Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

Environmental controls on bioluminescent dinoflagellate density,

in Laguna Grande, Fajardo, Puerto Rico

Yogani Govender1*; https://orcid.org/ 0009-0006-7216-1075

Mark R. Jury2; https://orcid.org/0000-0002-6871-403X

1. Climate & Ecological Studies Lab, Interamerican University of Puerto Rico, Metropolitan Campus, San Juan, PR, USA;

ygovender@metro.inter.edu (Correspondence*)

2. Physics Department, University of Puerto Rico, Mayaguez, PR, USA; mark.jury@upr.edu

Received 27-IX-2023. Corrected 27-II-2024. Accepted 24-VII-2024.

ABSTRACT

Introduction: Bio bays in Puerto Rico play an important socio-economic role and declines in dominant biolumi-

nescent dinoflagellate Pyrodinium bahamense are concerning. Studies show erratic blooms with is weak correla-

tion to in situ environmental factors. Our study examines shorter field and longer proxy records on dinoflagellate

density at Laguna Grande de Fajardo (LGF).

Objectives: To quantify temporal changes in dinoflagellate density in a long-term monitoring study, understand

how the marine environment modulates those changes, and determine the wider impacts of a fluctuating climate

and extreme events on proxies for dinoflagellate density.

Methods: Bimonthly samples were collected from 2016 to 2021 at three sampling sites in LGF. Dinoflagellates

density was estimated by Sedgewick Rafter counting cells. Environmental conditions were obtained from Rio

Fajardo 5007100 station and NOAA buoy 41056. Marine climate and biotic proxies were obtained from remote

sensing measurements. Kruskal Wallis, Spearman correlations and cross-correlations in the shorter field and

longer proxy records were used to evaluate environmental controls on LGF dinoflagellate blooms.

Results: Six years of field monitoring densities found a low period in 2016-2017, frequent and intense blooms in

2018-2021 punctuated by hurricanes. Generally low values were recorded in late winter in contrast with higher

values in late summer (Aug-Nov). Light winds and mixed layer response to seasonal warming in the form of high

tides and low salinity, were found to sustain dinoflagellate reproduction.

Conclusions: Bioluminescent dinoflagellates are vital to coastal tourism and require resource management. LGF

results show that: 1) dinoflagellate counts fluctuate widely, 2) fluorescing dinoflagellates are sensitive to environ-

mental conditions because of limited seasonality and narrow physiological range, 3) hurricanes play a role by

‘raking and refreshing’ the coastal lagoon for subsequent biotic reproduction, and 4) intra-seasonal fluctuations

of density and proxies relate to air-sea thermodynamic conditions, the salinity budget and sea level.

Key words: bioluminescent dinoflagellates; coastal lagoon; Puerto Rico.

RESUMEN

Controles ambientales sobre la densidad de dinoflagelados bioluminiscentes,

en Laguna Grande, Fajardo, Puerto Rico

Introducción: Las bahías bioluminiscentes en Puerto Rico desempeñan un papel socioeconómico importante

y disminuciones del dinoflagelado bioluminiscente dominante Pyrodinium bahamense son preocupantes. Los

estudios muestran proliferaciones erráticas débilmente correlacionadas con factores ambientales locales.

https://doi.org/10.15517/rev.biol.trop..v72i1.56729

TEMÁTICA ACUATIC ECOLOGY

2Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

INTRODUCTION

While dinoflagellates around the world are

associated by harmful algal blooms (HAB’s)

(Badylak & Phlips, 2009; Morquecho, 2019),

in Puerto Rico, the bioluminescence of the

dominant species Pyrodinium bahamense has

resulted in a lucrative eco-tourism industry.

Bioluminescent bays of Puerto Rico include

Laguna Grande de Fajardo on the Northeastern

tip, Mosquito Bay on the offshore island of

Vieques, and La Parguera on the Southwestern

coast (Bahía Fosforescente). They are surround-

ed by nature reserves under the Department of

Environment and Natural Resources (DNER,

2013), managed to protect the fringing man-

groves, access channels, water quality, and stim-

ulate adventure tourism (Weaver et al., 1999).

Commercial tour operators reported ~34 000

kayaks rented to visit Laguna Grande in 2018

earning approximately $1.7 million (DNER,

2013). Due to the socio-economic benefits of

the dominant bioluminescent dinoflagellate P.

bahamense, numerous impact assessments for

bio bays have been made. Sustained blooms

involve a delicate balance of environmental

conditions that control the flux of nutrients and

biological uptake (O’Connell et al., 2007). Their

density fluctuates over months and years due

to flushing action around the coastal lagoon’s

fringing coral reefs and mangrove vegetation,

controlled by sea level, wind-driven currents

and waves, sunlight and turbidity, sea tempera-

tures in the range 24-29 °C, salinity, run off and

organic content (Sastre et al., 2013; Soler-López

& Santos, 2010).

At La Parguera Puerto Rico, Soler-Figueroa

& Otero (2014) and Soler-Figueroa & Otero

(2016) measured sea temperature, salinity,

nutrient concentration, rainfall, wind veloc-

ity, and dinoflagellate count at multiple sites

during wet and dry seasons. In the wet season

(Aug-Nov), they found P. bahamense abun-

dance increased under warmer sea temperature

and lower salinity. Sastre et al. (2013) studied

dinoflagellates at Laguna Grande Fajardo (here-

after: LGF) and found higher P. bahamense

densities in summer and lower densities in

winter (Dec-Mar), but erratic fluctuations of

Ceratium (Tripos) furca. They found weak cor-

relations between dinoflagellate density and

the in-situ marine environment and nutrients

(phosphate, nitrate), except for a negative cor-

relation between fluorescence and salinity.

Objetivos: Cuantificar cambios en la densidad de dinoflagelados a largo plazo en la Laguna Grande de Fajardo

(LGF), comprender cómo el ambiente marino modula esos cambios y determinar los impactos de eventos climá-

ticos extremos en la densidad de dinoflagelados.

Métodos: Se recolectaron muestras bimestrales del 2016-2021 en tres sitios de la LGF. La densidad se estimó

mediante el conteo de células de Sedgewick Rafter. Condiciones ambientales se obtuvieron de la estación Río

Fajardo 5007100 y de la boya NOAA 41056. El clima marino y los indicadores bióticos se obtuvieron por medi-

ciones de teledetección. Se utilizaron correlaciones Kruskal Wallis, Spearman y las correlaciones cruzadas en los

registros de campo corto y proxy largos para evaluar los controles ambientales en las proliferaciones de dinofla-

gelados LGF.

Resultados: En seis años de monitoreo se encontraron densidades bajas entre el 2016-2017 y proliferaciones

frecuentes e intensas en 2018-2021 marcadas por huracanes. Se registraron valores bajos a finales del invierno, en

contraste con valores más altos a finales del verano (agosto-noviembre). Vientos ligeros y la respuesta de capas

mixtas al calentamiento estacional en forma de mareas altas y baja salinidad, mantienen la reproducción de los

dinoflagelados.

Conclusiones: Los dinoflagelados bioluminiscentes son vitales para el turismo costero y requieren manejo de los

recursos. En general: 1) los recuentos de dinoflagelados fluctúan ampliamente, 2) los dinoflagelados fluorescentes

son sensibles a las condiciones ambientales debido a su estacionalidad limitada y estrecho rango fisiológico, 3)

los huracanes juegan un rol de rastrillar y refrescar la laguna costera para su posterior reproducción, y 4) fluc-

tuaciones intra-estacionales se relacionan con las condiciones termodinámicas aire-mar, el balance de salinidad

y el nivel del mar.

Palabras clave: dinoflagelados bioluminiscentes; laguna costera; Puerto Rico.

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

LGF is surrounded by a nature reserve on

the Northeast tip of Puerto Rico, with a depth

of 3 m (Soler-López & Santos, 2010) and mean

annual rainfall ~1 500 mm. Freshwater runoff

to LGF is sporadic and secondary to seawater

exchange, especially during high tides from

August to November. The freshwater flux is

confined to infrequent storms and does not

represent an effective water-renewal mecha-

nism for LGF. Yet floods and drought do alter

salinity, sea temperature, and pH (Sastre et al.,

2013; Soler-López & Santos, 2010).

The loss to tourism-related businesses due

to the 2018 Florida HAB’s was estimated to

be $2.7 billion USD (Alvarez et al., 2024),

which implies that HABs and their impact on

eco-tourism can be considered as a potential

‘billion-dollar’ disaster for bio bays in Puerto

Rico. In 2003 and 2013, low seasons of biolumi-

nescence raised public and ecological concern

(Pagán, 2021; Xei, 2013). The DNER (2013)

attributed the decreases to lower sea-level/

runoff/nutrients in the ecosystem. Conseversly,

shifts in marine climate resulting in increased

storm events and hurricanes have caused sig-

nificant shifts in dinoflagellates communities

and density in Florida (Fiorendino et al., 2021;

Lopez et al., 2021; Phlips et al., 2020). Thus,

our study of ocean influences during an active

hurricane season on dinoflagellates in LGF was

conceived to 1) Quantify temporal changes in

bioluminescent dinoflagellate density in a six

year monitoring study, 2) Understand how the

marine environment modulates those changes,

and 3) Determine the wider impacts of a fluctu-

ating climate and extreme events on proxies for

dinoflagellate density over two decades.

MATERIAL AND METHODS

Dinoflagellate sampling points were estab-

lished in LGF: Pt.1 in Southwest shallows (0 m

depth), pt.2 in the Middle (2 m depth), pt.3 in

mid-Northern shallows (0 m), and pt.4 (from

2018) at the Southeast canal entrance (Fig. 1A,

Fig.1B, Fig. 1C). Each point was sampled twice

a month from July 2016 to December 2021

just after sunset 19:00 + LST, so to observe

bioluminescence on a 0-3 scale: None-to-high.

LGF watercolor ranged from 0: dark, 1: pale

white, 2: blue hue, 3: bright turquoise, with

1 and 2 requiring water stirring. In addition

to such visual observations, three 1 l water

samples were collected by hand in the shal-

lows, and with a Van Dorn depth sampler at

point 2 (at 2 m). Sampling was disrupted from

Oct-Dec 2017 by hurricanes and Mar-Apr 2020

by the COVID-19 pandemic. Water samples

were preserved with 20 ml of 2 % formalin and

transported to the Climate & Ecological Studies

Lab at the Interamerican University of Puerto

Rico, San Juan.

Organisms settled to the bottom of the 1 l

bottle (Huber, 2012) and were marked by a line

at 20 ml. The remaining water was decanted

after 48 hours using an automated 100 ml

volumetric pipette and filtered through a 20 μm

Nitex nylon mesh to isolate phytoplankton in

suspension (Sastre et al., 2013). The unfiltered

20 ml sample and cells retained by the mesh

were combined, washed into a 50 ml Falcon

centrifuge tube, and stored in our lab. The

concentrated plankton samples were swirled

in a 50 ml tube for counting, and a 1 ml sub-

sample was transferred to a gridded Sedwick-

Rafter (S-R) (Huber, 2012). The S-R counting

cell was covered with glass, and all identifi-

able dinoflagellate taxa were counted using a

Leica CME or Nikon Eclipse microscope at

100-200X magnification.

Since Sastre et al. (2013) found little cor-

relation between the lagoon environment and

dinoflagellate density, open ocean data were

collected from a NOAA buoy 41056: 18.26°N &

65.46°W (CariCOOS, 1996). Time series were

obtained during the field monitoring for wind

velocity, air and sea temperature and salinity.

Fresh water inputs to LGF were derived from

U.S. Geological Survey (2021) river discharge

at Rio Fajardo 5007100. Using ~ 66 months

of environmental data, a statistical analysis

was conducted by averaging bioluminescent

dinoflagellate density in LGF to monthly. Com-

parative work employed the non-parametric

Kruskal Wallis and Tukey Tests, Spearman

cross-correlation and multi-variate regression.

4Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

The non-parametric tests were used because

dinoflagellate density did not meet the normal-

ity assumption even after attempts of trans-

formation. The Kruskal Wallis Test and Tukey

Test is used to compare dinoflagellates density

by month, year, and sampling station. Correla-

tion analysis was completed for wind direction,

wind speed (m/s), air pressure (hPa), marine air

temperature (°C), salinity (‰), sea temperature

(°C), Fajardo River discharge (ft3/s) and dino-

flagellate density (ind/l).

After review, a second independent analy-

sis was conducted over a wider area and longer

period 2002-2022 using satellite and coupled

reanalysis fields based on assimilation of in-situ

and remote sensing measurements (Balmaseda

et al., 2015; Storto et al., 2019). Daily time series

were extracted at LGF: 18.39°N & 65.61°W, and

maps of the surrounding marine climate were

constructed. MODIS and VIIRS 4 km resolu-

tion satellite color radiance around LGF (Fig

1B) was extracted in the form of green-band

Fig. 1. A. Puerto Rico’s coast with shelf 0-60 m shaded; arrow points to LGF, buoy = x. B. Aerial view of LGF from SW to

NE. C. Close-up of Fajardo Peninsula and the coastal lagoon, sampling points, elevation labels, canal exchange with ocean

denoted by arrow.

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

chlorophyll (0.44 μm) and red-band fluo-

rescence (0.67 μm). Other satellite data sets

include satellite visible-band 4 km land veg-

etation color fraction, infrared 4 km GHR sea

surface temperature (SST), 25 km resolution

EC multi-satellite altimeter sea level, net out-

going longwave radiation (OLR) to represent

atmospheric convection, and MODIS radiom-

eter light extinction by atmospheric dust. The

Hybrid Coordinate Ocean Model 10 km reso-

lution coupled reanalysis (HYCOM3), (Chas-

signet et al., 2009) described surface ocean

currents, salinity, mixed layer depth (MLD) and

Ekman transport. Wavewatch 3 (W3) 25 km

reanalysis (Tolman, 2002) was used for surface

wave characteristics, based on buoy and satellite

assimilation in a multi-spectral model (Chawla

et al., 2013).

Marine weather at LGF was character-

ized by 25 km resolution European Reanalysis

(ERA5) and NOAA Coupled Forecasts (CFS2)

(Hersbach et al., 2020; Saha et al., 2014) for

surface winds, total heat balance and net solar

radiation. Environmental impacts of the 2017

hurricanes were studied via hourly gauge sea

level, buoy SST, daily wind time series and

satellite vegetation color fraction. Currents and

salinity were analyzed for 2018-2019 during

dinoflagellate blooms in LGF, as Hovmoller

plots on a N-S slice 65.6W in 2018-2019, and as

maps for 10 Oct. 2019 and 9 Nov. 2020.

Within the longer records, we found that

satellite red-band radiance (Jury, 2020) was out-

of-phase with our field data, so it was relegated.

Instead satellite green-band chlorophyll aligned

with dinoflagellate counts, while environmental

conditions were represented by sea level and

salinity. Pair-wise cross-correlations with satel-

lite and reanalysis variables (cf. Table 4) were

calculated to establish environmental influenc-

es over 2002-2022 and 2016-2021 respectively.

Serial persistence limits the degrees of free-

dom so |R| > 0.20 (longer record) and > 0.38

(shorter record) reaches 98 % confidence. To

understand inter-annual climate influence, 20

year time series were filtered to remove cycles

below 18 months (de-seasonal). Detrended

cross-correlations were analyzed for sea level,

zonal wind, and salinity. Wavelet spectra were

calculated to determine significant periodicity.

Late summer (Aug-Nov) chlorophyll content

was regressed onto fields of ERA5 surface

wind, satellite SST and meridional currents

around Puerto Rico 14.5-22°N & 70-60°W in

the period 2002-2022.

Point measurements from buoys, gaug-

es and stations should be supplemented with

coastal weather and ocean fields over a longer

time to fully explain the environmental controls

on bioluminescence in a coastal lagoon. These

methods distinguish our work from most field

studies with narrow sampling windows con-

fined to in-situ measurements.

RESULTS

Characteristics of field data: Field data

(Table 1, Table 2, Fig. 2A) exhibit wide fluc-

tuations of dinoflagellate density in LGF, with

monthly counts ranging from 4 388 cells/l in

Jun 2018 to 2.19 106 cells/l in Nov 2019. Sta-

tistical tests quantify the large intra-seasonal

variability: Kruskal Wallis: H = 105.43, P < .001

(Table 3, Fig. 2B). Annual dinoflagellate mini-

ma/maxima were noted in the cool dry / warm

wet season. Densities were low in 2016-17 com-

pared with 2018-20 (Fig. 2C), yielding signifi-

cant inter-annual variability (Kruskal Wallis:

Table 1

Descriptive statistics of dinoflagellate density (ind/l) by month 2016-2021 at LGF.

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Mean 262 523 306 823 334 224 77 687 78 500 302 931 144 706 690 132 379 409 831 808 1 062 000 736 072

Std. Dev. 380 777 321 594 416 978 63 775 91 233 548 597 280 487 797 234 620 360 915 352 940 367 588 353

Min. 15 667 17 944 5 474 36 727 22 556 4 833 7 328 10 000 23 556 142 278 31 000 128 000

Max. 701 050 629 300 877 300 151 167 183 778 1 124 000 646 250 1 654 000 1 475 000 2 160 000 2 190 000 540 000

6Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

H = 182.26 P < .001). Spatial heterogeneity

between collection sites in the 50-ha saltwater

coastal lagoon was significant (Kruskal Wallis:

H = 33.05, P < .001) (Table 4, Fig. 2D).

Comparison with in-situ environmental

variables: Discharge and buoy data time series

tended to have skew distributions requiring

non-parametric tests. Marine environmental

conditions off Fajardo (Fig. 3A, Fig. 3B, Fig. 3C,

Fig. 3D, Fig. 3E, Fig. 3F) exhibited noteworthy

seasonality: lowest air pressure, weakest aver-

age winds; warmest air and sea temperatures

and lowest salinity all occurred in the Aug-Nov

period. River discharge exhibited large inter-

annual fluctuations (Tukey Test: T > 2.3 P <

0.05) unrelated to offshore salinity. Through-

out 2016 and early 2017 salinity was high,

then declined in 2018 and remained low in

subsequent years. Most marine environmental

variables were poorly correlated with dinofla-

gellate density in LGF (Table 5), except salinity

Table 3

Tukey test to compare dinoflagellate density by month.

Months Months Mean Diff. SE t P tukey

1 11 -817 016 168 621 -4.845 < .001

2 11 -729 254 161 995 -4.502 < .001

3 10 -567 362 163 138 -3.478 0.026

3 11 -822 945 162 654 -5.059 < .001

4 8 -610 321 174 368 -3.5 0.025

10 -748 680 180 020 -4.159 0.002

11 -1.004e +6 179 582 -5.592 < .001

12 -655 815 179 582 -3.652 0.015

5 8 -604 420 176 914 -3.416 0.032

10 -742 779 182 487 -4.07 0.003

11 -998 362 182 055 -5.484 < .001

12 -649 914 182 055 -3.57 0.019

6 10 -529 537 159 435 -3.321 0.044

11 -785 120 158 941 -4.94 < .001

12 -436 672 158 941 -2.747 0.205

7 8 -538 008 143 392 -3.752 0.01

10 -676 367 150 213 -4.503 < .001

11 -931 951 149 688 -6.226 < .001

12 -583 502 149 688 -3.898 0.006

8 10 -359 182 156 738 -2.292 0.483

11 -614 765 156 234 -3.935 0.005

9 11 -614 765 156 234 -3.935 0.005

P-value adjusted for 11 degrees of freedom.

Table 2

Descriptive Statistics of dinoflagellate density by year at LGF (months sampled).

2016 (6) 2017 (9) 2018 (11) 2019 (12) 2020 (10) 2021 (10)

Mean 135 425 41 559 538 557 781 167 513 659 287 464

Std.Dev. 235 878 48 196 795 056 665 465 468 503 237 243

Min. 7 328 5 474 4 833 22 556 29 950 103 301

Max. 598 390 151 167 2 176 000 2 200 000 1 591 000 857 000

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

(R –0.56), explaining 1/3 of variance (Fig. 4A).

Salinity adjustments within LGF may be related

to runoff, but current advection may control

open ocean conditions (Fig. 4B).

Wider context with satellite reanalysis

data: Mapping the long-term average marine

conditions Fig. 5A, Fig. 5B and Fig. 5C we find

HYCOM3 salinity gradients between Atlantic

35.9 ‰ and Caribbean 35.4 ‰ waters. Cur-

rents sweep Southwestward through the Virgin

Islands and turn Northwestward 0.1 m/s near

Fajardo. Satellite SST are warmer in the Carib-

bean (28 °C), leeward of the Antilles Islands.

Trade winds prevail at ~ 6 m/s and induce

Ekman transport that draws seawater North-

ward off Fajardo. Time series of daily sea level

(Fig. 6A) show rising trends and annual cycling

that crests in late October. Higher tides are evi-

dent in 2012, 2014, 2015, 2018, 2019, 2021, that

tend to flush LGF. Time series of daily SST and

salinity (Fig. 6B) show a delayed inverse rela-

tionship: warmer sea temperatures are followed

by lower salinity. SST shows little trend, while

salinity at Fajardo declined after the drought in

2015 and salty spell of 2016-17. Salinity below

34.5 ‰ was evident during late summer (Oct-

Nov) of 2019-22, due to an influx of South

Table 4

Tukey Test comparison of dinoflagellate density by

sampling site.

Station  Mean Diff. SE t P tukey

1 2 -512 391 85 151.3 -6.017 < .001

3-496 792 85 151.3 -5.834 < .001

4-482 426 165 677 -2.912 0.019

2 3 15 599 85 230.2 0.183 0.998

429 965 165 717 0.181 0.998

3 4 14 366 165 717 0.087 1

P-value adjusted for 4 degrees of freedom.

Fig. 2. Variability of dinoflagellate density over six year sampling period by: A. Time series of twice-monthly in-situ

dinoflagellate density at LGF. B. Month. C. Year. D. Station. Box-whisker plots with median, upper/lower quartiles and

outliers.

8Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

Fig. 3. Box-whisker plots (median, quartiles, outliers) indicating seasonal cycles and inter-monthly variability of: A. Wind

speed (m/s). B. Air pressure (hPa). C. marine air temperature (°C). D. salinity (‰). E. Sea temperature (°C). F. Fajardo River

discharge (ft3/s).

Table 5

Correlation of Fajardo discharge and buoy time series with monthly dinoflagellate density 2016-2021.

Variable Spearman

Coefficient

Dinoflag.

(ind/l)

WDIR

(o)

WSPD

(m/s)

WGST

(m/s)

Pressure

(hPa)

airTemp

(oC)

se aTemp

(oC)

Salinity

(‰)

WDIR R-value -0.218 -0.538

 P-value 0.15 <.001

Pressure R-value -0.204 -0.466 0.625 0.619

 P-value 0.179 <.001 <.001 <.001

airTemp R-value 0.113 0.043 -0.306 -0.337 -0.552

 P-value 0.461 0.763 0.029 0.016 <.001

se aTemp R-value 0.121 0.095 -0.417 -0.445 -0.635 0.979

 P-value 0.451 0.557 0.008 0.004 <.001 <.001

Salinity R-value -0.559 0.028 0.278 0.231 0.305 -0.139 -0.212

 P-value <.001* 0.864 0.083 0.151 0.056 0.392 0.172

Discharge R-value -0.068 0.030 -0.056 -0.047 -0.384 0.309 0.328 -0.097

 P-value 0.644 0.833 0.697 0.742 0.006 0.028 0.032 0.534

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

American river plumes. Time series of daily

winds and waves (Fig. 6C) are pulsed by tropi-

cal cyclones and trade wind surges. Over the

20-year period, 2014-16 had low waves while

2004-10 and 2019-22 saw high wave heights.

Although hurricanes generate peak events at

Fajardo (Sep 2017), wave action along this coast

is diminished by upstream islands and shallow

bathymetry.

Turning our attention to seawater color,

the satellite green-band radiance (CHLa) map

for Oct 2019 is illustrated in Fig. 7A. High val-

ues hug the coast around LGF, and in leeward

zones of the Virgin Islands and Puerto Rico.

Mean annual cycles are investigated in Fig. 7B,

Chlorophyll rises from Jul-Nov (warm wet sea-

son) and reaches a minimum in Jan-Apr (cool

dry season). Northward currents are bi-modal

and peak in summer and winter as trade winds

induce Ekman transport. Sea level dips in Mar-

Apr and crests broadly from Aug-Nov during

infiltration of South American river plumes.

Fig. 4. A. Regression of buoy salinity and dinoflagellate density at LGF. B. Temporal cycles of salinity during the monitoring

period, lines indicate monthly salinity per year, shaded is the average.

10 Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

Monthly satellite chlorophyll time series and

LGF dinoflagellate counts are plotted in Fig. 7C.

The green-band radiance shows high values in

2004, 2009-10, and low values from 2012-2017

that correspond with LGF data. Sustained high

chlorophyll content in 2019-2022 coincides

with dinoflagellate density above 106/l.

Pair-wise cross-correlations of the Fajar-

do chlorophyll time series with continuous

monthly satellite reanalysis variables 2002-2022

(A) (Table 6) reveals significant associations

with SST, atmospheric convection (net OLR)

and land vegetation color which reflect how

seasonal warming enhances Aug-Nov rainfall

Fig. 5. Long-term mean maps of: A. HYCOM surface salinity (‰). B. Surface currents (largest 0.1 m/s), dashed arrow

highlights inflow. and C. GHR satellite sea surface temperature and ERA5 winds (largest vector 7 m/s).

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

and canopy growth. The cross-correlation of

LGF dinoflagellates with most environmental

variables is weak (B) (Table 6), only sea level is

significantly associated, followed by vegetation,

solar radiation and salinity. Although satellite

chlorophyll and in-situ dinoflagellate counts are

weakly associated, they share relationships with

many environmental variables.

With sea level being significantly corre-

lated with bioluminescent algal blooms at LGF,

it is inferred that flushing action by high tides

in Aug-Nov is beneficial. However, extreme

mechanical stress during storm surges may

not be. Having earlier noted high dinoflagel-

late counts in late 2018 and 2019, we focus on

environmental conditions during that era by

analysis of salinity and currents.

A case of Northward currents and low

salinity is mapped in Fig. 8A. This pattern

coincides with bioluminescent dinoflagellate

Fig. 6. Daily time series of A. Sea level and anomalous high tides at Fajardo, B. SST and salinity, and C. Wind speed and wave

height, identifying key features and storm events, field monitoring bracketed, annual tick marks in July.

12 Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

blooms in LGF. Currents in the Atlantic sector

is Eastward and decline near Fajardo, where

they meet Northward currents coming from

the Caribbean sector. A strong salinity gradient

is sustained. The canal connecting LGF with

the open ocean faces the Caribbean and thus

entrains low salinity seawater. Dinoflagellate

density during this time reached 5-7 million /l.

A hovmoller plot of salinity along the East coast

(Fig 8B) shows that pulses of low salinity from

the Caribbean in Aug-Nov in 2018 and 2019.

Surplus precipitation over evaporation and the

Northward spread of river plumes from South

America account for these changes.

Analysis of inter-annual variability: To

understand inter-annual variability, the 20 year

records of Fajardo Sea level, wind and salinity

were filtered for oscillations above 18 months.

Fig. 9A, Fig. 9B and Fig. 9C illustrate time

series that exhibit some coherence. Simultane-

ous cross-correlations between sea level and

salinity are -0.74 and between sea level and U

wind -0.59. These outcomes say that increased

Fig. 7. A. VIIRS satellite chlorophyll map for Oct 2019 (mg/m3) when blooms were recorded in LGF. B. Mean annual cycle

of chlorophyll and Northward currents (left) and sea level and salinity. C. Time series of CHLa at Fajardo 2003-2022, field

monitoring bracketed on y-axis, annual tick marks on July. (CHLa and salt use right-hand axes in b).

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

trade winds (-U) draw low saline water from

the Caribbean during high tides.

Wavelet spectral analysis (not shown)

found significant oscillations at ~ 3.5 year for

sea level and zonal wind, and ~ 7 year periods

for CHLa and salinity. These rhythms derive

from the Atlantic Walker Circulation (Jury &

Nieves Jiménez, 2020). A striking feature in

Fig. 9B, Fig. 9C is the downtrend in U wind

and salinity since 2016 and sustained high

tides since 2018. Increased Northward Ekman

transport is linked to faster Atlantic trade winds

and account for the uptick in bioluminescent

dinoflagellates after 2017.

To characterize spatial patterns of inter-

annual variability, the chlorophyll time series

was regressed onto Aug-Nov environmental

fields 2002-2022 (Fig. 9D, Fig. 9E). A N-S

dipole pattern emerges: phytoplankton blooms

relate to weaker trade winds / warmer SST in

the Caribbean, and stronger trade winds/cooler

SST in the Atlantic. Aug-Nov chlorophyll con-

centrations at Fajardo are stimulated by North-

ward currents off the East coast of Puerto Rico

that entrain Caribbean waters (Fig. 9E). The

N-S dipole pattern and trends (cf. Fig. 9B, Fig.

9C) suggest a Northward expansion of trade

winds consistent with climate change impacts.

Hurricanes: Northeastern Puerto Rico

experienced eight tropical cyclones during the

field campaign 2016-2021; hurricanes in 2017

impacted the LGF environment via 6 m waves,

0.8 m tidal surge, and 32 m/s winds that

stripped vegetation (Fig. 10A, Fig. 10B, Fig.

10C). Hurricane Maria changed structural and

functional features in the lagoon. It dropped

local SST from 30.5° to 28.3 °C and salinity

from 35.5 ‰ to 34.4 ‰ in Sep 2017, creating

thermodynamic conditions for dinoflagellate

blooms in LGF - following recovery from initial

mechanical stress (Fig. 1D and Fig. 2B).

DISCUSSION

During this study we found widely fluc-

tuating dinoflagellate densities at LGF: rising

in late summer and declining in late winter,

similar to seasonal patterns of P. bahamense

in Tampa Bay, where sea temperatures modu-

late cyclical blooms. The greatest dinoflagel-

late densities at LGF coincided with high tides

and low salinity from Aug-Nov. While our

short-term monitoring uncovered frequent

and intense blooms in 2018-2021, (Hinder et

al., 2012) found a down-trend of dinoflagel-

late densities in the Northeast Atlantic due to

increasing SST and wind speed. That study

challenged the view that algal blooms are on

the rise and found that previously abundant

dinoflagellates C. furca and Protoperidinium

Table 6

Cross-correlation of: A. Monthly green-band radiance (CHLa) at Fajardo and marine climate variables from coupled

reanalysis 2002-2022, B. Correlation of monthly dinoflagellate density and the same variables 2016-2021.

A CHLa  CHLa B Dinoflag. Dinoflag.

MLD -0.13 U wind -0.06 MLD 0.01 U wind 0.06

salinity -0.11 V wind 0.09 Salinity -0.24 V wind -0.03

SST 0.27 Waves 0.18 SST 0.15 Waves 0.09

U curr 0.18 net heat -0.09 U curr 0.16 net heat -0.13

V curr 0.05 Euph Z -0.19 V curr -0.04 Euph Z -0.12

sea level 0.17 dust 0.17 sea level 0.4 dust -0.18

net OLR -0.27 Ek V -0.05 net OLR 0.06 Ek V -0.22

Veget 0.36 solar -0.19 Veget 0.27 solar -0.26

  CHLa 0.15

Bold values have P < .02; acronyms: MLD mixed layer depth, SST sea surface temp, U V vector components, curr. currents, n

OLR net outgoing longwave radiation, net heat balance of components, Euph Z euphotic depth or water clarity, optical dust

extinction, Ek V Northward Ekman transport, and net solar radiation.

14 Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

spp. have diminished. Pre-hurricane results at

LGF showed a similar effect, when stakeholders

reported bioluminescent ‘blackouts’. After the

storm surges in 2017, our study found sustained

blooms that boosted eco-tourism.

Similarly, Hurricanes of 2017 in Florida’s

coastal waterways caused enhanced nutrient

loads which drove P. bahamense blooms in the

Indian River Lagoon and Lake Okeechobee

(Philips, 2020). This study reported blooms

immediately after Hurricanes due to exception-

ally high rainfall resulting from large inflows

of nutrient-rich water to the waterbodies

through increases in dissolved inorganic nitro-

gen, intense re-suspension of muddy floccu-

lent bottom sediments, resulting in high total

Fig. 8. A. Hycom surface currents (vector largest 0.2 m/s) and salinity (‰) on 10 Oct 2019 during a dinoflagellate bloom in

LGF. B. Hycom surface salinity hovmoller plot N-S on 65.6W off Fajardo in 2018-19, arrow identifies map date.

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

suspended solids concentrations and increased

nutrient concentrations through sediment re-

suspension and re-mineralization of nutrients

from damaged aquatic vegetation (e.g. sea-

grasses) and damaged terrestrial biomass in

the watershed. These key processes were also

observed in LGF following Hurricanes Irma

and Maria explaining the P. bahamense in the

following years, i.e.2018 and 2019.

Increased sea level and lower salin-

ity favored dinoflagellate growth in the longer-

term, and these two are linked (Fig. 11A).

Fig. 9. Inter-annual filtered satellite reanalysis time series at Fajardo: A. Sea level. B. U wind. C. salinity, with field monitoring

bracketed, annual tick marks on July. Inter-annual regression of CHLa time series onto Aug-Nov fields 2002-2022: D. SST

(shaded /°C) Fajardo , and E. meridional current (shaded + Northward, /m s-1) and wind (vector anomalies, largest 2 m/s).

16 Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

Given that plumes of seawater come from

the Orinoco and Amazon Rivers; the salinity

budget depends on the relative flux of Carib-

bean versus Atlantic seawater. Tropical climate

patterns modulate Fajardo daily sea level, as

evidenced by regression onto Aug-Nov fields of

netOLR 2002-2022 (Fig. 11B) which show that

a dry climate in the East Pacific contributes to

high tides which support bioluminescence.

The late summer peak in dinoflagellate

counts at LGF is like the higher latitudes (Hin-

der et al., 2012; Marcinko et al., 2013, Lopez et

Fig. 10. A. Fajardo gauge sea level during Sep 2017 hurricanes; inset: buoy data at Maria landfall 20 Sep 2017. B. Daily wind

stick vectors during Sep 2017 and buoy SST (red labels). C. Before (left) and after changes in satellite vegetation color.

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

al., 2021). Light winds and mixed layer response

to seasonal heating promote algal blooms. Nat-

urally these macro-scale studies do not capture

local effects such as the dip in LGF dinoflagel-

lates during mid-summer, when strong Easterly

winds bring Saharan dust plumes (Angeles et

al., 2010; Jury & Nieves Jiménez, 2020).

The environmental health of tidal lagoons

is vital to coastal tourism and ongoing resource

management. Our LGF results revealed that:

1) dinoflagellate counts fluctuate widely–often

surprising our monitoring team, 2) biolumi-

nescent dinoflagellates are sensitive to small

changes in environmental conditions–subtropi-

cal islands experience limited seasonality so

biota have a narrow physiological range and few

opportunities to adapt, 3) hurricanes ‘rake and

refresh’ the marine environment for subsequent

Fig. 11. A. Scatterplot of daily EC sea level and HYCOM salinity 2002-2022 and regression fit; arrows point to conditions

favoring bioluminescence in LGF. B. Regression of daily sea level at Fajardo onto Aug-Nov 2002-2022 field of net OLR (/W

m-2). South American river discharge and plume edge shown by blue contours > 1000 m3/s (land) and < 35‰ (sea).

18 Revista de Biología Tropical, ISSN: 2215-2075 Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

bioluminescent blooms following recovery–

an inference requiring further data, and 4)

intra-seasonal fluctuations of LGF dinoflagel-

late density relate to ocean-atmosphere ther-

modynamic conditions, the salinity budget and

sea level. We recommend continued monthly

in-situ monitoring of LGF biota and on-going

support for marine observations to guide con-

servation strategies of this natural resource,

amidst a changing climate.

Ethical statement: the authors declare that

they all agree with this publication and made

significant contributions; that there is no con-

flict of interest of any kind; and that we fol-

lowed all pertinent ethical and legal procedures

and requirements. All financial sources are fully

and clearly stated in the acknowledgments sec-

tion. A signed document has been filed in the

journal archives.

ACKNOWLEDGMENTS

Project funding was provided by Para la

Naturaleza and Interamerican University of

Puerto Rico, Institutional Funds. The authors

acknowledge family, students, and colleagues

who collected and processed samples in the Cli-

mate and Ecological Studies Laboratory from

2016 to 2021. Websites used in our data analysis

include Climate Explorer KNMI, IRI Climate

Library, University of Hawaii APDRC, NOAA

NDBC.

REFERENCES

Álvarez, S., Brown, C. E., García Diaz, M., O’Leary, H., &

Solís, D. (2024). Non-linear impacts of harmful algae

blooms on the coastal tourism economy. Journal of

Environmental Management, 351, 119811. https://doi.

org/10.1016/j.jenvman.2023.119811

Angeles, M. E., González, J. E., Ramírez-Beltrán, N. D.,

Tepley, C. A., & Comarazamy, D. E. (2010). Origins

of the Caribbean rainfall bimodal behavior. Journal

of Geophysical Research: Atmospheres, 115(D11106).

https://doi.org/10.1029/2009jd012990

Badylak, S., & Phlips, E. J. (2009). Pyrodi-

nium Bahamense one the most signifi-

cant harmful https://www.researchgate.net/

publication/330565577_Pyrodinium_bahamense_

One_the_Most_Significant_Harmful_Dinoflagella-

te_in_Mexico

Balmaseda, M. A., Hernandez, F., Storto, A., Palmer, M. D.,

Alves, O., Shi L., Smith, G. C., Toyoda, T., Valdivieso,

M., Barnier, B., Behringer, D., Boyer, T., Chang, Y. S.,

Chepurin, G. A., Ferry, N., Forget, G., Fujii, Y., Good,

S., Guinehut, S., & Gaillard, F. (2015). The Ocean

Reanalyses Intercomparison Project (Ora-IP). Journal

of Operational Oceanography, 8(sup1). https://doi.org

/10.1080/1755876x.2015.1022329

CariCOOS (1996). NDBC station page. National Data Buoy

Center. https://www.ndbc.noaa.gov/station_page.

php?station=41056

Chassignet, E., Hurlburt, H., Metzger, E. J., Smedstad, O.,

Cummings, J., Halliwell, G., Bleck, R., Baraille, R.,

Wallcraft, A., Lozano, C., Tolman, H., Srinivasan,

A., Hankin, S., Cornillon, P., Weisberg, R., Barth, A.,

He, R., Werner, F., & Wilkin, J. (2009). US GODAE:

Global ocean prediction with the hybrid coordinate

ocean model (HYCOM). Oceanography, 22(2), 64–75.

https://doi.org/10.5670/oceanog.2009.39

Chawla, A., Tolman, H. L., Gerald, V., Spindler, D., Spindler,

T., Alves, J. G. M., Cao, D., Hanson, J. L., & Devalie-

re, E. (2013). A multigrid wave forecasting model:

A new paradigm in operational wave forecasting.

Weather and Forecasting, 28(4), 1057–1078. https://

doi.org/10.1175/WAF-D-12-00007.1

Department of Environment and Natural Resour-

ces. (2013). Comunicado de Prensa DRNA Y JCA

revelan resultados sobre laguna bioluminiscente

de Fajardo. https://www.drna.pr.gov/wp-content/

uploads/2019/10/r-CP-6-de-diciembre-de-2013-DR-

NA-y-JCA-revelan-resultados-sobre-laguna-biolumi-

niscente-de-Fajardo.pdf

Fiorendino, J. M., Gaonkar, C. C., Henrichs, D. W., & Cam-

pbell, L. (2021). Drivers of microplankton commu-

nity assemblage following tropical cyclones. Journal

of Plankton Research, 45(1), 205–220. https://doi.

org/10.1093/plankt/fbab073

Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi,

A., Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu,

R., Schepers, D., Simmons, A., Soci, C., Abdalla, S.,

Abellan, X., Balsamo, G., Bechtold, P., Biavati, G.,

Bidlot, J., Bonavita, M., …& Thépaut, J. (2020). The

ERA5 global reanalysis. Quarterly Journal of the Royal

Meteorological Society, 146(730), 1999–2049. https://

doi.org/10.1002/qj.3803

Hinder, S. L., Hays, G. C., Edwards, M., Roberts, E. C.,

Walne, A. W., & Gravenor, M. B. (2012). Changes in

marine dinoflagellate and diatom abundance under

climate change. Nature Climate Change, 2(4), 271–

275. https://doi.org/10.1038/nclimate1388

Revista de Biología Tropical, ISSN: 2215-2075, Vol. 72: e56729, enero-diciembre 2024 (Publicado Ago. 13, 2024)

Huber, N. (2012). Estuaries -- Phytoplankton Protocol Wash-

ington. http://courses.washington.edu/uwtoce12/

methods/sops/phytoplankton.pdf

Jury, M. R. (2020). Sand transport in the northeastern

Caribbean characterized by wind-wave-current data.

Ocean & Coastal Management, 198, 105363.

https://doi.org/10.1016/j.ocecoaman.2020.105363

Jury, M. R., & Nieves Jiménez, A. T. (2020). Tropical Atlan-

tic dust and the zonal circulation. Theoretical and

Applied Climatology, 143(3–4), 901–913. https://doi.

org/10.1007/s00704-020-03461-4

Lopez, C. B., Tilney, C. L., Muhlbach, E., Bouchard, J. N.,

Villac, M. C., Henschen, K. L., Markley, L. R., Abbe, S.

K., Shankar, S., Shea, C. P., Flewelling, L., Garrett, M.,

Badylak, S., Phlips, E. J., Hall, L. M., Lasi, M. A., Parks,

A. A., Paperno, R., Adams, D. H., … & Hubbard,

K. A. (2021). High-resolution spatiotemporal dyna-

mics of harmful algae in the Indian River Lagoon

(florida)—a case study of aureoumbra lagunensis,

Pyrodinium Bahamense, and pseudo-nitzschia. Fron-

tiers in Marine Science, 8. https://doi.org/10.3389/

fmars.2021.769877

Marcinko, C. L. J., Painter, S. C., Martin, A. P., & Allen, J. T.

(2013). A review of the measurement and modelling

of Dinoflagellate bioluminescence. Progress in Ocea-

nography, 109, 117–129. https://doi.org/10.1016/j.

pocean.2012.10.008

Morquecho, L. (2019). Pyrodinium bahamense one the

most significant harmful dinoflagellate in Mexi-

co. Frontiers in Marine Science, 6, 1. https://doi.

org/10.3389/fmars.2019.00001

O’Connell, S., Martini, A., & Ku, T. (2007). The hydrodyna-

mics and biogeochemistry of bioluminescent. https://

keckgeology.org/files/pdf/symvol/20th/puertorico/

oconnell.pdf

Pagán, G. (2021). Esfuerzo multisectorial para Atender

Situación de emergencia en Laguna Bioluminiscente

en Fajardo. DRNA. https://www.drna.pr.gov/noticias/

esfuerzo-multisectorial-para-atender-situacion-de-

emergencia-en-laguna-bioluminiscente-en-fajardo/

Phlips, E. J., Badylak, S., Nelson, N. G., & Havens, K. E.

(2020). Hurricanes, El Niño and harmful algal blooms

in two sub-tropical Florida estuaries: Direct and

indirect impacts. Scientific Reports, 10(1). https://doi.

org/10.1038/s41598-020-58771-4

Saha, S., Moorthi, S., Wu, X., Wang, J., Nadiga, S., Tripp,

P., Behringer, D., Hou, Y.-T., Chuang, H., Iredell, M.,

Ek, M., Meng, J., Yang, R., Mendez, M. P., van den

Dool, H., Zhang, Q., Wang, W., Chen, M., & Becker,

E. (2014). The NCEP climate forecast system version

2. Journal of Climate, 27(6), 2185–2208. https://doi.

org/10.1175/jcli-d-12-00823.1

Sastre, M. P., Sánchez, E., Flores, M., Astacio, S., Rodríguez,

J., Santiago, M., Olivieri, K., Francis, V., & Núñez, J.

(2013). Population fluctuations of Pyrodinium baha-

mense and Ceratium furca (Dinophyceae) in Laguna

Grande, Puerto Rico, and environmental variables

associated during a three-year period. Revista de Bio-

logía Tropical, 61(4), 1799–1813.

Soler-Figueroa, B. M., & Otero, E. (2014). The influence of

rain regimes and nutrient loading on the abundance

of two dinoflagellate species in a tropical biolumines-

cent bay, Bahía Fosforescente, La Parguera, Puerto

Rico. Estuaries and Coasts, 38(1), 84–92. https://doi.

org/10.1007/s12237-014-9827-0

Soler-Figueroa, B. M., & Otero, E. (2016). Seasonal changes

in bioluminescence and dinoflagellate composition in

a tropical bioluminescent bay, Bahía Fosforescente,

La Parguera, Puerto Rico. Journal of Experimental

Marine Biology and Ecology, 483, 120–129. https://doi.

org/10.1016/j.jembe.2016.07.008

Soler-López, L. R., & Santos, C. R. (2010). Selected hydro-

logic, water-quality, biological, and sedimentation cha-

racteristics of Laguna Grande, Fajardo Puerto Rico,

March 2007-February 2009. U.S. Geological Survey,

Scientific Investigations Report 2010-5071.

Storto, A., Alvera-Azcárate, A., Balmaseda, M. A., Barth,

A., Chevallier, M., Counillon, F., Domingues, C.

M., Drevillon, M., Drillet, Y., Forget, G., Garric, G.,

Haines, K., Hernandez, F., Iovino, D., Jackson, L. C.,

Lellouche, J.-M., Masina, S., Mayer, M., Oke, P. R., &

Zuo, H. (2019). Ocean reanalyses: Recent advances

and unsolved challenges. Frontiers in Marine Science,

6. https://doi.org/10.3389/fmars.2019.00418

Tolman, H. L. (2002). User manual and system documen-

tation of WAVEWATCH III. https://polar.ncep.noaa.

gov/mmab/papers/tn276/MMAB_276.pdf

U.S. Geological Survey. (2021). National Water Information

System for Rio Fajardo, PR (river discharge data. water-

data.usgs.gov/pr/nwis/uv?site_no=50070900

Weaver, P.L., Ramirez, J. L., & Rivera, J. L.C. (1999).

Las Cabezas de San Juan Nature Reserve. Inception

Report, U.S. Dept. Agriculture, Washington D.C. (pp.

1–27).

Xei, l. (2013). A puzzling “blackout” at Puerto Rico’s famous

Bioluminescent Bay. https://www.bloomberg.com/

news/articles/2013-11-22/a-puzzling-blackout-at-

puerto-rico-s-famous-bioluminescent-bay.