrasses and other linked marine systems.
A). Existence and attainment of applicable water quality standards for all pollutants for which relaxation from secondary requirements is requested. (301(h)(1) and 40 CFR 125.61)
Water quality standards certainly exist, either directly or indirectly, for all of the traditional pollutants for which PRASA has requested relaxation: dissolved oxygen (BOD), turbidity (TSS) and pH. However, the water quality standards for turbidity, and those for fecal and total coliform bacteria (and occasionally some toxicants) are certainly and frequently violated in the immediate receiving waters, which are heavily influenced by the Rio Culebrinas. That river and its discharge zone in Bahia de Aguadilla and the Atlantic Ocean are significantly degraded by nonpoint source pollution from the watershed (as made very evident from examination of recent aerial photography [submitted as part of our colleagues’ comments], and as corroborated by commonwealth reports [PRASA, 1987; 1999 and 2000]). This degradation is exacerbated by the subject discharge and by spills and overflows from the sewage collection system for the Aguada plant.
The decision document (US EPA, August 10, 1999) inaccurately characterizes and minimizes the existing water quality problems in the immediate receiving system, citing only one violation of the PR ambient turbidity standard, and making the remarkable claim that that violation is in response to resuspension of bottom sediments. The truth is that the plume of riverine influence which routinely surrounds the discharge point almost certainly is characterized by frequent and serious exceedances of the PR turbidity standard. To the extent that bottom sediments contribute to violation of ambient water quality standards, those sediments arise from the Rio Culebrinas watershed. Direct evidence of this is scattered throughout the record (as presented by our colleagues; examples include: PRASA, 1987 II-C1.39 ["high level of siltation which prevails in the area due to the Rio Culebrinas influence"] and II-C2.10 ["Total suspended solids were as high as 1,800 mg/l in August 1984."]).
A fair and accurate characterization of receiving water quality must occur before EPA can consider options allowable under the circumstances. We believe that such an analysis would demonstrate that not only are large portions of the Bahia de Aguadilla not meeting standards, but that part of these waters should be considered for listing as impaired (and then for remediation) under CWA Sections 303(d) and 305(b). We ask EPA to show cause why these waters should not be listed as impaired. They are certainly "stressed" in the sense of 40 CFR 125.62(f).
In addition, we believe that the high-degree of riverine influence (apparent from aerial photographs examined by us, as well as water quality, sediment and invertebrate data) creates estuarine conditions in large portions of the Bahia de Aguadilla. In fact, the influence of the Rio Culebrinas is undoubtedly the single dominant ecological factor in the immediate discharge zone. The area clearly suffers from periodically reduced salinity, and significant riverine inputs of sediment and other pollutants.
We submit that even fulfilling the criteria under section 301(h) does not release the subject facility from meeting water quality standards and other applicable federal laws and regulations required to be met under Section 301(b)(1)(C). Specifically, the discharge of excessive nitrogen and/or phosphorus into nutrient sensitive coastal waters threatens ecological processes in coral reef and related ecosystems at a variety of scales. While threats to water quality standards for nutrients and/or surrogate standards for plant pigments (e.g. chlorophyll a) might be able to be maintained with large-scale dilution of wastewater, we adamantly maintain that excessive loading of nutrients into these systems generally – and the Western Puerto Rico environment in particular – threatens the ecological integrity of the system, especially essential fish habitats (EFH) and habitat areas of particular concern (HAPCs) designated by the Caribbean Fisheries Management Council (CFMC, 1998). These habitats are required to be protected against federal actions, including the approval of waivers under Section 301(h) (NMFS, 1999).
Specific effects of nutrient overenrichment of tropical marine ecosystems include the induction of excessive growth of phytoplankton, the alteration of phytoplankton community structure, the induction of excessive growth and/or alteration of epibenthic algal communities, and the ecological cascades they induce. Alteration of benthic algal distribution patterns may well have serious implications for invertebrate distribution and abundance patterns, settlement success of fish and invertebrate larvae, and a host of other effects. In addition, direct competition for space among sessile forms like corals, sponges and other live-bottom organisms is a major problem known to be associated with nutrient overenrichment in these systems.
Further, excessive nutrient delivery into shallow-water soft-sediment marine habitats may directly threaten the extent, production and ecological value as nursery grounds and habitats of vascular plant beds, including seagrasses (National Research Council, 1993; National Academy of Science, 2000). While the extent of seagrass beds in the zone of influence is clearly affected by wave energy and by sediment pollution from the Rio Culebrinas, there is nonetheless seagrass habitat in the area (that is, appropriate depth and appropriate substrate type). More careful mapping is needed for seagrasses and potential seagrass habitat.
Moreover, prolonged and excessive turbidity and other standard violations may well interfere with critical larval development processes that occur at both the air-water interface and the sediment-water interface.
Together, these changes threaten the value of this designated essential fish habitat and associated HAPCs.
Not only is primary treatment inappropriate for WWTPs in close conjunction to coral reef and related hard-bottom systems, so is secondary treatment. Full protection of the uses of these waters for the propagation of fish and invertebrates, as required by federal water quality regulations, and protection of essential fish habitat, requires at least tertiary treatment to remove nutrients. Non-discharge options should be given careful consideration. The recent report of the National Academy of Sciences on nutrient impacts in coastal waters specified removal of both nitrogen and phosphorus for tropical coastal systems (NAS, 2000). The anti-degradation policy requires that existing uses be maintained, whether or not waters are classified to protect such uses (40 CFR 125). The growth and maintenance of coral reef ecosystems, the function of EFH and EFH-HAPCs, and the production of fish are existing uses of these water bodies that are not fully protected by the suit of current water quality standards. In any event, protection must be fully achieved before any waiver is considered.
Finally, we echo our colleagues’ concerns about the inadequacy of the protection afforded local citizens caused by the absence of an appropriate surrogate standard for human pathogens. This issue is discussed well in a technical appendix in the National Research Council’s "Managing Wastewater in Coastal Urban Environments" (1993).
B). Maintenance of water quality necessary for balanced indigenous populations of shellfish, fish and wildlife, and protection of water supplies and public recreation. (301(h)(2) and 40 CFR 125.62)
We vehemently disagree with EPA’s determination that adequate evidence has been presented that the subject discharge is compatible with the requirements to protect ecological integrity (including the maintenance of a balanced indigenous population of shellfish, fish and wildlife) and human uses of the receiving waters. In addition to the arguments cited above, we provide the following analysis of the monitoring results from this plant in support of this contention.
1). General deficiencies of the monitoring program
The sweeping conclusions of the decision document (US EPA, 1999) and the quarterly reports from the demonstration studies (PRASA, 1999 and 2000) with respect to the impact on populations in the receiving waters are insupportable, and based largely on flawed monitoring program design and faulty interpretation. The low power of hypothesis tests associated with this monitoring program derive primarily from that inadequate design. Simply put, the lack of apparent impacts comes largely from not looking hard enough. Unfortunately, EPA’s interpretation also fails to provide a prudent and fair characterization of the meager data that exist.
The problems in the design of the monitoring program are considerable (the CWA restriction of monitoring requirements to "those . . . necessary to study the effects of the proposed discharge" [33 USC 1311(h)(3)] notwithstanding). The few reference stations chosen are inadequate to provide the comparisons required, as they are all influenced both by the discharge and – more importantly – by the degrading influence of the Rio Culebrinas. That is, the negative effects of polluted waters from the watershed are built in as baseline effects, which obscure impacts associated with the discharge. All of the far-field stations are significantly different in character, precluding direct comparisons. Moreover, the station arrangement is inadequately replicated, preventing any meaningful comparisons among putatively affected and non-affected stations. Finally, sampling protocols are inadequately replicated to allow meaningful comparisons among stations. Our overall conclusion is that the monitoring program design has inadequate power to allow meaningful assessments; all conclusions from it are at best dubious.
To make matters worse, a number of serious problems exist in data analysis in the various government documents. The PRASA 1987 document and the PRASA quarterly reports all make seemingly rote findings that the results comply with the regulations, even in cases where they clearly do not. The EPA decision document (US EPA, 1999) apes those earlier reports, with no significant further analysis – even when the earlier reports are simply wrong. The biggest problem is one of context: waivers from national standards should be big deals, with serious requirements for resource protection. EPA seems bound and determined to find the plant in compliance with 301(h) requirements, irrespective of the consequences. The documents simply do not honestly reflect the status of what is known about the impacts of the Aguada plant. Specific concerns are provided below.
2). Fish studies
The fish data presented in the various quarterly monitoring reports in no way suggests that a balanced indigenous population of fishes exists in the receiving waters. In fact, the apparent rarity of all fishes, in all studies, would be interpreted to support the stressed and depauperate nature of the fish community in the entire Bahia de Aguadilla. While a regionally depressed fish fauna is conceivable, given the twin impacts of overfishing and habitat impacts of the riverine plume and discharge, the videography from the coral reef areas and hard grounds tell a different story. Reef fishes are abundant (as they were in 1987) on proper habitat. Given that reef fishes often utilize many different habitats during their life histories, it is clear that the distributions of those habitats needs to be mapped, and the impacts of the discharge and the riverine plume assessed. In any event, EPA’s determination related to impacts on fish populations is purely spurious.
In fact, important shallow water fish habitats, critical to early life history stages of managed species and identified as HAPCs by the CFMC exist in the zone of influence of the discharge, although benthic habitat distributions remain – unforgivably – unmapped. Instead, a small array of sites have been tested with low intensity, with sparse results. Habitat maps for much of the Puerto Rico Shelf are currently in the draft stage (Mark Monaco, NOAA, personal communication), which should be obtained, analyzed for EFH potential in the region, and supplemented with detailed bottom mapping. Without such directed studies, the distribution and potential for impacts on EFH and HAPCs cannot reliably be determined. However, evidence exists that HAPCs are present close to the discharge point, which must be mapped and assessed (examples in the record include the "beachrock" listed in 1987 III-C1-39, and the hard bottoms and promontories listed in PRASA 1987, 1999 and 2000). The reports and the decision document both inaccurately under-represent the extent of important fish habitats in the region, and the likely degradation that is and has occurred. That is precisely why EFH consultation with NMFS is required.
The data that have been collected and presented are completely inadequate to support the conclusions made. Simple and effective visual census methods of the type used to census coral cover in the 2000 survey have become routine elsewhere, and are badly needed here. (Obviously, river plumes would impede visual censusing at some sites at some times.) However, trap and trawl results which yield such poor numbers of fish allow no comparative statements to be made. In fact, the discrepancy between the results from the coral videography and the fish trap data confirms the poor quality of the trap data. The 1987 collection yielded a total of 36 fish by trawl at only three stations, and must simply be discarded. (The finding in the 1987 report [III-D3.1] that no impacts were detected, that models suggest no impacts, and that added nutrients might actually benefit fish populations are both spurious and dangerous. In fact, nutrient additions in nutrient poor ecosystems tend to initiate ecological cascades caused by alteration of primary producer patterns, and are typically harmful to fish production.)
The October 1999 sampling netted a total of 43 fish, dominated by Lane snapper (Lutjanus synagris), including many caught on hook and line. Soak times varied; no catch per unit effort numbers are presented. The only statistics presented pertain to largely irrelevant issues (length and weight within a species), but constitute the only variables for which adequate data are available to even conduct a test. The number of stations, the replication among station type, and the replication in samples at a station are not adequate to test the impact of the discharge. The conclusions in the various reports reflect not bona fide "no difference" but design flaws in the monitoring program. (In fact the program should have been based on preliminary sampling that demonstrated the adequacy of data collected, and the a priori design of statistical tests, rather than post hoc data analysis.)
The January 2000 sampling again resulted in sparse data; the capture of a school of 23 tomtates at station A7 doubled the total take. Only eleven fish were taken in traps. These data are meaningless.
A major deficiency in the sampling regime is the failure to sample with appropriate gear, and in appropriate habitats and seasons for recently settled and juvenile fishes. The wire mesh traps and fishing hooks systematic exclude such critically important and sensitive life forms.
All in all, few conclusions are possible from the data presented to date, except that the overall fish community may be generally depleted, may be depauperate in the river-influenced zone, and may be missing certain expected components, either through fishing or through habitat degradation or absence. Even those conclusions are tentative, until a through assessment of fish and fish habitats is conducted. The conclusions in the quarterly reports and the decision document are certainly and emphatically not supported – in fact would be laughed out of any scientific setting!
The conclusion in the decision document that fish are not affected rests on the poor performance of the monitoring program, the putative paucity of fish (at best an indicator of stress, if true), modeled compliance with water quality standards and limited toxicity testing. None of this demonstrates that the discharge as proposed would be fully protective of sensitive early life history stages, including impacts on fish larvae at or near the surface in and near the discharge zone. None of this addresses the high likelihood of ecologically induced changes mediated through nutrient enrichment and resultant trophic cascades. EPA has accepted evidence that is shaky at best, and certainly inadequate to support the overall determination of support of balanced indigenous fish populations.
3). Benthic Invertebrates
The benthic invertebrate data suffer from similar deficiencies, with the difference that invertebrates are much more common overall than fishes, as is always the case. Inadequate reference site selection and replication, low sample replication, and low sampling frequency make the data nearly worthless in terms of testing the hypothesis of no difference induced by the discharge. Benthic community structure is inherently variable, both in time and space; such limited sampling at such a low number of stations guarantees results of limited value. Changes in monitoring protocols also make assessments difficult between the earlier sampling and the more recent.
The determination in the decision document that "benthic community structure seems to have remained similar from 1985 to 1987" (p. 33) is not supported by the sparse data presented, and should be stricken. In addition, the use of diversity indicators to drive conclusions such as "The community indices for the January 1987 survey demonstrate that all of the stations are diverse, indicating a healthy community" (PRASA 1987, II-c1.45) is pure hogwash.
Nonetheless, some generalities are possible. First, the impact of the Rio Culebrinas is very evident in the samples, with alteration of sediment size distributions and organic fractions being clearly reflected in the invertebrate community, with the domination of taxa generally considered to be pollution tolerant taxa such as Mediomastus, Notomastus and other capitellids, and certain amphipods. Similarly, the cluster analysis in 1987 clearly demonstrates the greatest differences among stations with increasing distance from the river plume (but also the discharge point).
Analyses of the distribution of pollution tolerant taxa in the 1987 report make clear that the stations farthest from the discharge (and from the Rio Culebrinas mouth) have the least pollution tolerant presence (stations 1 and 7 and 8). However, this conclusion is tempered by the fact that the degree of pollution tolerance among many of the taxa collected has simply not been determined for the Caribbean. Hence, EPA’s conclusion that "low to moderate abundances of these species do not suggest that pollution stress is occurring" (1999, p. 34) is patently absurd. The species known to be tolerant are in highest and relatively high overall abundance in the river-influenced areas and near the discharge, just what one would expect. The uncertain status of the rest in no way diminishes this finding. The distribution of presumably pollution-sensitive amphipods (minus Corophium, which is generally believed to be pollution tolerant) shows a similar decline inside the river plume and near the discharge – as one would expect.
All of these conclusions clearly support a determination that much of the Bahia de Aguadilla is negatively influenced by sediment pollution from the Rio Culebrinas, and probably exacerbated by the discharge. (The decision document’s use of the term "natural" [1999, p. 34] in describing this sediment pollution effect is unfortunate and incorrect, given EPA’s national determination that sediment pollution represents the major cause of water quality degradation.)
This conclusion is verified through the subsequent quarterly reports (to the extent that such meager sampling can discern any patterns, given the differences in depth and other relevant factors). In January 2000, the farthest field sampling stations had radically different sediment distribution patterns, including gravel, absent in the riverine plume. Station A5 had the most sand at the greatest distance; stations A 7 and A3, the next most at the average plume edges. TKN was also lowest at the widest distance. The pollution tolerant taxa (Notomastus plus the spionids) were dramatically more abundant near the discharge and in the plume.
EPA’s determination of the existence of a "well-balanced" community in the vicinity of the discharge is wishful thinking at best.
4) Corals
The 1987 survey of two coral reef areas in the Bahia de Aguadilla is among the better studies in the series, conducted by Dr. John Ogden. (A similar approach, but updated using underwater videography as with the 1999 report, should be used for benthic habitats and fishes in the rest of the study area.) However, the surveys presented cover only a small portion of the area impacted by the discharge, and cannot stand in for systematic surveys of benthic habitats, which have yet to be conducted.
Subsequent studies have gone to different sites, making comparisons impossible (see PRASA 1999 monitoring report). They are described as being "as close to the discharge as practicable" but also "typical reef communities present in the vicinity of the discharge" (PRASA, 1999, p 6-1). (Note that the maps and verbal descriptions of the station locations do not match up: both are pictured as NE of the discharge site, but Alt. AGS3 is described as being SW.) Our cursory inspection of aerial photographs provided by our colleagues convinces us that other hard-bottom areas exist in the zone of influence. The 1999 report notes that AGS3 is "one of many rock promontories that rise from a mostly sandy mud sediment base at 20 meters to a fairly uniform terrace 12 meters" (1999, p. 6-7). The 1987 report also notes dead corals to southwest of the discharge site, and the increasing erosion of living corals, likely associated with increasing sediment loads from the Rio Culebrinas (1987, p. ii-C2.10). (Note: silty conditions impaired diver visibility in October 1999, verifying the large-scale of riverine influence.)
Coral bleaching was noted in underwater videography, as further evidence of stress in this benthic community. The decision document notes dead coral reefs to the south. The 1987 report also determined that the coral communities of the region were subject to a "state of continued degradation" and identify the proximity to the Rio Culebrinas as the major threat to recovery (PRASA, 1987, II-C2.10). The 1999 surveys of hard substrate areas found high frequency of filamentous algae, which can be an indicator of nutrient enrichment (CFMC, 1998; US CRTF, 1999; and Alina Szmant, UNC-W, personal communication). Overgrowth of dead coral by encrusting organisms does not provide evidence of coral at the edge of its range, as EPA asserts, but most likely damage caused by environmental conditions (US EPA, 1999, p. 37).
These hard substrate communities are clearly in trouble, but have been and remain important fish habitats. Their "poor development" (US EPA, 1999, p. 38) does not diminish their importance to fishes and as EFH and HAPCs. They must be mapped and protected from nutrient discharge and from sedimentation. Simply monitoring them will not be enough.
5) Phytoplankton
We dismiss as ridiculous EPA’s contention that the chlorophyll a data presented provides evidence that phytoplankton communities are not affected by the discharge. No algal community studies have been conducted.
EPA’s overall conclusion that the continued existence of a large (and enlarging) primary discharge is compatible with a "balanced indigenous population" in this area is based on sparse evidence and erroneous interpretation, and is simply ludicrous. We maintain that the record for this decision contains abundant evidence that in fact the receiving waters are already "stressed" (in the meaning of 40 CFR 125.62(f)), and that the discharge will actively exacerbate that condition and impede the recovery of the waters.
6) Human Health Impacts
We have not had the opportunity yet to fully assess the implications of the proposed action on human health. We note the well-known and serious problem with indicators of fecal contamination in tropical marine ecosystems (NRC, 1993). Based on the limited evidence we have seen, and anecdotal evidence from local citizens, we believe that the current collection and treatment system certainly threatens human recreational uses. We endorse our colleague groups’ comments in this regard.
C). Adequacy of monitoring systems. (301(h)(3) and 40 CFR 125.63)
Monitoring plans for this discharge are grossly inadequate; plans associated with the waiver are seriously deficient. We disagree with EPA’s determination in this regard, for the reasons stated above. We observe that adequate monitoring is a serious matter in these complex ecosystems, one that PRASA has not adequately addressed.
D). No additional impacts on other point or non-point sources. (301(h)(4) and 40 CFR 125.64)
We strongly disagree with EPA’s finding that no significant additional restrictions are likely to result from waiver issuance. The discharge is located right in the middle of a seriously degraded river plume, which should be considered for listing and remediation under CWA Sections 303(d) and 305(b). Issuance of the subject waiver will clearly make compliance with water quality standards here much harder, and more costly to nonpoint sources in the watershed. We note, in addition, that reliance on primary treatment would undoubtedly limit the utility of the WWTP to any future industry, and result in significant additional burdens on such industries. We also note that significant non-point source problems in the watershed are certainly exacerbated by failures within the existing sewage collection and treatment system. We believe that failure to address this problem now will guarantee much more draconian measures in the future.
E). Urban area pretreatment programs and toxics control programs. (301 (h)(5-6) and 40 CFR 125.65-66)
We have not reviewed this issue in depth, and offer no comments on this aspect of the proposed waiver.
F). No increase in flows and discharge loads from the plant. (301(h)(7) and 40 CFR 125.67)
We vigorous disagree with EPA’s determination that the issuance of the waiver will not result in the increase in pollution loads or flows. The region’s population is increasing fairly rapidly, and new construction is rampant in and near the service area. The intent is specifically to increase the flow volume to 16 MGD from 8 MGD. We believe that waiver issuance would result in enhanced delivery of pollution loads and volumes into sensitive waters with detrimental effects. Certainly nutrient loads would increase through time, with serious potential consequences. In addition, the issuance of the waiver will actively facilitate further coastal development that will result in increased pollution loading (including the same materials for which the relaxation of secondary standards are requested). The proposed waiver clearly fails this test.
G). Primary or equivalent treatment; water quality criteria met after mixing. (301(h)(9) and 40 CR 125.60)
We echo our colleagues’ comments on the issue of meeting primary standards, and adopt those by reference. In brief, we note violations of minimum treatment levels. More importantly, we reiterate our previous comment that primary treatment is simply inadequate for this facility. A large-scale ocean discharge at this site should only be possible if non-discharge options are shown to be impossible, and if a more complex system of smaller on-site systems is proven infeasible. Even so, such a large-scale discharge is patently ineligible for a primary waiver, and would absolutely require at least tertiary treatment with advanced nutrient removal in order to protect EFH and human health.
References
Caribbean Fisheries Management Council. 1998. Essential Fish Habitat (EFH) generic amendment to the Fishery Management Plans (FMPs) of the U.S. Caribbean including a draft environmental assessment. San Juan, Puerto Rico. 2 vols.
Lindeman, K.C., R. Pugliese, G.T. Waugh and J.S. Ault. 2000. Developmental patterns within a multispecies reef fishery: management applications for essential fish habitats and protected areas. Bull. Mar. Sci. 66(3):929-956.
National Academy of Science. 2000. Clean coastal waters: understanding and reducing the effects of nutrient pollution. National Academy Press, Washington, DC.
National Marine Fisheries Service. 1999. Essential fish habitat: new marine fish habitat conservation mandates for federal agencies. Habitat Conservation Division, Southeast Regional Office.
National Research Council. 1993. Managing wastewater in coastal urban environments. National Academy Press, Washington, DC. 477 pp.
Puerto Rico Aqueduct and Sewer Authority. 1987. Application for a waiver of secondary treatment for the Aguadilla Regional Wastewater Treatment Plant.
PRASA. 1999. First quarterly report for the Aguadilla RWWTP 301(h) waiver demonstration studies. 2 vols.
PRASA. 2000. Second quarterly report for the Aguadilla RWWTP 301(h) waiver demonstration studies. 2 vols.
US Coral Reef Task Force. 1999. Recommendations of the Water and Air Quality Working Group to US Coral Reef Task Force. 42 pp.
US Environmental Protection Agency. 1999. Analysis of the Section 301(h) secondary waiver application for the Aguadilla Wastewater Treatment Plant NPDES No. PR 0023736 Puerto Rico. Decision Document. 54 pp.