Site Descriptions:
Three areas in Florida Bay north of the Middle Keys were chosen for sampling based on previous work conducted between 2012 and 2015 by colleagues at the University of North Carolina (UNC) and FWC-M. Site MB1 is located in the middle of our Mystery Basin site (24°56'30.42"N 80°49'31.80"W), located in West Florida Bay (WFB) and is physically representative of other interior Florida Bay sites. This interior area is characterized by numerous small basins created through shoaling carbonate sediments, resulting in lowered water exchange. The series of MB sampling locations were intended to provide a characteristic data suite. Burnt Point (BP) is located in the Back Key Basin (BKB) region and represents nearshore embayments, a common feature along the Florida Bay side of the Keys island chain. Station E 01 is located between the MB stations and BP. The BP field site underwent a sponge biomass loss of approximately 95 ± 5% during Hurricane Irma (Bollinger et al. 2018). The last station J01 (24°49'54.30"N, 80°48'44.82"W) is located slightly offshore from Long Key, but is not in a physically protected embayment like BP. Together, the J01 and BP stations, are representative of the range of nearshore environments along the southern portion of Florida Bay. Site locations are depicted in the attached Supplemental File ("Fig. 1.jpg").

Sample Preparation:
We measured DIN and Chl-a concentrations at each of the three stations via monthly water sampling from October 2017 through December 2018. Each month, water samples were collected 10 centimeters (cm) below the water surface at each station, shipped overnight on dry ice, and vacuum-filtered through a 47-millimeter (mm) Whatman GF/F filter. Samples collected by FWC-M personnel on most dates could only be provided in frozen, unfiltered form due to the limited availability of boating operations and sample handling personnel at their laboratory. The UNC team worked out of the FWC-M laboratory and collected water samples on December 5, December 11, April 3, April 9, and September 27. Samples from these four dates were filtered prior to freezing and shipping to UNC. Cell lysis undoubtedly occurred in the frozen samples collected by FWC-M personnel and shipped unfiltered to UNC, thereby suggesting that their Chl-a measurements represent minimum concentration values.

Sample Analysis:
The entire volume of the samples was vacuumed filtered through a 47 mm Whatman GF/F, before conducting analyses. The filters were placed in combusted foil for chlorophyll-a (Chl-a) measurements. NH4+ concentration was determined using acidified fluorometry (Holmes et al. 1999) using a Turner Design TD-700 laboratory fluorometer with a near UV lamp (310-390 nM). Discrete samples were analyzed for NOx- using a Spectrophotometric Elemental Analysis System (SEAS) configured for benchtop applications (Adornato et al. 2005, 2007). Lastly, phytoplankton abundance was quantified using Chl-a as a proxy; measurements were made on a Turner Design 10-AU Fluorometer using the acidified fluorometry method (Yentsch and Menzel 1963).

Sediment Suspension and Core Experiments:
Qualitative sediment suspension experiments using freshly collected bulk wet sediment were performed in April and June 2018 to mimic potential storm resuspension that might release DIN stored in porewaters, primarily in the form of NH4+. An acrylic box was placed open face down, on the sediment surface at station BP. The box was brushed back and forth in one-foot lengths along the surface of the sediment twenty times, at sufficient speed to resuspend the upper few centimeters of sediment. Without removing the box from the sediment surface water in the box was then sampled by a 60 milliliter (mL) polypropylene syringe and stored on ice in an amber HDPE bottle, to eliminate exposure to light. This crude "shake experiment" is not quantitative but provides a qualitative estimate of the potential for sudden porewater release to elevate water column DIN concentration. A more quantitative core experiment was conducted in June 2018 in sediment cores with depths of 2.5, 3.5, and 5 cm each with equal headspaces of 6.5 cm. Sediments were collected to each depth in BP, then shaken into the headspace twenty times, and the core was analyzed for NH4+.