**6. Discussion**

#### **6.1 Circum-reef circulation defining the reefal bay**

Hydrodynamic modelling showed that circulation around the Wreck Bay and Sand Hills Bay reef parabola continuously looped the reef as circum-reef circulation (CRC). The CRC was considered "closed" when fore-reef currents fed water back into the back-reef and "open" when main fore-reef flow continued along-shore (Figure 9). Channel surge currents were responsible for the propagation of inner bay waters seawards, and encouraged open CRC. Tracking models revealed the longevity and spatial spread of this flow, simulating the patterns first observed in these bays by field drogues and fixed measurements that depicted continuous current flow around reef arms at surface and depth (Maxam & Webber, 2010). The presence of the reef induced this persistence and localized the (CRC). The lack of reefs in Engine head bay supported this premise as gyre formation and localization was not evident in the non-reefal bay. This was confirmation that open bays did not facilitate recycling of their inside waters from the outside as reefal bays do. In the absence of prominent reef arms, the CRC cannot exist.

#### **6.2 Reef arm crc dominance and cycling**

Simulations of new particles introduced into the bay on an hourly basis revealed that under particular tide and wind regimes, one reef's circulation was strengthened while the other abated in the same bay (Figures 10, 11). This simulated the dynamics that prevented field drogues from entering the weaker reef gyre while trapped in the dominant one (Maxam & Webber, 2010). The dominant gyre was responsible for the greatest extensions of the bay system, and so the presence of two prominent reef arms resulted in regular switching of dominance.

Full development of both reef arm gyres occurred in one tidal cycle. The reef gyre downstream the main long-shore flow appeared strengthened on the rising tide while the adjacent reef up-shore was strengthened by the falling tide. It is important to note that these simulations accurately portray the importance of the tidal influence in a micro-tidal environment where it was otherwise expected to be overwhelmed by wind- and waveinduced stresses. In the absence of large tidal fluctuations, as during a neap tide, the upshore gyre was too weak to be developed and the down-shore gyre dominated. Up-shore reef arms were more reliant on tidal changes to effect gyre formation than down-shore reefs. The sea-breeze aided in strengthening both gyres during simulation, agreeing with longterm field data that showed this correlation (Maxam & Webber, 2010). This wind regime

western edge of the east gyre. When the west reef circulation emanated, gyre extensions were smaller and did not exceed 75 %. West reef gyres were most developed at low-to-rising tide during land-breeze emanation and coincided with the lowest current component speeds recorded in the channel at that time. The longest duration of this closed western gyre was

Sand Hills Bay had its largest extension at 198 % during the combination of a rising tide and when the sea-breeze emanated. This was due to the south reef gyre that also tended to be more closed than the north reef's. The north reef gyre was most developed at the rising-tohigh tide (also when the sea-breeze emanated) and had its largest extension at 154 %. In the absence of large tidal changes and developed wind regimes, the south gyre dominated the

Hydrodynamic modelling showed that circulation around the Wreck Bay and Sand Hills Bay reef parabola continuously looped the reef as circum-reef circulation (CRC). The CRC was considered "closed" when fore-reef currents fed water back into the back-reef and "open" when main fore-reef flow continued along-shore (Figure 9). Channel surge currents were responsible for the propagation of inner bay waters seawards, and encouraged open CRC. Tracking models revealed the longevity and spatial spread of this flow, simulating the patterns first observed in these bays by field drogues and fixed measurements that depicted continuous current flow around reef arms at surface and depth (Maxam & Webber, 2010). The presence of the reef induced this persistence and localized the (CRC). The lack of reefs in Engine head bay supported this premise as gyre formation and localization was not evident in the non-reefal bay. This was confirmation that open bays did not facilitate recycling of their inside waters from the outside as reefal bays do. In the absence of

Simulations of new particles introduced into the bay on an hourly basis revealed that under particular tide and wind regimes, one reef's circulation was strengthened while the other abated in the same bay (Figures 10, 11). This simulated the dynamics that prevented field drogues from entering the weaker reef gyre while trapped in the dominant one (Maxam & Webber, 2010). The dominant gyre was responsible for the greatest extensions of the bay system, and so the presence of two prominent reef arms resulted in regular switching of

Full development of both reef arm gyres occurred in one tidal cycle. The reef gyre downstream the main long-shore flow appeared strengthened on the rising tide while the adjacent reef up-shore was strengthened by the falling tide. It is important to note that these simulations accurately portray the importance of the tidal influence in a micro-tidal environment where it was otherwise expected to be overwhelmed by wind- and waveinduced stresses. In the absence of large tidal fluctuations, as during a neap tide, the upshore gyre was too weak to be developed and the down-shore gyre dominated. Up-shore reef arms were more reliant on tidal changes to effect gyre formation than down-shore reefs. The sea-breeze aided in strengthening both gyres during simulation, agreeing with longterm field data that showed this correlation (Maxam & Webber, 2010). This wind regime

observed during 15 hrs of some of the smallest tidal changes recorded.

**6.1 Circum-reef circulation defining the reefal bay** 

prominent reef arms, the CRC cannot exist.

**6.2 Reef arm crc dominance and cycling** 

extension.

dominance.

**6. Discussion** 

Fig. 9. Diagrams depicting closed and open circum-reef circulation (CRC) simulated from RMATRK discrete particle tracking modelling. The closed CRC displaying recirculation were evident for Wreck Bay west reef (A) and east reef (B) arms, as well as Sand Hills Bay south reef (C) and north reef (D) arms particularly during wind calms. Open CRC is also displayed in Wreck Bay west reef (E) and east reef (F) arms, and again around Sand Hills Bay south reef (G) and north reef (H) arms particularly during increased channel flow.

The Hydrodynamic Modelling of Reefal Bays –

Placing Coral Reefs at the Center of Bay Circulation 171

Fig. 11. Dominant west reef CRC in Wreck Bay is shown here typically occurring during neap periods when bay extension was due primarily to wind and over-the-reef forcing. CRC

is displayed as circled area in model particle tracks (A) and model vectors (B). CRC formation on the opposing reef arm is weakened during dominance of the other.

induced more flow over the reef due to increased heights of waves impinging on the reef and at higher frequencies (Roberts et al., 1992). Breaking would occur and the rapid energy transferred caused an increase in water level, driving strong back-reef surge currents and increasing current speeds in the northern part of the gyre. These surges, however, reduced the retention times of these gyres.

This cycle of emanation and contraction is characteristic of the reefal bay system, giving the reefal bay a spatial pulse that is dependent on prevailing wind and tidal regimes. The reefal bay does not have a static bay area but instead will be at a minimum when the CRC is most contracted and at a maximum when the CRC is most extended. At its minimal spatial extent, the horizontal area of the hydrodynamic reefal bay is dependent on the size of the reef. The larger reef in Wreck Bay, the east reef arm, gave the lager dominant gyre resulting in the greater seaward extensions of the bay. The same was observed in Sand Hills Bay where the south reef was the larger reef and therefore gave the greater extensions (Figure 12).

Fig. 10. Dominant east reef CRC in Wreck Bay due to large falling tide range is displayed in A and B as circled area in model particle tracks (A) and model vectors (B). CRC formation on the opposing reef arm is weakened during dominance of the other.

induced more flow over the reef due to increased heights of waves impinging on the reef and at higher frequencies (Roberts et al., 1992). Breaking would occur and the rapid energy transferred caused an increase in water level, driving strong back-reef surge currents and increasing current speeds in the northern part of the gyre. These surges, however, reduced

This cycle of emanation and contraction is characteristic of the reefal bay system, giving the reefal bay a spatial pulse that is dependent on prevailing wind and tidal regimes. The reefal bay does not have a static bay area but instead will be at a minimum when the CRC is most contracted and at a maximum when the CRC is most extended. At its minimal spatial extent, the horizontal area of the hydrodynamic reefal bay is dependent on the size of the reef. The larger reef in Wreck Bay, the east reef arm, gave the lager dominant gyre resulting in the greater seaward extensions of the bay. The same was observed in Sand Hills Bay where the

south reef was the larger reef and therefore gave the greater extensions (Figure 12).

Fig. 10. Dominant east reef CRC in Wreck Bay due to large falling tide range is displayed in A and B as circled area in model particle tracks (A) and model vectors (B). CRC formation

on the opposing reef arm is weakened during dominance of the other.

the retention times of these gyres.

Fig. 11. Dominant west reef CRC in Wreck Bay is shown here typically occurring during neap periods when bay extension was due primarily to wind and over-the-reef forcing. CRC is displayed as circled area in model particle tracks (A) and model vectors (B). CRC formation on the opposing reef arm is weakened during dominance of the other.

The Hydrodynamic Modelling of Reefal Bays –

Placing Coral Reefs at the Center of Bay Circulation 173

Kingston Harbour to the north during a flood event undergo retention along the Hellshire shoreline all through the tidal cycle, particularly during wind clams, but alternating in these reefal bays depending on the stage of the cycle prevailing. The longest retention time derived from field data was 9 hrs (Maxam & Webber, 2010) and compares well with model

Simulations also show that CRC presence is characterized by increased fluctuations in the retention of particles. Model simulations depicted that after 6 hrs, Wreck Bay and Sand Hills Bay showed the greatest variation in number of particles remaining and Engine Head Bay the least variation across all conditions. Therefore, those conditions that facilitated greater particle retention in the reefal bays, particularly wind calms (Maxam & Webber, 2010), significantly increased retention times over that of the open Engine Head Bay. The same is true for those conditions that facilitated decreased particle retention in the reefal bays where

Provided wind conditions did not dominate, Engine Head Bay produced similar retention times as particles oscillated back and forth inside the bay arc with the change of the tidal regime. This oscillation, however, did not extend outside the bay arc, unlike with the reefal bays. Reefal bays therefore display the ability to not only trap particles throughout tidal cycles, but also create a wider trapping area (extended seawards) than open bays along the

ii. in the increased recirculation rate of particles resulting from increased gyre current

The topography of the reefal bay allows it to produce signature dynamics driven primarily by over-the-reef flow, wind and tidal forcings. Waves break over the reef and the generated flow feed reef-parallel currents that in turn supply a major channel outflow. The channel (Figure 13) features significantly in this system and its prominence is the main bathymetric difference from other more popularly studied reef systems such as atolls, platform and

iii. in the broad spatial extent of the CRC occupying a greater portion of the bay area. Hence, the CRC is considered persistent because it continuously loops the reef, and is strengthened when gyres are closed and it broadens horizontally. This closing recirculation demonstrates very well the connectivity and continuity of the channel outflow re-entering the bay over the reef, and therefore best confirms the reef as the circulatory centre of the bay. Model simulations did not produce a reversal in back-reef currents at any time, evidence that the CRC is never completely reversed but instead may become severely weakened, usually coinciding with very rare events of channel reversal at depth (recorded by field instruments in Maxam & Webber, 2010). The functional bay is therefore seen to exist around the reef such that the reef parabola are the center of the system. Increased flow over the reef, especially during the sea-breeze regime, caused surges in channel currents that would increase the speed of the current loop and result in faster flushing times. Reefal bay flushing and retention regimes have direct implications on the dynamics of vulnerable planktonic species important to reef establishment (Wolanksi & Sarsenski, 1997), and the ability of these bays to draw in, retain and flush pollutants (Black

results that showed the longer retention of particles ranging from 6 to 9 hrs.

these were significantly lower than in the non-reefal bay.

Hellshire shoreline.

speeds; and

CRC strengthening was therefore evident i. with the closure of the looping circulation;

et al., 1990; Lasker & Kapela, 1997).

**6.5 Bathymetric characteristics necessary for promoting CRC** 

Fig. 12. RMTRK Tracking model outputs depicting gyre dominance cycling in Wreck Bay and Sand Hills Bay. Closed gyre formation is dominant on the down-shore reef during rising tide regimes, abate at high tide, then re-form on the up-shore reef during falling tide. The larger reef in both bays produced the larger dominant gyre resulting in the greater seaward extensions of the bay. The east reef for Wreck Bay and the south reef at Sand Hills Bay therefore expanded the bays the most.

### **6.3 Reef CRC persistence between paired reef arms**

Persistence of one reef CRC over another was observed with the reef pairs and was characteristic of one reef only, unlike reef dominance that alternated between reefs. Persistence of a reef arm CRC occurred when, during conditions that caused the least change in current flow, the CRC was continuously propagated on that reef. This was observed during a combination of decreased over-the-reef flow and small changes in tidal amplitude, when the Wreck Bay west reef arm and the Sand Hills Bay south reef arm displayed continuous CRC while the other reef arms in the pair showed none, even during changing tidal cycles. This persistence, along with the larger west reef flow, has led to the west reef contributing more than the east reef overall to the channel flow in Wreck Bay.

#### **6.4 Variability in retention**

Sand Hills Bay retained particles longer than Wreck Bay in model simulations, with retention controlled mostly by the dominant reef gyre. The dominant reef gyre is maintained in Sand Hills Bay during the rising tide, while that of Wreck Bay is well-formed during the falling tide. This presents the likelihood that waterbourne particles flushed out of

Fig. 12. RMTRK Tracking model outputs depicting gyre dominance cycling in Wreck Bay and Sand Hills Bay. Closed gyre formation is dominant on the down-shore reef during rising tide regimes, abate at high tide, then re-form on the up-shore reef during falling tide. The larger reef in both bays produced the larger dominant gyre resulting in the greater seaward extensions of the bay. The east reef for Wreck Bay and the south reef at Sand Hills

Persistence of one reef CRC over another was observed with the reef pairs and was characteristic of one reef only, unlike reef dominance that alternated between reefs. Persistence of a reef arm CRC occurred when, during conditions that caused the least change in current flow, the CRC was continuously propagated on that reef. This was observed during a combination of decreased over-the-reef flow and small changes in tidal amplitude, when the Wreck Bay west reef arm and the Sand Hills Bay south reef arm displayed continuous CRC while the other reef arms in the pair showed none, even during changing tidal cycles. This persistence, along with the larger west reef flow, has led to the west reef contributing more than the east reef overall to the channel flow in Wreck Bay.

Sand Hills Bay retained particles longer than Wreck Bay in model simulations, with retention controlled mostly by the dominant reef gyre. The dominant reef gyre is maintained in Sand Hills Bay during the rising tide, while that of Wreck Bay is well-formed during the falling tide. This presents the likelihood that waterbourne particles flushed out of

Bay therefore expanded the bays the most.

**6.4 Variability in retention** 

**6.3 Reef CRC persistence between paired reef arms** 

Kingston Harbour to the north during a flood event undergo retention along the Hellshire shoreline all through the tidal cycle, particularly during wind clams, but alternating in these reefal bays depending on the stage of the cycle prevailing. The longest retention time derived from field data was 9 hrs (Maxam & Webber, 2010) and compares well with model results that showed the longer retention of particles ranging from 6 to 9 hrs.

Simulations also show that CRC presence is characterized by increased fluctuations in the retention of particles. Model simulations depicted that after 6 hrs, Wreck Bay and Sand Hills Bay showed the greatest variation in number of particles remaining and Engine Head Bay the least variation across all conditions. Therefore, those conditions that facilitated greater particle retention in the reefal bays, particularly wind calms (Maxam & Webber, 2010), significantly increased retention times over that of the open Engine Head Bay. The same is true for those conditions that facilitated decreased particle retention in the reefal bays where these were significantly lower than in the non-reefal bay.

Provided wind conditions did not dominate, Engine Head Bay produced similar retention times as particles oscillated back and forth inside the bay arc with the change of the tidal regime. This oscillation, however, did not extend outside the bay arc, unlike with the reefal bays. Reefal bays therefore display the ability to not only trap particles throughout tidal cycles, but also create a wider trapping area (extended seawards) than open bays along the Hellshire shoreline.

CRC strengthening was therefore evident


Hence, the CRC is considered persistent because it continuously loops the reef, and is strengthened when gyres are closed and it broadens horizontally. This closing recirculation demonstrates very well the connectivity and continuity of the channel outflow re-entering the bay over the reef, and therefore best confirms the reef as the circulatory centre of the bay. Model simulations did not produce a reversal in back-reef currents at any time, evidence that the CRC is never completely reversed but instead may become severely weakened, usually coinciding with very rare events of channel reversal at depth (recorded by field instruments in Maxam & Webber, 2010). The functional bay is therefore seen to exist around the reef such that the reef parabola are the center of the system. Increased flow over the reef, especially during the sea-breeze regime, caused surges in channel currents that would increase the speed of the current loop and result in faster flushing times. Reefal bay flushing and retention regimes have direct implications on the dynamics of vulnerable planktonic species important to reef establishment (Wolanksi & Sarsenski, 1997), and the ability of these bays to draw in, retain and flush pollutants (Black et al., 1990; Lasker & Kapela, 1997).
