Land-Based Nutrient Enrichment of the Buccoo Reef Complex and

Fringing Coral Reefs of Tobago, West Indies

B. E. Lapointe1, R. Langton2, O. Day2, B. Bedford1, C. Hu3 and Arthur Potts4

1 Center for Marine Ecosystem Health, Harbor Branch Oceanographic Institute, 5600 US

Highway One, Fort Pierce, FL, 34946 U.S.A.

2 Buccoo Reef Trust, Cowie’s Building, Auchenskeoch

Road, Carnbee Junction, Carnbee, Tobago, W. I.

3 Institute for Marine Remote Sensing, College of Marine

Science, University of South Florida, St. Petersburg, FL

33701 U.S.A.

4 Centre for Marine Sciences, Institute of Marine Affairs,

Hill Top Lane, Chaguaramas, P.O. Box 3160, Carenage,

Trinidad and Tobago, W.I.

*Corresponding author; Tel.: 772-465-2400 ext 276; fax: 772-468-0757

e-mail address: lapointe@hboi.edu

Key words: coral reef, macroalgae, eutrophication, sewage, nitrogen, phosphorus,

Orinoco

Abstract

Land-based runoff represents a major source of marine pollution in the Caribbean Sea.

The fringing coral reefs of Tobago, including the Buccoo Reef Complex (BRC), can be

affected locally by wastewater discharges and storm water runoff, as well as regionally

by the floodwaters of the Orinoco River during the wet season. We examined the effects

of land-based nutrient pollution by measuring dissolved water column nutrients (DIN,

SRP), and chlorophyll a, C:N:P and 15N contents of macroalgae, and % cover of coral

and macroalgae at a variety of sites on the Buccoo Reef and fringing reefs (FR) in the dry

(April - June) and wet (September -October) seasons of 2001. Concentrations of DIN

(mostly ammonium) increased at all stations from the dry to wet seasons, in contrast to

SRP that decreased from the dry to wet season. These seasonal shifts in nutrient

concentrations resulted in a significant increase in the DIN:SRP ratio from the dry to wet

season. In April, C:N ratios of macroalgae were significantly lower in the BRC vs. FR

sites (15.8 vs. 21.0), in contrast to higher N:P ratios in the BRC (49.5 vs. 34.6), indicating

greater N-enrichment of the reefs within the Buccoo Reef itself. Overall, the 15N

contents of macroalgae averaged + 6.05 ± 1.59 o/oo and did not vary with season,

although the more remote and less populated sites off Little Tobago island did increase

significantly from the dry to wet season. However, the 15N contents of macroalgae in

the BRC (+ 6.62 o/oo) were significantly higher than the FR sites (+ 5.53 o/oo),

indicating a higher degree of wastewater derived N on the Buccoo Reef. Concurrent

satellite imagery also suggested that although the near-shore waters of Tobago during the

wet season were under significant influence of the Orinoco River plume, the influence

was primarily through riverine colored dissolved organic matter rather than

phytoplankton pigments. The % cover of coral correlated inversely with macroalgae

among the reef sites, with the highest % cover of macroalgae occurring throughout the

Buccoo Reef Complex. These data suggest that local inputs of sewage are a primary

source of nutrient stress on Tobago’s reefs, especially throughout the Buccoo Reef, and

that improved collection and treatment of wastewater could lead to improved water

quality and reef health.

1. Introduction

Coral reefs are biologically diverse and economically valuable ecosystems that

provide an array of ecological services to coastal communities throughout their range.

Reef ecosystems occur in oligotrophic waters and are susceptible to eutrophication as a

result of low level nutrient pollution (Johannes, 1975; NRC 2000). Studies in Kaneohe

Bay, Hawaii, clearly showed how reef corals can be overgrown by macroalgae when

small increases in nutrient concentrations (nitrogen and/or phosphorus) occur as a result

of sewage pollution (Banner, 1974; Smith et al. 1981).

Pollution from land-based sources is now considered the single most important threat

to the marine environment of the Caribbean and impediment to sustainable use of its

resources (UNEP, 1994). Sources of nutrient pollution include sewage outfalls, septic

tanks, storm water runoff, fertilizers, deforestation (Likens, 2001; Lapointe and Thacker

2002) and atmospheric deposition (Barile and Lapointe, 2005).

The island of Tobago, often described as the “Jewel of the Caribbean”, is located at

the southern end of the Lesser Antilles off the coast of Venezuela. Tobago is primarily a

volcanic island but with considerable coral reef growth, especially in the southwestern

area where extensive coral reef development has resulted in a shallow limestone platform

known as the Buccoo Reef Complex (BRC; Fig. 1). The Buccoo Reef Complex

encompasses an area of ~ 7 km-2 and was officially designated as a marine protected area,

The Buccoo Reef Marine Park, in 1973 (Laydoo, 1991). The Buccoo Reef Maine Park is

the best example of contiguous coral reef – seagrass – mangrove ecosystem in the

Republic of Trinidad & Tobago. The BRC is also a major economic asset, attracting

some tens of thousands of visitors per year.

Like most marine protected areas in the Caribbean region, creation of the Buccoo

Reef Marine Park has not protected the area from water quality degradation. Many

coastal dwellings in the contiguous Buccoo Village area, for example, have septic tanks

and “soak-away” systems built in the carbonate-rich limestone that overlay the volcanic

geology. This sewage disposal technology provides little removal of nitrogen, which is

highly mobile in carbonate-rich systems and flows subsurface, via groundwater, into

coastal waters (Lapointe et al. 1990, Costa et al. 2000, Lapointe and Thacker 2001).

Similarly, secondarily treated sewage, a technology that does not remove nutrients

sufficiently to protect coral reef ecosystems, enters Buccoo Bay and the Bon Accord

Lagoon via several sewage treatment plants that service subdivisions near Buccoo

Village (Coral Garden Estates) and in Bon Accord. Sewage pollution in Buccoo Bay

presents a significant risk to public health as indicated by consistently elevated

concentrations of coliform bacteria and fecal streptococci (John 1996).

Although there has long been concern about sewage pollution in the Buccoo Reef

Complex, little is known about the dynamics by which nutrient pollution fosters

eutrophication and degradation of these coral reef communities. Kenny (1976) provided

one of the first qualitative studies of coral communities in the BRC and noted that “care

should be taken not to increase run-off from the adjacent land and to restrict the entry of

pollutants”. Laydoo and Heileman (1987) sampled effluents from sewage treatment

plants servicing subdivisions on the watershed of the BRC and recommended upgrading

the treatment facilities throughout the Buccoo Reef watershed. In 1994, the Institute of

Marine Affairs completed a management plan for the Buccoo Reef Marine Park (IMA,

1994) that specifically recognized sewage impacts (i.e. fecal coliform contamination) in

the inshore waters of Buccoo Bay, but the plan did not address the chronic ecological

impacts of nutrient pollution of the coral reef communities.

Like the Buccoo Reef Complex, fringing reefs around Tobago’s coast could also be

receiving increasing nutrient loads from poorly or inadequately treated sewage effluents.

Both large and small communities rely on “soak-aways” as on the BRC catchment or less

than adequate (i.e. tertiary level) treatment of waste water. Where centralized collection

and treatment systems for sewage have been developed they only provide only

secondary-level treatment at best (no nutrient removal) and are usually inadequate for the

end-volume, resulting in nutrient-rich discharges into rivers or wetlands that eventually

flow into coastal waters.

To specifically assess the chemical and ecological impacts of sewage pollution in the

Buccoo Reef Complex and the fringing reefs around Tobago, we undertook a water

quality sampling and coral reef monitoring project in 2001. The study involved

measurement of water column dissolved inorganic nitrogen (DIN), soluble reactive

phosphorus (SRP), and chlorophyll a during the dry (April - June) and wet (September –

October) seasons. This sampling design was necessary to address the seasonal effects of

the Orinoco River floodwaters that impact Tobago’s coastal waters during the wet season

(Laydoo, 1991). Underwater video of benthic reef communities was used to quantify

cover of hard corals, octocorals, macroalgae, turf algae, coralline algae, and sponges.

Stable nitrogen isotope ratios (15N) in benthic macroalgae were used to discriminate

among natural (N-fixation) versus anthropogenic (e.g. sewage) nitrogen sources fueling

algal blooms and eutrophication in the seasonal samplings.

2. Materials and methods

2.1 Study sites and sampling rationale - A total of six sites in the Buccoo Reef Complex

were sampled to assess water quality and benthic community structure along an inshore

to offshore gradient, to address the importance of land-based nutrient discharges. The

sites included a sewage outfall emptying into Buccoo Bay, a site immediately off the

shore in a location near the house where Prince Margaret honeymooned that we refer to

as Princess Reef and off Buccoo Point, near Buccoo Village, all in the inner BRC;

Walkabout Reef, where glass-bottom boat tour operators allow people to disembark and

walk on the backreef, and Nylon Pool, both in the middle of the BRC; and Coral Gardens

and Outer Reef in the offshore areas of the BRC (Fig. 1). We predicted that the inshore

sites of the Buccoo Reef Complex would show the strongest evidence of nutrient

enrichment from land-based runoff and that there would be a decreasing impact with

distance from shore. In addition, a variety of fringing reefs around Tobago’s Caribbean

and Atlantic coasts were sampled, which included a series of well known dive sites: Mt.

Irvine “Wall”, the Maverick (a sunken ship-artificial reef), Diver’s Dream, a site off shore

from the Tobago Vanguard Hotel, Diver’s Thirst, Culloden, Arnos Vale, Englishman’s

Bay, Three Sisters and two sites off Little Tobago Island -- Kelliston Drain and Black

Jack Hole. We hypothesized that nutrient enrichment would be relatively minimal off

Little Tobago Island, the site furthest offshore and flushed with oceanic “blue water”

waters of the Guyana Current and conversely, greater nutrient enrichment on the fringing

reef sites more directly impacted by land-based runoff from Scarborough and the more

urbanized areas of southwest Tobago, which includes sites off the Vanguard Hotel, Mt.

Irvine, the Maverick, Diver’s Dream and Diver’s Thirst. Finally, we sampled all sites in

both the dry and wet seasons of 2001 (Fig. 2) to assess the role of land-based storm water

runoff in driving seasonal shifts in nutrient concentrations and eutrophication.

2.2 Analysis of seawater for DIN, SRP, and chlorophyll a - SCUBA divers collected

water samples from the various reef sites between April 10 and June 20, 2001 in the dry

season and between September 24 and October 23, 2001 during the wet season. Water

temperatures and depths at the reef sites were measured using Oceanic Datamax Pro Plus

dive computers. Divers used clean, one liter polyethylene Nalgene bottles to collect

replicate (n = 4) seawater samples ~ 0.5 m off the bottom at the various reef sites. The

water samples were held on ice in the dark in a cooler until return to shore where aliquots

of the samples were filtered through a 0.45 m GF/F filter and frozen. The samples were

analyzed for DIN ( = NH4+plus NO3- plus NO2-) and SRP on a Bran and Luebbe

TRAACS Analytical Console at Harbor Branch Oceanographic Institution’s (HBOI)

Environmental Laboratory in Ft. Pierce, FL. The methods for collection, handling, and

processing of all nutrient samples followed a quality assurance/quality control protocol

developed by HBOI’s Environmental Lab. This plan prevented problems associated with

sample contamination and excessive holding times. Salinity of the water samples was

measured to ± 1.0 psu using a Bausch and Lomb refractometer. The GF/F glass fiber

filters used for filtering the water samples were frozen and analyzed for chlorophyll a

(chl a) as a measure of phytoplankton biomass. The filters were extracted for 30 minutes

using 10 ml of dimethyl sulfoxide and then with an added 15 ml of 90% acetone at 5 oC

overnight. The samples were measured fluorometrically before and after acidification for

determination of chlorophyll a and phaeopigment concentrations. Fluorescence

measurements were made using a Turner Designs 10-000R fluorometer equipped with a

infrared-sensitive photomultiplier and calibrated using pure chlorophyll a.

2.3 Analysis of macroalgae for C:N:P and 15N – SCUBA divers collected replicate (n =

2) composite samples of macroalgae from the various reef sites into nylon mesh bags. At

least 3-6 separate plants were collected for each composite sample of each species to

ensure representativeness. Immediately following collection, the macroalgae were

cleaned of debris and transferred to plastic zip-loc baggies and held in a cooler during

transport to the lab. In the lab, the samples were identified, rinsed briefly (3-5 s) in

deionized water to further remove debris. The plants were placed in plastic drying dishes

and dried in a lab oven at 65 oC for 48 h. The dried macroalgae were ground to a fine

powder using a mortar and pestle and stored in plastic vials until analysis. Samples of

dried, powdered macroalgae were analyzed for stable nitrogen isotope ratios with a

Carlo-Erba N/A 1500 Elemental Analyzer and a VG Isomass mass spectrometer using

Dumas combustion (by Isotope Services, Los Alamos, New Mexico). The standard used

for stable nitrogen isotope analysis was N2 in air. 15N values, expressed as o/oo (ppm),

were calculated as [(Rsample/Rstandard) – 1] x 103, with R equal to 15N/14N.

Subsamples of the powdered macroalgae samples from the dry season sampling were

analyzed for C:N:P contents at Nutrient Analytical Services, Chesapeake Biological

Laboratory, University of Maryland System, Solomons, MD. Tissue C and N were

measured on an Exeter Analytical, Inc. (EAI) CE-440 Elemental Analyzer, whereas P

was measured following the methodology of Asplia et al. (1976) using a Technicon

Autoanalyzer II with an IBM-compatible, Labtronics, Inc. DP500 software data

collection system (D’Elia et al. 1997).

2.4 Quantification of reef biota - SCUBA divers used an underwater digital video

camcorder (Sony TRV 900 in an Amphibico Navigator housing, 100 degree spherical

lens) to record imagery along two replicate 50 m long belt transects at each reef site.

Divers obtained the imagery by holding the camcorder perpendicular to the reef surface

(0.5 m off bottom, 0.4 m-2 image area) and slowly swimming along the transects to obtain

clear, steady images. A second close-up oblique video transect was also recorded to

allow identification of algal taxa and to distinguish among the fine-scale (1-5 cm)

changes in biotic composition among these diverse and often multi-layered communities.

The digital video images were anaylzed on a high resolution color monitor using the

random point-count method. This method provided a relatively unbiased estimate of the

percent cover of hard corals, macroalgae (> 2 cm tall), algal turf (< 2 cm tall), coralline

algae, sponges, and octocorals (Lapointe and Thacker, 2002). Two independent scorers

used the point-count method (with ten randomized dots superimposed over the video

image) to score ten randomly selected images from each 50 m reef transect (200 point

counts/reef site x 2 scorers = 400 total point counts per reef site).

2.5 Satellite remote sensing – Tobago is not isolated, but connected to remote rivers

including the Orinoco and Amazon (Hu et al., 2004). To help differentiate if the study

sites were under influence of primarily local wastewater/storm water or the Orinoco

River plume, data from the MODIS (MODerate resolution Imaging Spectroradiometer)

instrument onboard the satellite Terra (1999 – present) were analyzed. The Level-1 data

for the year of 2001 were obtained from the NASA GSFC (Goddard Space Flight

Center), and processed using the default algorithms in the software package SeaDAS

(version 5.1). The calibrated radiance data were first corrected for the atmospheric effects

and solar/viewing geometry, then used in a blue/green band-ratio algorithm (O’Reilly et

al., 2000) to estimate chlorophyll-a concentrations (Chl, mg m-3 ) in surface waters. The

underlying assumption was that the blue light was modulated primarily by phytoplankton.

However, because most (>80-90%) of the blue-light absorption was due to colored

dissolved organic matter (CDOM) in the Gulf of Paria and adjacent waters (especially in

the Orinoco River plume, Odriozola et al., 2007), such derived “Chl” was only used as a

color index to show spatial patterns of different water masses, including the Orinoco

River plume. In contrast, the MODIS fluorescence line height (FLH) product, derived

from the red bands using a baseline subtraction algorithm (Letelier et al., 1996), was

nearly immune to CDOM influence (Hu et al., 2005). Hence, FLH was used as a better

index to show spatial patterns of biomass.

3. Results

3.1 Seawater DIN, SRP, and chl a

DIN concentrations measured during this study averaged 1.58 ± 1.08 (n = 107)

overall and varied both temporally (p < 0.0001) and spatially (p = 0.005) among the sites

(Fig. 3). The mean DIN concentration of all sites was lower (p < 0.0001) in the dry

season (0.66 ± 0.47 µM; n = 51) than the wet season (2.42 ± 0.73 µM; n = 56). The

seasonal increase in DIN was mostly the result of increased (p < 0.0001) ammonium

from the dry season (0.35 ± 0.20 µM; n = 51) to wet season (1.55 ± 0.57 µM; n = 56; Fig.

3). Nitrate also increased (p < 0.0001) from the dry season (0.31 ± 0.42 µM; n =51) to

the wet season (0.86 ± 0.51 µM; n = 56) but was a relatively minor component of the

DIN pool compared to ammonium in both seasons as indicated by consistently low f-

ratios (0.32 – 0.35). The mean DIN of the BRC sites was higher (p = 0.012) than the

fringing reef sites in the dry season (0.86 ± 0.51 µM vs. 0.52 ± 0.39 µM) but not the wet

season (2.35 vs. 2.53 µM). The lowest mean concentrations of both DIN (1.04 ± 1.09)

and ammonium (0.66 ± 0.70) occurred at Black Jack Hole whereas the highest DIN (2.42

± 0.73 µM) and ammonium (1.95 ± 0.88 µM) occurred at Buccoo Point in the inner BRC.

SRP concentrations during the study averaged 0.23 ± 0.15 (n = 107) overall and

varied temporally (p = 0.034) and spatially (p < 0.0001; Fig. 4). The mean SRP

concentration of all sites was higher (p < 0.011) in the dry season (0.26 ± 012 µM; n =

51) than the wet season (0.20 ± 0.17 µM; n =56). Mean SRP concentrations of the BRC

sites were not significantly different than the fringing reefs sites either overall or in the

dry or wet seasons. The lowest SRP concentrations occurred at Black Jack Hole (0.18 ±

0.09 µM) and the highest at Buccoo Point (0.50 ± 0.61 µM), which was higher (p <

0.014) than all other sites.

The DIN:SRP ratios averaged 8.7 ± 6.5 (n = 51) during the study and varied both

temporally and spatially (Fig. 4). Overall, the mean DIN:SRP ratio was higher (p <

0.0001) in the wet season (13.9 ± 4.2; n = 56) than the dry season (2.9 ± 2.6; n = 51).

The DIN:SRP ratio was higher (p < 0.0001) in the BRC (4.0 ± 2.9; n = 20) than the

fringing reefs (2.3 ± 2.6; n = 31) in the dry season. However, in the wet season, the

DIN:SRP ratio was higher (p = 0.008) on the fringing reefs (14.6 ± 4.1; n = 44) than the

BRC (11.9 ± 3.7; n =26).

Chl a averaged 0.32 ± 0.19 µg/l at all stations throughout the study and varied

significantly temporally (p = 0.038) and spatially (p = 0.031; Fig. 5). Overall, chl a

averaged 0.28 ± 0.02 µg/l (n = 51) in the dry season and 0.35 ± 0.17 µg/l (n = 70) in the

wet season. The lowest values in the dry season were at Coral Gardens (0.085 ± 0.02

µg/l) and the highest in the wet season were in the Bon Accord Lagoon (1.13 ± 0.16

µg/l). In the dry season, the mean chl a values were similar for the Buccoo Reef Complex

(0.23 ± 0.22 µg/l) and the Fringing Reefs (0.31 ±0.22 µg/l) whereas, in the wet season,

the mean chl a of the BRC (0.41 ± 0.24 µg/l) was greater (p = 0.031) than that of the FR

(0.32 ±0.10).

3.2 C:N:P and 15N in macroalgae

Macroalgae in the Buccoo Reef Complex were enriched in N but not P compared

to the fringing reef sites during the dry season (Table 1). The mean C:N (15.80 ± 5.78; n

= 16) for macroalgae in the BRC was lower (p = 0.014) than the Fringing Reefs (21.0 ±

6.58; n = 24) in contrast to the mean C:P values that were statistically similar (770 ± 294

vs. 759 ± 385). These changes co-occurred with a higher (p = 0.002) N:P ratio in the

BRC (49.5 ± 15.6; n = 16) compared to the Fringing Reefs (34.6 ± 11.5; n = 24). The

lowest C:N ratio (8.0) occurred at Buccoo Point in the BRC and the highest (25.7) at

Black Jack Hole off Little Tobago Island. The lowest C:P (444) occurred at Buccoo

Point and the highest (1238) at Coral Gardens in the BRC. The lowest N:P (27.5)

occurred at Kelliston Drain off Little Tobago Island and the highest (61.6) at Coral

Gardens in the Buccoo Reef Complex.

The 15N values of macroalgae averaged + 5.85 ± 1.23 o/oo (n = 174) throughout

the study and varied temporally, spatially, and by taxa (Fig. 6). The mean 15N value in

the dry season (+ 5.52 ± 1.24 o/oo; n=82) was lower (p = 0.01) than the wet season (+

6.14 ± 1.16 o/oo; n = 92). Overall, the mean 15N value of the fringing reef macroalgae

(+5.54 ± 1.29 o/oo; n=94) was lower (p < 0.0001) than that of the Buccoo Reef Complex

(+6.21 ± 1.06 o/oo; n = 80). The lowest 15N values in reef macroalgae occurred at

Black Jack Hole (+ 2.87 o/oo) and the highest at Princess Reef (+7.25 o/oo) in the inner

BRC. However, the highest 15N values (~ + 12 o/oo) of macroalgae were in nutrient

indicator chlorophytes (Ulva lactuca) and rhodophytes (Gracilaria tikvahiae, Agardhiella

subulata) collected adjacent to a wastewater outfall in Buccoo Bay (Fig. 6; Table 1).

When grouped by class, the 15N values varied significantly (p < 0.001) with the highest

values in rhodophytes (+7.0 ± 2.32 o/oo; n = 30), followed by chlorophytes (+ 6.23 ±

1.33 o/oo; n = 82) with the lowest in phaeophytes (+ 5.47 ± 1.23 o/oo; n = 54).

3.3 Quantification of reef biota

The reef sites in the inner BRC showed a distinct pattern of increased cover of

macroalgae and reduced coral cover (Fig. 7). The highest % cover of macroalgae (up to

40 %) occurred at Princess Reef and Buccoo Point where blooms of Halimeda opuntia,

Bryopsis spp., and Caulerpa spp. were observed (Fig 8 a,b,c). Coral cover was < 10 %

on these two sites compared to the more offshore sites (Coral Gardens, Outer Reef)

where coral cover ranging from 20 to 40 % was comprised of the genera Acropora,

Montastrea, Diploria, and Siderastrea (Figure 7). Algal turf was a major component of

these reef communities in the Buccoo Reef Complex with values ranging from up to ~ 50

% cover. The zoanthid Palythoa was another significant component of the BRC reef

communities and ranged up to > 20 % cover at Buccoo Point where it was observed

physically overgrowing the brain coral Diploria (Fig. 8d). The fringing reef sites had the

highest % hard coral cover (Culloden Reef, > 45 % cover), and generally lower frondose

macroalgae (< 10 % cover) but high cover of coralline algae (up to ~ 40 % cover; Figs. 7,

8d).

4. Discussion

Our water column nutrient data from Tobago are the first to document the role of

land-based nitrogen runoff in fueling varying degrees of eutrophication on the island’s

fringing coral reefs. Previous studies in Trinidad & Tobago have emphasized the role of

land-based phosphorus pollution from industrial and domestic wastewaters (Kumarsingh

et al. 1998a,b). However, the relatively high background SRP concentrations throughout

the entire study area (~ 0.23 M) are greater than reported for many coastal waters in the

Caribbean (Lapointe, 2004) and, combined with relatively low DIN:SRP ratios (<15:1),

suggest N rather than P limitation of algal growth (Lapointe et al., 1992). Such high

background SRP concentrations, low DIN:SRP ratios, and a tendency towards N-

limitation are indicative of a siliciclastic geology. This contrasts with the carbonate-rich

environments in Jamaica, the Bahamas, and the Florida Keys where low SRP

concentrations and high DIN:SRP ratios supports strong P-limitation (Lapointe et al.

1992). Hence, the natural volcanic geology of Tobago may predispose its coastal waters

for nitrogen rather than phosphorus limitation.

The differential effects of carbonate-rich vs. siliciclastic geology on N:P

availability in Tobago is apparent in both the Buccoo Reef Complex and the Fringing

Reef data in our study. Although the extensive limestone formation surrounding the

Buccoo Reef provides a sink for natural and anthropogenic SRP through adsorption onto

calcium carbonate surfaces (Berner, 1974; DeKanel and Morse, 1978), such is not the

case for DIN (Lapointe et al. 1990). This phenomenon explains how relatively high N:P

inputs from the Buccoo Reef watershed results in higher water column DIN:SRP and

macroalgal N:P ratios in the Buccoo Reef Complex compared to the Fringing Reef sites.

The elevated DIN:SRP ratios in the BRC were most pronounced during the wet season,

when wastewater loads from “soak-aways” and other sources in the watershed would be

maximal. Such high N:P inputs lead to P-limitation of primary production in carbonate-

rich coastal waters (Lapointe, 1987) and explains why many Caribbean coastal waters are

limited primarily by P rather than N (Lapointe et al. 1992; Corredor et al., 1999). The

reduced “limestone” effects at the Fringing Reef sites outside of the Buccoo Reef

Complex results in lower DIN:SRP and N:P ratios, where N-limitation would be more

important.

Multiple lines of evidence from our study indicate that the reef at Buccoo Bay in

the Buccoo Reef Complex was the most nutrient impacted reef site in our study. The

highest DIN, ammonium, and SRP concentrations of the entire study occurred at Buccoo

Point, a reef directly impacted by wastewater discharges from the adjacent upland

watershed. The lowest C:N and C:P ratios of all macroalgae in the study also occurred at

Buccoo Point, demonstrating impacts of these elevated water column DIN and SRP

concentrations on reef biota. These observations are consistent with previous research

using cylindrical cores of the coral Montastrea annularis in the BRC vs. Culloden reef to

compare historical trends in P availability (Kumarsingh et al. 1998b). These researchers

found higher fractions of organic P in the coral cores in the BRC than Culloden reef, and

that temporal patterns in both total and organic P in the BRC coral cores correlated with

human activities on the watershed. Our results also support previous findings by the IMA

(1996) that the highest nutrient concentrations in southwest Tobago occurred in Buccoo

Bay and the Bon Accord Lagoon, both of which have received increased wastewater

nutrient loadings in recent decades.

The effects of high nutrient concentrations with high DIN:SRP ratios in the BRC

were evident in the increased biomass and C:N:P contents of macroalgae in the BRC

compared to the FR. Macroalgae in the BRC were enriched in N compared to the FR

sites, resulting in a lower C:N (15 vs. 21) and higher N:P (50 vs. 35) ratio. The lower

C:N ratio reflects reduced N-limitation of growth, and explains the overall higher

macroalgal biomass in the BRC compared to the FR sites. The relatively high N

enrichment in the BRC drives the macroalgae to increased P-limitation, evidenced by the

higher N:P ratio of macroalgae in the BRC compared to FR. Blooms of the chlorophytes

Bryopsis, Caulerpa, and Halimeda were observed overgrowing reef biota and

frameworks in the BRC, a phenomenon that was not observed at any of the FR sites. The

phenomenon by which high N:P inputs from the watershed drive macroalgal blooms on

coral reefs towards increased P-limitation has been reported for other carbonate-rich

waters in the Caribbean, including the Florida Keys (Lapointe, 1987; Lapointe and Clark,

1992; Lapointe et al. 2004), southeast Florida (Lapointe et al. 2005a), Jamaica (Lapointe,

1997; Lapointe and Thacker), Martinique (Littler et al. 1992), and Bermuda (Lapointe

and O’Connell, 1989).

The identification of wastewater as the source of elevated DIN in the BRC was

apparent from the elevated 15N values of macroalgae in the BRC compared to the FR

sites. Numerous studies have shown how 15N values in macroalgae faithfully

discriminate among various natural and anthropogenic nitrogen sources (see Risk et al.

2009 for a review). For example, France et al. (1998) showed that 21 species of tropical

macroalgae in coastal waters of southwestern Puerto Rico had a mean 15N of + 0.3 ±

1.0 o/oo, a value close to the atmospheric signature of 0 o/oo and indicative of nitrogen

fixation. In contrast, macroalgae exposed to varying degrees of urban wastewater

pollution in Negril, Jamaica (Lapointe and Thacker, 2002), southeast Florida (Lapointe et

al. 2005b), southwest Florida (Lapointe and Bedford, 2007), Florida Keys (Lapointe et al.

2004), and Moreton Bay, Australia (Costanzo et al. 2001) all reported 15N values

ranging from ~ + 3.0 – + 15 o/oo, with higher values > 8 o/oo in close proximity to the

wastewater source. This pattern was consistent with our results where the highest 15N

values of ~ + 12 o/oo were in “nutrient indicator species” (Ulva lactuca, Gracilaria

tikvahiae, Agardhiella subulata) near a wastewater outfall in Buccoo Bay; the lower

mean value of + 7.25 o/oo for Princess Reef in inner Buccoo Bay and + 6.21 o/oo for

macroalgae throughout the BRC would be expected from increasing dilution of land-

based wastewater N throughout the BRC. These 15N data support the hypothesis that

the entirety of the BRC has been significantly impacted by DIN derived from

wastewaters via submarine groundwater discharge of leachate from “soak-aways” or

direct discharges from sewage outfalls. Lower, but relatively elevated 15N values

occurred in macroalgae at other reef sites in southwest Tobago, especially off the Hilton

Hotel downstream of Scarborough, indicating widespread sewage DIN enrichment

downstream of urban areas. The lowest 15N values of + 2.5 o/oo for macroalgae at

Black Jack Hole off Little Tobago Island in the dry season were indicative of more

oligotrophic and “natural” nitrogen sources, although these values increased significantly

during the wet season as a result of more widespread dispersion of land-based sewage

pollution via stormwater runoff.

Elevated chl a concentrations during the wet season provides further evidence of

watershed-driven eutrophication in the BRC. In both seasons, maximum chl a correlated

with decreased DIN concentrations, but not SRP. This pattern results from rapid uptake

of DIN -- the primary limiting nutrient-- by phytoplankton in coastal waters receiving

land-based sewage pollution. These findings support previous observations in Kaneohe

Bay, Hawaii (Laws and Redalje 1979) and the Florida Keys (Lapointe and Clark 1992)

that chl-a provides a good “integrated” index of eutrophication in tropical and subtropical

coastal waters. The generally low DIN concentrations on Tobago’s fringing reefs (< 0.3

M), especially off Little Tobago Island, not only limits phytoplankton biomass, but also

the growth of macroalgae. Mean DIN concentrations > 0.5 M are needed to support

“blooms” of benthic macroalgae, such as those we observed on reef sites in the inner

BRC. Classic “indicator species” of nutrient enrichment, including the macroalgae Ulva,

Halimeda, Bryopsis, Caulerpa, and Sargassum were abundant on reefs in the BRC that

had DIN concentrations > 0.5 M. In contrast, macroalgae were less abundant on the

fringing reefs around Tobago that had lower DIN concentrations.

The MODIS image series added additional evidence to differentiate the origin of

elevated nutrients and biomass. Fig. X shows several examples of the MODIS “Chl” and

FLH images during both the dry and wet seasons. While the “Chl” images clearly showed

the influence of the Orinoco River plume on the near-shore waters of Tobago (the green-

yellow-red colors) during the wet season, the corresponding FLH images suggested that

this influence was primarily through CDOM as opposed to phytoplankton biomass. The

absence of elevated biomass in the Orinoco River plume near Tobago suggested that

riverine nutrients might be depleted when the plume reached Tobago. Hence, the elevated

chl a and nutrient concentrations from our field sampling during the wet season were

primarily of local origin rather than from the Orinoco River plume.

The loss of hard coral cover and replacement by macroalgae on Jamaican coral reefs

over the past two decades has been repeatedly attributed to the “top-down” effects of

overfishing of herbivorous fish stocks and die-off of the long-spined urchin Diadema

(Hughes, 1994; Jackson et al., 2001). However, similar loss of hard corals and

macroalgal blooms have occurred over the past two decades in the BRC, a MPA with

abundant parrotfish (including large adults), surgeonfish, and chubs. However, as

described here, the BRC has also been polluted by land-based sewage inputs. The most

sewage impacted inshore reef in the BRC (Princess Reef) not only had abundant

macroalgae (Caulerpa, Halimeda) but also dense populations of reef urchins

(Echinometra viridis). These observations point to “bottom-up” control (i.e. nutrients) of

not only the macroalgal biomass, but the echinoid populations as well. Hunter and Price

(1992) suggested that a “bottom-up template” is the most realistic approach for

understanding trophic dynamics in ecosystems because plants have obvious primacy in

food webs. We concur and suggest that sewage nitrogen subsidies over recent decades

may have increased grazer populations (both echinoids and herbivorous fishes) in the

BRC where growth is typically limited by dietary nitrogen (Lapointe, 1999). Reef health,

biodiversity and the ecological services have been reduced by sewage-driven

eutrophication in the BRC despite abundant populations of grazers.

In summary, our results support the conclusion of Siung-Chang (1997) for the

Caribbean region and Agard and Gobin (2000) for Trinidad & Tobago that the most

important source of marine pollution is domestic sewage, which appears to be impacting

not only the BRC but also other fringing coral reefs “downstream” of Tobago’s populated

southwest coast.

Acknowledgements – We gratefully acknowledge Mr. Ray DeBeer and the World of

Watersports for providing dive boat support that made this research possible. This

research was supported by the Buccoo Reef Trust and the Harbor Branch Oceanographic

Institution, Inc. This is contribution number from the Harbor Branch Oceanographic

Institute at Florida Atlantic University, Ft. Pierce, FL.

5. References

Agard, J. B. R. and J. F. Gobin. 2000. The Lesser Antilles, Trinidad and Tobago. pp. 627-

641, In: Seas at The Milleneum: An Environmental Evaluation. C. Sheppard (ed.).

Asplia, I., Agemian, H., Chau, A.S.Y., 1976. A semi-automated method for the

determination of inorganic, organic, and total phosphate in sediments. Analyst

101, 187-197.

Barile, P.J., Lapointe, B.E, 2005. Atmospheric nitrogen deposition from a remote source

enriches macroalgae in a coral reef ecosystem at Green Turtle Cay, Abacos,

Bahamas. Mar. Poll. Bull. 50, 1262-1272.

Costa, O. S., Z. M. A. N. Leao, M. Nimmo, and M. J. Attrill. 2000. Nutrification impacts

on coral reefs from northern Bahia, Brazil. Hydrobiologia 440:307-315.

D’Elia, C.F., Connor, E.E., Kaumeyer, N.L., Keefe, C.W., Wood, K.V., Zimmerman,

C.F., 1997. Nutrient analytical services: standard operating procedures., Technical

Report Series No. 158-97. Chesapeake Biological Laboratory, Solomons, MD.

France, R., J. Holmquist, M. Chandler and A. Cattaneo. 1998. 15 N evidence for

nitrogen fixation associated with macroalgae from a seagrass-mangrove-coral reef

ecosystem. Marine Ecology Progress Series. 167: 297-299.

Hu, C., E. Montgomery, R. Schmitt, and F. E. Muller-Karger (2004). The Amazon and

Orinoco River plumes in the tropical Atlantic: Observation from space and S-

Floats. Deep Sea Res-II. 51:1151-1171.

Hu, C., F. E. Muller-Karger, C. Taylor, K. L. Carder, C. Kelble, E. Johns, and C. Heil

(2005). Red tide detection and tracing using MODIS fluorescence data: A

regional example in SW Florida coastal waters. Remote Sens. Environ., 97:311-

321.

Hughes, T. 1994. Catastrophes, phase-shifts, and large-scale degradation of a Caribbean

coral reef. Science 265: 1547-1551.

Hunter, M. D. and P. W. Price. 1992. Playing chutes and ladders: heterogeneity and the

relative roles of bottom-up and top-down forces in natural communities. Ecology

73: 724.

IMA, 1994. The formulation of a management plan for the Buccoo Reef Marine Park.

Volume 3. Biological investigations and water quality monitoring. Document

prepared for the Tobago House of Assembly. Institute of Marine Affairs,

Trinidad.

IMA, 1996. Report of phase I: an assessment of the biophysical environment of

southwest Tobago. Document prepared for the Organization of American States.

Technical Report. Institute of Marine Affairs, Trinidad.

John, G. M. 1996. Quantitative characteristics of land-based sources of marine pollution

in the vicinity of the Buccoo Reef Marine Park, Tobago. MS Thesis, University of

the West Indies, Department of Biology, Cave Hill Campus, Barbados.

Kenny, J.S. 1976. A preliminary study of the Buccoo Reef/ Bon Acord Complex with

special reference to development and Management. Department of Biological

Sciences, University of the West Indies. St. Augustine, 123pp.

Kumarsingh, K., L. A. Hall, A. M. Siunng-Chang, and V. A. Stoute. 1998a. Phosphorus

in sediments of a shallow bank influenced by sewage and sugar factory effluents

in Trinidad, West Indies. Marine Pollution Bulletin 36: 185-192.

Kumarsingh, K., R. Laydoo, J. K. Chen, and A. M. Siunng-Chang. 1998b. Historic

records of phosphorus levels in the reef-building coral Montastrea annularis from

Tobago, West Indies. Marine Pollution Bulletin 36: 1012-1018.

Lapointe, B.E., J.D. O’Connell and G.S. Garrett. 1990. Nutrient couplings between on-

site sewage disposal systems, groundwaters, and nearshore surface waters of the

Florida Keys. Biogeochemistry 10: 289-308.

Lapointe, B. E., M. M. Littler, and D. S. Littler. 1992. Nutrient availability to marine

macroalgae in siliciclastic versus carbonate-rich coastal waters. Estuaries 15:75-

82.

Lapointe, B.E. 1997. Nutrient thresholds for bottom-up control of macroalgal blooms on

coral reefs in Jamaica and southeast Florida. Limnology and Oceanography 42:

1119-1131.

Lapointe, B.E. 1999. Simultaneous top-down and bottom-up forces control macroalgal

bloom on coral reefs. Limnology and Oceanography 44: 1586-1592

Lapointe, B.E. and K. Thacker. 2002 Community-based water quality and coral reef

monitoring in the Negril Marine Park, Jamaica: Land-based nutrient inputs and

their ecological consequences. pp. 939-963, In The Everglades, Florida Bay, and

Coral Reefs of the Florida Keys (J.W. Porter and K.G. Porter, eds.) CRC Press,

Boca Raton, FL.

Laydoo, R.S. and L. Heileman. 1987. Environmental impacts of the Buccoo and Bon

Accord Sewage Treatment plants, southwestern Tobago. A preliminary report.

Institute of Marine Affaires and Crusoe Reef Society, 27pp.

Laydoo, 1991. A guide to the coral reefs of Tobago. Institute of Marine Affairs and the

Asa Wright Nature Centre, Republic of Trinidad and Tobago, 43 pp.

Letelier, R. M., and Abott, M. R. (1996). An analysis of chlorophyll fluorescence

algorithms for the Moderate Resolution Imaging Spectrometer (MODIS). Remote

Sens. Environ. 58:215– 223.

Lindau, C. W., R. D. Delaune, W. H. Patrick, Jr., and E. N. Lambremont. 1989.

Assessment of stable nitrogen isotopes in fingerprinting surface water inorganic

nitrogen sources. Water Air Soil Pollut. 48:489-496.

Likens, G. E. 2001. Biogeochemistry, the watershed approach: some uses and

limitations. Marine Freshwater Research 52: 5.

McClelland, J. and I. Valiela. 1998. Changes in food web structure under the influence

of increased anthropogenic nitrogen inputs to estuaries. Marine Ecology Progress

Series 168:259-271.

NRC (National Research Council), 2000. Clean coastal waters: understanding and

reducing the effects of nutrient pollution. National Academy Press, Washington

DC.

Odriozola, A. L., R. Varela, C. Hu, Y. Astor, L. Lorenzoni, and F. E. Muller-Karger

(2007). On the absorption of light in the Orinoco River plume. Cont. Shelf Res.

27:1447-1464.

O'Reilly, J. E., Maritorena, S., O'Brien, M. C., Siegel, D. A., Toole, D., Menzies, D., et

al. (2000). SeaWiFS postlaunch calibration and validation analyses: Part 3. SeaWiFS

postlaunch technical report series, 11. In S. B. Hooker, & E. R. Firestone (Eds.),

NASATechnical Memorandum 2000-206892 49 pp.

Siung-Chang, Avril. 1997. A review of marine pollution issues in the Caribbean.

Environmental geochemistry and Health 19:45-55.

UNEP, 1994. Regional overview of land based sources of pollution in the wider

Caribbean region, in Caribbean Environment Program Technical Report # 33,

UNEP Caribbean Environment Program, Kingston.

Figure 1. Map of Tobago showing locations of monitoring sites in the Buccoo Reef

Complex and fringing reefs.

Figure 2. Cumulative monthly rainfall measured at Crown Point, Tobago, during 2001.

Figure 3. Concentrations of ammonium, nitrate, DIN, and the f-Ratio for monitoring sites in the Buccoo Reef Complex (left side) and fringing reefs (right side) in the dry (April – June) and wet (September – October) seasons, 2001. Values represent means ± 1 SD (n = 4).

Figure 4. Concentrations of soluble reactive phophorus (SRP and the DIN:SRP ratio for monitoring sites in the Buccoo Reef Complex (left side) and fringing reefs (right side) in the dry (April – June) and wet (September – October) seasons, 2001. Values represent means ± 1 SD (n = 4).

1.13

0.0

0.2

0.4

0.6

Bon

Acc

ord

Lago

onPr

ince

ss R

eef

Bucc

oo P

oint

Nyl

on P

ool

Cor

al G

arde

nsBu

ccoo

Ree

fH

ilton

Ree

fD

iver

s Thi

rst

Div

ers D

ream

Mt.

Irvi

ne W

all

Mav

eric

kA

rnos

Val

eC

ullo

den

Ree

fEn

glish

man

's Ba

yTh

ree

Sist

ers

Kel

lest

on D

rain

Blac

k Ja

ck H

ole

Chl a

Apr-Jun Sep-Oct

Figure 5. Concentrations of chlorophyll a for monitoring sites in the Buccoo Reef Complex (left side) and fringing reefs (right side) in the dry (April – June) and wet (September – October) seasons, 2001. Values represent means ± 1 SD (n = 4).

Figure 6. 15N values of macroalgae at monitoring sites in the Buccoo Reef Complex (left side) and fringing reefs (right side) in the dry (April) and wet (September – October) seasons, 2001. Values represent means ± 1 SD (n = 4).

Figure 7. Percent cover of reef biota at reef sites in the Buccoo Reef Complex (Princess Reef, Buccoo Point, Coral Gardens, Buccoo Reef) and fringing reefs (Arnos vale, Culloden, Englishman’s Bay, and Three Sister’s). Values represent means ± 1 SD (n=4).

Figure 8. Photos of reef biota (a,b,c,d).

Fig. X. MODIS/Terra imagery showing that near-shore waters of Tobago are under influence of the Orinoco River plume during the wet season (August – October), but most of the influence is through the riverine colored dissolved organic matter (CDOM) rather than phytoplankton pigments (chlorophyll). The left panels show the MODIS “chlorophyll-a” concentrations (Chl, mg m-3 ), derived from the color ratio between the

MODIS/Terra Chl MODIS/Terra FLH

Apr 28, 2001

Aug 27, 2001

Nov 2, 2001

Apr 28, 2001

Aug 27, 2001

Nov 2, 2001

10oN

11oN

62oW 61oW

TOBAGOTRINIDAD

blue (443 or 488 nm) and green (551 nm) bands. Because both chlorophyll-a and CDOM strongly absorb the blue light, the Chl values actually represent a combined effect of the two. The right panels show the MODIS chlorophyll fluorescence line height (FLH, mW cm-2 m-1 sr-1 ), which is much less affected by riverine CDOM. Despite the significant noise (striping, cloud-edge effect, sun glint effect), the spatial patterns of FLH and their contrast to the left panels suggest that most of the river plume signal near Tobago are not due to chlorophyll-a, but rather due to riverine CDOM.

Table 1. Tissue C:N:P contents of macroalgae from reef sites in the Buccoo Reef Complex (BRC) and fringing reefs around Tobago.

LANGTON paragraphs for discussion??

Macroalgal overgrowth on coral reefs clearly reflects a well documented phase shift resulting in the demise of the reef itself (references??), which raises the question as to the

cause and potential controls behind such shifts. In the current study we found, in general terms, a nutrient gradient from inshore to offshore and/or from the least populated end of the island, in the northeast, to the most highly populated southwest Tobago. It is clear that human activity is impacting the near shore coral reefs and land-based nutrients are fueling a shift from coral to macroalgae. The solution is to remove the nutrients prior to wastewater discharge into the sea, and although that is technologically possible and under considered, the more urgent question is whether or not there is an ecological mechanism to control the proliferation of macroalgae while the existing methods of waste water treatment are our only option.

In the case of Tobago, as virtually all of the Caribbean, the major invertebrate macroalgal predator, Diadema antillarum was lost several decades ago and continues to be slow to recover (Frey, 2004). Vertebrate herbivores include a number of fish species and variety of feeding guilds, but studies on fish feeding (eg. McClanahan et al., 2000; Ledlie et al 2007) have documented the avoidance of macroalgae as a food choice by many of the more common herbivorous reef fish. This begs the question as to how to slow or reverse a phase shift to favor coral recovery. Mumby et al. (2006) found a negative correlation between grazing by parrot fish and macroalgal cover, but this is somewhat counterintuitive considering that macroaglae is not a preferred prey item by most herbivorous fish (Ledlie et al 2007). Presumably, what is happening is that as the intensity of grazing by parrot fish on epilithic algae within the confines of a Marine Protected Area increases, due to a complex predator/prey relationship, this substrate disturbance reduces the settlement and growth of macroalgae, as has been reported by Lewis, 1986; Williams et al. 2001 and Paddack et al. 2006. Although fish are often opportunistic feeders, and there is a well documented case, of what was termed the identification of a “sleeping functional group” (Bellwood et al. 2006) where an invertebrate and plankton feeding batfish shifted its diet to consume macroalgae, there is little evidence to suggest that this mechanism offers a universal solution to the problem of nutrient overload on coral reefs. Despite a relatively healthy and abundant herbivorous fish population, as found in the Buccoo Reef complex (Frey, 2004), we still find the proliferation of macroalgae reflecting an overload of land-based nutrients and no really viable mechanism for reversing this trend.