Precipitation, submarine groundwater discharge of nitrogen, and red tides along the southwest Florida Gulf coast

Abstract

Blooms of the dinoflagellate Karenia brevis occur almost every year along the southwest Florida Gulf coast. Long-duration blooms with especially high concentrations of K. brevis, known as red tides, destroy marine life through production of neurotoxins. Current hypotheses are that red tides originate in oligotrophic waters far offshore using nitrogen (N) from upwelling bottom water or, alternatively, from blooms of Trichodesmium, followed by advection to nearshore waters. But the amount of N available from terrestrial sources does not appear to be adequate to maintain a nearshore red tide. To explain this discrepancy, we hypothesize that contemporary red tides are associated with release of N from offshore submarine groundwater discharge (SGD) that has accumulated in benthic sediment biomass by dissimilatory nitrate reduction to ammonium (DNRA). The release occurs when sediment labile organic carbon (LOC), used as the electron donor in DNRA, is exhausted. Detritus from the resulting destruction of marine life restores the sediment LOC to continue the cycle of red tides. The severity of individual red tides increases with increased bloom-year precipitation in the geographic region where the SGD originates, while the severity of ordinary blooms is relatively unaffected.

Fig. 1

(a) Florida and the West Florida Shelf. (b) Detail of the southwest part of the Florida peninsula where red tides are most prevalent (county names in bold), showing the range in latitude (Lat) of the study region and sections (see Materials and methods).

Fig. 2

(a) Detail of the part of the study region that includes the LTA. Circled letters mark terrestrial recharge (R) and discharge (D) areas. The terrestrial crosshatched areas mark monitoring well locations. The offshore crosshatched area marks an area where submarine groundwater discharge from the Lower Tamiami Aquifer is likely. (b) Approximate hydrogeological section at marked location (S).

Fig. 3

(a) Time series of Karenia brevis (Kb) concentrations for the entire study region, north latitude (N Lat) 27.54–25.80, and individual region sections from 1994 to 2020. (b) Example bloom severity calculation using the 2015 bloom time series of Karenia brevis concentrations. Bloom severity is the product of bloom average Kb concentration in million cells L⁻¹ and bloom duration in days, in this case 0.222 × 169 = 38.

Fig. 4

(a) COCO1 normalized precipitation and groundwater (GW) levels in well C-1059 for 1999 through 2009 (blue lines mark end-of-year, dashed red lines illustrate synchronization between precipitation events and groundwater levels). (b) Comparison of GW levels between different monitoring wells for 1990 through 2020 (dashed red lines illustrate synchronization between groundwater levels in different wells). Levels are above mean sea level. Red arrows mark a precipitation event in Lee County not detected by the COCO1 precipitation gauge. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

Hypothesized DNRA process for benthic accumulation and release of N to support a red tide, showing seepage of submarine groundwater into benthic sediment and subsequent reactions. Blue arrows represent accumulation of N, dashed red arrows represent release, black arrows represent possible competing reactions. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

Sample locations (blue) and locations with Kb concentration ≥1 × 10⁶ cells L⁻¹ (red) for the 1994–2020 study period; circles mark hotspots that suggest possible karst SGD-N features; table shows number of samples and % samples ≥1 × 10⁶ cells L⁻¹ for each latitude (lat) band; results in red are local maxima. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 7

(a) Values of R-squared for bloom severity (S_B) regressed against bloom-year precipitation in counties potentially contributing groundwater flows to the study region. (b–e) Bloom severities vs. bloom-year precipitation for Lee, Charlotte, DeSoto, and Sarasota counties. (f) Cumulative bloom severity vs. cumulative Lee County precipitation (red line segments mark red tides). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)