Document Type


Publication Date



Historically, Acropora spp. are the major reef building corals seen throughout the Caribbean and parts of the Western Atlantic that can grow relatively rapidly in dominant mono-specific stands. Their rapid growth and fragmentation allows them to out compete other benthic organisms and form the major framework for entire reef zones. They are the most abundant and important species for reef accretion. Their branching morphologies provide important habitat for many other reef species and no other Caribbean coral species fills these ecosystem functions. Acroporids were once the dominant reef builder in the Caribbean and provided the majority of live coral cover, but have had extensive population declines. Despite the recent declines, dense patches of Acropora have been reported in several areas throughout the Caribbean. Perhaps the most surprising of these locations is southeast Florida (SE). SE Florida reefs are a higher latitude system that transitions from a subtropical to temperate climate and is in close proximity to about 6 million people. These are some of, if not, the largest dense patches of A. cervicornis in the continental United States and offer a unique opportunity to evaluate population demographic structure and condition in a growth form (dense patches) which was once dominant but now rare.

In the 1990’s seven large high-cover Acropora patches were identified and characterized at 6 meters depth or less in Broward County. In 2014, an additional twenty-eight new patches were found covering an area of approximately 110,000 m². The patch delineations were not ideal due to mapping resolution and that they need to be mapped with higher precision. The threatened ESA status requires a plan to facilitate the recovery of the species back to historical levels. Thus, understanding the current population extents and condition is necessary to establish a reference baseline condition. These data were needed to determine if management strategies are necessary, which to employ, and reasonable success criteria for management actions. Hence, this study was conducted to provide these data.

Mean total cover between all patches was 56.5% ± 14.9. Live and dead cover were similar. Mean rubble was 12.5% (± 9.2). Mean disease cover was low (0.8% ± 0.7). Mean fireworm predation density was 1.4 m-² ± 1.09. On average, there was one damselfish garden every 5.9 square meters (0.17 m-² ± 0.14). And the mean density of disease occurrences was 0.91 m-² ± 0.84. Multivariate analyses of percent live, dead, rubble, and disease Acropora cervicornis at the densest portion of each patch indicated three main categories: Good (2 sites) – high amounts of live tissue; Moderate (20 sites) – similar amounts of live tissue and standing dead framework; Poor (13 sites) – high amounts of dead framework and rubble. The Poor group had an average of 26% cover of dead framework, 21% rubble, and 10% live cover. The Moderate group had an average of 20% cover of dead framework, 8% rubble, and 26% live cover. The Good group had an average of 13% cover of dead framework, 4% rubble, and 62% live cover.

Twenty-three perimeters were mapped around 35 dense patches. The imagery indicated that the dense patches are still distinctly different, however, the in situ surveys indicate that several dense patches are spread out and connected to adjacent patches. The diver GPS perimeter mapping yielded a total patch area of approximately 826,609 m² (204 acres).

This study found that six percent of the dense patches are in Good condition, fifty-seven percent in Moderate condition, and thirty-seven percent in Poor condition. Without having previous data on most of the patches, not much can be said about their condition trajectories or what caused their declines into the Moderate and Poor states. Little disease was recorded during this study indicating that disease was not a big factor of present patch condition. However, the large amounts of dead framework measured in our study indicate a relatively recent decline in condition. Due to the lack of frequent monitoring, it is unknown how much past disease events contributed to the amount of the present dead framework and rubble cover. Although not significant, Gillliam and Larson (2014) previously found that Rapid Tissue Loss (RTL) disease coincided with decreases in live cover, especially after hurricane Sandy and tropical storm Isaac. Presumably, this could have affected the condition of many of the SE FL dense patches. The cause of increased RTL after these storms is unknown and should be established to mitigate for future impacts to A. cervicornis live cover.

Glimpses at patch condition trajectories were possible for a few sites based on historical data from a few longer-term studies. Vargas-Angel et al. (2003) patch categorization contained three groups: A, B, and C. Group A, their mostly-dead site (Coral Ridge), which has since disappeared, was not evident in recent aerial photographs. The exact timing of its disappearance is unknown, but it was before 2007 and is thought to have been due to a strong storm event; perhaps Hurricane Wilma. Group B, defined by relatively high coral cover and greatest A. cervicornis density (Commercial I, Commercial II, and Dave), have persisted through time. The Dave patch was renamed as FTL6 in the Broward County annual reef monitoring and BCA in the Southeast Coral Reef Evaluation and Monitoring Program. This patch has been studied extensively. Group C patches (Oakland I, Cervicornis II, and Oakland II) have increased in A. cervicornis. The Oakland I patch was renamed Scooter and has been monitored at least semi-annually since 2007.

SE FL is presently in a hurricane drought. The last strong storm to hit the direct area was hurricane Wilma in 2005. Hurricanes Sandy and Matthew came close along the eastern seaboard but they were mostly rain events for south Florida. Increases in live cover have been measured over periods in between storms that may be related to low storm activity (Gilliam and Larson 2014). This needs more investigation as the correlation is not obvious and it is unknown if this is due to lower RTL prevalence or reduced physical impacts. Events like hurricane Sandy and tropical storm Isaac may have catalyzed RTL outbreaks (Gilliam and Larson, 2014), but were not strong enough to move large amounts of framework. A direct hit from a hurricane could spread the patches of mostly dead framework off the reefs leaving little to no live fragments behind to maintain dense patch status similar to the Coral Ridge patch. This scenario could drastically affect the number of dense patches, their condition, and extents.

The patch mapping efforts, funded by NSU, show that spreading continues at both BCA and Scooter however, the densest areas in the patches still exist in the original locations. These patches are in Poor condition. In terms of the percent live cover to total Acropora measured in each patch, Scooter ranked 26 out of 35 sites and BCA ranked 34 out of 35. We estimated live cover at 9.7% which is similar to other recent results. After revisiting BCA and mapping the perimeter during this study, it is clear that live cover has decreased and not just moved away.

Scooter was a similar story to BCA in that live cover had decreased through time at the densest areas in the patch to about 15% and did not significantly change from 2011 to 2013 (Gilliam and Larson, 2014). Changes in live cover occurred between 2008 and 2011 where the majority of live cover shifted away from the densest framework areas. After visiting Scooter it was obvious the densest portions were degraded, however because the site is so large, shifting of live cover to a new area was not obvious or investigated. A visual comparison of 2013 aerial photography and 2017 ESRI satellite imagery do not show obvious differences, but the ESRI imagery cannot be statistically analyzed.

The perimeter surveys showed that most of the patches are much larger than originally visualized in the 2013 aerial imagery. It appears they have also spread across the reef-scape through time. The densest areas are still the areas with the most concentrated colonies, but many of the perimeters span between these areas.

This study elucidated new data on the extent and condition of the dense patches of Acropora cervicornis in SE FL. Approximately 20% of the dense patches were previously known before Walker and Klug (2014) and only two were previously mapped. This study statistically analyzed dense patch conditions and binned them into three groups based on the amount of live, dead, disease, and rubble cover. The GPS diver mapping identified the spreading of dense patches and increased total area of dense A. cervicornis to 826,609 m² (204 acres), an increase of over 500% from previous estimates. This new information highlights more critical gaps in our knowledge of regional A. cervicornis distributions and population distribution, demographics, and status.

Below are a series of recommendations to help fill those knowledge gaps:

  • Conduct A. cervicornis mapping and condition assessments more frequently to determine cause of live tissue declines.
  • Establish a cause of increased RTL after storm events to mitigate for future impacts to live cover.
  • Analyze historical imagery to determine the timing of dense A. cervicornis patch inception and persistence over time.
  • Collect regular, periodic regional standardized imagery to elucidate the dynamics of dense patches and document the current extent of nearshore resources.
  • Investigate the genetic diversity of the dense A. cervicornis patches to determine if they are genetically similar to each other and other local populations.
  • Monitor fecundity and reproduction to identify if environmental factors and patch conditions are related to reproductive success.


Completed in Fulfillment of PO AF19EE for Florida Department of Environmental Protection, Coral Reef Conservation Program, 1277 N.E. 79th Street Causeway, Miami, FL, 33138

Additional Comments

NOAA National Ocean Service Award #: NA14NO24190053