- Email: [email protected]
Assessing ecological integrity for impaired waters decisions in Chesapeake Bay, USA
Marine Pollution Bulletin 59 (2009) 48–53
Contents lists available at ScienceDirect
Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul
Assessing ecological integrity for impaired waters decisions in Chesapeake Bay, USA Roberto J. Llansó a,*, Daniel M. Dauer b, Jon H. Vølstad a,c a b c
Versar Inc., Ecological Sciences and Applications, 9200 Rumsey Road, Columbia, Maryland 21045, USA Department of Biological Sciences, Old Dominion University, Norfolk, Virginia 23529, USA Institute of Marine Research, P.O. Box 1870 Nordnes, 5817 Bergen, Norway
a r t i c l e
i n f o
Keywords: Ecological integrity Benthic community condition Impaired waters assessment Biological criteria Chesapeake Bay
a b s t r a c t To meet the requirements of the Clean Water Act, the States of Maryland and Virginia are using benthic biological criteria for identifying impaired waters in Chesapeake Bay and reporting their overall condition. The Chesapeake Bay benthic index of biotic integrity (B-IBI) is the basis for these biological criteria. Working together with the states and the US Environmental Protection Agency, we developed a method for impairment decisions based on the B-IBI. The impaired waters decision approach combines multiple benthic habitat-dependent indices in a Bay segment (equivalent to water bodies in the European Water Framework Directive) with a statistical test of impairment. The method takes into consideration uncertainty in reference conditions, sampling variability, multiple habitats, and sample size. We applied this method to 1430 probability-based benthic samples in 85 Chesapeake Bay segments. Twenty-two segments were considered impaired for benthic community condition. The final decision for each segment considers benthic condition in combination with key stressors such as dissolved oxygen and toxic contaminants. Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction In the United States, the Clean Water Act (CWA) of 1972 (US Code title 33, Sections 1251–1387) directs states to develop water quality standards, and to design and implement a plan to restore waters with impaired water quality. The standards define the water quality needed for designated water uses based on support for aquatic life and societal benefits. If a water body exceeds the limits of pollution allowed by the water quality standards, it does not satisfy the requirements of Section 303(d) of the CWA, and is then listed as not supporting one or more of its designated uses. The water quality standards may include biological criteria, which are defined as ‘‘numerical values or narrative expressions that describe the reference biological integrity of aquatic communities inhabiting waters of a given designated aquatic life use” (USEPA, 1990). Section 305(b) of the CWA requires states to produce a biennial report on the condition of their surface waters (USEPA, 1997). There are several advantages to using biological data in the assessment of water quality and identification of impaired waters. Biological assessment data provide direct measures of aquatic life use support, integrate temporally variable environmental conditions and the effects of multiple types of environmental stress, and can be used to assess the effects of habitat alteration, verify
* Corresponding author. E-mail address: [email protected] (R.J. Llansó). 0025-326X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2008.11.011
impacts of point source discharges, and capture episodic or nonpoint source pollution (Fausch et al., 1990; Diaz, 1992; Barbour et al., 2000; Simon, 2000). Also, biological assessment data can help identify sources of impacts to the aquatic community and assist in establishing total maximum daily loads (TMDLs). A TMDL defines the limit of a pollutant that a water body can receive while still meeting the water quality standards, and can be used to set restoration priorities. Recognizing that there was a wealth of benthic community data being collected by Chesapeake Bay monitoring programs, the States of Maryland and Virginia decided to use benthic biological criteria to identify impaired waters in Chesapeake Bay and report their overall condition. The Chesapeake Bay benthic index of biotic integrity (B-IBI) (Weisberg et al., 1997) is the basis for these biological criteria. This paper outlines a method for deciding whether Chesapeake Bay tidal waters are biologically impaired based on the B-IBI. The method uses reference conditions to identify degradation and designate impairment, but unlike most other biological listing methods (USEPA, 2002), it incorporates uncertainty into the impairment decision. The approach makes full use of the probabilistic design of the monitoring programs. Together with the index of biotic integrity, the probabilistic design allows decisions to be made on the extent of degraded area at multiple spatial scales (Dauer and Llansó, 2003). We applied the method at the level of Bay segments, equivalent to the water bodies of the European Union Water Framework Directive (WFD). Similar to the CWA, the
49
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
WFD establishes a framework for the protection and improvement of all European surface and ground waters (Borja, 2005). The WFD provides guidance for assessing the ecological status of estuarine and coastal waters (EC, 2003), but it leaves to member states the development of specific methodologies, such as those for assessing benthic quality status (Borja et al., 2004). The approach taken here may therefore also be useful for determining benthic impairments in estuaries and coastal waters of Europe or other regions where benthic indices are used to evaluate ecological integrity. 2. Methods
Table 1 Number of reference (degraded and undegraded) samples by habitat in the calibration and validation data sets used in the development of the Chesapeake Bay benthic index of biotic integrity (B-IBI). For salinity and silt-clay characteristics, see Weisberg et al. (1997). Habitat
Degraded
Undegraded
Tidal freshwater Oligohaline Low mesohaline High mesohaline sand High mesohaline mud Polyhaline sand Polyhaline mud
136 92 49 5 81 7 136
75 32 20 14 39 39 24
2.1. Benthic index of biotic integrity The Chesapeake Bay B-IBI is a multi-metric index that evaluates the ecological condition of a sample by comparing key benthic invertebrate community attributes based on species composition, abundance, and biomass to reference values expected under non-degraded conditions in similar habitat types. It is therefore a measure of departure from reference conditions. Reference conditions are estimated from benthic samples taken at sites that satisfy abiotic reference criteria, and are therefore best described as least disturbed conditions (Stoddard et al., 2006). The metrics and thresholds of the B-IBI were developed separately for seven habitat types (tidal freshwater, oligohaline, low mesohaline, high mesohaline sand, high mesohaline mud, polyhaline sand, and polyhaline mud). For each habitat, metrics are scored 1, 3, or 5 and averaged. Although the B-IBI can be regarded as consisting of seven benthic habitat-dependent indices, the scoring system is the same for all habitats and, therefore, B-IBI scores from different habitat types have the same meaning. The B-IBI was described by Weisberg et al. (1997), and further detail can be found in Alden et al. (2002) and Llansó et al. (2003). 2.2. Reference sites The method described here compares B-IBI scores from assessment sites with B-IBI scores from reference sites, and assumes that the reference sites are a representative sample from a ‘‘super population” of all possible reference sites. Assessments are then made at the level of Chesapeake Bay segments and subsegments containing benthic sites. Segments are Chesapeake Bay regions having similar salinity and hydrographic characteristics (USEPA, 2004a). In Virginia, segments were sub-divided into smaller units consisting of the main stem of rivers and bays and the smaller systems draining into the main stem. The reference sites were the calibration and validation sites used in the development of the B-IBI. There were two types of reference sites, undegraded and degraded. Undegraded sites had good dissolved oxygen and did not exceed sediment contaminant, toxicity, and organic carbon criteria. Degraded sites had low dissolved oxygen or exceeded any of the sediment criteria (see Weisberg et al., 1997 for criteria). Undegraded sites were used to develop the B-IBI, and both types of sites were used to test and validate the B-IBI. The term ‘‘reference” is usually applied to the undegraded sites, and reference condition is defined as that of sites determined to be relatively free of anthropogenic stress. In this paper, we define reference condition in the same way, but we broadly distinguish between the undegraded and the degraded reference sites, as both types of sites are used in the present method. The degraded reference sites are assumed to represent the expected biological characteristics of all sites with degraded benthic condition. Table 1 shows the number of reference samples for the seven habitat types.
2.3. Assessment sites The assessment sites were selected within Chesapeake Bay in a stratified random manner, and were sampled in either August or September, 2000 to 2004. For the present study, they were poststratified into segments using a Geographic Information System. A total of 1430 benthic grab samples (one sample per site) were collected, including 750 Maryland benthic monitoring program samples, 500 Virginia benthic monitoring program samples, 150 Elizabeth River biological monitoring program samples, and 10 samples each in Mobjack Bay, the Mattaponi River, and the Nansemond River. Each sample was collected with a Young grab over a sediment surface area of 440 cm2. Samples were sieved on a 0.5mm screen and preserved in the field. The organisms retained on the screen were sorted, enumerated, and identified to the lowest possible taxon. Ash-free dry weight biomass was determined for each taxon.
2.4. Assessment method The assessment method consisted of two steps: first, we calculated for each segment the proportion of sites with low B-IBI scores below a threshold (i.e., the proportion failing), and second, we determined whether this proportion differed from what would be expected from chance alone. That is, if a segment were in good condition (had no low dissolved oxygen, contaminant, toxicity, or organic enrichment problems), we would still expect a small proportion of sites to have low scores because of natural variability; this proportion under the null hypothesis was defined as 5%. Usually, a B-IBI score of <3.0 is characterized as failing the Chesapeake Bay benthic community restoration goals (Ranasinghe et al., 1994). However, because the B-IBI scores for the undegraded and degraded reference sites overlapped in low salinity habitats (Alden et al., 2002), thresholds were set for each of these habitats as the 5th percentile B-IBI score of the undegraded reference sites. For habitats where there was no overlap, the threshold was set as the maximum B-IBI score of the degraded reference sites. The 5th percentile score and its variance were estimated by bootstrap resampling of the undegraded reference sites, with replacement (Efron and Tibshirani, 1998). For each bootstrap iteration (i), the assessment sites were then compared to the thresholds established as above, by habitat (most segments had sites in more than one habitat), and the proportion of sites below threshold (Pi) was determined as the weighted average of proportions across habitats. The final proportion of sites below threshold in a segment (P) was the average over all the iterations. The purpose of these procedures was to incorporate uncertainty in reference conditions and balance Type I and Type II errors. In the second step of the assessment method, segments were declared impaired if P was greater than expected under the null hypothesis (Schenker and Gentleman, 2001):
50
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
P Po > 0 with 95% confidence:
P Po 1:64
pffiffi ðVarP þ VarPo Þ
where Po is the proportion of sites failing under the null hypothesis (5%). Thus, for a segment to be declared impaired, the lower bound of the 95% confidence interval of the proportion of sites in a segment below threshold had to be higher than 5%. To estimate the confidence interval, the variance from the bootstrap resampling (due to estimating thresholds) was added to the binomial variance associated with sampling variability within a segment:
Varp ¼
i¼5000 X i¼1
ðPi PÞ2 þ ðpq=N 1Þ 5000 1
where p and q are parameters in the binomial distribution and N is the number of samples in a segment. 3. Results Eighty-five Chesapeake Bay segments contained benthic data. The number of samples within each segment varied from 1 to 112. Thus, the sample size of some segments was insufficient for reliable impairment determinations. Segments and subsegments with less than 10 samples were considered inconclusive relative to the listing process. Of the segments with at least 10 samples, 22 were considered impaired for benthic community condition and 17 were judged not impaired (Table 2). The proportion of sites below threshold for the impaired segments ranged from 28% to 76%, indicating that at least a quarter of the area of these segments was degraded. For most of these segments, the average B-IBI score was <3.0. Greatest degradation (Table 2) was found in segments with known nutrient (lower Chester and Choptank Rivers), dissolved oxygen (lower Patuxent, Potomac, and Rappahannock Rivers; Maryland main stem), or contaminant problems (Patapsco River estuary, Elizabeth River) (Dauer et al., 2000). The estimates for the main stem segments CB4MH and CB5MH exclude the deep trough (>12 m) of the Chesapeake Bay, which is not monitored by the benthic programs because this area is affected by summer anoxia and consistently has been found to be azoic. 4. Discussion The process used to determine if the segments of the Chesapeake Bay are biologically impaired was based on the B-IBI. The B-IBI is based on biologically defined metric thresholds (Weisberg et al., 1997). An index value of 3 or more (on a 1 to 5 scale) is considered to indicate good benthic community condition indicative of good habitat quality. This quantitative benchmark, however, does not account for uncertainty in reference conditions. In low salinity habitats, overlap in B-IBI score distributions for the degraded and undegraded reference sites produced a class of uncertain condition (Alden et al., 2002). Uncertainty arises from natural stressors that mask the signal from pollution effects (Diaz, 1989), patchiness in the spatial distribution of macroinvertebrates (Dauer, 1993), and small sample size (Vølstad et al., 2003). Our approach is unique among impaired waters methods in that we use confidence limits to account for variability in reference conditions. The use of B-IBI criteria that take into consideration habitat specificity and uncertainty in reference conditions should produce results with fewer Type I errors (calling a segment impaired when it is not). With decreasing sampling size, the 5th percentile of the reference site distribution is less precisely estimated. Therefore, we used a bootstrap routine to get a reliable estimate of this value and its associated confidence interval. The width of that confidence
interval widens as sample size decreases. This variability was added to that of the estimated proportion of sites degraded within a segment to create confidence intervals for impairment decisions. The criterion for impairment decisions was based on statistical significance; however, the extent of degradation in a segment that is considered impaired is a management objective. Decision rules could be established based on specific thresholds of protection that the states may wish to implement. These thresholds could be based on biological or ecological endpoints, or on policy decisions that are based on publicly unacceptable conditions. Such thresholds based on management objectives can clearly be used in our approach, in addition to or in place of a statistical threshold. In an evaluation of water quality assessment methodologies used by various states to determine impairment of surface waters, Keller and Cavallaro (2008) found that listing decisions by the states were often based on insufficient data, or data that were not spatially or temporally representative of the conditions of the water body being assessed. The need for sufficient and adequate data was viewed as one of the most important elements of the listing process. In agreement with this view, a minimum sample size of 10 for impairment assessments in Chesapeake Bay is suggested. This target sample size should incorporate temporal variability in the assessment of benthic condition, and is consistent with the sample size required for assessing non-tidal waters in the Chesapeake Bay watershed (Vølstad et al., 2003). Accounting for temporal variability is important because in Chesapeake Bay benthic condition is strongly associated with oxygen depletion (Llansó et al., 2007), which arises from interactions between biological and physical processes and varies with inter-annual patterns of precipitation and freshwater inflow (Kemp et al., 2005). Consistency among tidal and non-tidal assessments is also important because of the need for integrated reporting that links watershed development to benthic condition across aquatic ecosystems. It might be argued that incorporating uncertainty in reference condition and a required sample size is too cautious given the obligations under the CWA. However, our approach focuses in minimizing error that might occur from the inherent variability of benthic communities, the effects of natural stressors, and sampling and methodological error. We believe that where impairments are common, it is prudent to err on the side of limiting Type I errors. In the regulatory realm, avoiding unnecessary listing and remediation will help focus restoration efforts where they are most needed. This is not to say that the tidal waters of the Chesapeake Bay should not be protected, but rather that impairment designations should be used to initiate effective restoration in critically degraded waters. Our method could be enhanced by setting additional protection thresholds, such as, for example, requiring that one half of the confidence interval be less than 25% for a segment to be designated as not impaired (i.e., supporting aquatic life use). Our approach focuses on the amount (percentage) of area that is impaired, which is a measure of the extent of degradation. This focus differs from that of methods that measure average degradation within an area. Using the average degradation, a segment could conceivably meet criteria and be degraded over more than 50% of its area. Although an impairment decision made on this basis would clearly be undesirable, the percent area approach also has disadvantages. For example, the severity of degradation within a segment could increase significantly without changing the percent area degraded. The B-IBI scores in a segment, however, can be inspected to determine magnitude of degradation. Within the WFD, ecological integrity is assessed by the ecological quality ratio (EQR), which compares biological index values in a water body to those of reference conditions (Borja et al., 2004). EQR derivation methods vary by member state, from a simple ratio of two numbers (Rosenberg et al., 2004) to more complex methods involving multivariate analysis (Muxika et al., 2007). Classifications of
51
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
Table 2 Impaired waters results for 85 Chesapeake Bay segments and subsegments for the period 2000–2004. Impaired segments were sorted according to the lower 95% bound of the confidence interval of the difference between the proportion of sites in the segment with B-IBI scores below threshold (P) and the proportion of sites failing under the null hypothesis (Po). The mean B-IBI value in each segment is shown for comparison. CLL = lower 95% confidence limit bound. CLU = upper 95% confidence limit bound. * = Inconclusive; sample size insufficient to make a reliable determination. State of Virginia’s subsegments are identified by lower case letters. Segment
Water body name
Sample size
P
Po
P Po
CL–L (P Po)
CLU (P Po)
Impaired
POTMH SBEMHa CB4MH PATMH PAXMH CHSMH EBEMHa ELIMHa CB3MH RPPMHa YRKMHa CHOMH2 JMSMHa JMSMHb YRKPHa CB5MH PMKOHa MAGMH ELIPHa MOBPHa WBEMHa JMSOHa LAFMHa CB7PHa TANMH SEVMH GUNOH MPNOHa POTOH CB6PHa CB2OH CB1TF CB8PHa MPNTFa MANMH POTTF RPPTFa JMSTFa JMSPHa WICMH POCMH BSHOH NANMH SOUMH CHOMH1 CRRMHa ELKOH POCOH RHDMH PAXOH LCHMH CHSOH PAXTF HNGMH CHOOH RPPOHa BIGMH CHKOHa EASMH BACOH NANOH FSBMH NORTF PMKTFa JMSPHd JMSMHd JMSMHc MIDOH SASOH RPPMHm MOBPHh MOBPHf
Potomac River mesohaline Southern Branch of Elizabeth River Maryland main stem Patapsco River Patuxent River mesohaline Chester River mesohaline Eastern Branch of Elizabeth River Elizabeth River main stem Maryland main stem Rappahannock River mesohaline York River mesohaline Choptank River mesohaline James River mesohaline Nansemond River York River polyhaline Maryland main stem Pamunkey River oligohaline Magothy River Elizabeth River main stem Mobjack Bay Western Branch of Elizabeth River James River oligohaline Lafayette River Virginia main stem Tangier Sound Severn River Gunpowder River Mattaponi River oligohaline Potomac River oligohaline Virginia main stem Maryland upper main stem Maryland upper main stem Virginia main stem Mattaponi River tidal fresh Manokin River Potomac River tidal fresh Rappahannock River tidal fresh James River tidal fresh James River polyhaline Wicomico River Pocomoke Sound Bush River Nanticoke River mesohaline South River Choptank River Corrotoman River Elk River Pocomoke River oligohaline Rhode River Patuxent River oligohaline Little Choptank River Chester River oligohaline Patuxent River tidal fresh Honga River Choptank River oligohaline Rappahannock River oligohaline Big Annemessex River Chickahominy River Eastern Bay Back River Nanticoke River oligohaline Fishing Bay Northeast River Pamunkey River tidal fresh Willoughby Bay Warwick River Chuckatuck River Middle River Sassafras River Totuskey Creek East River Ware River
91 47 28 49 112 33 15 37 61 98 64 22 46 16 29 44 11 17 17 20 19 22 27 43 48 13 15 11 21 18 40 19 15 13 13 12 11 14 10 9 9 9 9 8 8 8 8 7 7 7 6 6 6 5 5 5 5 5 4 4 4 4 4 4 3 3 3 3 3 2 2 1
0.76 0.70 0.67 0.52 0.49 0.53 0.57 0.48 0.44 0.37 0.43 0.41 0.37 0.45 0.38 0.32 0.46 0.41 0.39 0.36 0.36 0.28 0.31 0.15 0.13 0.26 0.22 0.25 0.16 0.15 0.00 0.10 0.00 0.00 0.10 0.08 0.07 0.02 0.12 0.42 0.29 0.24 0.13 0.88 0.38 0.23 0.15 0.49 0.43 0.40 0.31 0.17 0.17 0.21 0.21 0.06 0.03 0.00 0.75 0.44 0.25 0.25 0.08 0.00 0.91 0.34 0.33 0.00 0.00 0.50 0.00 1.00
0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05
0.71 0.65 0.62 0.47 0.44 0.48 0.52 0.43 0.39 0.32 0.38 0.36 0.32 0.40 0.33 0.27 0.41 0.36 0.34 0.31 0.31 0.23 0.26 0.10 0.08 0.21 0.17 0.20 0.11 0.10 0.05 0.05 0.05 0.05 0.05 0.03 0.02 0.03 0.07 0.37 0.24 0.19 0.08 0.83 0.33 0.18 0.10 0.44 0.38 0.35 0.26 0.12 0.12 0.16 0.16 0.01 0.02 0.05 0.70 0.39 0.20 0.20 0.03 0.05 0.86 0.29 0.28 0.05 0.05 0.45 0.05 0.95
0.62 0.51 0.45 0.33 0.33 0.32 0.27 0.26 0.26 0.21 0.18 0.17 0.17 0.16 0.15 0.14 0.12 0.12 0.10 0.10 0.08 0.05 0.01 0.01 0.03 0.04 0.05 0.06 0.06 0.09 0.11 0.14 0.15 0.15 0.16 0.17 0.17 0.18 0.29 – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
0.80 0.78 0.79 0.61 0.55 0.65 0.77 0.60 0.52 0.43 0.58 0.56 0.48 0.65 0.52 0.40 0.69 0.59 0.58 0.52 0.55 0.41 0.52 0.21 0.18 0.45 0.39 0.47 0.28 0.28 0.01 0.23 0.05 0.05 0.25 0.22 0.20 0.12 0.43 – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No No No No No No No No No No No No No No * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
Mean B-IBI
1.7 2.0 2.3 2.4 2.4 2.6 2.2 2.5 2.7 2.6 2.5 2.9 2.7 2.4 3.0 2.7 2.6 2.3 2.8 3.0 2.4 2.9 2.4 3.3 3.2 2.7 2.9 2.6 3.4 3.3 3.8 3.1 3.4 3.5 3.1 3.1 3.5 3.2 3.4 2.8 2.6 2.6 3.0 2.1 2.6 2.4 3.2 2.5 2.9 2.8 2.5 3.2 2.6 2.8 2.7 3.5 2.9 3.7 2.1 2.1 3.5 3.6 3.2 3.9 1.7 2.8 3.1 3.4 3.2 3.1 2.7 1.3 (continued on next page)
52
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
Table 2 (continued) Segment
Water body name
Sample size
P
Po
P Po
CL–L (P Po)
CLU (P Po)
Impaired
Mean B-IBI
MOBPHg RPPMHd WSTMH YRKPHd MATTF YRKMHb YRKPHe CHSTF CHOTF APPTFa BOHOH MOBPHe POCTF
North River Robinson Creek West River Sarah Creek Mattawoman Creek Queen Creek Timberneck Creek Chester River tidal fresh Choptank River tidal fresh Appomattox River Bohemia River Severn Creek Pocomoke River tidal fresh
1 1 1 1 1 1 1 1 1 1 1 1 1
1.00 1.00 1.00 1.00 0.73 0.73 0.37 0.19 0.00 0.00 0.00 0.00 0.00
0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05
0.95 0.95 0.95 0.95 0.68 0.68 0.32 0.14 0.05 0.05 0.05 0.05 0.05
– – – – – – – – – – – – –
– – – – – – – – – – – – –
* * * * * * * * * * * * *
1.7 1.7 2.2 1.3 1.7 1.7 2.7 2.0 3.0 3.0 4.0 2.7 2.5
ecological status for single sites are produced, but procedures that integrate these results into a decision based approach are generally lacking. Water quality assessments under the CWA are based on the ability of the waters to support designated uses. Support is defined as meeting the criteria for each use. Once water bodies have been assessed, they are classified into 3 categories: fully supporting, inconclusive, and impaired. If a water body is classified as impaired, a stressor identification procedure is used to determine if a TMDL is necessary. State and federal agencies with jurisdiction over the Chesapeake Bay have established criteria for dissolved oxygen (DO), contaminants, water clarity, and chlorophyll for specific uses, such as open water and shallow water submerged aquatic vegetation (USEPA, 2003, 2004b). At present, the B-IBI is evaluated independently of these parameters; i.e., benthic community impairment is an independent and sufficient cause for listing. However, decision models in which the B-IBI is corroborative are being considered by the State of Maryland. In the decision process illustrated in Fig. 1, the final decision for each segment considers benthic condition in combination with key stressors such as DO
and toxic contaminants. In this process, TMDL development would be required when there is a likely cause for the benthic community impairment. Similar decision processes have already been developed for non-tidal waters in the Chesapeake Bay region (MDE, 2008). Our assessment of Chesapeake Bay segments produced impairment classifications that were generally consistent with expectations based on known patterns of human disturbance in the watershed (Dauer et al., 2000). We take this as confirmation of the general utility of the method in estuarine management decisions, and to recommend further study of the novel features of the method. Because benthic condition indices are used to evaluate estuarine ecological integrity in both the US and Europe, water quality decision procedures incorporating this approach may be widely applicable. Acknowledgements We greatly appreciate the computer programming support from Ed Weber (Versar Inc.). This paper benefited from reviews
NO START
Data sufficient?
Score sample
YES
YES
YES
Is segment impaired for DO numeric criteria?
Aquatic life fails Cause: DO B-IBI corroborative
Develop TMDL to correct low DO DO corrected
Aquatic life unknown
Test segment
Is segment degraded for B-IBI?
NO Evaluate B-IBI for other stressors
YES
Other stressors identified?
NO
Insufficient data
?
Does segment meet WQ criteria?
Additional monitoring information needed
NO Aquatic life fails Cause: DO, etc.
YES NO
Aquatic life supported
Aquatic life fails Cause: Pollutants B-IBI corroborative
Develop TMDL to correct pollutants
Aquatic life fails Cause: Pollution Unknown source No TMDL required
Pollutants corrected
Fig. 1. Proposed decision process for assessing impaired waters in Chesapeake Bay.
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
by Leska Fore (Statistical Design Inc.), Jeroen Gerritsen and Lei Zheng (Tetra Tech Inc.), and Charles Hawkins (Utah State University) on an earlier report submitted to the state agencies. The comments of Don Strebel and Mark Southerland (Versar Inc.), and two anonymous reviewers helped improve the paper. References Alden III, R.A., Dauer, D.M., Ranasinghe, J.A., Scott, L.C., Llansó, R.J., 2002. Statistical verification of the Chesapeake Bay benthic index of biotic integrity. Environmetrics 13, 473–498. Barbour, M.T., Swietlik, W.F., Jackson, S.K., Courtemanch, D.L., Davies, S.P., Yoder, C.O., 2000. Measuring the attainment of biological integrity in the USA: a critical element of ecological integrity. Hydrobiologia 422 (423), 453–464. Borja, A., 2005. The European Water Framework Directive: a challenge for nearshore, coastal and continental Shelf Research. Continental Shelf Research 25, 1768–1783. Borja, A., Franco, J., Valencia, V., Bald, J., Muxica, I., Belzunce, M.J., Solaun, O., 2004. Implementation of the European water framework directive from the Basque country (northern Spain): a methodological approach. Marine Pollution Bulletin 48, 209–218. Dauer, D.M., 1993. Biological criteria, environmental health and estuarine macrobenthic community structure. Marine Pollution Bulletin 26, 249–257. Dauer, D.M., Llansó, R.J., 2003. Spatial scales and probability based sampling in determining levels of benthic community degradation in the Chesapeake Bay. Environmental Monitoring and Assessment 81, 175–186. Dauer, D.M., Ranasinghe, J.A., Weisberg, S.B., 2000. Relationships between benthic community condition, water quality, sediment quality, nutrient loads, and land use patterns in Chesapeake Bay. Estuaries 23, 80–96. Diaz, R.J., 1989. Pollution and tidal benthic communities of the James River Estuary, Virginia. Hydrobiologia 180, 195–211. Diaz, R.J., 1992. Ecosystem assessment using estuarine and marine benthic community structure. In: Burton, A. (Ed.), Contaminated Sediment Toxicity Assessment. Lewis Publishers, Boca Raton, Florida, pp. 67–85. EC, 2003. Common Implementation Strategy for the Water Framework Directive (2000/60/EC), Guidance Document No. 5, Transitional and Coastal Waters – Typology, Reference Conditions and Classification Systems. Directorate General Environment of the European Commission, Brussels, Belgium. Efron, B., Tibshirani, R.J., 1998. An Introduction to the Bootstrap. CRC Press, Boca Raton, Florida. Fausch, K.D., Lyons, J., Karr, J.R., Angermeier, P.L., 1990. Fish communities as indicators of environmental degradation. American Fisheries Society Symposium 8, 123–144. Keller, A.A., Cavallaro, L., 2008. Assessing the US Clean Water Act 303(d) listing process for determining impairment of a waterbody. Journal of Environmental Management 86, 699–711. Kemp, W.M., Boynton, W.R., Adolf, J.E., Boesch, D.F., Boicourt, W.C., Brush, G., Cornwell, J.C., Fisher, T.R., Glibert, P.M., Hagy, J.D., Harding, L.W., Houde, E.D., Kimmel, D.G., Miller, W.D., Newell, R.I.E., Roman, M.R., Smith, E.M., Stevenson, J.C., 2005. Eutrophication of Chesapeake Bay: historical trends and ecological interactions. Marine Ecology Progress Series 303, 1–29. Llansó, R.J., Dauer, D.M., Vølstad, J.H., Scott, L.C., 2003. Application of the benthic index of biotic integrity to environmental monitoring in Chesapeake Bay. Environmental Monitoring and Assessment 81, 163–174.
53
Llansó, R.J., Dew, J., Scott, L.C., 2007. Chesapeake Bay Water Quality Monitoring Program: long-term Benthic Monitoring and Assessment Component, Level 1 Comprehensive Report (July 1984-December 2006). Report prepared for the Maryland Department of Natural Resources by Versar Inc., Columbia, Maryland. MDE, 2008. Part G: appendix II – Listing Methodologies. G.1 Biological Assessment of Water Quality for Non-Tidal Streams. Maryland Department of the Environment, Baltimore, Maryland. Muxika, I., Borja, A., Bald, J., 2007. Using historical data, expert judgement and multivariate analysis in assessing reference conditions and benthic ecology status, according to the European Water Framework Directive. Marine Pollution Bulletin 55, 16–29. Ranasinghe, J.A., Weisberg, S.B., Dauer, D.M., Schaffner, L.C., Diaz, R.J., Frithsen, J.B., 1994. Chesapeake Bay Benthic Community Restoration Goals. Report prepared for the USEPA Chesapeake Bay Program Office, the Governor’s Council on Chesapeake Bay Research Fund, and the Maryland Department of Natural Resources by Versar Inc., Columbia, Maryland. Rosenberg, R., Blomqvist, M., Nilsson, H.C., Cederwall, H., Dimming, A., 2004. Marine quality assessment by use of benthic species-abundance distributions: a proposed new protocol within the European Water Framework Directive. Marine Pollution Bulletin 49, 728–739. Schenker, N., Gentleman, J.F., 2001. On judging the significance of differences by examining the overlap between confidence intervals. The American Statistician 55, 182–186. Simon, T.P., 2000. The use of biological criteria as a tool for water resource management. Environmental Science and Policy 3 (Suppl. 1), 43–49. Stoddard, J.L., Larsen, D.P., Hawkins, C.P., Johnson, R.K., Norris, R.H., 2006. Setting expectations for the ecological condition of streams: the concept of reference condition. Ecological Applications 16, 1267–1276. USEPA, 1990. Biological Criteria: National Program Guidance for Surface Waters. EPA 440/5-90-004, April 1990. US Environmental Protection Agency, Office of Water, Washington D.C. USEPA, 1997. Guidelines for the Preparation of the Comprehensive State Water Quality Assessments (305[b] reports). EPA 841-B-97-002A. US Environmental Protection Agency, Office of Water, Washington D.C. USEPA, 2002. Summary of Biological Assessment Programs and Biocriteria Development for States, Tribes, Territories, and Interstate Commissions: Streams and Wadeable Rivers. EPA 822-R-02-048. US Environmental Protection Agency, Washington, D.C. USEPA, 2003. Ambient Water Quality Criteria for Dissolved Oxygen, Water Clarity and Chlorophyll a for the Chesapeake Bay and Its Tidal Tributaries. EPA 903-R03-002, April 2003. US Environmental Protection Agency, Chesapeake Bay Program Office, Annapolis, Maryland. USEPA, 2004a. Chesapeake Bay Program Analytical Segmentation Scheme. Revisions, Decisions and Rationales 1983-2003. EPA 903-R-04-008 CBP/TRS 268/04, October 2004. US Environmental Protection Agency, Chesapeake Bay Program Office, Annapolis, Maryland. USEPA, 2004b. Ambient Water Quality Criteria for Dissolved Oxygen, Water Clarity and Chlorophyll a for the Chesapeake Bay and Its Tidal Tributaries. 2004 Addendum. EPA 903-R-03-002, October 2004. US Environmental Protection Agency, Chesapeake Bay Program Office, Annapolis, Maryland. Vølstad, J.H., Neerchal, N.K., Roth, N.E., Southerland, M.T., 2003. Combining biological indicators of watershed condition from multiple sampling programs–a case study from Maryland, USA. Ecological Indicators 3, 13–25. Weisberg, S.B., Ranasinghe, J.A., Dauer, D.M., Schaffner, L.C., Diaz, R.J., Frithsen, J.B., 1997. An estuarine benthic index of biotic integrity (B-IBI) for Chesapeake Bay. Estuaries 20, 149–158.
Contents lists available at ScienceDirect
Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul
Assessing ecological integrity for impaired waters decisions in Chesapeake Bay, USA Roberto J. Llansó a,*, Daniel M. Dauer b, Jon H. Vølstad a,c a b c
Versar Inc., Ecological Sciences and Applications, 9200 Rumsey Road, Columbia, Maryland 21045, USA Department of Biological Sciences, Old Dominion University, Norfolk, Virginia 23529, USA Institute of Marine Research, P.O. Box 1870 Nordnes, 5817 Bergen, Norway
a r t i c l e
i n f o
Keywords: Ecological integrity Benthic community condition Impaired waters assessment Biological criteria Chesapeake Bay
a b s t r a c t To meet the requirements of the Clean Water Act, the States of Maryland and Virginia are using benthic biological criteria for identifying impaired waters in Chesapeake Bay and reporting their overall condition. The Chesapeake Bay benthic index of biotic integrity (B-IBI) is the basis for these biological criteria. Working together with the states and the US Environmental Protection Agency, we developed a method for impairment decisions based on the B-IBI. The impaired waters decision approach combines multiple benthic habitat-dependent indices in a Bay segment (equivalent to water bodies in the European Water Framework Directive) with a statistical test of impairment. The method takes into consideration uncertainty in reference conditions, sampling variability, multiple habitats, and sample size. We applied this method to 1430 probability-based benthic samples in 85 Chesapeake Bay segments. Twenty-two segments were considered impaired for benthic community condition. The final decision for each segment considers benthic condition in combination with key stressors such as dissolved oxygen and toxic contaminants. Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction In the United States, the Clean Water Act (CWA) of 1972 (US Code title 33, Sections 1251–1387) directs states to develop water quality standards, and to design and implement a plan to restore waters with impaired water quality. The standards define the water quality needed for designated water uses based on support for aquatic life and societal benefits. If a water body exceeds the limits of pollution allowed by the water quality standards, it does not satisfy the requirements of Section 303(d) of the CWA, and is then listed as not supporting one or more of its designated uses. The water quality standards may include biological criteria, which are defined as ‘‘numerical values or narrative expressions that describe the reference biological integrity of aquatic communities inhabiting waters of a given designated aquatic life use” (USEPA, 1990). Section 305(b) of the CWA requires states to produce a biennial report on the condition of their surface waters (USEPA, 1997). There are several advantages to using biological data in the assessment of water quality and identification of impaired waters. Biological assessment data provide direct measures of aquatic life use support, integrate temporally variable environmental conditions and the effects of multiple types of environmental stress, and can be used to assess the effects of habitat alteration, verify
* Corresponding author. E-mail address: [email protected] (R.J. Llansó). 0025-326X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2008.11.011
impacts of point source discharges, and capture episodic or nonpoint source pollution (Fausch et al., 1990; Diaz, 1992; Barbour et al., 2000; Simon, 2000). Also, biological assessment data can help identify sources of impacts to the aquatic community and assist in establishing total maximum daily loads (TMDLs). A TMDL defines the limit of a pollutant that a water body can receive while still meeting the water quality standards, and can be used to set restoration priorities. Recognizing that there was a wealth of benthic community data being collected by Chesapeake Bay monitoring programs, the States of Maryland and Virginia decided to use benthic biological criteria to identify impaired waters in Chesapeake Bay and report their overall condition. The Chesapeake Bay benthic index of biotic integrity (B-IBI) (Weisberg et al., 1997) is the basis for these biological criteria. This paper outlines a method for deciding whether Chesapeake Bay tidal waters are biologically impaired based on the B-IBI. The method uses reference conditions to identify degradation and designate impairment, but unlike most other biological listing methods (USEPA, 2002), it incorporates uncertainty into the impairment decision. The approach makes full use of the probabilistic design of the monitoring programs. Together with the index of biotic integrity, the probabilistic design allows decisions to be made on the extent of degraded area at multiple spatial scales (Dauer and Llansó, 2003). We applied the method at the level of Bay segments, equivalent to the water bodies of the European Union Water Framework Directive (WFD). Similar to the CWA, the
49
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
WFD establishes a framework for the protection and improvement of all European surface and ground waters (Borja, 2005). The WFD provides guidance for assessing the ecological status of estuarine and coastal waters (EC, 2003), but it leaves to member states the development of specific methodologies, such as those for assessing benthic quality status (Borja et al., 2004). The approach taken here may therefore also be useful for determining benthic impairments in estuaries and coastal waters of Europe or other regions where benthic indices are used to evaluate ecological integrity. 2. Methods
Table 1 Number of reference (degraded and undegraded) samples by habitat in the calibration and validation data sets used in the development of the Chesapeake Bay benthic index of biotic integrity (B-IBI). For salinity and silt-clay characteristics, see Weisberg et al. (1997). Habitat
Degraded
Undegraded
Tidal freshwater Oligohaline Low mesohaline High mesohaline sand High mesohaline mud Polyhaline sand Polyhaline mud
136 92 49 5 81 7 136
75 32 20 14 39 39 24
2.1. Benthic index of biotic integrity The Chesapeake Bay B-IBI is a multi-metric index that evaluates the ecological condition of a sample by comparing key benthic invertebrate community attributes based on species composition, abundance, and biomass to reference values expected under non-degraded conditions in similar habitat types. It is therefore a measure of departure from reference conditions. Reference conditions are estimated from benthic samples taken at sites that satisfy abiotic reference criteria, and are therefore best described as least disturbed conditions (Stoddard et al., 2006). The metrics and thresholds of the B-IBI were developed separately for seven habitat types (tidal freshwater, oligohaline, low mesohaline, high mesohaline sand, high mesohaline mud, polyhaline sand, and polyhaline mud). For each habitat, metrics are scored 1, 3, or 5 and averaged. Although the B-IBI can be regarded as consisting of seven benthic habitat-dependent indices, the scoring system is the same for all habitats and, therefore, B-IBI scores from different habitat types have the same meaning. The B-IBI was described by Weisberg et al. (1997), and further detail can be found in Alden et al. (2002) and Llansó et al. (2003). 2.2. Reference sites The method described here compares B-IBI scores from assessment sites with B-IBI scores from reference sites, and assumes that the reference sites are a representative sample from a ‘‘super population” of all possible reference sites. Assessments are then made at the level of Chesapeake Bay segments and subsegments containing benthic sites. Segments are Chesapeake Bay regions having similar salinity and hydrographic characteristics (USEPA, 2004a). In Virginia, segments were sub-divided into smaller units consisting of the main stem of rivers and bays and the smaller systems draining into the main stem. The reference sites were the calibration and validation sites used in the development of the B-IBI. There were two types of reference sites, undegraded and degraded. Undegraded sites had good dissolved oxygen and did not exceed sediment contaminant, toxicity, and organic carbon criteria. Degraded sites had low dissolved oxygen or exceeded any of the sediment criteria (see Weisberg et al., 1997 for criteria). Undegraded sites were used to develop the B-IBI, and both types of sites were used to test and validate the B-IBI. The term ‘‘reference” is usually applied to the undegraded sites, and reference condition is defined as that of sites determined to be relatively free of anthropogenic stress. In this paper, we define reference condition in the same way, but we broadly distinguish between the undegraded and the degraded reference sites, as both types of sites are used in the present method. The degraded reference sites are assumed to represent the expected biological characteristics of all sites with degraded benthic condition. Table 1 shows the number of reference samples for the seven habitat types.
2.3. Assessment sites The assessment sites were selected within Chesapeake Bay in a stratified random manner, and were sampled in either August or September, 2000 to 2004. For the present study, they were poststratified into segments using a Geographic Information System. A total of 1430 benthic grab samples (one sample per site) were collected, including 750 Maryland benthic monitoring program samples, 500 Virginia benthic monitoring program samples, 150 Elizabeth River biological monitoring program samples, and 10 samples each in Mobjack Bay, the Mattaponi River, and the Nansemond River. Each sample was collected with a Young grab over a sediment surface area of 440 cm2. Samples were sieved on a 0.5mm screen and preserved in the field. The organisms retained on the screen were sorted, enumerated, and identified to the lowest possible taxon. Ash-free dry weight biomass was determined for each taxon.
2.4. Assessment method The assessment method consisted of two steps: first, we calculated for each segment the proportion of sites with low B-IBI scores below a threshold (i.e., the proportion failing), and second, we determined whether this proportion differed from what would be expected from chance alone. That is, if a segment were in good condition (had no low dissolved oxygen, contaminant, toxicity, or organic enrichment problems), we would still expect a small proportion of sites to have low scores because of natural variability; this proportion under the null hypothesis was defined as 5%. Usually, a B-IBI score of <3.0 is characterized as failing the Chesapeake Bay benthic community restoration goals (Ranasinghe et al., 1994). However, because the B-IBI scores for the undegraded and degraded reference sites overlapped in low salinity habitats (Alden et al., 2002), thresholds were set for each of these habitats as the 5th percentile B-IBI score of the undegraded reference sites. For habitats where there was no overlap, the threshold was set as the maximum B-IBI score of the degraded reference sites. The 5th percentile score and its variance were estimated by bootstrap resampling of the undegraded reference sites, with replacement (Efron and Tibshirani, 1998). For each bootstrap iteration (i), the assessment sites were then compared to the thresholds established as above, by habitat (most segments had sites in more than one habitat), and the proportion of sites below threshold (Pi) was determined as the weighted average of proportions across habitats. The final proportion of sites below threshold in a segment (P) was the average over all the iterations. The purpose of these procedures was to incorporate uncertainty in reference conditions and balance Type I and Type II errors. In the second step of the assessment method, segments were declared impaired if P was greater than expected under the null hypothesis (Schenker and Gentleman, 2001):
50
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
P Po > 0 with 95% confidence:
P Po 1:64
pffiffi ðVarP þ VarPo Þ
where Po is the proportion of sites failing under the null hypothesis (5%). Thus, for a segment to be declared impaired, the lower bound of the 95% confidence interval of the proportion of sites in a segment below threshold had to be higher than 5%. To estimate the confidence interval, the variance from the bootstrap resampling (due to estimating thresholds) was added to the binomial variance associated with sampling variability within a segment:
Varp ¼
i¼5000 X i¼1
ðPi PÞ2 þ ðpq=N 1Þ 5000 1
where p and q are parameters in the binomial distribution and N is the number of samples in a segment. 3. Results Eighty-five Chesapeake Bay segments contained benthic data. The number of samples within each segment varied from 1 to 112. Thus, the sample size of some segments was insufficient for reliable impairment determinations. Segments and subsegments with less than 10 samples were considered inconclusive relative to the listing process. Of the segments with at least 10 samples, 22 were considered impaired for benthic community condition and 17 were judged not impaired (Table 2). The proportion of sites below threshold for the impaired segments ranged from 28% to 76%, indicating that at least a quarter of the area of these segments was degraded. For most of these segments, the average B-IBI score was <3.0. Greatest degradation (Table 2) was found in segments with known nutrient (lower Chester and Choptank Rivers), dissolved oxygen (lower Patuxent, Potomac, and Rappahannock Rivers; Maryland main stem), or contaminant problems (Patapsco River estuary, Elizabeth River) (Dauer et al., 2000). The estimates for the main stem segments CB4MH and CB5MH exclude the deep trough (>12 m) of the Chesapeake Bay, which is not monitored by the benthic programs because this area is affected by summer anoxia and consistently has been found to be azoic. 4. Discussion The process used to determine if the segments of the Chesapeake Bay are biologically impaired was based on the B-IBI. The B-IBI is based on biologically defined metric thresholds (Weisberg et al., 1997). An index value of 3 or more (on a 1 to 5 scale) is considered to indicate good benthic community condition indicative of good habitat quality. This quantitative benchmark, however, does not account for uncertainty in reference conditions. In low salinity habitats, overlap in B-IBI score distributions for the degraded and undegraded reference sites produced a class of uncertain condition (Alden et al., 2002). Uncertainty arises from natural stressors that mask the signal from pollution effects (Diaz, 1989), patchiness in the spatial distribution of macroinvertebrates (Dauer, 1993), and small sample size (Vølstad et al., 2003). Our approach is unique among impaired waters methods in that we use confidence limits to account for variability in reference conditions. The use of B-IBI criteria that take into consideration habitat specificity and uncertainty in reference conditions should produce results with fewer Type I errors (calling a segment impaired when it is not). With decreasing sampling size, the 5th percentile of the reference site distribution is less precisely estimated. Therefore, we used a bootstrap routine to get a reliable estimate of this value and its associated confidence interval. The width of that confidence
interval widens as sample size decreases. This variability was added to that of the estimated proportion of sites degraded within a segment to create confidence intervals for impairment decisions. The criterion for impairment decisions was based on statistical significance; however, the extent of degradation in a segment that is considered impaired is a management objective. Decision rules could be established based on specific thresholds of protection that the states may wish to implement. These thresholds could be based on biological or ecological endpoints, or on policy decisions that are based on publicly unacceptable conditions. Such thresholds based on management objectives can clearly be used in our approach, in addition to or in place of a statistical threshold. In an evaluation of water quality assessment methodologies used by various states to determine impairment of surface waters, Keller and Cavallaro (2008) found that listing decisions by the states were often based on insufficient data, or data that were not spatially or temporally representative of the conditions of the water body being assessed. The need for sufficient and adequate data was viewed as one of the most important elements of the listing process. In agreement with this view, a minimum sample size of 10 for impairment assessments in Chesapeake Bay is suggested. This target sample size should incorporate temporal variability in the assessment of benthic condition, and is consistent with the sample size required for assessing non-tidal waters in the Chesapeake Bay watershed (Vølstad et al., 2003). Accounting for temporal variability is important because in Chesapeake Bay benthic condition is strongly associated with oxygen depletion (Llansó et al., 2007), which arises from interactions between biological and physical processes and varies with inter-annual patterns of precipitation and freshwater inflow (Kemp et al., 2005). Consistency among tidal and non-tidal assessments is also important because of the need for integrated reporting that links watershed development to benthic condition across aquatic ecosystems. It might be argued that incorporating uncertainty in reference condition and a required sample size is too cautious given the obligations under the CWA. However, our approach focuses in minimizing error that might occur from the inherent variability of benthic communities, the effects of natural stressors, and sampling and methodological error. We believe that where impairments are common, it is prudent to err on the side of limiting Type I errors. In the regulatory realm, avoiding unnecessary listing and remediation will help focus restoration efforts where they are most needed. This is not to say that the tidal waters of the Chesapeake Bay should not be protected, but rather that impairment designations should be used to initiate effective restoration in critically degraded waters. Our method could be enhanced by setting additional protection thresholds, such as, for example, requiring that one half of the confidence interval be less than 25% for a segment to be designated as not impaired (i.e., supporting aquatic life use). Our approach focuses on the amount (percentage) of area that is impaired, which is a measure of the extent of degradation. This focus differs from that of methods that measure average degradation within an area. Using the average degradation, a segment could conceivably meet criteria and be degraded over more than 50% of its area. Although an impairment decision made on this basis would clearly be undesirable, the percent area approach also has disadvantages. For example, the severity of degradation within a segment could increase significantly without changing the percent area degraded. The B-IBI scores in a segment, however, can be inspected to determine magnitude of degradation. Within the WFD, ecological integrity is assessed by the ecological quality ratio (EQR), which compares biological index values in a water body to those of reference conditions (Borja et al., 2004). EQR derivation methods vary by member state, from a simple ratio of two numbers (Rosenberg et al., 2004) to more complex methods involving multivariate analysis (Muxika et al., 2007). Classifications of
51
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
Table 2 Impaired waters results for 85 Chesapeake Bay segments and subsegments for the period 2000–2004. Impaired segments were sorted according to the lower 95% bound of the confidence interval of the difference between the proportion of sites in the segment with B-IBI scores below threshold (P) and the proportion of sites failing under the null hypothesis (Po). The mean B-IBI value in each segment is shown for comparison. CLL = lower 95% confidence limit bound. CLU = upper 95% confidence limit bound. * = Inconclusive; sample size insufficient to make a reliable determination. State of Virginia’s subsegments are identified by lower case letters. Segment
Water body name
Sample size
P
Po
P Po
CL–L (P Po)
CLU (P Po)
Impaired
POTMH SBEMHa CB4MH PATMH PAXMH CHSMH EBEMHa ELIMHa CB3MH RPPMHa YRKMHa CHOMH2 JMSMHa JMSMHb YRKPHa CB5MH PMKOHa MAGMH ELIPHa MOBPHa WBEMHa JMSOHa LAFMHa CB7PHa TANMH SEVMH GUNOH MPNOHa POTOH CB6PHa CB2OH CB1TF CB8PHa MPNTFa MANMH POTTF RPPTFa JMSTFa JMSPHa WICMH POCMH BSHOH NANMH SOUMH CHOMH1 CRRMHa ELKOH POCOH RHDMH PAXOH LCHMH CHSOH PAXTF HNGMH CHOOH RPPOHa BIGMH CHKOHa EASMH BACOH NANOH FSBMH NORTF PMKTFa JMSPHd JMSMHd JMSMHc MIDOH SASOH RPPMHm MOBPHh MOBPHf
Potomac River mesohaline Southern Branch of Elizabeth River Maryland main stem Patapsco River Patuxent River mesohaline Chester River mesohaline Eastern Branch of Elizabeth River Elizabeth River main stem Maryland main stem Rappahannock River mesohaline York River mesohaline Choptank River mesohaline James River mesohaline Nansemond River York River polyhaline Maryland main stem Pamunkey River oligohaline Magothy River Elizabeth River main stem Mobjack Bay Western Branch of Elizabeth River James River oligohaline Lafayette River Virginia main stem Tangier Sound Severn River Gunpowder River Mattaponi River oligohaline Potomac River oligohaline Virginia main stem Maryland upper main stem Maryland upper main stem Virginia main stem Mattaponi River tidal fresh Manokin River Potomac River tidal fresh Rappahannock River tidal fresh James River tidal fresh James River polyhaline Wicomico River Pocomoke Sound Bush River Nanticoke River mesohaline South River Choptank River Corrotoman River Elk River Pocomoke River oligohaline Rhode River Patuxent River oligohaline Little Choptank River Chester River oligohaline Patuxent River tidal fresh Honga River Choptank River oligohaline Rappahannock River oligohaline Big Annemessex River Chickahominy River Eastern Bay Back River Nanticoke River oligohaline Fishing Bay Northeast River Pamunkey River tidal fresh Willoughby Bay Warwick River Chuckatuck River Middle River Sassafras River Totuskey Creek East River Ware River
91 47 28 49 112 33 15 37 61 98 64 22 46 16 29 44 11 17 17 20 19 22 27 43 48 13 15 11 21 18 40 19 15 13 13 12 11 14 10 9 9 9 9 8 8 8 8 7 7 7 6 6 6 5 5 5 5 5 4 4 4 4 4 4 3 3 3 3 3 2 2 1
0.76 0.70 0.67 0.52 0.49 0.53 0.57 0.48 0.44 0.37 0.43 0.41 0.37 0.45 0.38 0.32 0.46 0.41 0.39 0.36 0.36 0.28 0.31 0.15 0.13 0.26 0.22 0.25 0.16 0.15 0.00 0.10 0.00 0.00 0.10 0.08 0.07 0.02 0.12 0.42 0.29 0.24 0.13 0.88 0.38 0.23 0.15 0.49 0.43 0.40 0.31 0.17 0.17 0.21 0.21 0.06 0.03 0.00 0.75 0.44 0.25 0.25 0.08 0.00 0.91 0.34 0.33 0.00 0.00 0.50 0.00 1.00
0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05
0.71 0.65 0.62 0.47 0.44 0.48 0.52 0.43 0.39 0.32 0.38 0.36 0.32 0.40 0.33 0.27 0.41 0.36 0.34 0.31 0.31 0.23 0.26 0.10 0.08 0.21 0.17 0.20 0.11 0.10 0.05 0.05 0.05 0.05 0.05 0.03 0.02 0.03 0.07 0.37 0.24 0.19 0.08 0.83 0.33 0.18 0.10 0.44 0.38 0.35 0.26 0.12 0.12 0.16 0.16 0.01 0.02 0.05 0.70 0.39 0.20 0.20 0.03 0.05 0.86 0.29 0.28 0.05 0.05 0.45 0.05 0.95
0.62 0.51 0.45 0.33 0.33 0.32 0.27 0.26 0.26 0.21 0.18 0.17 0.17 0.16 0.15 0.14 0.12 0.12 0.10 0.10 0.08 0.05 0.01 0.01 0.03 0.04 0.05 0.06 0.06 0.09 0.11 0.14 0.15 0.15 0.16 0.17 0.17 0.18 0.29 – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
0.80 0.78 0.79 0.61 0.55 0.65 0.77 0.60 0.52 0.43 0.58 0.56 0.48 0.65 0.52 0.40 0.69 0.59 0.58 0.52 0.55 0.41 0.52 0.21 0.18 0.45 0.39 0.47 0.28 0.28 0.01 0.23 0.05 0.05 0.25 0.22 0.20 0.12 0.43 – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No No No No No No No No No No No No No No * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
Mean B-IBI
1.7 2.0 2.3 2.4 2.4 2.6 2.2 2.5 2.7 2.6 2.5 2.9 2.7 2.4 3.0 2.7 2.6 2.3 2.8 3.0 2.4 2.9 2.4 3.3 3.2 2.7 2.9 2.6 3.4 3.3 3.8 3.1 3.4 3.5 3.1 3.1 3.5 3.2 3.4 2.8 2.6 2.6 3.0 2.1 2.6 2.4 3.2 2.5 2.9 2.8 2.5 3.2 2.6 2.8 2.7 3.5 2.9 3.7 2.1 2.1 3.5 3.6 3.2 3.9 1.7 2.8 3.1 3.4 3.2 3.1 2.7 1.3 (continued on next page)
52
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
Table 2 (continued) Segment
Water body name
Sample size
P
Po
P Po
CL–L (P Po)
CLU (P Po)
Impaired
Mean B-IBI
MOBPHg RPPMHd WSTMH YRKPHd MATTF YRKMHb YRKPHe CHSTF CHOTF APPTFa BOHOH MOBPHe POCTF
North River Robinson Creek West River Sarah Creek Mattawoman Creek Queen Creek Timberneck Creek Chester River tidal fresh Choptank River tidal fresh Appomattox River Bohemia River Severn Creek Pocomoke River tidal fresh
1 1 1 1 1 1 1 1 1 1 1 1 1
1.00 1.00 1.00 1.00 0.73 0.73 0.37 0.19 0.00 0.00 0.00 0.00 0.00
0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05
0.95 0.95 0.95 0.95 0.68 0.68 0.32 0.14 0.05 0.05 0.05 0.05 0.05
– – – – – – – – – – – – –
– – – – – – – – – – – – –
* * * * * * * * * * * * *
1.7 1.7 2.2 1.3 1.7 1.7 2.7 2.0 3.0 3.0 4.0 2.7 2.5
ecological status for single sites are produced, but procedures that integrate these results into a decision based approach are generally lacking. Water quality assessments under the CWA are based on the ability of the waters to support designated uses. Support is defined as meeting the criteria for each use. Once water bodies have been assessed, they are classified into 3 categories: fully supporting, inconclusive, and impaired. If a water body is classified as impaired, a stressor identification procedure is used to determine if a TMDL is necessary. State and federal agencies with jurisdiction over the Chesapeake Bay have established criteria for dissolved oxygen (DO), contaminants, water clarity, and chlorophyll for specific uses, such as open water and shallow water submerged aquatic vegetation (USEPA, 2003, 2004b). At present, the B-IBI is evaluated independently of these parameters; i.e., benthic community impairment is an independent and sufficient cause for listing. However, decision models in which the B-IBI is corroborative are being considered by the State of Maryland. In the decision process illustrated in Fig. 1, the final decision for each segment considers benthic condition in combination with key stressors such as DO
and toxic contaminants. In this process, TMDL development would be required when there is a likely cause for the benthic community impairment. Similar decision processes have already been developed for non-tidal waters in the Chesapeake Bay region (MDE, 2008). Our assessment of Chesapeake Bay segments produced impairment classifications that were generally consistent with expectations based on known patterns of human disturbance in the watershed (Dauer et al., 2000). We take this as confirmation of the general utility of the method in estuarine management decisions, and to recommend further study of the novel features of the method. Because benthic condition indices are used to evaluate estuarine ecological integrity in both the US and Europe, water quality decision procedures incorporating this approach may be widely applicable. Acknowledgements We greatly appreciate the computer programming support from Ed Weber (Versar Inc.). This paper benefited from reviews
NO START
Data sufficient?
Score sample
YES
YES
YES
Is segment impaired for DO numeric criteria?
Aquatic life fails Cause: DO B-IBI corroborative
Develop TMDL to correct low DO DO corrected
Aquatic life unknown
Test segment
Is segment degraded for B-IBI?
NO Evaluate B-IBI for other stressors
YES
Other stressors identified?
NO
Insufficient data
?
Does segment meet WQ criteria?
Additional monitoring information needed
NO Aquatic life fails Cause: DO, etc.
YES NO
Aquatic life supported
Aquatic life fails Cause: Pollutants B-IBI corroborative
Develop TMDL to correct pollutants
Aquatic life fails Cause: Pollution Unknown source No TMDL required
Pollutants corrected
Fig. 1. Proposed decision process for assessing impaired waters in Chesapeake Bay.
R.J. Llansó et al. / Marine Pollution Bulletin 59 (2009) 48–53
by Leska Fore (Statistical Design Inc.), Jeroen Gerritsen and Lei Zheng (Tetra Tech Inc.), and Charles Hawkins (Utah State University) on an earlier report submitted to the state agencies. The comments of Don Strebel and Mark Southerland (Versar Inc.), and two anonymous reviewers helped improve the paper. References Alden III, R.A., Dauer, D.M., Ranasinghe, J.A., Scott, L.C., Llansó, R.J., 2002. Statistical verification of the Chesapeake Bay benthic index of biotic integrity. Environmetrics 13, 473–498. Barbour, M.T., Swietlik, W.F., Jackson, S.K., Courtemanch, D.L., Davies, S.P., Yoder, C.O., 2000. Measuring the attainment of biological integrity in the USA: a critical element of ecological integrity. Hydrobiologia 422 (423), 453–464. Borja, A., 2005. The European Water Framework Directive: a challenge for nearshore, coastal and continental Shelf Research. Continental Shelf Research 25, 1768–1783. Borja, A., Franco, J., Valencia, V., Bald, J., Muxica, I., Belzunce, M.J., Solaun, O., 2004. Implementation of the European water framework directive from the Basque country (northern Spain): a methodological approach. Marine Pollution Bulletin 48, 209–218. Dauer, D.M., 1993. Biological criteria, environmental health and estuarine macrobenthic community structure. Marine Pollution Bulletin 26, 249–257. Dauer, D.M., Llansó, R.J., 2003. Spatial scales and probability based sampling in determining levels of benthic community degradation in the Chesapeake Bay. Environmental Monitoring and Assessment 81, 175–186. Dauer, D.M., Ranasinghe, J.A., Weisberg, S.B., 2000. Relationships between benthic community condition, water quality, sediment quality, nutrient loads, and land use patterns in Chesapeake Bay. Estuaries 23, 80–96. Diaz, R.J., 1989. Pollution and tidal benthic communities of the James River Estuary, Virginia. Hydrobiologia 180, 195–211. Diaz, R.J., 1992. Ecosystem assessment using estuarine and marine benthic community structure. In: Burton, A. (Ed.), Contaminated Sediment Toxicity Assessment. Lewis Publishers, Boca Raton, Florida, pp. 67–85. EC, 2003. Common Implementation Strategy for the Water Framework Directive (2000/60/EC), Guidance Document No. 5, Transitional and Coastal Waters – Typology, Reference Conditions and Classification Systems. Directorate General Environment of the European Commission, Brussels, Belgium. Efron, B., Tibshirani, R.J., 1998. An Introduction to the Bootstrap. CRC Press, Boca Raton, Florida. Fausch, K.D., Lyons, J., Karr, J.R., Angermeier, P.L., 1990. Fish communities as indicators of environmental degradation. American Fisheries Society Symposium 8, 123–144. Keller, A.A., Cavallaro, L., 2008. Assessing the US Clean Water Act 303(d) listing process for determining impairment of a waterbody. Journal of Environmental Management 86, 699–711. Kemp, W.M., Boynton, W.R., Adolf, J.E., Boesch, D.F., Boicourt, W.C., Brush, G., Cornwell, J.C., Fisher, T.R., Glibert, P.M., Hagy, J.D., Harding, L.W., Houde, E.D., Kimmel, D.G., Miller, W.D., Newell, R.I.E., Roman, M.R., Smith, E.M., Stevenson, J.C., 2005. Eutrophication of Chesapeake Bay: historical trends and ecological interactions. Marine Ecology Progress Series 303, 1–29. Llansó, R.J., Dauer, D.M., Vølstad, J.H., Scott, L.C., 2003. Application of the benthic index of biotic integrity to environmental monitoring in Chesapeake Bay. Environmental Monitoring and Assessment 81, 163–174.
53
Llansó, R.J., Dew, J., Scott, L.C., 2007. Chesapeake Bay Water Quality Monitoring Program: long-term Benthic Monitoring and Assessment Component, Level 1 Comprehensive Report (July 1984-December 2006). Report prepared for the Maryland Department of Natural Resources by Versar Inc., Columbia, Maryland. MDE, 2008. Part G: appendix II – Listing Methodologies. G.1 Biological Assessment of Water Quality for Non-Tidal Streams. Maryland Department of the Environment, Baltimore, Maryland. Muxika, I., Borja, A., Bald, J., 2007. Using historical data, expert judgement and multivariate analysis in assessing reference conditions and benthic ecology status, according to the European Water Framework Directive. Marine Pollution Bulletin 55, 16–29. Ranasinghe, J.A., Weisberg, S.B., Dauer, D.M., Schaffner, L.C., Diaz, R.J., Frithsen, J.B., 1994. Chesapeake Bay Benthic Community Restoration Goals. Report prepared for the USEPA Chesapeake Bay Program Office, the Governor’s Council on Chesapeake Bay Research Fund, and the Maryland Department of Natural Resources by Versar Inc., Columbia, Maryland. Rosenberg, R., Blomqvist, M., Nilsson, H.C., Cederwall, H., Dimming, A., 2004. Marine quality assessment by use of benthic species-abundance distributions: a proposed new protocol within the European Water Framework Directive. Marine Pollution Bulletin 49, 728–739. Schenker, N., Gentleman, J.F., 2001. On judging the significance of differences by examining the overlap between confidence intervals. The American Statistician 55, 182–186. Simon, T.P., 2000. The use of biological criteria as a tool for water resource management. Environmental Science and Policy 3 (Suppl. 1), 43–49. Stoddard, J.L., Larsen, D.P., Hawkins, C.P., Johnson, R.K., Norris, R.H., 2006. Setting expectations for the ecological condition of streams: the concept of reference condition. Ecological Applications 16, 1267–1276. USEPA, 1990. Biological Criteria: National Program Guidance for Surface Waters. EPA 440/5-90-004, April 1990. US Environmental Protection Agency, Office of Water, Washington D.C. USEPA, 1997. Guidelines for the Preparation of the Comprehensive State Water Quality Assessments (305[b] reports). EPA 841-B-97-002A. US Environmental Protection Agency, Office of Water, Washington D.C. USEPA, 2002. Summary of Biological Assessment Programs and Biocriteria Development for States, Tribes, Territories, and Interstate Commissions: Streams and Wadeable Rivers. EPA 822-R-02-048. US Environmental Protection Agency, Washington, D.C. USEPA, 2003. Ambient Water Quality Criteria for Dissolved Oxygen, Water Clarity and Chlorophyll a for the Chesapeake Bay and Its Tidal Tributaries. EPA 903-R03-002, April 2003. US Environmental Protection Agency, Chesapeake Bay Program Office, Annapolis, Maryland. USEPA, 2004a. Chesapeake Bay Program Analytical Segmentation Scheme. Revisions, Decisions and Rationales 1983-2003. EPA 903-R-04-008 CBP/TRS 268/04, October 2004. US Environmental Protection Agency, Chesapeake Bay Program Office, Annapolis, Maryland. USEPA, 2004b. Ambient Water Quality Criteria for Dissolved Oxygen, Water Clarity and Chlorophyll a for the Chesapeake Bay and Its Tidal Tributaries. 2004 Addendum. EPA 903-R-03-002, October 2004. US Environmental Protection Agency, Chesapeake Bay Program Office, Annapolis, Maryland. Vølstad, J.H., Neerchal, N.K., Roth, N.E., Southerland, M.T., 2003. Combining biological indicators of watershed condition from multiple sampling programs–a case study from Maryland, USA. Ecological Indicators 3, 13–25. Weisberg, S.B., Ranasinghe, J.A., Dauer, D.M., Schaffner, L.C., Diaz, R.J., Frithsen, J.B., 1997. An estuarine benthic index of biotic integrity (B-IBI) for Chesapeake Bay. Estuaries 20, 149–158.