Oyster Habitat Evaluation Using Hydrocoast Salinity Data and Two Approaches to Suitability Analysis in the Pontchartrain Basin, Southeast Louisiana

Aimee Préau, M.Sci. Patrick Smith, Ph.D. John Lopez, Ph.D.

Lake Pontchartrain Basin Foundation June 2016

SaveOurLake.org

Acknowledgements

The authors would like to acknowledge Thomas Soniat, Ph.D. for his assistance in preparing this report and reviewing a previous draft, Theryn Henkel, Ph.D., for providing useful comments and editing a previous draft, and the Walton Family Foundation for providing funding.

Cover Graphic - Location of Oyster Beds of St. Bernard Parish 1912 by Frank Payne Colorization added: Blue = Dense growths of oyster Green = Scattered growths of oysters Orange = Very scattered growths of oyster

Page 2 of 85

Table of Contents

Acknowledgements ...... 2 List of Figures ...... 5 List of Tables ...... 6 Executive Summary ...... 7 Introduction ...... 10 Life History Traits ...... 10 Oyster Fishery in Louisiana ...... 11 Oyster Reefs for Coastal Protection ...... 12 Oysters in Louisiana’s Comprehensive Master Plan for a Sustainable Coast ...... 12 Estimating Oyster Habitat Suitability ...... 13 Materials and Methods ...... 13 Study Area Description ...... 13 Hydrocoast Salinity ...... 14 Chatry Optimal Oyster Salinity (COOS) Regime ...... 15 Soniat Optimal Oyster Salinity (SOOS) Regime ...... 16 V2- Mean salinity during spawning season ...... 17 V3 -Minimum monthly mean salinity ...... 17 V4- Historic mean salinity ...... 17 Map Classifications of COOS and SOOS Results ...... 18 Comparison between Biloxi Marsh and Breton Sound Sub-Basins ...... 19 Hypoxia Polygons and Contours ...... 20 Other Oyster Suitability Indicators and Restoration Targets ...... 21 Proposed USACE Oyster Targets...... 21 Historic Oyster Reefs...... 22 State Master Plan Projects ...... 22 Salinity Gradient Points ...... 22 Results ...... 23 Study Area Salinity Suitability Results ...... 23 Multi-Year Average Salinity Suitability (2013-2015) ...... 26

Page 3 of 85

Biloxi Marsh and Breton Sound Sub-Basins...... 28 Overlay Maps ...... 30 Annual Analysis Results with State Master Plan Projects (Figures 11-16) ...... 30 Annual Analysis Results with Salinity Graphs along Transects (Figures 17-22) ...... 37 Results with Historic Oyster Reefs (Figures 23-28) ...... 44 Results with USACE Previously Proposed Oyster Targets (Figures 29-34) ...... 51 Discussion ...... 58 General Findings ...... 58 Biloxi Marsh and Breton Sound Sub-Basins...... 59 Comparison between COOS and SOOS ...... 59 Limitations and Scope ...... 60 Implications for the Oyster Fishery ...... 62 Implications for Coastal Restoration Now and in the Future ...... 63 Baseline Conditions and Future Change ...... 63 Diversion Operations ...... 63 Historic Reef Restoration ...... 64 Barrier Reef Construction ...... 64 Future Work ...... 64 References ...... 65 Appendix 1. Steps and Parameters for Spatial Analysis ...... 71 Appendix 2. Guide to ArcMap GIS Analysis Implementation ...... 76

Page 4 of 85

List of Figures

Figure 1. USGS Landsat imagery of the Hydrocoast salinity study area ...... 14 Figure 2. Optimal salinity regime associated with high seed production (Chatry et al. 1983) .... 16 Figure 3. Oyster HSI (SOOS) variables ...... 18 Figure 4. Areal extent of Biloxi Marsh and Breton Sound Sub-Basins...... 19 Figure 5. Frequency of hypoxia occurrence in the Pontchartrain Basin ...... 21 Figure 6. Salinity suitability surfaces for each year and each methodology...... 25 Figure 7. Multiyear average salinity suitability for COOS (2013-2015) ...... 27 Figure 8. Multiyear average salinity suitability for SOOS (2013-2015) ...... 27 Figure 9. Yearly maps showing results of sub-basin analyses ...... 28 Figure 10. Areal coverage measures for sub-basin, year, and class ...... 29 Figure 11. 2013 COOS results with 2012 State Master Plan projects ...... 31 Figure 12. 2013 SOOS results with 2012 State Master Plan projects...... 32 Figure 13. 2014 COOS results with 2012 State Master Plan projects ...... 33 Figure 14. 2014 SOOS results with 2012 State Master Plan projects...... 34 Figure 15. 2015 COOS results with 2012 State Master Plan projects ...... 35 Figure 16. 2015 SOOS results with 2012 State Master Plan projects...... 36 Figure 17. 2013 COOS results with transect points, Ford and Palmisano Lines ...... 38 Figure 18. 2013 SOOS results with transect points, Ford and Palmisano Lines ...... 39 Figure 19. 2014 COOS results with transect points, Ford and Palmisano Lines ...... 40 Figure 20. 2014 SOOS results with transect points, Ford and Palmisano Lines ...... 41 Figure 21. 2015 COOS results with transect points, Ford and Palmisano Lines ...... 42 Figure 22. 2015 SOOS results with transect points, Ford and Palmisano Lines ...... 43 Figure 23. 2013 COOS results with Louisiana 1910 oyster reef locations ...... 45 Figure 24. 2013 SOOS results with Louisiana 1910 oyster reef locations ...... 46 Figure 25. 2014 COOS results with Louisiana 1910 oyster reef locations ...... 47 Figure 26. 2014 SOOS results with Louisiana 1910 oyster reef locations ...... 48 Figure 27. 2015 COOS results with Louisiana 1910 oyster reef locations ...... 49 Figure 28. 2015 SOOS results with Louisiana 1910 oyster reef locations ...... 50 Figure 29. 2013 COOS results with several USACE salinity restoration targets ...... 52 Figure 30. 2013 SOOS results with several USACE salinity restoration targets ...... 53 Figure 31. 2014 COOS results with several USACE salinity restoration targets ...... 54 Figure 32. 2014 SOOS results with several USACE salinity restoration targets ...... 55 Figure 33. 2015 COOS results with several USACE salinity restoration targets ...... 56 Figure 34. 2015 SOOS results with several USACE salinity restoration targets ...... 57

Page 5 of 85

List of Tables

Table 1. Location, size, and total cost of planned reefs from the 2012 CMP...... 13 Table 2. Optimal oyster salinity values by month from Chatry et al. (1983) ...... 15 Table 3. Areal calculations by class, year, and methodology ...... 24

Page 6 of 85

Executive Summary

Eastern oysters (Crassostrea virginica; hereafter oysters) are bivalves found in estuarine and coastal waters from the Gulf of St. Lawrence in Canada, down the western Atlantic coastline through the Gulf of Mexico into the Caribbean and down to the Brazilian coastline (Gunter 1951, Buroker 1983). They are common in sounds, bays, tidal creeks and bayous from depths ranging from intertidal to 30 m (Galtsoff, 1964). Some have referred to oysters as ecosystem engineers because of the ecosystem services they provide (e.g., Gutiérrez et al., 2003; Grabowski and Peterson, 2007).

Because oysters are commercially harvested within the Pontchartrain Basin, and are of ecologic, commercial and cultural significance, the Lake Pontchartrain Basin Foundation (LPBF) mapped oyster salinity suitability in the basin. LPBF used two techniques derived from approaches taken by Mark Chatry and others (1983) and by Thomas Soniat (2012). Chatry and others (1983) identified an ideal salinity regime for each month based upon empirical data collected from public seed ground within the Breton Sound Basin. The model used by Soniat (2012) encompasses four parameters that characterize optima for salinity and substrate based on theoretical values found in literature and a field validation. These two approaches are referred to as Chatry Optimal Oyster Salinity (COOS) Regime and Soniat Optimal Oyster Salinity (SOOS) Regime and were applied to surface water salinity information from LPBF Hydrocoast Maps of the Pontchartrain Basin to identify the areas with the most optimal oyster salinities for each year from 2013 through 2015(saveourlake.org/coastal-hydromap.php).

Hydrocoast Maps and a written report are produced biweekly, and can be found at SaveOurLake.org. These maps are popular with local fishers who use them for real-time fishing efforts and with coastal scientists who use them for understanding the shifting baseline conditions of the estuary. The Hydrocoast Maps characterize estuarine conditions in the basin and include a salinity, habitat, weather, water quality and biological map. For this study, biweekly salinity surfaces based on isohalines were used for both the COOS and SOOS approaches to estimate the annual oyster suitability for 2013, 2014 and 2015 in the Pontchartrain Basin.

Overall, there was good agreement between the resulting most suitable categories for the SOOS and COOS methodologies, with the majority of the COOS most suitable areas co-occurring near or within the SOOS most suitable areas. Where there were geographic discrepancies, the SOOS approach suggested ideal oyster conditions extended lower in the estuary (i.e., toward the Gulf of Mexico) than the COOS method. These discrepancies are due to differences in the methodologies. Chatry et al. (1983) ideal salinities were based on long-term data sets of oyster production and salinity. Monthly salinities of years with greatest oyster production were used to define an optimum annual salinity profile. Soniat’s (2012) index is based on the requirement that

Page 7 of 85

the reefs be self-sustaining and thus incorporates the demands of a higher salinity for optimal reproduction and for long-term survival.

Some observed spatial differences among years should be expected as average salinities within estuaries vary across years. Resulting maps indicated an up-estuary (i.e., more inland) shift in most suitable areas from 2013 to 2015 for both methodologies. Salinity in the Pontchartrain Basin estuary depends on area rainfall, river discharges, freshwater diversion operation, and storms. There were no large freshwater diversion openings (e.g., Bonnet Carré Spillway) and no large tropical storms in this area during our study period. Therefore, a closer inspection of area rainfall and river discharges may be warranted to see if they are associated with the observed up estuary shift.

Our findings also indicated more suitable oyster salinity within the Biloxi Marsh than Breton Sound, with the SOOS methodology indicating more than four times greater area of the highest classification in Biloxi Marsh than Breton Sound. These findings corroborate other studies that indicated better recent oyster resources and harvests in Biloxi Marsh than Breton Sound. For instance, in 2014, approximately 15 times more seed oysters and seven times more sack oysters were harvested in the management area including Biloxi Marsh than Breton Sound (LDWF 2014). In contrast, from 1992 through 2001, the management area that includes Breton Sound had much higher harvesting rates (up to three orders of magnitude for some between year comparisons) than 2011-2014.

Several factors other than salinity influence oyster populations, and individual growth and mortality. Other studies on oyster habitat suitability from this area typically include a measurement of bottom character (Soniat et al. 2012, Swannick et al. 2014). The results presented here may not indicate areas where oysters occur, because of the exclusion of a bottom character parameter. Presence of hard bottom (cultch) is necessary for oyster larvae settlement and growth, and quality of hard bottom, such as vertical relief, may also be important (Galtsoff 1964, Schulte et al. 2009). That said, “good” areas identified in this analysis could be interpreted as good oyster salinity habitat and therefore where oysters may occur given suitable substrate. Additionally, restoration and management (e.g., cultch planting) activities could be planned using these results. Hypoxia, or low bottom dissolved oxygen levels, within the Pontchartrain Basin, has been suggested as having a negative impact on oyster resources in the Pontchartrain Basin estuaries (LDWF 2011). Therefore, hypoxia should be considered when classifying oyster habitat suitability. The periodic occurrence of hypoxia in Chandeleur, Breton, and Mississippi Sounds may limit the extent of oyster productivity (Lopez et al. 2010, Henkel et al. 2012, Moshogianis et al. 2012, Moshogianis et al. 2013). In addition to low bottom dissolved oxygen, hypoxic waters are also typically vertically stratified. This indicates that surface salinities, in some areas, are different than bottom salinities. Surface salinities were used for all of the analyses, when bottom salinities would have been better to characterize oyster habitat. Therefore,

Page 8 of 85

complete vertical mixing was an assumption in this study that may not hold true uniformly across the study’s area and time.

Louisiana’s Comprehensive Master Plan for a Sustainable Coast, authored by the Coastal Protection and Restoration Authority of Louisiana, calls for bioengineered oyster reefs in the Biloxi Marsh to improve the fishery and to serve as a breakwater against shoreline erosion (CPRA 2012). This plan suggested that vertical reefs would be particularly useful against storm surge in areas where sea level rise and subsidence prevent land-building as a tenable strategy. Creation of over 52 km of oyster reef to serve as a barrier to wave and surge action was included in the plan (CPRA 2012). In Louisiana, over 400,000 ha of coastal land have been lost since 1932 (Couvillion 2011). Due to the large-scale coastal land loss across the entire coast of Louisiana, the state has proposed sediment diversions by creating artificial outlets from the Mississippi River (CPRA 2012). These diversions would deliver much needed sediment, but would freshen and may promote stratification and hypoxia in parts of the Pontchartrain Basin estuary and in turn affect the oyster fishery. These results may provide baseline oyster salinity conditions before large planned Mississippi River sediment diversions within the study area come online (CPRA 2012). This also represents a new application of Hydrocoast salinity data produced by LPBF since 2012.

Page 9 of 85

Introduction

Two techniques are utilized to map oyster habitat suitability in the Pontchartrain Basin within coastal Louisiana. The Lake Pontchartrain Basin Foundation (LPBF) considered approaches taken by Chatry et al. (1983) and Soniat (2012), who modeled ideal salinity conditions for oysters. The two approaches differ in that one is derived from historical spatfall and seed production observations in three Louisiana oyster seed grounds (Chatry et al. 1983); whereas the other is based on theoretically ideal conditions for self-sustaining reefs (Soniat 2012). These two approaches were applied to surface water salinity information from LPBF Hydrocoast maps (http://saveourlake.org/coastal-hydromap.php) to identify areas best suited to oyster production and habitat.

Life History Traits The eastern oyster (Crassostrea virginica; hereafter oyster) is a bivalve found in estuarine and coastal waters from the Gulf of St. Lawrence in Canada down the western Atlantic coastline through the Gulf of Mexico into the Caribbean and south to the Brazilian coastline (Gunter 1951, Buroker 1983). They are common in sounds, bays, tidal creeks and bayous from depths ranging from intertidal to 30 m (Galtsoff 1964). Some have referred to oysters as “ecosystem engineers” because of the ecosystem services they provide (e.g., Gutiérrez et al. 2003, Grabowski and Peterson 2007). Adults are sessile and adhere to hard substrate forming large conglomerates that function as reefs, providing important habitat for many aquatic species (Zimmerman et al. 1989). They feed by drawing water across their gills which can remove pollutants, excess nutrients, and reduce turbidity in the water column (Newell et al. 2005).

Oysters have at least four different life stages (Galtsoff 1964):

 Fertilized egg (Zygote) o Short lived  Larvae o Planktonic, free swimming, about 2 weeks o Stage where oysters distribute across estuaries o Pediveliger is oldest stage where they can crawl and attach  Juvenile oyster o Has shell, is attached to hard substrate, and filter feeds o Not sexually mature, but other than this functions similarly to adults o May only be a few months o Sometimes referred to as “spat” (< 2.54 cm in shell length) or “seed” (< 7.62 cm)  Adult o Same as juvenile, but is sexually mature

Page 10 of 85

o Can reach sexual maturity in less than 6 months for some populations o Sometimes referred to as “seed” (< 7.62 cm) or “sack” (> 7.62 cm)

Oysters in Louisiana typically spawn when water temperatures increase to 25° C, and in waters with extended periods above 25° C they have been observed to spawn multiple times (Soniat, personal correspondence, 2016). Using 25° C as an indicator, the spawning season can last four months in Louisiana (waterdata.usgs.gov). Settled oysters are commonly divided into three size groups. Spat are recently settled juveniles, at approximately 2.54 cm shell length they are referred to as seed oysters, and at 7.62 cm shell length they are referred to as sack oysters, which is market size in Louisiana (LDWF 2011).

Oyster growth and populations are regulated by both internal and external factors. Previous studies often relate these factors to abiotic water conditions (Galtsoff 1964). They have a broad tolerance of environmental conditions and have been observed in waters ranging from brackish (5 ppt) to full seawater (35 ppt) and in water temperatures from 2° C to 36° C (Galstoff 1964). They can close their shell, decrease their metabolism and enter periods of dormancy that can last several months when conditions are poor (Andrews 1966). A group of carnivorous gastropods, commonly referred to as “oyster drills”, probably have the largest predacious impact on oyster populations (Galtsoff 1964). Oyster drills can cause high oyster mortalities, with mortalities above 90% for small, recently settled often reported (Gosselin and Qian 1997). Oyster drills can be a number of different species (e.g., Galstoff 1964 mentions 10), of which Stramonita haemastoma may be of greatest concern to oysters in Louisiana (Garton and Stickle 1980). Typically S. haemastoma do not occur in waters with salinity below 15 ppt, and higher oyster predation rates have been associated with higher salinities, up to 30 ppt (Garton and Stickle 1980). Furthermore, their predation rates on oysters are also influenced by temperature and interactions between temperature and salinity (Garton and Stickle 1980). Perkinsus marinus, often called Dermo, is a protozoan oyster pathogen whose infestations in oysters increase as water temperature increases until approximately 30° C (Chu and Greene 1989). P. marinus is probably the primary oyster pathogen that can cause oyster deaths. Winter temperatures can decrease P. marinus infection rates in oysters, but in some areas, such as Louisiana, winters may not always get cold enough for P. marinus to die back (Chu and La Peyre 1993).

Oyster Fishery in Louisiana The oyster fishery in Louisiana is an important fishery locally and nationally with Native American roots that precede European settlement. Historically, oysters were harvested on wild reefs and used locally. Today, Louisiana’s oyster fishery primarily involves cultivation and the utilization of different habitats for different life stages and sizes. Starting in the 1840s-50s, private oyster beds were set aside for cultivation (Deseran and Riden 2000). Typically, oyster larvae, oyster cultch, juvenile, and/or adult oysters are transplanted from public reefs to private water bottoms. After this, oysters grow and are harvested from leased water bottoms. The fishery represents the state’s oldest largest commercial fishery until 1925 (Deseran and Riden 2000). In

Page 11 of 85

2014, the Louisiana oyster fishery ranked 5th in terms of landings weight with 5,757,013 kg (behind Gulf Menhaden, white shrimp, brown shrimp, and blue crabs) and 4th in dock value at $67,481,540 (behind white shrimp, brown shrimp, and Gulf Menhaden; st.nmfs.noaa.gov). From 1982 to 2014 Louisiana’s oyster fishery produced more oysters than any other state by weight in 29 of 33 years (st.nmfs.noaa.gov). Two of the lower production years were 2005 and 2006, when Hurricane’s Katrina and Rita devastated coastal Louisiana and another was 2010, the year of the Deepwater Horizon Oil Spill.

Oyster Reefs for Coastal Protection In addition to providing habitat for many aquatic organisms and an important commercial fishery, oysters can increase coastal protection. For instance, oyster reefs can stabilize shorelines (Piazza et al. 2005), attenuate and dampen waves (Meyer et al. 1997), transfer sediment and particulate organic material from the water column to the benthos (Coen et al. 2007), and improve water quality (Coen et al. 2007). As such, oyster reef construction has been included in many coastal protection plans (e.g., LPBF 2006, Borsje et al. 2011, CPRA 2012). However, careful consideration when planning to use oysters in coastal restoration is warranted. Oysters will not thrive in all coastal environments (Galtsoff 1964), are not effective for all projects (Pomeroy et al. 2006, Coen et al. 2007, Borsje et al. 2011), and reef size can be limited by resource availability and habitat (Borsje et al. 2011). Location, purpose, and scale along with clear goals and long term evaluations are all important factors when using oysters for coastal protection (Borsje et al. 2011, La Peyre et al. 2014). Studies, such as this one, examining oyster habitats, may be used to better plan for future restoration efforts involving oyster reefs.

Oysters in Louisiana’s Comprehensive Master Plan for a Sustainable Coast In Louisiana, more than 400,000 hectares of coastal land have been lost since 1932 (Couvillion 2011). Planning and implementing risk reduction and coastal restoration have been and continue to be important (Peyronnin et al. 2013). In 2007, the first Louisiana’s Comprehensive Master Plan for a Sustainable Coast, authored by the Coastal Protection and Restoration Authority of Louisiana (CPRA), was produced, with an update in 2012, hereafter referred to as CMP (CPRA 2007 and 2012). These documents provide a comprehensive, science-based plan for restoration of coastal Louisiana. Oysters are mentioned in both documents in two different ways. The first mentions using a Habitat Suitability Index (HSI) to understand oyster habitat in future planning scenarios, the second calls for engineering and establishing new oyster reefs (CPRA 2012, Soniat et al. 2013). Specifically, 50 km of oyster reef construction is planned for three locations at a cost of $100,000,000 (Table 1; CPRA 2012, amendment A2). The goals of these projects are to improve oyster propagation and decrease wave attenuation and models suggest collectively they could increase land area by as much as 800 hectares by 2050 (CPRA 2012, amendment A2).

Page 12 of 85

Table 1. Location, size, and total cost over a 50- year period (includes construction and maintenance) of the three planned reefs from the 2012 CMP. Location Reef Size (km) Total cost East Cote Blanche Bay 9.14 $21,826,000 West Cote Blanche Bay 8.53 $20,322,000 Biloxi Marsh 34.44 $83,732,000 Total 52.12 $125,880,000

Estimating Oyster Habitat Suitability Identifying the optimal conditions for oyster propagation, growth and harvesting is a key factor in delineating areas best suited for oyster reef restoration and production. Determining optimal habitat conditions for oysters has been a goal of scientists and resource managers dating back to the 1800s, with a comprehensive report published by the US Fish and Wildlife Service’s (USFWS) Fishery Bureau in 1964 (Galstoff 1964). The USFWS authored a series of HSIs for many different species in the 1980s, with a goal of determining hypothetical quantitative models of species-habitat relationships. An oyster HSI was created by the USFWS covering two life stages, larvae and adult (Cake 1983). A field evaluation of Cake’s (1983) HSI suggested that modification of some variables should be considered (Soniat and Brody 1988). For determining restoration targets at larger spatial scales, a four parameter model using three salinity parameters was found to be useful for both data rich and data poor estuaries (Swannick et al. 2014), suggesting salinity may be the most influential parameter.

In a study completed in 1983 for the Louisiana Department of Wildlife and Fisheries (LDWF), Chatry et al. (1983) used historical oyster seed production data to examine the relationship between salinities and seed oyster production in three sites in southeast Louisiana. This research resulted in the establishment of an optimal salinity regime for 12 calendar months using salinities observed prior to good seed production years. Salinity in the setting year, particularly in the summer, was found to be a prime determinant of seed production in the ensuing year.

A more recent method for evaluating conditions ideal for oyster habitat was developed by Soniat and is available online at http://oystersentinel.org (Soniat, n.d., Soniat and Brody 1988, Soniat 2012). This HSI modifies the work of Cake (1983) and encompasses four parameters that characterize optima for salinity and substrate based on theoretical values found in literature and field validation (Soniat and Brody 1988). These two methods were applied to Hydrocoast surface water salinities to determine areas most aligned with optimal salinities.

Materials and Methods

Study Area Description The Pontchartrain Basin is an estuary in southeast Louisiana characterized by tidal water with higher salinities interacting with freshwater riverine discharge and contains fresh, intermediate, brackish and saline environments. Sampled and gauged salinity measures show this area has

Page 13 of 85

salinities ranging from near 0 ppt on the far western side of the basin, to 37 ppt and higher near the most seaward boundary. For purposes of oyster propagation and growth, salinity is important and duration of certain salinity regimes can be critical (Galstoff 1964, Cake 1983, Soniat and Brody 1988). Salinity within the basin is influenced by many factors, including tides, wind, river stage, precipitation, and channelization. The study area for the suitability analyses is comprised primarily of the basin’s water and wetland areas covered by Hydrocoast salinity mapping during the 2013-2015 calendar years (Figure 1).

Figure 1. USGS Landsat imagery of the Hydrocoast salinity study area is shown with USGS 2013 vegetation type (Sasser et al. 2014). Hydrocoast Salinity LPBF produces a biweekly set of maps characterizing conditions in the basin, including surface water salinity. These five maps are called Hydrocoast maps, and the set includes salinity, habitat, weather, water quality, and biological maps. The maps are produced using both primary and secondary data compiled from fieldwork, and federal and state agencies. During each Hydrocoast mapping period, contours representing the surface water salinity gradient in the basin are created using salinity data from fixed stations and from supplemental data collected by LPBF. Isohalines are manually delineated using GIS software. Isohaline generation takes into account coastal

Page 14 of 85

processes, topography, hydrology, rainfall, wind characteristics, tides, currents, and bathymetry (Lopez 2015).

As part of long-term Hydrocoast data analysis, LPBF created 70 continuous raster (grid-based) surfaces of salinity in the basin, one for each mapping period. Using GIS software, continuous raster surfaces were created for Hydrocoast salinity data from Jan 2013 – Dec 2015. These surfaces, generated by interpolating biweekly salinity contours and sampled data in the basin, were subsequently analyzed using two oyster salinity suitability models to generate a “best oyster area” map for each approach and each calendar year.

Chatry Optimal Oyster Salinity (COOS) Regime Using spat and seed production observations from 1971-1981, Chatry et al. (1983) identified an ideal salinity regime for each calendar month. The regime reflected salinities during eight “good” (>20 seed oysters per square meter) seed production years per location. Chatry et al. (1983) observed and documented oyster setting, seed and salinity over a ten-year period at three locations in Breton Sound. Optimal monthly salinity values documented were applied to Hydrocoast data to identify locations most aligned with ideal values (Table 2). The monthly salinity mean and range associated with “good” production years can be seen in Figure 2. For the purposes of this analysis, the mean salinity value was used to identify good oyster salinity areas in the study area.

Table 2. Optimal oyster salinity values by month (ppt), from Chatry’s LDWF study (Chatry et al. 1983, USACE 1984). Standard error varies by month. Optimum Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Salinity (ppt) 16.4 14.4 11.6 8.0 7.0 12.5 12.7 15.7 17.0 16.8 16.1 15.7

Page 15 of 85

Figure 2. Optimal salinity regime associated with high seed production (Chatry et al. 1983).

Interpolated Hydrocoast salinity surfaces were averaged for each month. These mean monthly salinity surfaces were evaluated against Chatry’s monthly optima. To represent total divergence from optimal monthly salinity, regardless of positive or negative direction, absolute values of the difference between each Hydrocoast monthly average and Chatry’s optimal salinity for that month were calculated. The monthly differences were then summed to give a total yearly divergence from optimal for each year.

The resulting summed surface represents the total deviation of that year’s mean monthly surface salinity from optimal oyster conditions based on Chatry’s observations. This methodology is referred to as the Chatry Optimal Oyster Salinity (COOS) Regime. Lower values indicate more suitable areas for oyster production based on this approach. Resulting summed divergences were mapped according to the relative level of divergence from optimal conditions.

Soniat Optimal Oyster Salinity (SOOS) Regime Thomas Soniat’s HSI offers another approach for evaluating conditions for oyster habitat. Soniat (2012) uses three salinity variables and one substrate variable to determine an area’s suitability for oyster propagation. The premise of this approach is that the primary parameters of good oyster habitat are suitable salinity over suitable cultch, defined as hard substrate (Soniat 2012).

Page 16 of 85

Parameters are percent suitable cultch cover, mean annual salinity, mean spawning season salinity and minimum monthly salinity. This model differs from Cake’s (1983) in that it combines larvae and adult salinity requirements into a single component, and does not include historical oyster stock, disease prevalence or predator density. Thus it is a more simplistic approach that may lend itself better to limited data availability (Soniat 2012; Swannack et al. 2014). Soniat developed linear curves that relate salinity values to a dimensionless suitability index that ranges from 0 (unsuitable) to 1 (ideally suitable; Figure 3). All variables have equal weight, and the composite HSI index is the geometric mean of all variables. For this analysis, in the absence of detailed bottom information, we assumed substrate (V1 - cultch) was 100% suitable coverage in all areas, and assigned it an index value of 1, in effect negating its contribution to the spatial analysis undertaken here. In addition we assumed each raster grid cell was 0% land. This study focused solely on salinity to identify good oyster propagation areas according to the remaining three variables:

Variable 2 (V2) - Salinity during the May – September spawning season Variable 3 (V3) – Minimum monthly salinity Variable 4 (V4) - Average annual salinity (surrogate for historic mean salinity)

To apply the HSI to Hydrocoast data, biweekly Hydrocoast salinity data was interpolated for 2013-2015. Then, using Microsoft Excel (v. 14) and ESRI ArcGIS (v. 10.2) linear relationships were applied to the salinity surfaces to get an index surface with values ranging from 0 to 1 for each of the three salinity variables. Finally, the geometric mean of those three surfaces was calculated to get a composite HSI surface for each year.

V2- Mean salinity during spawning season Hydrocoast surface salinity grids were selected for the time periods between May 1 and September 30 for each year. Values for those grid surfaces were averaged to create one 500- meter gridded surface representing mean spawning season salinity. A V2 index surface was derived from that surface by applying Soniat’s linear equations. V3 -Minimum monthly mean salinity Hydrocoast salinity grids were averaged for each month. The twelve monthly-average surfaces were compared within GIS software and the minimum value for each grid cell was selected to generate a new, single surface representing the minimum monthly average for that year. Linear relationships provided by Soniat were applied to create a V3 index surface. This was calculated for each year. V4- Historic mean salinity The HSI uses average annual salinity as a surrogate for historic mean. This variable was calculated for 2013-2015 Hydrocoast data using GIS software. Again, the HSI equations for V4 were applied to the resulting annual mean surface to calculate a 0-to-1 index surface for this variable.

Page 17 of 85

Figure 3. Oyster HSI variables: V1-Cultch, V2-Mean salinity during spawning season, V3-Minimum salinity, and V4-Historic mean salinity (mean annual salinity; Soniat et al. n.d., and personal correspondence).

Salinity Suitability Index Composite Finally, the three index surfaces were combined by calculating the geometric mean to create the final salinity suitability surface for each year, using the following formula: SSI = (V2 * V3 * V4)1/3

Final surfaces for each year were mapped according to their level of suitability for oysters.

Map Classifications of COOS and SOOS Results For both COOS and SOOS methods, resulting yearly raster data was classified into five ordinal classes (1-5). Class 1 indicates the most suitable salinity and Class 5 indicates the least suitable. The yearly classified result for each the COOS and SOOS method were averaged over the entire 3-year study period. A single map for each method was generated showing the average classified suitability. On all maps the classes are represented by a gradient of dark green for Class 1 to dark brown for Class 5. It is important to understand that the 1-5 class scale relates to COOS and SOOS independently and so are not necessarily equal to each other in suitability.

Page 18 of 85

Comparison between Biloxi Marsh and Breton Sound Sub-Basins Results of the COOS and SOOS analysis were extracted for two sub-basins, the Biloxi Marsh Sub-Basin and the Breton Sound Sub-Basin (

Figure 4). The Biloxi Marsh Sub-Basin is approximately bounded by the Louisiana/Mississippi state line to the north, the Breton National Wildlife Refuge to the east, the southwestern extent of the MRGO to the south, and Lake Borgne to the west. The Breton Sound Sub-Basin’s northern border is from the juncture of the Inner Harbor Navigation Canal – Lake Borgne Surge Barrier and the Mississippi River Levee in Caernarvon, LA, along the southernmost part of the Inner Harbor Navigation Canal – Lake Borgne Surge Barrier to the southwestern terminus of the Biloxi Marsh Sub-Basin on the Mississippi River Gulf outlet near Doulutus Canal. The eastern border is of this Basin is the MRGO, the southern border is the northern extent of the Louisiana Department of Health and Hospital’s Oyster Harvest Area 8 (which is currently closed to oyster harvest) and the northwestern edge of the Breton Sound National Wildlife Refuge, and the western border is the eastern edge of the Mississippi River.

Figure 4. Areal extent of two Sub-Basins considered for further analysis. The approximate boundaries for the Biloxi Marsh Sub-Basin are the Louisiana/Mississippi state line to the north, Breton National Wildlife Refuge (BNWR) to the east, the Mississippi River Gulf Outlet (MRGO) to the south, and the western extent of the Biloxi Marsh to the west. The Breton Sound Sub-Basin’s approximate boundaries are the Inner Harbor Navigation Canal – Lake Borgne Surge Barrier’s southern extent to the north, the MRGO to the east, the BNWR and the Louisiana Department of Health and Hospitals’ Oyster Harvest Area 8 (currently closed to oyster harvest) to the south, and the east bank of the Mississippi River to the west. See text for more details.

Page 19 of 85

Hypoxia Polygons and Contours Hypoxia can affect the distribution of benthic estuarine organisms. To better understand the spatial extent of its impacts, two different visualizations of hypoxia within our study area were created. Hypoxia is defined here as water that has less than 2 mg/l of dissolved oxygen, and often occurs near the bottom of the water column. Since 2008, LPBF and other researchers have periodically monitored hypoxia within parts of the Pontchartrain Basin (Lopez et al. 2010, Henkel et al. 2012, Moshogianis et al. 2012, Moshogianis et al. 2013). During LPBF surveys, a YSI hand held water quality meter with a 30 m cable was used to measure dissolved oxygen concentrations at approximately two feet above bottom, mid-depth, and 2 feet below the surface. Two transects were selected which crossed the deepest axial through portions of Mississippi, Chandeleur, Breton Sounds (Moshogianis et al, 2013). One transect extends through the Cat Island Channel in Mississippi Sound to a deep portion near the northern end of Chandeleur Sound. Another transect extends from eastern Ship Island southward to the east-west center of Chandeleur Sound. These locations were monitored periodically by LPBF scientists.

Hypoxic areas identified from these surveys were examined for recurrence frequency. Polygons delineating hypoxic areas from all survey years (2008 and 2010-2015) were converted to raster layers. Each raster’s cell values were set equal to 1, indicating hypoxia was observed at that location that year. All rasters were combined into a single raster by summing cell values using the “raster calculator” tool in ESRI ArcMap 10.3. The resulting raster represents the geographic extent of all areas where hypoxia was observed, with each cell value indicating the number of times it was observed. This count raster was converted to a percent raster using the formula: years observed / years surveyed * 100. Note that Chandeleur Sound was surveyed over 7 years, 2008-2015 (no surveys were conducted in 2009), whereas Breton Sound was surveyed over 3 years, 2013-2015. Hypoxia occurred in both regions. There is a 33% - 66% occurrence rate in Breton Sound, and 14% - 100% in Chandeleur Sound (Figure 5).

Page 20 of 85

Figure 5. Frequency of hypoxia occurrence in Pontchartrain Basin, based on surveys by LPBF and other researchers. Frequency contours are depicted on map with State Master Plan projects (Figures 11-16). General area of hypoxia occurrence is shown on remaining overlay maps (Figures 17-34).

Other Oyster Suitability Indicators and Restoration Targets Proposed USACE Oyster Targets Analyses for suitable oyster salinity were examined in the context of oyster reef restoration goals and diversion operations. Maps were created showing analysis results with two salinity restoration targets.

Chatry Line. United States Army Corps of Engineers (USACE) identified the “Chatry Line” as the salinity target line for MRGO ecosystem restoration in the Biloxi Marsh (USACE 2012). It represents a linear location at which meeting Chatry’s optima four years out of ten, or 40% of the time, was a goal adopted by USACE for several projects, such as the Bonnet Carré Spillway and MRGO ecosystem restoration (USACE 1984, USACE 2012). Meeting this goal at the Chatry Line was considered a determinant as to whether the historic salinity regime had been restored (USACE 2012). As of 2008, the Biloxi Marsh had not met that threshold (Van den Heuvel 2010).

Page 21 of 85

USACE Primary Zone. USACE identified a primary “with-project productive oyster zone” for restoring habitat to conditions optimal for oyster habitat in the Biloxi Marsh (USACE 1984). Within this zone, the USACE assumed a production rate of 20 oysters per square-meter cultch.

Other targets. Other potential targets include the Ford Line and the Palmisano Line, both addressed below under “Salinity Gradient Points”.

Historic Oyster Reefs Examining current conditions in the context of historic reef locations can be instructive in both where reef impacts from changing salinities have occurred, and where restoration might focus (LPBF 2006, USACE 1984). Spatial data depicting early 20th century oyster reef locations were mapped atop the Chatry and Soniat analysis results. Both 1910 surveyed reefs in St. Bernard Parish (LBOC 1912) and data from Mississippi (Mississippi Department of Marine Resources) showing historic locations in Mississippi Sound were displayed.

State Master Plan Projects Project types and locations from Louisiana's Comprehensive Master Plan for a Sustainable Coast, including oyster reef construction, were overlaid atop the 2013-2015 Chatry and Soniat analysis results. Hypoxia contours were also included on these maps and represent areas where occasional or frequent hypoxia may have occurred.

Salinity Gradient Points Reference points were selected along the salinity gradient in the basin. Resulting values from each analysis were extracted for the point locations and graphed against Chatry’s and Soniat’s optimal salinity regimes. These were created with the Ford and Palmisano lines in order to spatially compare each method with a single average condition, and to see how those conditions changed during the three years analyzed.

Ford Line. Annual mean salinity of 15 ppt line established as a general salinity target by Theodore B. Ford with LDWF (Chatry et al. 1983) because that salinity level prohibits proliferation of the oyster drills. This line is hereafter referred to as the Ford Line and was a salinity target considered by the USACE for restoration of oyster habitat at historical reef locations (USACE 1984). This study did not recommend continuous maintenance of 15 ppt at the Ford Line, but rather allowing for some salinity variation mimicking historical conditions occurring with Mississippi River bank overflow (USACE 1984). In fact, rather than a constant salinity target, the Ford Line represents a reference location at which ideal salinities aligned with Chatry’s regime for oysters should be achieved using diversion operations (Chatry et al. 1983, USACE 1984).

Palmisano Line. This line is the easternmost boundary line with a maximum salinity of 15 ppt and was established as another salinity target. It may be considered the “brackish-saline marsh contact” and a starting line for identifying salinity requirements for purposes of marsh plant growth and related wildlife productivity (USACE 1984).

Page 22 of 85

USACE proposed maintaining 15 ppt at the Palmisano Line October - March, and at the Ford Line April – September (USACE 1984).

Results

The ensuing maps show 2013-2015 surface water salinity in the Lake Pontchartrain Basin categorized according to suitability for oyster production and habitat. Using each model (COOS and SOOS) and year (2013-2015; and multiyear average 2013-15), a set of maps was generated depicting 500-meter-grids of continuous suitability surfaces. The level of salinity suitability is classified from Class 1 to Class 5 which corresponds to a range of most suitable salinity (dark green tone; Class 1) to least suitable (brown tone; Class 5). These grid surfaces were mapped showing the geographic extent of the salinity suitability regimes over the time period studied. Additional relevant data sets were mapped to provide a context for habitat suitability, oyster management, and reef creation and restoration planning. Maps are organized under these headings:

1. Overall Salinity Suitability – shows geographic extent of suitability classes for each year/model combination (Table 3, Figure 6) 2. Multi-Year Average Salinity Suitability – shows suitability class for each model averaged over the entire study period, 2013-2015 (Figures 7-8) 3. Biloxi Marsh and Breton Sound Sub-Basins – shows salinity suitability extracted for each sub-basin for each year/model combination (Figure 8-9) 4. Overlay Maps – shows yearly model results overlaid with: a) State Master Plan Projects – State Master Plan 2012 project type and location (Figures 11-16) b) Salinity Along Transect Points – Point locations along salinity gradient extracted salinity values graphed against model ideals, with the Ford and Palmisano lines (Figures 17-22) c) Historic Oyster Reefs – Early 20th century reef locations (Figures 23-28) d) Target Areas –Restoration targets identified by USACE in previous studies (Figures 29-34)

Study Area Salinity Suitability Results The major observations and trends for this section are summarized below:

 SOOS indicated much larger Class 1 areal coverage (Table 3)

 SOOS indicated better habitat was more down estuary than COOS (Figure 5)

 In 2013, all COOS Class 1 regions overlap in part with SOOS Class 1 band

 In 2014, three of five COOS Class 1 regions overlap in part with SOOS Class 1 band

Page 23 of 85

 In 2015, all COOS Class 1 regions overlap in part with SOOS class 1 band

 In general, Class 1 and 2 regions from COOS overlap near up-estuary border of SOOS Class 1 and 2 bands

Table 3. Areal calculations for each suitability class, year, and methodology across the entire study area. Note: classifications for each methodology (COOS and SOOS) were developed independently and are not necessarily equal in suitability. Chatry Classes; 1=best (COOS) 2013 km2 2014 km2 2015 km2 1 40 65 63 2 2,368 2,306 1,914 3 2,631 2,139 2,200 4 2,156 2,186 2,111 5 12,523 13,021 13,429 Soniat Classes; 1=best (SOOS) 2013 km2 2014 km2 2015 km2 1 3,437 3,678 3,338 2 3,215 3,337 2,607 3 4,209 5,379 4,764 4 2,025 1,321 3,342 5 6,831 6,001 5,666 Study Area Size (km2) 19,717 19,717 19,717

Page 24 of 85

Figure 6. Salinity suitability surfaces generated using Hydrocoast surface water salinity for each year (2013- 2015) and each methodology (Chatry et al. 1983, Soniat 2012). See methods section for more information.

Page 25 of 85

Multi-Year Average Salinity Suitability (2013-2015) The results of the averages across years indicated that the two methodologies identified different areas with most suitable oyster salinities (Figure 8 & Figure 8). Like each yearly result, the multi-year average for the COOS methodology did not indicate a clear band of Class 1 area (dark green in Figure 7). Eight COOS Class 1 areas occurred in non-contiguous regions. The SOOS methodology indicated a contiguous band of Class 1 salinity habitat that at least partially includes five of the eight COOS Class 1 regions. These COOS Class 1 regions occurred at or near the up estuary extrema of the SOOS Class 1 band. The remaining three COOS Class 1 regions occurred up estuary of the SOOS Class 1 band. Overall, the SOOS averages indicated better oyster salinity habitat more down estuary than the COOS averages. It also indicated a significantly larger Class 1 area.

Page 26 of 85

Figure 7. Multiyear average salinity suitability for the COOS methodology using Hydrocoast surface salinities from 2013-2015.

Figure 8. Multiyear average salinity suitability for the SOOS methodology using Hydrocoast surface salinities from 2013-2015.

Page 27 of 85

Biloxi Marsh and Breton Sound Sub-Basins In the Biloxi Marsh, the SOOS methodology suggested more good oyster habitat overall with Class 1 having the most areal coverage within this sub basin, across all years (Figure 9 & Figure 10). At least four times as much Class 1 area was indicated in the Biloxi Marsh Sub-Basin than the Breton Sound Sub-Basin for each year analyzed. Furthermore, more than 40% of the most suitable habitat for the entire Basin occurred within the Biloxi Marsh Sub-Basin. For both the Biloxi Marsh and Breton Sound Sub-Basins in 2013-2015, the COOS methodology indicated that Class 1 had the lowest areal coverage (Figure 9 & Figure 10). The COOS methodology indicated that Class 2 had the second largest areal coverage for all year-sub-basin combinations, except for 2013 for the Biloxi Marsh Sub-Basin where it had the most areal coverage. More Class 1 and Class 2 coverage was observed in the Biloxi Marsh Sub-Basin than the Breton Sound Sub-Basin for each year.

Figure 9. These maps show yearly results of the sub-basin areas. Both the COOS and SOOS methodologies were applied to the Biloxi Marsh and Breton Sound Sub-Basins for all years from 2013-2015.

Page 28 of 85

COOS COOS Biloxi Marsh Sub-Basin Breton Sound Sub-Basin 1800 1800 1600 1600 1400 1400 1200 1200 1000 1000 km2 km2 800 800 600 600 400 400 200 200 0 0 2013 2014 2015 2013 2014 2015

SOOS SOOS Biloxi Marsh Sub-Basin Breton Sound Sub-Basin 1800 1800 1600 1600 1400 1400 1200 1200 1000 1000 km2 km2 800 800 600 600 400 400 200 200 0 0 2013 2014 2015 2013 2014 2015

Figure 10. This figure shows areal coverage in km2 by sub-basin, year, and salinity suitability classification. The SOOS methodology classifications correspond with index values, with Class 1 being > 0.75, Class 2 between 0.5 and 0.75, Class 3 between 0.25 and 0.5, Class 4 being 0 to 0.25, and Class 5 being an index value of 0. The COOS methodology classifications correspond to total annual deviations from Chatry and others (1983) monthly salinity optima, with Class 1 being < 25 ppt, Class 2 being 25 – 50 ppt, Class 3 50 – 75 ppt, Class 4, 75 – 100 ppt, and Class 5 > 100 ppt.

Page 29 of 85

Overlay Maps Annual Analysis Results with State Master Plan Projects (Figures 11-16) The maps below show results of each annual analysis with the SMP projects overlaid. In 2013, the oyster reef restoration project, depicted as a blue line along the northeastern edge of the Biloxi Marsh, lies on the downstream perimeter of the COOS Class 2 band. In 2014 and 2015 none of this project is located within either of the top two classifications for this methodology. The SOOS methodology indicates that at least some of this area lies within the best classification for all years, with the entire project being Class 1 for 2014 and 2015.

Page 30 of 85

Figure 11. Results from the 2013 COOS analysis overlaid with State Master Plan projects. Dark green areas indicate most suitable salinities. State Master Plan projects shown include the construction of an oyster barrier reef in Biloxi Marsh.

Page 31 of 85

Figure 12. Results from the 2013 SOOS analysis overlaid with State Master Plan projects. Dark green areas indicate most suitable salinities. State Master Plan projects shown include the construction of an oyster barrier reef in Biloxi Marsh.

Page 32 of 85

Figure 13. Results from the 2014 COOS analysis overlaid with State Master Plan projects. Dark green areas indicate most suitable salinities. State Master Plan projects shown include the construction of an oyster barrier reef in Biloxi Marsh.

Page 33 of 85

Figure 14. Results from the 2014 SOOS analysis overlaid with State Master Plan projects. Dark green areas indicate most suitable salinities. State Master Plan projects shown include the construction of an oyster barrier reef in Biloxi Marsh.

Page 34 of 85

Figure 15. Results from the 2015 COOS analysis overlaid with State Master Plan projects. Dark green areas indicate most suitable salinities. State Master Plan projects shown include the construction of an oyster barrier reef in Biloxi Marsh.

Page 35 of 85

Figure 16. Results from the 2015 SOOS analysis overlaid with State Master Plan projects. Dark green areas indicate most suitable salinities. State Master Plan projects shown include the construction of an oyster barrier reef in Biloxi Marsh.

Page 36 of 85

Annual Analysis Results with Salinity Graphs along Transects (Figures 17-22) The annual analysis maps with graphs of salinity values at transect point locations can be used to better understand the similarities and differences between the COOS and SOOS methods. The maps show the transect point locations overlaid on the annual analysis results while graphs on the map show the salinities for each transect location over the year. Thus both the spatial extent of the annual suitability as well as the monthly and seasonal salinity variations can be seen together. The graphs on the COOS maps show the monthly deviations from the Chatry et al. (1983) optimal for each transect point allowing for interpretation of what monthly mean salinities may be driving the overall divergence. The graphs on the SOOS maps show the salinity variation for each variable at each point location, which may give further insight into how each location fared compared to the optimal range for each of the three Soniat (2012) salinity variables. For example, on the COOS 2013 map (Figure 17), points 5 and 6 both fall in the same suitability class. The graph shows location 6 experiences higher than optimal salinity in April and November whereas location 5 experiences lower than optimal salinity in January and February. On the SOOS 2015 map (Figure 22), locations 5 and 6 both fall in the most suitable salinity class. However, as seen in the graphs, location 5 is closer to ideal values for minimum monthly mean and annual mean salinities, whereas location 6 is closer to ideal spawning season salinities. The graphs show a finer temporal resolution of how salinities at the point locations vary from optimal for each monthly or seasonal variable.

Page 37 of 85

Figure 17. Results from the 2013 COOS analysis overlaid with transect point locations. Dark green areas indicate most suitable areas. Monthly Hydrocoast salinities were extracted at point locations and graphed against ideal salinities. Potential Ford and Palmisano Lines are also shown (USACE 1984).

Page 38 of 85

Figure 18. Results from the 2013 SOOS analysis overlaid with transect point locations. Dark green areas indicate most suitable salinities. Salinity values for each parameter were derived from Hydrocoast salinity data and graphed against the HSI optimal range. Potential Ford and Palmisano Lines are also shown (USACE 1984).

Page 39 of 85

Figure 19. Results from the 2014 COOS analysis overlaid with transect point locations. Dark green areas indicate most suitable areas. Monthly Hydrocoast salinities were extracted at point locations and graphed against ideal salinities. Potential Ford and Palmisano Lines are also shown (USACE 1984).

Page 40 of 85

Figure 20. Results from the 2014 SOOS analysis overlaid with transect point locations. Dark green areas indicate most suitable salinities. Salinity values for each parameter were derived from Hydrocoast salinity data and graphed against the HSI optimal range. Potential Ford and Palmisano Lines are also shown (USACE 1984).

Page 41 of 85

Figure 21. Results from the 2015 COOS analysis overlaid with transect point locations. Dark green areas indicate most suitable areas. Monthly Hydrocoast salinities were extracted at point locations and graphed against ideal salinities. Potential Ford and Palmisano Lines are also shown (USACE 1984).

Page 42 of 85

Figure 22. Results from the 2015 SOOS analysis overlaid with transect point locations. Dark green areas indicate most suitable salinities. Salinity values for each parameter were derived from Hydrocoast salinity data and graphed against the HSI optimal range. Potential Ford and Palmisano Lines are also shown (USACE 1984).

Page 43 of 85

Results with Historic Oyster Reefs (Figures 23-28) Our results indicated much of the historic (early 20th century) oyster reefs had Class 1 or Class 2 salinity suitability for both methodologies and all three years. Overall, the reefs in Mississippi State waters had high variability both spatially within years and from year to year with respect to salinity suitability using both methods. The St. Bernard Parish historic reefs, observed in 1910, were more consistently within Class 1 or 2 suitability for year/methodology combinations. For example, the SOOS methodology indicated some of these reefs belonging to Class 1 for each year. Some inter-annual variation is noted here, with the most SOOS Class 1 salinity suitability observed for 2015, and the least in 2013. If the top two classifications are considered, 2013 indicated more suitable salinity area than in 2015 for the St. Bernard Parish historic reefs according to the COOS methodology.

Page 44 of 85

Figure 23. Results from the 2013 COOS analysis overlaid with historic oyster reef locations.

Page 45 of 85

Figure 24. Results from the 2013 SOOS analysis overlaid with historic oyster reef locations.

Page 46 of 85

Figure 25. Results from the 2014 COOS analysis overlaid with historic oyster reef locations.

Page 47 of 85

Figure 26. Results from the 2014 SOOS analysis overlaid with historic oyster reef locations.

Page 48 of 85

Figure 27. Results from the 2015 COOS analysis overlaid with historic oyster reef locations.

Page 49 of 85

Figure 28. Results from the 2015 SOOS analysis overlaid with historic oyster reef locations.

Page 50 of 85

Results with USACE Previously Proposed Oyster Targets (Figures 29-34) The Primary Oyster Zone Target Area and the Chatry Line established by the USACE coincided with some Class 1 and 2 salinities by both methodologies. However, some differences across years were apparent. Similar to the patterns observed across the entire study area, the majority of the SOOS methodology’s most suitable habitat was down estuary (i.e., more seaward) in 2013 and moved up estuary in the subsequent years. The Primary Oyster Zone Target Area had the most area in Class 1 for 2015 according to both methodologies (COOS, 55 km2; SOOS, 755 km2). However, if we consider the top two classifications, the COOS methodology indicated that 2014 had the largest area (668 km2 in 2014 vs. 628 km2 for 2013 and 648 km2 for 2015). Across years the SOOS methodology indicated more Class 1 salinities within the USACE primary zone target area.

Page 51 of 85

Figure 29. Results from the 2013 COOS analysis overlaid with two USACE salinity restoration targets (USACE 1984, USACE 2012).

Page 52 of 85

Figure 30. Results from the 2013 SOOS analysis overlaid with two USACE salinity restoration targets (USACE 1984, USACE 2012).

Page 53 of 85

Figure 31. Results from the 2014 COOS analysis overlaid with several USACE salinity restoration targets (USACE 1984, USACE 2012).

Page 54 of 85

Figure 32. Results from the 2014 SOOS analysis overlaid with several USACE salinity restoration targets (USACE 1984, USACE 2012).

Page 55 of 85

Figure 33. Results from the 2015 COOS analysis overlaid with several USACE salinity restoration targets (USACE 1984, USACE 2012).

Page 56 of 85

Figure 34. Results from the 2015 SOOS analysis overlaid with several USACE salinity restoration targets (USACE 1984, USACE 2012).

Page 57 of 85

Discussion

General Findings This study represents a practical interpretation of Hydrocoast surface water salinities. Here, two different oyster salinity suitability models were applied to the Pontchartrain Basin estuary using Hydrocoast salinities. These results are useful to better understand the spatial and temporal variation in oyster salinity suitability by comparing results over the entire estuary for three years. Some of the conclusions and general findings follow below:

 Up-estuary, landward shift in Class 1 and 2 salinities from 2013 to 2015 for both methods  COOS classifications were up estuary from SOOS  Biloxi Marsh Sub-Basin had more suitable salinities than the Breton Sound Sub-Basin  Hydrocoast surface salinity data was useful with inclusion of historical hypoxia and vertical stratification information  CMP oyster projects within our study area overlapped with Class 1 and 2 salinities for both methods  Historic oyster reefs observed in the early 20th century (pre-MRGO construction) overlapped with Class 1 and 2 salinities for both methods  USACE targets overlapped with Class 1 and 2 salinities for both methods  Average suitability maps can be used to better understand what “normal” oyster salinity suitability conditions may be in the study area  These results and subsequent updates may provide important background information before, during, and after coastal restoration activities, such as sediment diversions, come online.

Combining all of this information may help oyster fishers and fishery managers more effectively harvest and manage the resource. The results of these models were compared to other coastal plans, historical data, and management criteria, like the SMP projects, early 20th century oyster reef locations, and USACE criteria, allowing for the possibility of comparing them to more current conditions. Hypoxia observational data were also overlaid with the modeling results. Hypoxia location and frequency of occurrence can affect oyster health and distributions (LDWF 2011, LDWF 2013, LDWF 2014). Synthesizing all of this provides information on the current state of oyster suitability, which could be used as a baseline for impending changes (both natural and anthropogenic).

Generally, both models identified highly suitable salinity areas for oysters throughout the Basin. Mapped results show COOS best areas as smaller regions in somewhat fresher areas of the basin that may be ideally suited for seed production. The SOOS method, on the other hand, identified a large swath of best salinity conditions for self-sustaining oyster reefs from upper Biloxi Marsh across MRGO and down to the Mississippi bird’s foot delta.

Page 58 of 85

Differences in where the salinity suitabilites were observed among years using both the SOOS and COOS methods (e.g., the up-estuary shift in Class 1 and 2 suitabilities noted above). This was expected as salinity in the Pontchartrain Basin estuary is highly variable and depends on area rainfall, river discharges, diversion operation, and storms. There were no large diversion openings (e.g., Bonnet Carré Spillway) and no large tropical storms in this area for these years. Therefore, a closer inspection of area rainfall and river discharges may be warranted to see if they are associated with the spatial differences observed among yearly tabulated salinity suitability maps. The Bonnet Carré Spillway was opened in 2016 and similar analyses have been planned that could allow for better understanding of how these events affect the spatial distribution of suitable oyster salinities.

Biloxi Marsh and Breton Sound Sub-Basins Our findings suggest that the Biloxi Marsh had more suitable oyster salinity habitat than the Breton Sound Sub-Basin, which corroborates other studies. Previous Louisiana oyster stock assessment reports that occurred during the same temporal period as the current study noted better oyster resources and larger harvests in Biloxi Marsh than Breton Sound. The LDWF 2014 Oyster Stock Assessment Report (LDWF 2014) found much higher sack and seed oyster production from their management region that includes the Biloxi Marsh (CSA-1N) than their management region corresponding with the Breton Sound Sub-Basin (CSA-1S). In 2014, 15 times more seed oysters and seven times more sack oysters were harvested in CSA-1N than CSA-1S (LDWF 2014). This finding was not unusual, as preliminary data from the 2015 season indicated similar relationships and the stock assessments from 2011 and 2013 observed much more oyster harvest from CSA-1N than CSA-1S (LDWF 2011, LDWF 2013, Beck 2016). Additionally, results using a shell budget model indicated that harvesting oysters in Breton Sound at these rates did not maintain current reef quality, as defined by density, suggesting that harvest rates have not been sustainable for most years from 2002-2009 and 2015 (Soniat et al. 2012, Beck 2016). Our sub-basin results provide further evidence that recently the Biloxi Marsh area may have experienced better oyster salinities than the Breton Sound area. Low harvest rates and unsustainable fishing rates have not always been observed in this area. From 1992 through 2001 CSA-1S had much higher harvesting rates (up to three orders of magnitude for some between year comparisons) than it did from 2011-2014. Furthermore, CSA-1S had much higher harvesting rates than CSA-1N during this time period and had some of the highest harvest rates in Louisiana (LDWF 2014).

Comparison between COOS and SOOS Some of differences in results between the two methods presented here may be attributed to their differences in approach and the methodologies used when they were developed. Chatry et al. (1983) determined the ideal salinity regime used in COOS for public seed grounds in areas downstream of a proposed freshwater diversion near Breton Sound, Louisiana. Data were collected over a decade in a region that lies within our study area, so these results are geographically relevant. However, there are some limitations to this approach. It was intended to

Page 59 of 85

determine optimal salinity targets for oyster seed production to guide operation of a freshwater diversion. Other than salinity, no variables were mentioned or controlled for in the experimental design; therefore the differences in seed production may not have been attributable to differences in salinity. In addition, this did not test whether higher or lower salinities would have improved seed production for any given year and/or area. Thus the “good” year salinities identified may not be the best possible. In comparison, Soniat and Brody (1988) developed their model based upon literature values, field, and laboratory work on oysters from across their native range in the US. This method attempted to account for successful reproduction and long-term survival of adult oysters. To do this, it used three different salinity variables. In theory this would make the HSI approach more robust.

Mathematically, many of the observed differences could be due to the differences in ideal April, May, June, and July salinities. These months represented the largest divergence between the two ideal salinity regimes with the COOS method preferring lower salinities. The occurrence of large discrepancies (e.g., Chatry’s optima is less than half of Soniat’s for May) is due to Soniat’s variable of mean summer salinity (Chatry et al. 1983, Soniat 2012). This variable penalized salinities lower than 18 ppt during the summer and is included in Soniat’s HSI, because studies suggested higher salinities are more optimal for larval settlement and spawning (Cake 1983, Soniat 2012). The salinity time series used to define Chatry’s optimum represented the best seed conditions measured compared to more theoretical-based requirements of the Soniat model.

Limitations and Scope In our study, only surface water salinity was considered when determining oyster habitat suitability, which limits the strength of our findings. Several factors other than salinity influence oyster populations, and individual growth and mortality. Many of these are difficult to quantify, especially within such a large spatial area. Other studies on oyster habitat suitability from this area typically include character of bottom (Soniat 2012, Swannick et al. 2014). These results should not be interpreted as indicators of where the best oyster habitats are, because of the exclusion of a bottom character parameter. Presence of hard bottom, or cultch, has been indicated as necessary for oyster larvae settlement and growth, and quality of hard bottom, such as vertical relief, may also be important (Galtsoff 1964, Schulte et al. 2009). That said, “good” areas identified in this analysis could be interpreted as places where oysters are more likely to occur, given some hard substrate. To manage for oyster production and restore oyster reefs, hard substrate such as cutch material is often planted and the results here could be used in part to guide these activities.

Our study does not include historical salinity trends. We produced the results here using one year of data per map. Historical salinity trends (~ 3 years of data) have been suggested as an important consideration of oyster habitat quality (Cake 1983), but is sometimes excluded (Soniat et al. 2013). We had a limited amount of Hydrocoast salinity data. While we could have included

Page 60 of 85

a total of three years of salinity data for the 2015 model, this would have involved including a different methodology which would have made the results across years less comparable.

Neither methodology considers oyster metapopulation dynamics, which can be an important factor. The relationships and genetic exchange among different oyster reefs has been shown to influence reef success and these effects were not considered for either method (Munroe et al. 2012). Different parts of estuaries have different oyster populations with varying levels of connectivity through dispersal of larvae. Butler (1954) considered four different oyster communities within estuaries, with some having higher reproductive potential and low population densities. These could potentially be a source or larvae for other reefs that have low reproductive potential, but high growth rates or population densities. Metapopulation dynamics have been demonstrated to affect gene flow in oyster populations and excluding it reduces the applicability of our results (Munroe et al. 2012).

The primary limitations of Chatry and others’ (1983) approach included a lack of data and elements inherent in the methodology itself. These aspects are covered in the preceding section. In addition, higher or lower than mean optimal salinities will have different levels of impacts during different months, and there was no clear relationship how deviations from optimal mean monthly salinities could affect seed production. We assumed that deviation from Chatry and others’ (1983) mean salinity optimum had a negative linear relationship with respect to habitat suitability. Other studies have suggested that a range of salinities for any given month could be ideal and that deviations from this ideal salinity range are non-linear (Cake, 1983). We chose not to make these assumptions for three main reasons: 1) the USACE uses Chatry and others’ (1983) optimal salinity regime to be able to quantify habitat management goals (USACE 1984, 2012), 2) a simple relationship between the optimal salinity described by Chatry and others (1983) and Hydrocoast salinity outputs was desirable, and 3) making any assumptions that would strongly affect the results was undesirable. Based on previous studies the linear decrease in suitability assumption made here is probably not indicative of reality (Butler 1954, Cake 1983, Soniat and Brody 1988).

Hydrocoast surface salinity data also had some limitations. Surface salinities were the only depth available and were used for all of the analyses, when bottom salinities would have been better to characterize oyster habitat. Therefore, a major assumption here is that the estuary was fully mixed vertically. Vertical stratification has been observed in this estuary during our study period (Moshogianis et al. 2013). This phenomenon has at least two major effects impacting our results; 1) at certain times and locations, bottom salinities were different than surface salinities and 2) vertical stratification has been associated with hypoxia which can have a tremendous negative impact on oysters (LDWF 2011, Moshogianis et al. 2013). To remediate this limitation, we indicated areas with frequently observed vertical stratification and hypoxia by including historical hypoxia observation on some maps. It is also important to recognize that Hydrocoast isohaline generation is a manual and subjective process that has improved over time with increasing understanding of the Pontchartrain Basin’s hydrodynamics. This process may result in

Page 61 of 85

some erroneously identified “changes” in the basin that were in fact improvements in the mapping process due to more comprehensive sampling and process refinement.

The color schemes used to map the results of both methodologies were designed to indicate relative performance of areas. Category thresholds are not representative of biologically relevant demarcations. In other words, the difference between the dark green, or most suitable habitat, and the light green areas were not determined to indicate a biologically important change in habitat quality, but meant to visually represent relative, within-model, differences in habitat quality. That is, dark green areas indicated better habitat than light green, but the degree to which this affects oysters was unknown at the time of publication. The SOOS methodology used a geometric mean and was non-dimensional, making it difficult to relate values to actual deviations from an ideal salinity range. The COOS methodology did allow for a direct comparison between mapped values and their deviation from Chatry and others’ (1983) optimal salinity regime. While this did give meaning to the different COOS zones, the biological and ecological relevance was unknown.

Implications for the Oyster Fishery Based upon the approach of the two methodologies, the results of both models could help better manage oyster stocks within the Pontchartrain Estuary. The COOS method was based on research of oyster seed production (Chatry et al. 1983). Given this, the maps generated using this methodology could be useful for some oyster management strategies. Long term COOS trends could be used to determine the areas with the best chance for high seed production. Deviations during some months (e.g., the month(s) during and following potential spatfall events) may be more influential than others. Examining the COOS Annual Analysis Results with Salinity Graphs along Transects figures could be used to show which months are closer to the Chatry ideal (Chatry et al. 1983) among areas that have similar yearly deviations. The approach used by Soniat and Brody (1988) to develop the HSI used in the SOOS methodology took into account the entire oyster’s lifecycle and so may be used to best identify areas for self-sustaining reefs. These results could be useful for oyster fishers when they buy or lease new water bottoms for bedding and producing market sized oysters.

Ground-truthing and further model refinement would be required to distinguish which method performs best for all of the applications previously mentioned in this section. Ideally, measurements of oyster densities, growth rates, settlement rates, and spawning potential across the Pontchartrain Basin would be used to determine the effectiveness of both models, but even this can be misleading. Fisheries activities can significantly alter the composition of a reef, making it difficult to distinguish natural versus anthropogenic influences. However, these activities may be useful and more cost efficient than biological sampling. Comparison between model results and oyster fishing fleet densities from the Hydrocoast Biological maps could be used as a proxy for both bedding and seed ground success. High fleet density over leased or privately owned oyster grounds could indicate good conditions for self-sustaining reefs or

Page 62 of 85

bedding areas. High fleet density over a cultch plant on a public seed ground may indicate high densities of seed oysters. LDWF periodically samples public seed ground reefs and cultch plants and these data could also be used to test which model performs best. However, both of these methods could have influences from the oyster fishery and fishery management. Part of coastal Louisiana has been closed to oyster fishing for the past several years. Sampling in this area could provide further validation for the current models and/or provide insight into model refinement without fishery impacts.

Implications for Coastal Restoration Now and in the Future The methods and results presented here serve several functions related to coastal restoration planning. The maps provide recent baseline information on relative salinity conditions for oysters throughout the estuary. The methods can help identify potential impacts on oysters from salinity changes caused by large-scale river diversions. The State of Louisiana is planning to use diversions to reduce coastal land loss (CPRA 2012). Both methods can be useful for guiding diversion operation strategies. Previous restoration plans suggest using an adaptive management approach to freshwater diversion operation to control downstream salinities (USACE 1984). Finally, the salinity suitability can be used to help set geographic targets for historic reef restoration and to optimally locate barrier reef construction projects. Several plans for the Louisiana coast and the Pontchartrain Basin include construction or restoration of oyster reefs (LPBF 2006, Lopez 2010, CPRA 2012).

Baseline Conditions and Future Change The 2013-2015 maps provide an assessment of current oyster salinity conditions before large planned Mississippi River sediment diversions impacting the study area come online (CPRA 2012). Most restoration plans, including the state CMP, suggest developing large-scale freshwater and sediment Mississippi River diversions (capable of 50,000 ft3/s discharge or more). These diversions and their management will affect the salinities in surrounding waters. Without diversions, coastal geomorphology and hydrology is still predicted to change (CPRA 2012). Baseline oyster salinity habitat conditions, like those presented here, provide a benchmark to measure how these changes might affect oysters. Soniat (2013) applied his HSI to modeled future salinities to identify impacts geographically over time. As the salinity regime changes over time, the most suitable areas for oysters will shift location (Soniat et al. 2013). These methods can help delineate those areas, and determine how they align with geographic restoration targets.

Diversion Operations Both methods of analysis are applicable to examining how adaptive management strategies might achieve targeted salinities for oyster habitat. For instance, the USACE (1984) suggested an adaptive management approach could be employed to operate existing freshwater diversions to benefit oyster fisheries by setting salinity targets. When setting salinity targets, the USACE encouraged including monthly and seasonal salinity ideals, and emulating natural conditions present historically with Mississippi River overflow (USACE 1984). Results and analyses like

Page 63 of 85

those presented here could be used to better understand the outcomes of management strategies in the context of planned salinity targets.

Historic Reef Restoration Salinity suitability analysis results could be used to help guide oyster restoration site selection. The Comprehensive Habitat Management Plan for the Lake Pontchartrain Basin (CHMP; LPBF 2006), and the Multiple Lines of Defense Strategy to Sustain Coastal Louisiana (MLODS; Lopez 2010) both suggest that restoring historic oyster reef areas should be a priority. Specifically, the CHMP (LPBF 2006) suggested optimizing salinity conditions for oysters in the Biloxi Marsh where historic oyster reefs occurred (LBOC 1912). For all years, the COOS method results show no Class 1 areas coinciding with historic surveyed reefs from 1910. However, the majority of the historic reefs do fall within Class 2 for all three years. For all years the SOOS method indicated some of the historic reef locations fall within the most suitable classification.

Barrier Reef Construction The CHMP called for re-establishment of conditions suitable for oysters in Biloxi Marsh and near Mississippi Sound by creating hard ground, protecting existing reefs, enhancing reefs with structure such as reefballs, and developing new reefs (LPBF 2006). These self-sustaining structures may have greater vertical height than managed oyster beds and could serve as part of a larger Biloxi Marsh barrier against wave and storm surge action. The state CMP included bioengineered oyster reefs to improve the fishery and serve as a breakwater against shoreline erosion (CPRA 2012). This plan suggested that vertical reefs would be particularly useful against storm surge in areas where sea level rise and subsidence prevent land-building as a tenable strategy. Creation of 34,000 m of oyster reef to serve as a barrier to wave and surge action was included in the plan (CPRA 2012). Results of these salinity analyses, along with applying HSIs to forecasted salinity data (e.g., Soniat et al. 2013) can help identify the most advantageous areas in which to build substrate or restore oyster reefs. Additionally, using objective approaches such as Soniat’s HSI (2012) and the Chatry et al. (1983) optimal regime to best locate and restore reefs helps fulfill the CMP ethic of efficient use of resources.

Future Work Future spatial analyses of most suitable areas for oyster reef creation and restoration should include additional data not investigated in this study, such as temperature, bottom conditions, water mixing, and diversion modeling. Ultimately this research informs locating restoration projects involving public and private seed grounds and reefs for both fishery and breakwater functions, and can inform diversion operations for optimizing oyster habitat, all within a larger context of sustaining Louisiana’s coast.

Page 64 of 85

References

Andrews, J.D. 1966. Oyster mortality studies in Virginia. V. Epizootiology of MSX, a protistan pathogen of oysters. Ecology. 47: 19-31.

Beck, S. 2016. 2015/2016 Oyster Season Summary. Presentation at the May 5, 2016 Oyster Task Force of Louisiana Meeting. Retrieved Online May 2016. http://www.wlf.louisiana.gov/sites/default/files/oyster_season_summary_presentation_5. 17.16.pdf

Borsje, B.A., B.K. van Wasenbeeck, F. Dekker, P. Paalvast, T.J. Bouma, M.M. van Katwijk, and M.B. de Vries. 2011. How ecological engineering can serve in coastal protection. Ecological Engineering. 37: 113-122.

Barnes, T.K., A.K. Volety, K. Chartier, F.J. Mazzotti, and L. Peralstine. 2007. A habitat suitability index model for the Eastern oyster (Crassostrea virginica), a tool for restoration of the Caloosahatchee Estuary, Florida. Journal of Shellfish Research. 26(4): 949-959.

Buroker, N.E. 1983. Population genetics of the American oyster Crassostrea virginica along the Atlantic coast and the Gulf of Mexico. Marine Biology. 75: 99-112.

Butler, P.A. 1954. Summary of our knowledge of the oyster in the Gulf of Mexico. Fishery Bulletin. United States Fish and Wildlife Service. 89: 479-489.

Cake, E.W. 1983. Habitat Suitability Index Models: Gulf of Mexico American Oyster. US Department of the Interior Fish & Wildlife Service. FWS/OBS-82/10.57. 37pp.

Chatry, M., R.J. Dugas, and K.A. Easley. 1983. Optimum Salinity Regime for Oyster Production on Louisiana’s State Seed Grounds. Contributions in Marine Science. 26: 81-94.

Chu, F-L.E. and K.F. Greene. 1989. Effect of temperature and salinity on in vitro vulture of the oyster pathogen Perkinsus marinus (Apicomplexa: Perkinsea). Journal of Invertebrate Pathology. 53: 260-268.

Chu, F-L.E. and J.F. La Peyre. 1993. Perkinsus marinus susceptibility and defense-related activities in eastern oysters Crassostrea virginica: temperature effects. Diseases of Aquatic Organisms. 16: 223-234.

Coen, L.D., R.D. Brumbaugh, D. Bushek, R. Grizzle, M.W. Luckenbach, M.H. Posey, S.P. Powers, and S.G. Tolley. 2007. Ecosystem services related to oyster restoration. Marine Ecology Progress Series. 341: 303-307.

Page 65 of 85

Coastal Protection and Restoration Authority of Louisiana (CPRA). 2007. Louisiana's Comprehensive Master Plan for a Sustainable Coast. Baton Rouge, Louisiana: CPRA.

Coastal Protection and Restoration Authority of Louisiana (CPRA). 2012. Louisiana's Comprehensive Master Plan for a Sustainable Coast. Baton Rouge, Louisiana: CPRA, 188 p.

Couvillion B.R., J.A. Barras, G.D. Steyer, W. Sleavin, M. Fischer, H. Beck, N. Trahan, B. Griffin and D. Heckman. 2011. Land area change in coastal Louisiana from 1932 to 2010: US Geological Survey Scientific Investigations Map vol 3164, p 12, scale 1:265,000, Pamphlet.

Davis, H.C. 1958. Survival and growth of clam and oyster larvae at different salinities. Biological Bulletin. 114: 296-307.

Deseran, F.A. and C. Riden. 2000. “Louisiana Oystermen…Surviving in a Troubled Fishery.” Baton Rouge: Louisiana Sea Grant College Program, Louisiana State University.

Galtsoff, P.S. 1964. The American oyster Crassostrea virginica Gmelin. Fishery Bulletin. United States Fish and Wildlife Service. 64: 1–480 (United States Govt Printing Office).

Garton, D. and W.B. Stickle. 1980. Effects of salinity and temperature on the predation rate of Thais haemastoma on Crassostrea virginica spat. Biological Bulletin. 158: 49-57.

Gosselin, L.A. and P-Y. Qian. 1997. Juvenile mortality in benthic marine invertebrates. Marine Ecology Progress Series. 146: 265-282.

Grabowski, J.H. and C.H. Peterson. 2007. Restoring oyster reefs to recover ecosystem services. In: Cuddington, K., J.E. Byers, W.G. Wilson, A. Hastings (eds.) Ecosystem engineers: concepts, theory and applications. Elsevier-Academic Press, Amsterdam, p 281-298.

Gunter, G. 1951. The species of oysters of the Gulf, Caribbean and West Indian region. Bulletin of Marine Science of the Gulf and Caribbean. 1: 40-45.

Gutiérrez, J.L., C.G. Jones, D.L. Strayer, and O.O. Iribarne. 2003. Mollusks as ecosystem engineers: the role of shell production in aquatic habitats. Oikos. 101: 79-90.

Hayes, P.F. and R.W. Menzel. 1981. The reproductive cycle of early setting Crassostrea virginica (Gmelin) in the Northern Gulf of Mexico, and its implications for population recruitment. Biological Bulletin. 160: 80-88.

Page 66 of 85

Henkel T., J. Lopez, E. Boyd, A. Baker, D. Weathers, C. Cropp, and N.A. Robins. 2012. Recently observed seasonal hypoxia within Chandeleur Sound of Southeastern Louisiana and near coastal Mississippi within the Gulf of Mexico. Technical report. Retrieved May 2016 from http://saveourlake.org/PDF-documents/our-coast/Chandeleur-2011-Hypoxia- FINAL-Jan2012.pdf

Hidu, H. and H.H. Haksin. 1978. Swimming speeds of oyster larvae Crassostrea virginica in different salinities and temperatures. Estuaries and Coasts. 1(4): 252-255.

La Peyre, M., J. Furlong, L.A. Brown, B.P. Piazza, K. Brown. 2014. Oyster reef restoration in the northern Gulf of Mexico: Extent, methods and outcomes. Ocean and Coastal Management. 89: 20-28.

Lopez, J.A. 2006. The Multiple Lines of Defense Strategy to Sustain Coastal Louisiana. Technical report. Retrieved March 2016 from http://www.saveourlake.org/PDF- documents/MLODSfullpt2-06.pdfxtent

Lopez, J., A. Baker, E. Boyd. 2010. Water quality in Chandeleur Sound in 2008 and 2010. Technical report. Retrieved May 2016. from http://saveourlake.org/PDF-documents/our-coast/ChandeleurSound_WaterQuality_10-2010.pdf

Lopez, J.A., T.K. Henkel & P. Connor. 2015. Methodology for Hydrocoast Mapping of the Pontchartrain Basin: 2012 to 2015. Technical report. Retrieved March 2016, from http://www.saveourlake.org/PDF-documents/our- coast/CoastalFactSheets/HydrocoastMethodology_Final.pdf Lake Pontchartrain Basin Foundation (LPBF). 2006. Comprehensive Habitat Management Plan for the Lake Pontchartrain Basin. Technical report. Retrieved March 2016 from http://www.saveourlake.org/PDF-documents/our-coast/CHMP_final_22706.pdf

Louisiana Department of Wildlife and Fisheries (LDWF). 2011. Oyster stock assessment of the public oyster areas in Louisiana. Baton Rouge, LA: Louisiana Department of Wildlife and Fisheries. 92 pp. Retrieved April 2016 from http://www.wlf.louisiana.gov/sites/default/files/pdf/document/37757-stock- assessments/2011_oyster_stock_assessment.pdf

Louisiana Department of Wildlife and Fisheries (LDWF). 2013. Oyster stock assessment of the public oyster areas in Louisiana. Baton Rouge, LA: Louisiana Department of Wildlife and Fisheries. 92 pp. Retrieved April 2016 from http://www.wlf.louisiana.gov/sites/default/files/pdf/document/37757-stock- assessments/2011_oyster_stock_assessment.pdf

Page 67 of 85

Louisiana Department of Wildlife and Fisheries (LDWF). 2014. Oyster stock assessment of the public oyster areas in Louisiana. Baton Rouge, LA: Louisiana Department of Wildlife and Fisheries. 92 pp. Retrieved April 2016 from http://www.wlf.louisiana.gov/sites/default/files/pdf/document/37757-stock- assessments/2011_oyster_stock_assessment.pdf

Louisiana Board of Commissioners for the Protection of Birds, Game and Fish (LBOC). 1912. The 1912 Report of the Board of Commissioners – Including a 1910 Survey of Oyster Reefs in St. Bernard Parish by Frank T. Payne, Chief Surveyor (copy of report and map may be found at the Historic New Orleans Collection and the Louisiana State Museum).

Meyer, D.L., E.C. Townsend, and G.W. Thayer. 1997. Stabilization and erosion control value of oyster cultch for intertidal marsh. Restoration Ecology. 5(1): 93-99.

Moshogianis, A., J. Lopez, T. Henkel, E. Boyd, A. Baker, and E. Boyd. 2012. Recently observed hypoxia within the Gulf of Mexico in the Chandeleur Sound of Southeastern Louisiana and Mississippi Bight of coastal Mississippi. Technical report. Retrieved April 2016 from http://saveourlake.org/PDF-documents/our-coast/Chandeleur-2012-Hypoxia-Final.pdf

Moshogianis, A., J. Lopez, T. Henkel, E. Boyd, A. Baker, and E. Hillmann. 2013. Preliminary results of recently observed hypoxia development in the Chandeleur Sound and Breton Sound of southeastern Louisiana, east of the Mississippi River Delta. Technical report. Retrieved April 2016 from http://saveourlake.org/PDF-documents/our-coast/Chandeleur- 2013-Hypoxia-July2013.pdf

Munroe, D.M., J.M. Klinck, E.E. Hofmann, and E.N. Powell. 2012. The role of larval dispersal in metapopulation gene flow: Local population dynamics matter. Journal of Marine Research. 70: 441-467.

Newell R.I.E., T.R. Fisher, R.R. Holyoke, and J.C. Cornwell. 2005. Influence of eastern oysters on N and P regeneration in Chesapeake Bay, USA. In: Dame R, Olenin S (eds) The comparative roles of suspension feeders in ecosystems. Springer, Berlin, p 93–12.

North, E.W., Z. Schlag, R.R. Hood, M. Li, L. Zhong, T. Gross, V.S. Kennedy. 2008. Vertical swimming behavior influences the dispersal of simulated oyster larvae in a coupled particle-tracking and hydrodynamic model of Chesapeake Bay. Marine Ecology Progress Series. 359: 99-115.

Peyronnin, N., M. Green, C.P. Richards, A. Owens, D.J. Reed, J. Chamberlain, D.G. Groves, W.K. Rhinehart, K. Belhadjal 2013. Louisiana’s 2012 Coastal Master Plan: Overview of a Science-Based and Publicly Informed Decision-Making Process. Journal of Coastal Research. 67: 1-15.

Page 68 of 85

Piazza, B.P., P.D. Banks, M.K. La Peyre. 2005. The Potential for Created Oyster Shell Reefs as Sustainable Shoreline Protection Strategy in Louisiana. Restoration Ecology 13(3): 499- 506.

Pomeroy, L.R., C.F. D’Elia, L.C. Schaffner. 2006. Limits to top-down control of phytoplankton by oysters in Chesapeake Bay. Marine Ecology Progress Series. 325: 301-309.

Sasser, C. E., J. M. Visser, E. Mouton, J. Linscombe, and S. B. Hartley. 2014. Vegetation types in coastal Louisiana in 2013: U.S. Geological Survey Scientific Investigations Map 3290, 1 sheet, scale 1:550,000, http://dx.doi.org/10.3133/sim3290

Schulte D.M., R.P. Burke, R.N. Lipcius. 2009. Unprecedented restoration of a native oyster metapopulation. Science 325: 1124–1128.

Soniat, T.M. and M.S. Brody. 1988. Field Validation of a Habitat Suitability Index Model for the American Oyster. Estuaries 2: 87-95.

Soniat, T.M. 2012. Eastern Oyster Habitat Suitability Index Technical Report. In: Louisiana's Comprehensive Master Plan for a Sustainable Coast, Appendix D-13.

Soniat, T.M., J.M. Klinck, E.N. Powell, N. Cooper, M. Abdelguerfi, E.E. Hofmann, J. Dahal, S. Tu, J. Finigan, B.S. Eberline, J.F. La Peyre, M.K. La Peyre, and F. Qaddora. 2012. A shell-neutral modeling approach yields sustainable oyster harvest estimates: a retrospective analysis of the Louisiana state primary seed grounds. Journal of Shellfish Research. 31(4): 1103-1112.

Soniat, T.M., C.P. Conzelmann, J.D. Byrd, D.P. Roszell, J.L. Bridevaux, K.J. Suir, and S.B. Colley. 2013. Predicting the effects of proposed Mississippi River diversions on oyster habitat quality; application of an oyster habitat suitability index model. Journal of Shellfish Research. 32(3): 629-638.

Soniat, T.M. 2014. Freshwater diversion, oyster reef restoration and fisheries sustainability on the Primary Public Seed Grounds of Louisiana. Presentation at Advancing Ecological Modeling for Diversions and Hypoxia in the Northern Gulf of Mexico. Stennis Space Center. Retrieved February 2016 from http://www.ngi.msstate.edu/impact/resources/hypoxia2014/soniat.pdf

Soniat, T. & S. Ray. (n.d.). Oyster Sentinel. Retrieved September 3, 2015, from http://www.oystersentinel.org/

Soniat, T. June 27, 2016. Personal correspondence.

St.nmfs.noaa.gov. NOAA Office of Science and Technology: Commercial Fisheries Statistics. Retrieved March 2016 from https://www.st.nmfs.noaa.gov/commercial- fisheries/commercial-landings/annual-landings/index

Page 69 of 85

Swannack, T.M., M. Reif, and T.M. Soniat. 2014. A robust, spatially explicit model for identifying oyster restoration sites: case studies on the Atlantic and Gulf coasts. Journal of Shellfish Research. 33(2): 395-408.

US Army Corps of Engineers. 1984. Mississippi and Louisiana Estuarine Areas: Freshwater Diversion to Lake Pontchartrain Basin and Mississippi Sound Feasibility Study, Vol. 2, Technical Appendixes A, B, C, D.

US Army Corps of Engineers. 2012. Mississippi River Gulf Outlet (MRGO) Ecosystem Restoration Plan Final Feasibility Report. Final Environmental Impact Statement.

Van den Heuvel, S. 2010. Modeling the hydrodynamics and salinity of the Pontchartrain Basin. Master’s Thesis. Submitted to Delft University of Technology.

Villalba, A., K.S. Reece, M.C. Ordas, S.M. Casas, and A. Figueras. 2004. Perkinsosis in molluscs: a review. Aquatic Living Resources. 17, 411–432.

Waterdata.usgs.gov. USGS 30072208915010 Mississippi Sound near Grand Pass. Retrieved March 2016 from http://waterdata.usgs.gov/la/nwis/uv?site_no=300722089150100

Zimmerman, R., T.J. Minello, T. Baumer, and M. Castiglione. 1989. Oyster reef as habitat for estuarine macrofauna. Technical Memorandum NMFS-SEFC-249. National Oceanic and Atmospheric Administration, Washington, D.C., USA.

Page 70 of 85

Appendix 1. Steps and Parameters for Spatial Analysis

The following steps in ESRI ArcMap 10.2 and 10.3 software were used to analyze Hydrocoast data according to both the COOS and the SOOS methods of oyster habitat evaluation. Parameters are explicitly given below, along with any altered raster analysis environment settings, and can be applied using any GIS software.

Page 71 of 85

Data preparation 1. Assembled set of interpolated salinity rasters for each calendar year in planar projection (used UTM Zone 15 North). 2. In ArcMap, used “cell statistics” tool to calculate mean salinity raster surface for each calendar month. a. Inputs: interpolated rasters for that month based on when data was collected. b. Output: monthly mean raster. c. Statistic: mean. d. Ignored “NoData” in calculations. 3. Used “aggregate” tool in ArcMap to aggregate from 30m to 500m resolution (by default tool uses mean statistic). 4. Results: 12 monthly mean rasters for each calendar year with resolution of 500m square.

Spatial analysis steps using Chatry’s optimal regime 1. In ArcMap used “raster calculator” to create a new set of monthly rasters by calculating the absolute value of the difference between each Hydrocoast monthly mean and the corresponding Chatry monthly optimum. Repeated for each of the 12 calendar months in each year. a. Used batched process. b. Map algebra expression example: Abs(“Mean201401” – 16.4). c. Outputs: differences rasters. d. Environment setting: raster analysis cell size = 500m. 2. In ArcMap, used “raster calculator” to create a new raster by calculating the sum of the differences rasters (above step) for each calendar year. a. Inputs: differences rasters from Step 1. b. Map algebra: sum. c. Output: sum raster. 3. Reclassified sum raster to classes 1-5. a. 0-25 = 1. b. 25-50 = 2. c. 50-75 = 3. d. 75 – 100 = 4. e. > 100 = 5. 4. Mapped classified raster with appropriate symbology to indicate ordinal range.

Page 72 of 85

Steps to generate Soniat’s oyster salinity suitability index spatial data 1. Created a point feature class (can derive from raster; output = 500-meter spaced point dataset). Created a unique ID field. Created 7 additional floating point fields: a. SpawnMean: mean salinity during spawning season (May – Sept). b. MinMthMean: minimum monthly mean salinity during calendar year. c. AnnualMean: mean salinity for calendar year. d. S2Spawn: S2 index value in Soniat’s oyster HSI based on a.) above. e. S3Min: S3 index value in Soniat’s oyster HSI based on b.) above. f. S4AnnMean: S4 index value in Soniat’s oyster HSI based on c.) above. g. S2xS3xS4: product of salinity indices. h. HSI: geometric mean of salinity indices – HSI assuming cultch (S1) = 1.

Page 73 of 85

2. Using the monthly mean rasters and the interpolated Hydrocoast rasters, derived surfaces as follows using “cell statistics” tool in ArcMap: a. Spawning season mean (V2) – mean value from May through Sept for each calendar year. i. Inputs: original interpolated rasters May – Sept. ii. Statistic: mean. iii. Ignore NoData. iv. Output: spawning season salinity raster. b. Minimum monthly mean (V3) – minimum value for each cell from among all 12 monthly mean surfaces. i. Inputs: 12 monthly mean rasters for calendar year. ii. Statistic: mean. iii. Ignore NoData. iv. Output: minimum monthly mean raster. c. Annual mean (V4) – mean of all 12 months in each calendar year. i. Inputs: all original interpolated rasters for calendar year. ii. Statistic: mean. iii. Ignore NoData. iv. Output: annual mean raster.

3. For each salinity parameter (V2-V4) in the suitability index, used ArcMap “zonal statistics as table” tool to assign values to point feature class (Step 1). a. Input feature zone: point feature class. b. Zone field: unique ID field in point feature class. c. Input value raster: in turn, spawning season mean, minimum monthly mean, and annual mean. d. Ignore NoData. e. Statistic: any (because zones are point locations, all statistics will be identical and equal to the value at that location). f. Output: table for each of the three runs listed in c.) above. 4. One by one joined the zonal statistics results tables to the points attribute table using the unique ID field as the join field. Calculated attribute table fields accordingly (SpawnMean, MinMthMean, AnnualMean). Removed joins. 5. Referencing Soniat’s threshold values and linear formulas (see body text), ran a series of selections and field calculations to calculate S values. Example for spawning season mean salinity values between 5 and 10: a. Spawning mean – In the point attribute table, selected spawning mean values >= 5 and <= 10. b. For those selected records, calculated field S2 = (0.06*SpawnMean) – 0.3 (applying linear formulas supplied by Soniat’s relationships, see Figure 3).

Page 74 of 85

c. Ran select/calculate for all groups in Soniat’s spawning season mean worksheet, and repeated for each group in the other two salinity parameters. 6. Calculated S2xS3xS4 = product of the three corresponding fields. 7. Calculated HSI as cube root of Step 8. Note the assumption is that S1 (Cultch) = 1. 8. Converted points to raster. 9. Symbolized on map according to divisions of 0.25.

Page 75 of 85

Appendix 2. Guide to ArcMap GIS Analysis Implementation

The following provides a guide to implementing spatial analysis in this document using ArcMap 10.3 GIS software and Microsoft Excel.

Page 76 of 85

Soniat’s Approach Create 500-meter-spaced point feature class with fields:

Step 1. UniqueID - integer Create SpawnMean – floating point MinMthMean – floating point point AnnualMean – floating point features S2Spawn – floating point S3Min – floating point S4AnnMean – floating point S2xS3xS4 – floating point HSI – floating point

Step 2. Calculate V2, V3, V4 rasters

Use “cell statistics” to derive salinity index surfaces for S2, S3, and S4.

V2:

Spawning Season Mean: May - September

Input all original Hydrocoast salinity rasters May-Sept.

Page 77 of 85

V3:

Minimum Monthly Mean

Input all 12 monthly mean rasters (Jan-Dec).

V4:

Annual Mean

Input all original Hydrocoast salinity rasters for calendar year.

Page 78 of 85

Step 3. Create tables with V values to assign to point locations.

Use “zonal statistics” to assign raster values to point features.

Zone: Unique ID field

Repeat for each salinity suitability raster (spawning season mean, minimum monthly mean, and annual

mean

Statistic: any – because the zone is a single point location, all statistics will be identical and equal to the raster value at that location

Step 4. Join results table to point attributes and calculate values for V fields.

Page 79 of 85

Step 5. Calculate index values (0-1) based on V values.

For each V value, select range of values in point attribute table according to Soniat’s HSI.

Calculate field value for selected records using

appropriate linear formula from Soniat HSI.

(0.06*SpawnMean) – 0.3

Repeat for each range, then for each parameter.

Page 80 of 85

Step 6-7. Calculate final Salinity Suitability Index (“HSI” field).

(S2xS3xS4 = product of S2Spawn, S3Min and S4AnnMean.)

Step 8-9. Convert points with final HSI to raster and map with divisions of 0.25.

Calculated in Excel.

Page 81 of 85

Page 82 of 85

Chatry’s Approach

Step 1. Create “difference” rasters.

Use “raster calculator” to create a new set of monthly rasters: absolute value of the difference between each Hydrocoast monthly mean and the corresponding Chatry monthly optimum. Repeat for each month.

Page 83 of 85

Step 2. Sum difference rasters.

Use “raster calculator” to create a new raster by calculating the sum of the differences rasters (above step) for each calendar year.

Step 3. Reclassify.

Reclassify sum raster to classes 1-5.

Page 84 of 85

Step 4. Map classified raster showing ordinal range.

Page 85 of 85