Coral ECA Water Quality Assessment Data Analysis

Coral ECA Water Quality Assessment Data Analysis

Document: Coral ECA Water Quality Assessment Data Analysis

Document Type: Report

Author Name: Henry Briceño, Joseph N. Boyer, and Ian L. Drydan

Executive Summary

The Kristin Jacobs Coral Reef Ecosystem Conservation Area (Coral ECA) Water Quality Assessment (WQA) was designed in 2014 by a collaborating body of National Oceanic and Atmospheric Administration (NOAA) scientists, Florida Department of Environmental Protection’s Coral Reef Conservation Program (CRCP) staff, and partners from the Southeast Florida Coral Reef Initiative (SEFCRI). The goal of the WQA was to provide data for managers to assess the status of the Coral ECA, an area which historically did not have a consistent water quality monitoring program.

The focus of this study extends from the St. Lucie Inlet in the north to offshore Biscayne Bay in the south, containing nine major inlets, namely, St. Lucie (STL), Jupiter (JUP), Lake Worth (ILW), Boynton (BOY), Boca Raton (BOC), Hillsboro (HIL), Port Everglades (PEV), Baker’s Haulover (BAK), and Government Cut (GOC).

The overall goal of this assessment is to identify both the constituents and impacts of land-based sources of pollution on coral reef ecosystems and inform resource managers and decision-makers on the water quality status in the Southeast Florida coastal zone. Our objectives in that context are aimed to answer the following questions:

1. Does water quality differ among ICAs?

Yes, water quality is different among the various ICAs and is driven by geographical location, distance from shore, terrestrial runoff, freshwater input, etc. This is confounded by the composition of “Site Types” (Inlet, Outfall, & Reef) within each ICA. Some ICAs include Outfall sources (BAK, BOC, GOC, HIL, & PEV), some ICAs are dominated by large freshwater inlets (BOC, HIL, JUP, and especially STL). ILW sites are classified as Inlet and Reef but they are all low in nutrients with high optical clarity so look more like other Reef areas. Whereas, the STL sites have high nutrients, low salinity, and poor optical clarity and therefore look more like Inlet than Reef. The take home message is that we should let the data inform site classifications rather than comparing apples to oranges.

2. Does water quality differ among Site Types (Inlet, Outfall, and Reef)?

Yes, water quality is different among “Site Types”, but these categories have been subjectively applied to sampling sites. Objective analyses (PCA, MDS, and Cluster analysis) showed, with a few exceptions, that the classifications were distinct enough to be statistically different. Inlet sites all had similar water quality except for the STL, which was significantly different than other Inlets. The STL Reef sites clustered with other Inlet sites. Conversely, the ILW Inlet water quality clustered with other Reef sites. For future analyses, FDEP might want to reclassify some Site Types to be more consistent. 2

3. Does water quality differ between surface and bottom waters?

Yes, water quality is typically significantly different between surface and bottom samples from the same site. Variables most different tend to be those land-derived variables such as nutrients and sediments. These differences are complicated by current patterns and density stratification. Freshwater inputs from Inlets tend to remain at the surface rather than mix with depth. These areas also tend to have higher surface nutrient inputs. Outfall water sources enter at the bottom of the water column and are fresher than seawater. This means outfall sources tend to be buoyant and mix sewage nutrients throughout the water column.

4. How Do Available Water Quality Data Compare to Relevant Published Benchmarks, Especially Those of SE Florida Waters?

Scientific Consensus Approach: The purpose of this Task is to provide a scientifically defensible methodology to ultimately assist FDEP in the development of biogeochemically relevant benchmarks for the ECA. We provide a compendium of those studies which strived to summarize these effects and draw a ‘line in the water’ by establishing water quality benchmarks/thresholds/criteria to further the continued growth, development, and survival of coral reefs.

Early work in the 60’s and 70’s was mostly observational, much like Hart (1974). More refined empirical and nutrient dosing field studies provided the foundation for most governmental standards and criteria e.g., ANZECC, ARMCANZ (2000), GBRMPA (2009), FDEP (2013), Hawai’i State Department of Health (2021). Smith’s pioneering work in Kaneohe Bay (1977) fired the starting gun for more in situ and lab-based studies.

For laboratory-based, experimental studies, the question has always been, what is the best measure to assess ‘impact’? Some researchers have used rate of coral growth, some used mortality, and others used everything else in between. Bradley et al. (2010) provided a relevant discussion in relating the Clean Water Act to biocriteria as well as water quality criteria for coral reefs.

A recent meta-analysis on lab-based, experimental nutrient effects to corals by Nalley et al. (2023) attempted to address this multiple response problem by “classifying effects of DIN (nitrate and ammonium) and DIP (phosphate)” into nine physiological “coral holobiont responses”:

A. Photosynthetic responses of the coral endosymbiont 1. Zooxanthellae density

2. Chl-a concentration

3. Photosynthetic rate

4. Photosynthetic efficiency (max. quantum yield)

B. Coral growth and calcification

3

1. Growth rate

2. Calcification rate

1. Adult tissue and colony survival

2. Larval survival and settlement

3. Fertilization success

C. Mortality

We believe that the value of the many empirical studies and field dosing experiments, being more inclusive of the coral community, should not be discounted and have therefore given more weight to this body of work for benchmark development. Our ‘suggested’ Scientific Consensus of the published benchmarks is described in Task 2.1b and summarized in the following Table i.

EPA 75th Percentile Approach: EPA recommends a reference site approach for setting benchmarks (US-EPA 2001) where sites are selected based on minimal human influence. If values from other sites fall within an acceptable range of reference sites (typically 75th percentile) they are considered to meet the designated use. In some areas minimally disturbed reference conditions do not exist and may not be achievable. In these situations, “least disturbed sites” may be used (25th percentile) if they demonstrate that the existing biological community structure and function is representative of a sustainable, natural system. Results of 75th Percentile analysis are shown in Table i.

CUSUM Approach:

Given the large proportion of non-detects in the ECA water quality database, we could not derive benchmarks using the cumulative sum method (Briceño et al. 2010; Regier et al. 2019). Instead, we used data gathered during the 2009-2012 period for the same ECA study area, from the Southeast Florida Coral Reef Initiative (SEFCRI) project (Boyer 2012). This method entails plotting CUSUM-transformed CHLa data against potential drivers to extract meaning in the context of driver-response relationships.

Benchmark Approach Comparison: The three separate benchmark approaches are compared in Table i. The 75th percentiles for total ECA compare relatively well with the Scientific Consensus benchmarks except that DIN values were higher and Secchi depth was lower in the ECA than proposed by Scientific Consensus. CUSUM benchmarks compare well with Scientific Consensus and 75th Percentile approaches but were slightly lower overall. 4

Table i. Benchmark Approach Comparison of Scientific Consensus, 75th Percentile, and CUSUM from 4 distinct site clusters. Demarcation “a” means insufficient or no data.

5. Recommendations

We have three general recommendations concerning this project moving forward. The first reiterates a recommendation from our previous report (Briceño et al. 2022) that FDEP should work to reduce the number of non-detects in future laboratory analyses. This means either upgrading the existing laboratory sensitivity for low level nutrient analyses or by using a different contract laboratory with higher analytic sensitivity.

The second recommendation concerns treatment of non-detect data. Censored maximum likelihood estimation is an efficient method to estimate the distributions, taking account of the observations below the MDL. If more readings can be obtained over the MDL then different estimation methods will become more similar, and ideally the analysis should not be very sensitive to the choice of estimation method.

The third recommendation concerns derivation of water quality benchmarks. Each approach, the Scientific Consensus, 75th Percentile, and CUSUM, are valid methods in their own right. Surprisingly, the results of all three approaches for the ECA water quality were similar for NO3+NO2, PO4, CHLa, turbidity, and Secchi depth. The Scientific Consensus result was lower for NH4 but higher than the other two approaches for TN and TSS.

We believe that the choice of which benchmark approach to use for the ECA should be debated by the local coral reef scientific and regulatory community. There are advantages and disadvantages for each method.

Advantages of the Scientific Consensus are that there is considerable weight of evidence generated from the many peer-reviewed studies and that these results all fall within a relatively 5

narrow range. The disadvantages include inherent variability in laboratory analyses, fluctuations in geographical ambient nutrient levels, global differences in coral community structure, etc.

An advantage of the 75th Percentile approach is that it uses data generated from the local area of interest. It may also be used to mine historical data. Conversely, the main disadvantage of the 75th Percentile approach is that it relies on data collected from the local area of interest. If the area of interest is already impacted, the benchmarks will be overestimated. However, future benchmarking could be refined by more selective use of least-impacted ICAs. The PC/MDS/Cluster analyses showed that data from most Inlets should be excluded from the computation because they are very different than Reef sites. Conversely, the ILW might be included in REEF as its water quality was comparable with other Reef sites. Outfall sites are problematic because they typically occur farther from shore and in deeper waters than Reefs (2.1-17.1 m vs 16.8-54.9 m). In addition, there is no data collected at bottom Outfall sites where impacts may be expected.

The CUSUM approach also relies on local data but has the advantage in that it quantifies a driver-response threshold for increase in CHLa production (phytoplankton biomass). The disadvantage is that, currently, there are no other driver-response effects in place for coral reef impacts which occur at the same time scale of water quality sampling events. Lags between driver-response reduce the ability to resolve thresholds.