Spatial patterns of coral reef benthic communities vary across a range of broad-scale biogeographical levels to fine-scale local habitat conditions. This study described spatial patterns of coral reef benthic communities spanning across the 536-km coast of Kenya. Thirty-eight reef sites representing different geographical zones within an array of habitats and management levels were assessed by benthic cover, coral genera and coral colony size classes. Three geographical zones were identified along the latitudinal gradient based on their benthic community composition. Hard coral dominated the three zones with highest cover in the south and Porites being the most abundant genus. Almost all 15 benthic variables differed significantly between geographical zones. The interaction of habitat factors and management levels created a localised pattern within each zone. Four habitats were identified based on their similarity in benthic community composition; 1. Deep-Exposed Patch reef in Reserve areas (DEPR), 2. Deep-Exposed Fringing reefs in Unprotected areas (DEFU), 3. Shallow Fringing and Lagoon reefs in Protected and Reserve areas (SFLPR) and 4. Shallow Patch and Channel reefs (SPC). DEPR was found in the north zone only and its benthic community was predominantly crustose coralline algae. DEFU was found in central and south zones mainly dominated by soft corals, Acropora, Montipora, juvenile corals and small colonies of adult corals. SFLPR was dominated by macroalgae and turf algae and was found in north and central zones. SPC was found across all geographical zones with a benthic community dominated by hard corals of mostly large colonies of Porites and Echinopora. The north zone exhibits habitat types that support resistance properties, the south supports recovery processes and central zone acts as an ecological corridor between zones. Identifying habitats with different roles in reef resilience is useful information for marine spatial planning and supports the process of designing effective marine protected areas.
Citation: Karisa JF, Obura DO, Chen CA (2020) Spatial heterogeneity of coral reef benthic communities in Kenya. PLoS ONE 15(8): e0237397. https://doi.org/10.1371/journal.pone.0237397
Editor: Jose M. Riascos, Universidad de Antioquia, COLOMBIA
Received: March 7, 2020; Accepted: July 26, 2020; Published: August 26, 2020
Copyright: © 2020 Karisa et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: This work benefited from various funding sources: JK was a recipient of a scholarship funded by the Taiwan International Graduate Program–Biodiversity Program of Academia Sinica and National Taiwan Normal University, Taiwan. CAC was supported by funding from Biodiversity Research Centre, Academia Sinica and Ministry of Science and Technology, Taiwan. Fieldwork was partly funded by the Western Indian Ocean Marine Science Association (WIOMSA) through the Marine Research Grant (MARG) program under grant number MARG I Contract 006/2014 and The Nature Conservancy Africa Regional Office. Logistical support for fieldwork was provided by the Kenya Marine and Fisheries Research Institute (KMFRI), Kenya Wildlife Service (KWS) and World Wide Fund for Nature Kenya (WWF-Kenya). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Coral reefs are among the most biodiverse ecosystems in the world , providing an array of ecological services which are important for human well-being [2, 3]. However, coral reefs are declining at an alarming rate globally with climate change posing as one of the biggest threats . There are still limited global efforts to abate climate change  making reefs a highly threatened ecosystem with climate scenarios showing 99% of the reef could disappear in this century . The persistence of coral reefs will rely heavily on their ability to maintain a coral-dominated state and avoid shift to algal-dominated or other alternative stable states amid the inevitable effects of climate change [7–11].
The Kenya coast stretches along 536 km, between latitudes 10 and 50 S with narrow fringing reefs in the southern part and patchy reefs with low reef development in the north [12, 13], (Fig 1). The distribution of coral species in Kenyan reefs is influenced by the large-scale current dynamics with the East African Coastal Current (EACC) bringing coral larvae from the southern ‘center of diversity’ for the Western Indian Ocean (WIO) region . A cold-water system prevails in northern Kenya due to the convergence of the EACC with the seasonal Somali Current (SC) that is characterized by poor water conditions for reef development [15, 12]. The interaction of the EACC and the SC in the north creates a marginal, high-latitude and upwelling system with transitioning communities from the East African to Somali-Arabian fauna . This results in high coral species diversity in the southern parts of Kenya and a decrease in diversity towards the north . In addition, the presence of river systems in the central-northern region introduces small-scale influences in species distribution by creating environmental barriers that further limit the transport of larvae to the north .
(A) Locations of the study sites along the Kenyan coast divided into three geographical zones in the, (B) North (red), (C) Central (blue) and (D) South (green). Ocean base map credits- Esri, Garmin, GEBCO,NGDC, and other contributors.
There has been a substantial amount of coral reef studies in Kenya focusing on a range of issues including: the effect of MPAs in protecting habitats and biodiversity [18–20], impact of anthropogenic stressors, including the 1998/99 mass coral bleaching event, on coral and fish communities [21–25]; but without reference to spatial patterns of communities along the latitudinal gradient. These studies, which include coral reef monitoring with different levels of management programs (marine reserve vs. marine park), have been running for more than three decades showing long-term trends in coral reef communities [26, 2, 27] but have mostly focused on shallow lagoons within the south coast, leaving deeper forereefs and the north zone depauperate of studies. In addition, there is limited information on coral reef communities on reefs that are remote and not within conservation areas along the Kenyan coast. Knowing the spatial patterns of the coral reef benthic community can help identify habitats that offer resistance/protection to, and recovery after a disturbance including climate change and could help the persistence of coral reefs.
In this study, we follow a quantitative approach to characterize coral reef benthic communities along the Kenyan coast. Using extensive data collected at different geographical zones, reef habitats and management levels multivariate analysis was utilized to investigate the effects of the latitudinal gradient, local habitat and management in structuring spatial patterns in benthic communities.
Materials and methods
Kenya is located on the east coast of Africa between Somalia and Tanzania with a coastline running approximately 500km long between 10 and 50 S (Fig 1), . Two monsoons namely, southwest (December to March) and northeast (May to October) highly influence the climate and seasonality of this area. The Kenyan coast also experiences the EACC originating from Tanzania and flows northwards throughout the year. In addition, the SC mainly driven by northeast monsoon (NEM) counteracts the EACC when the wind blows southwards. Interaction of these two currents during the NEM causes cooling in the northern part of Kenya, and through mixing of nutrients makes this zone highly productive .
Coral reefs in Kenya are characterized by narrow fringing reefs in the southern part and patchy and discontinuous fringing reefs in the northern part . These reefs are found at different depths ranging from less than 1 m at low tide to over 20 m. There are different geomorphological characteristics of the reefs with channel reefs in highly sedimented areas of river inflows to ocean, lagoon reefs mostly occurring at leeward side of reef flat and fringing reefs on the seaward side and patch reefs occurring at offshore outcrops of reefs.
Sea surface temperature (SST) for two ‘normal years’ with no influence from extreme climate events (e.g.the Indian Ocean Dipole) in 2013 and 2014 show that the north zone had a low mean SST (27.45 0C) but with the highest variation with a standard deviation of 1.23 0C, a maximum of 30.16 0C and minimum of 25.27 0C. The south zone had the highest mean SST of 27.63 0C and the lowest variation (sd = 1.15 0C, max = 29.71 0C, min = 25.71 0C). The central zone had the lowest mean SST (27.26 0C) and an intermediate variation (sd = 1.21 0C, max = 29.58 0C, min = 25.41 0C). SST data were derived from MODIS satellite source by plotting georeferenced surveyed sites on a GIS system. Using ArcGis 10.8, points were created along vector lines at two pixels apart (one pixel = 4x4km) for each zone (north = 10points, central = 5points and south = 3points) and SST values were averaged over the two years. There are different levels of management for coral reefs in Kenya; ‘marine parks’ receive legal protection from fishing activities and collection of shells or corals. ‘Marine reserves’ receive partial protection with only traditional fishing methods allowed. ‘Unprotected’ areas are only regulated by fishing licensing and recently some areas receive fishing restrictions set by local communities. The southern zone has 2 parks and 2 reserves, and the central zone 1 park and 1 reserve in south zone. The rest of the reef is within unprotected areas. The northern part has only 1 reserve and the rest are unprotected reef areas.
A hierarchical sampling design (S1 Table) was used to collect benthic community data across Kenya coast. There were three pre-defined geographical zones (North, Central, South). In each geographical zone, sampling sites were selected to reflect different habitat factors: two depth levels (deep [6.5-18m], shallow [0-6m]), two levels of exposure to oceanic waves (exposed, sheltered) and four different reef types (channel, lagoon, fringing, patch). Additionally, three different management states were sampled; park (no-extraction), reserve (regulated extraction) and unprotected (extraction allowed). It should be noted that not all habitat types and management levels were found at the three geographical zones, for example the north zone had sites in reserve and unprotected levels but there is no park in this zone (S1 Table). Furthermore, the north zone had all the four levels of reef type while central and south zones did not. In the end, there were twenty sites in the north, ten sites in central and eight sites in the south. Choice of these study sites was based on historical monitoring of reef resilience , with an additional fourteen sites selected haphazardly and stratified within each geographical zone in order to cover habitats that had not been studied before.
Surveys for benthic communities were conducted during the north-east monsoon season during 2015 and 2016. A modified resilience assessment protocol  was used to record benthic cover and coral size class distributions. Benthic cover was recorded using photoquadrats with an underwater camera positioned at approximately 1 m above the seafloor with the aid of a 1 m PVC stick. Within each belt-transect, twenty five 50 x 50 cm photos were taken at an interval of 1 m. Cover was recorded in fifteen categories; hard corals (HC), coralline algae (CA), macroalgae (MA), turf algae (TA), soft corals (SC), Halimeda (HA), bare substrate (BS), dead standing coral (DSC), recently dead coral (RDC), invertebrates (INV), rubble (RB), sand (SD), seagrass (SG), silt (SL) and others (OT). HA was separated from macroalgae group as it is a calcifying algae that contributes to the production of carbonate and formation of reef sediments [30, 31]. Living coral cover (LCC) was recorded by coral genera.
The size class distribution of coral colonies was measured for 23 selected genera of adult corals using the belt transects , and for all juvenile corals by subsampling within the belt transects in six 1m2 quadrats that were 5 m apart along the 25m belt. Adult colonies were defined as those larger than 10cm in diameter and grouped into six size classes (11–20 cm, 21–40 cm, 41–80 cm, 81–160 cm, 161–320 cm and >320 cm). Juvenile corals were defined as those smaller than 10 cm in diameter and were classified into three size classes (1–2.5 cm, 2.6–5 cm and 6–10 cm).
Benthic photos were analysed using CPCe v4.0 software  by randomly placing 25 points on each photo. This number of points overlaid on the photo are within the optimum number required for accurately estimating coral cover in a coral reef that has about 30% cover . The type of benthic substrate was identified under each point.
Three datasets were analyzed: 1. benthic cover by major categories (S1 Table), 2. benthic cover by coral genera (S2 Table) and 3. coral size classes (S3 Table). A set of criteria was established to select only the key variables important for determining the spatial pattern as follows. First, only biotic variables from the major benthic categories were considered as these are influenced by latitudinal gradient and habitat environmental conditions. Only coral genera that had a > 1% cover were included in the analysis. In the second stage, analysis was done in PRIMER v7  where variables were square root transformed in order to reduce skewness and make the variances more homogenous. Each of the three datasets were examined by principle coordinate analysis (PCoA) in order to select those key variables that only contributed to the clustering of sites. Only those variables whose vector direction demonstrated influencing on clustering of sites were selected for further analyses. Variables whose vectors had similar direction were either pooled together or representative variables were selected to avoid analysing many variables that otherwise have similar effect on clustering sites. A permutational multivariate analysis of variance (PERMANOVA) , was conducted for each of the three datasets to determine if geographic zone had an effect on site clustering. The selected key variables from the three datasets were then pooled together and collectively analysed. A One-way ANOVA was done in excel to determine how each key variable differed across the geographic zones. PERMANOVA was done in R using the Adonis function within the vegan package  to test the effect of depth, reef type, exposure to oceanic waves, management and their interaction. Since geographic zone was found to have a significant effect on variables, this analysis was treated as a block design where all other factors were randomly assigned within each block (i.e. geographic zone).
Description of variables-overall
Among the major benthic categories assessed, there were ten biotic benthic communities with hard corals having the highest cover (24.1 ± 26.7%) followed by turf algae (16.2 ± 21.8%) and macroalgae (12.9 ± 22.7%), (S2 Table). Forty-three coral genera were recorded with only 13 having a cover of > 1%. Porites was the most abundant genus (7.7 ± 17.6%) followed by Acropora (2.1 ± 7.3), Echinopora (2.06 ± 10.07) and Montipora (1.64 ± 7.99) (S3 Table). Coral density (number of colonies per 100 m2) was higher among juvenile coral colonies and some of the adult colonies within small size classes. Density was highest in the 6–10 cm size class (197.1 ± 182.1 colonies) and lowest in the >320 cm size class (5.7 ± 10.4 colonies) (S4 Table).
Selection of key variables
Three PCoAs performed on the ten biotic benthic categories, 13 coral genera and ten coral size classes showed clustering of sites (Fig 2A–2C). Of the ten benthic categories, six (hard corals, coralline algae, turf algae, macroalgae, soft corals and dead standing corals), were selected for further analyses (Fig 2A). Halimeda was not included because their vectors pointed in the same direction, so only CCA was selected as the representative variable. Seagrass and ‘other-invertebrate’ categories were not included as key variables because they had very low mean percentage cover. Only four representative genera, Acropora, Montipora, Echinopora and Porites, were considered as key variables for further analysis based on the different sets of similar-effect vectors on the PCoA (Fig 2B) as well as their abundance (S6 Table). Vectors of Acropora and Montipora clearly pointed in opposite direction showing that they had distinct influence on site clustering. Vectors of Porites and Echinopora pointed in similar direction and most orthogonal to Acropora and Montipora, and were among the most abundant genera recorded in this study, so were selected as indicative of other genera. Vectors for 1–10 cm size classes were closely grouped, as well as vectors for > 40 cm size classes (Fig 2C), so a representative size class from each was selected. The other size-classes (11-20cm and 20-40cm) were considered separately based on their independence in clustering sites.
Principle Coordinate analysis (PCoA) analyses of (A) major benthic communities (B) coral genera and (C) coral size classes. Clusters of 38 reef sites based on Bray–Curtis index. Different colours differentiate sites based on the geographical zone (red = north, blue = central, green = south).
A PERMANOVA test on the effect of geography for each of the datasets showed significant differences between zones (S8 Table) and therefore further analysis was restricted to within geographic zones (Table 1).
Describing key variables
One-way ANOVA for the key variables (Fig 3) showed that almost all differed significantly (p<0.05) across geographical zones. There was significant difference in hard coral cover across the three geographical zones (F2,773 = 6.88, p = 0.00, Table 2) with the highest cover occurring in the south (32.06 ± 29.48) while north and central zones had similar cover (22.5 ± 26.6% and 23.0 ± 23.7%, respectively), (Fig 3). Crustose coralline algae cover was significantly higher in the north (8.9 ± 14.8%) than in the central (3.4 ± 8.0%) and south (2.3 ± 5.6%)(F2,773 = 21.02, p = 0.00). Turf algae had a significantly higher cover in north and central zones (17.8 ± 22.7% and 19.2 ± 22.5%, respectively) and lowest in the south (6.3 ± 12.8%)(F2,773 = 16.85, p = 0.00). Soft coral cover was significantly higher in the south (17.9 ± 23.1%) followed by central (8.4 ± 16.0%) and lowest in the north (3.7 ± 12.2%)(F2,773 = 16.85, p = 0.00). Dead standing coral was significantly higher (F2,773 = 15.19, p = 0.00) in the south (1.63 ± 4.5%) followed by central (0.7 ± 3.0%) and north (0.2 ± 1.6%). Macroalgae cover did not show significant variation across geographic zones.
Mean percentage cover of five major benthic cover categories, four coral genera and density (number of colonies per 100m2) of four coral size classes for north (upper, red), central (middle, blue) and south (lower, green) geographic zones. Error bars indicate standard error.
Results of One-way ANOVA test done to determine variation of key benthic communities across the 3 geographical zones.
There was significant difference in the cover of Porites across the three geographic zones (F2,774 = 3.6, p = 0.03, Table 2) with highest cover in the south (10.3 ± 17.9%) followed by north (7.9 ± 18.2%) and central (4.7 ± 15.0%) (Fig 3). Acropora cover was significantly different across zones (F2,774 = 24.8, p = 0.00) with highest cover in south (5.9 ± 11.2%) followed by central (2.6 ± 6.6%) and north zones (1.0 ± 5.7%). Montipora cover differed significantly across the zones (F2,774 = 5.13, p = 0.01) with the highest cover in the south (3.3 ± 11.5%) followed by north (1.6 ± 8.0%) and central zones (0.4 ± 2.2%). Echinopora cover did not show any difference across geographic zones.
A total of 64,850 coral colonies were recorded with an average density of 79 colonies per 100m2 across 4 size classes. The density of corals (number of colonies per 100m2) in 1-10cm size class differed significantly (F2,88 = 5.0, p = 0.01; Table 2) with the highest density in central and south zones (510.8 ± 333.0 colonies and 510.9 ± 352.8 colonies, respectively) (Fig 3). The density of 11-20cm size class was significantly higher in the south zone (208.3 ± 134.5 colonies) followed by central and north zones (108.6 ± 55.3 and 107.5 ± 92.8 colonies, respectively)(F2,88 = 9.5, p = 0.00). Density of corals in the 21-40cm size class differed significantly (F2,88 = 15.0, p = 0.00) with highest density in the south (143.3 ± 81.3 colonies) followed by north (66.4 ± 65.0 colonies) and central (57.1 ± 30.4 colonies). The size-class >40cm did not show significant difference among geographic zones (Table 2) but densities were higher in the north (86.7 ± 101.0 colonies).
Habitat factors and management status
A PERMANOVA test on the effect of habitat factors and management showed significant effect (p<0.05) of depth and management on benthic variables (Table 3). While exposure and reef type did not show significant effect, their interaction with depth and management was significant.
Embedded on the four quadrants of a PCoA for the key variables is an illustration of different combinations of habitat and management interactions at the three zones (Fig 4). The upper-left quadrant represents a habitat found only in the north zone with the benthic community highly dominated by crustose coralline algae. This habitat is characterized by Deep-Exposed Patch reefs that occur within Reserves (DEPR). The upper-right quadrant contains sites from central and south zones dominated by soft corals, Acropora and Montipora, and with a high density of juvenile corals (1-10cm) and small adult corals (11-20cm). This habitat is characterized by Deep-Exposed Fringing reefs within Unprotected areas (DEFU). The lower-left quadrant contains sites found mainly in the north and central zones. The benthic community is predominantly turf algae and macroalgae, and is characterized as Shallow Fringing and Lagoon reefs within Parks and Reserves (SFLPR). The quadrant to the lower-right contains sites from all geographic zones and has moderate levels of all variables. This represents a habitat that characterized as Shallow Patch and Channel reefs (SPC).
Clusters of 38 reef sites based on Bray–Curtis index. Embedded in each quadrant of the PCO is interpretation of the habitat factors and management levels associated with the clustering of sites.
Of the four habitats described from this study, the north and central zones contained three of them and south contained two habitats (S1 Fig). DEPR was found uniquely in the north, SFLPR was found in all geographic zones, DEFU was found in central and south zones and SPC was found in the north and central zones.
Two distinct patterns of coral reef benthic community structure were observed along the Kenyan coast. A latitudinal gradient was observed indicating a differentiation in benthic community composition between northern and southern sites. A second pattern was based on localised habitat and management factors prevailing within each geographic zone.
Latitudinal/geographic gradients in benthic community composition
The observed differences in benthic community composition among geographic zones found in this study is consistent with previous authors that have documented the biogeographic distribution of reef-building corals [37, 16].
The south zone had a higher LCC with most genera occurring in abundance but mainly dominated by Porites and Acropora. This area is known to have a high diversity of corals mainly due to its proximity to the WIO center of biodiversity which has been documented to be around the northern Mozambique channel (NMC) [37, 16]. The south zone is the receiving point of coral larvae into Kenyan waters as currents transporting larvae progress northwards  resulting in higher coral diversity compared to the north. In the north, a smaller number of genera was found with Porites being dominant, indicating a marginal coral community as it is farther from sources of larvae. The north zone is also an area of poor reef development due to the unfavorable conditions caused by the convergence of the colder nutrient rich SC and the EACC, which diminishes chances of high coral diversity in this zone. This convergence of currents makes the north an area of biogeographic transition from southern East African reef communities to Somali-Arabian Gulf communities to the north. This may mean that the north zone can act as an environmental corridor for species movements as the climate warms as has been suggested for other marginal communities . Other studies have reported low coral diversity on marginal reefs e.g. upwelling reefs of the Arabian gulf , but there is also some isolated cases of high coral cover and diversity in marginal reefs e.g., at the world’s southern-most coral reefs in subtropical Islands of the Pacific Ocean [40, 41]. Some of these marginal reefs have the potential to act as habitable spaces for corals as climate warms, creating climate refugia or ‘seed banks’ that can re-populate coral communities once climate stabilizes to favorable conditions .
The benthic communities in the north and south represent two alternate aspects of coral reef resilience. The comparably lower abundance of juvenile coral colonies in the north and comparably higher abundance of large colonies suggests poor coral recruitment but high survival of adult coral colonies. Low recruitment could be attributed to isolation of the north zone from the southern sources of larvae, with only local sources of larvae for recruitment. Higher abundance of large colonies in the north may be due to the coral community in the north being protected from bleaching mortality by the cooling effect of upwellings in the SC. On the other hand, higher abundance of juvenile corals in the south indicates high recruitment likely due to its proximity to sources of larvae [17, 16]. The low abundance of large colonies in the south implies low survival and only a few coral colonies surviving to large sizes. The south zone is known to face more stress from bleaching as it is closer to the warmest zones at 10–12 oS, and higher anthropogenic pressures due to higher human population densities as compared to the north. There have been several bleaching episodes since the 1998/99 mass coral bleaching which could have slowed down recovery of coral communities in the south.
Other studies in the WIO region have shown that the coral community in the north is very different from that in the south, with a higher diversity in the south comprising species more closely resembling the northern Tanzanian coral community [43, 16]. Studies from other regions have also shown that the latitudinal gradient drives differences in benthic community composition on coral reefs. Coral reefs in the Hawaiian Archipelago in the north Pacific Ocean show a latitudinal gradient contributing to three distinct reef regimes each dominated by either hard coral, turf algae or macroalgae . In a Caribbean-wide study, spatial and temporal differences in coral cover and macroalgae were found between subregions . Eastern Australia coral reefs showed a trend of decreasing cover of Acroporidae as latitude increased. The diversity of benthic communities differed along the latitudinal gradient and across geographic zones in Taiwan .
Influence of habitat factors and management on benthic community composition
Habitat factors and management influenced the pattern of benthic communities within each geographic zone. This offers the opportunity to propose ecological controls that structure benthic communities at this scale, and which may be used to guide resource management decisions. Similar findings have been reported in parts of the Pacific Ocean, where benthic communities were structured by depth, exposure to waves and reef types, in the Hawaiian Islands [47–49] and in French Polynesia [50–54].
The heterogeneity and uniqueness of habitat types in the north of Kenya can be attributed to a number of factors including the fact that it is composed of a set of small islands in an archipelago, presenting diverse reef types at different depths. Perpendicular to the shore, in a cross-shelf line, there are numerous channel reefs between the islands and mainland, then lagoons, fringing reefs and at the outermost edge, a deep ridge with patch reefs on it. In addition, the northern part of Kenya has been described as an ‘ecotone’ zone due to the presence of marginal/transitional coral reef communities composed of rare or regional endemic species . Other studies have also shown significant differences in benthic communities within archipelagos such as distinct coral species assemblages among different islands in Penghu in Taiwan , a cross-shelf difference in benthic community within the Spermonde Archipelago in the Coral Triangle  and differences in benthic communities across habitats in the Abrolhos archipelago of eastern Brazil .
The habitat type DEPR that was uniquely found in the north zone is dominated by coralline algae, which may be due to a high biomass of fish on the patch reefs [21, 58]. Fish herbivores graze away competitors of crustose coralline algae, and fish predators feed on sea urchins which are the greatest eroders of coralline algae . Also, the upwelling currents in this zone could be a source of nutrients stimulating growth of coralline algae. Previous studies have associated coralline algae as a suitable cue for coral settlement and recruitment [60, 61] making these habitats potential spaces for corals to grow. Being deep habitats in an area of upwelling-cooler waters could offer an ecological refugia to coral communities as the climate warms . However, light limitation would reduce the opportunity of a diverse coral community growing and dominating these spaces [63–65] and only a few taxa that can tolerate these depths would colonise this habitat.
A majority of deep- fringing reefs that are exposed to oceanic waves in the central and north zones were dominated by Acropora, soft corals and small-sized coral colonies (DEFU habitat type). This habitat type is characterized by gradual slopes on the fringing reefs creating an environment where Acropora can dominate competitively. Other studies have reported Acropora occupying similar environments of forereef areas, where there is low stress from sedimentation and high rates of development processes such as reproduction . Notably, the low abundance of large colonies within this habitat is worrying considering that the south zone is an area of high coral diversity. This could be indicative of failure to replenish standing stocks of large colonies in the central and south zone, likely due to bleaching mortality which has been recurrent over the past two decades [26, 67]. All zones were badly affected by the 1997/98 mass coral bleaching episode with a 50% - 80% average loss of coral cover with some individual sites loosing up to 100% cover [68, 58]. Recovery has been slow since then perhaps due to the occurrence of smaller-scale bleaching episodes in 2005, 2007, 2010 and 2016 [26, 23, 67]. In the south zone, this bleaching impact was likely compounded by other anthropogenic stresses originating from its high human settlement. A similar explanation goes for the central zone reef where this study found few large colonies likely due to a slowed down recovery from the mass coral bleaching. Post-bleaching reports indicate that there was an increase in macroalgae and a decrease in fish biomass followed by an explosion in sea urchin density probably as a response to the decrease of fish biomass which constitute their predators. Consequently, sea urchin grazed the reef substate to bare . The presence of poor water conditions due to proximity to river Sabaki effluents, high fishing pressure and recurrent bleaching episodes could have made perseverance of coral colonies to large sizes a great challenge [69, 23].
Shallow fringing or lagoon reefs in the north and central zones (SFLPR habitat type) were characterized by a dominance of macroalgae and turf algae. Most of these habitats are proximal to land and are highly influenced by land-use activities such as effluents from rivers as well as fishing. When hard corals die macroalgae and turf algae take over the space due to increased nutrient availability or reduced herbivory [70–72]. Turf algae are the main focus of grazing but their form allows for rapid regrowth  making them thrive even in protected areas where grazing is high. Other studies have found that macroalgae are not influenced by latitudinal effects, with more local-scale influences such as nutrient levels strongly driving their distribution . In this study, the abundance of macroalgae within these habitats indicates a worrying level of coral reef degradation since most coral cover and taxonomic diversity has always been thought to occur within shallow depths [75–77].
Shallow habitats on patch or channel reefs (SPC) were dominated by large colonies with an abundance of Porites and Echinopora. Similar observations were made in the Great Barrier Reef where large colonies of Porites were associated with shallow inshore reefs . The presence of large coral colonies in shallow reefs indicates a habitat type that supports the survival of corals probably due to acclimatization to bleaching . Also, the high fluctuation of water temperatures in shallow reefs creates a variable environment that may support different coral genotypes. A diversity of genotypes offers a chance for coral colonies to survive to larger sizes and persist over time as seen in an inshore reef of the Great Barrier Reef where Porites dominated through very persistent genotypes . In addition, Porites and Echinopora have been considered as resistant and generalist coral genera within the functional groups of corals, based on their response to bleaching . This makes them capable of inhabiting shallow areas where competitive but less tolerant corals cannot survive the fluctuating temperatures, bleaching events and proximity to human disturbance such as sedimentation.
Porites is a widespread genus that occurs at different habitats and thus reinforces its functional role as a resistant species [81–83]. The presence of very large colonies of Porites in the north compared to the south could imply that bleaching mortality of Porites is lower in the north due to the presence of habitat types that support resistance properties for that genus. In the south zone, the high percentage cover of Porites is mainly composed of small colonies. This signifies an area which could be facing high bleaching mortality, as evidenced by high percentage cover of dead standing corals, but experiences high recruitment that recovers the standing stock. This recovery feature may not be present in the north because Acropora communities failed to recover after the 1998 coral bleaching episode and recruitment is persistently low ).
Implications for conservation
In order to effectively manage and mitigate the degradation of coral reefs, information on community spatial patterns, habitat and management influences is necessary. Bleaching episodes have become more frequent and under the most recent IPCC report, 99% of coral reefs are expected to be lost this century if temperatures increase 2°C above the pre-industrial baseline . However, if temperatures can be stabilized at 1.5°C, then 10~30% of the reefs could be saved . These reefs could serve as refugia [62, 85] or ‘seed banks’ to repopulate other reefs once mechanisms of reducing climate change to desirable levels have been achieved .
Management strategies should ensure the protection of habitats that exhibit bleaching protection or recovery characteristics in order to enhance the resilience of coral communities. In addition, there should be mechanisms to reduce local impacts such as sedimentation and pollution. High sedimentation levels are experienced on near-shore coral reefs near rivers with high sediment load, particularly in the central zone due to the river Sabaki and river Tana . Most of the reefs in this area are within parks and reserves but the effect of sediment load is likely to still degrade these reefs. In the south zone, sediment loading is mainly experienced from smaller rivers and creeks, with increasing threat from logging of mangroves, thus exposing sediments to erosion. In all these areas, increasing urbanization elevates discharge of sewage and other pollutants, further exacerbating stress to corals.
In conclusion, the occurrence of a geographical and a local habitat pattern in coral benthic communities that is further mediated by the level of management offers a range of environments that support coral reef communities. The north zone presents a unique environment of a marginal reef that provides protection to coral communities from bleaching  even as the climate warms, expressing resistance properties of coral reef resilience . The south zone’s high diversity of corals  and high recruitment potentially offer a habitat that expresses recovery properties after a bleaching disturbance . The central zone is intermediate, providing a corridor for transfer of coral propagules between the north and south. Similar propositions on the importance of ecological corridors in conservation have been mentioned for large scale studies in the Eastern Tropical Pacific  and within the Gulf of Mexico .
The heterogeneity of coral reef habitat types in Kenya offers an opportunity for designing MPAs that protect a diverse range of functional traits of coral communities . As such, spatial heterogeneity in MPAs may reduce the risk of catastrophic regime shifts .
S1 Table. Sampling design.
A table of sites sampled according to their geographical zones, habitat factors and management levels.
S2 Table. Benthic cover data.
Percentage cover of major benthic categories recorded at all surveyed coral reef sites along the Kenyan coast.
S3 Table. Coral genera data.
Percentage cover of coral genera recorded at all surveyed coral reef sites along the Kenyan coast.
S4 Table. Coral colony size class data.
Density (no. of colonies per 100m2) of coral colonies recorded at all surveyed coral reef sites along the Kenyan coast.
S5 Table. Mean percentage cover and standard deviation of 15 major benthic categories.
Summary of all studied sites along the Kenyan coast.
S6 Table. Mean percentage cover and standard deviation of 44 coral genera.
Summary of all studied sites along the Kenyan coast.
S7 Table. Mean density of coral size classes (number of colonies per 100m2).
Summary of all studied sites along the Kenyan coast.
S8 Table. PERMANOVA main-test table of results.
S1 Fig. Illustration of the four habitat types occurrence at the three geographical zones.
North zone has three habitat types one of them uniquely to this zone (DEPR); Central zone has three habitat types, one it shares with all other zones (SPC), another with only north (SFLPR) and only south zone (DEFU); South zone has only two habitat types, one it shares with all zones (SCP) and the other with only Central zone (DEFU).
We would like to particularly thank Benjamin Cowburn for his assistance in data collection, James Mbugua for assisting the development of the study site map, Maxwel Azalli for advising on data analysis and Gladys Okemwa for helpful discussions during the development of this study.
- 1. Reaka-Kudla ML. The global biodiversity of coral reefs: a comparison with rain forests. Biodiversity II: Understanding and protecting our biological resources. 1997;2:551.
- 2. Moberg F, Folke C. Ecological goods and services of coral reef ecosystems. Ecological economics. 1999; 29(2):215–33.
- 3. Spurgeon JP. The economic valuation of coral reefs. Marine pollution bulletin. 1992; 24(11):529–36.
- 4. Hughes TP, Kerry JT, Baird AH, Connolly SR, Dietzel A, Eakin CM, et al. Global warming transforms coral reef assemblages. Nature. 2018; 556(7702):492–6. pmid:29670282
- 5. UNFCCC. Adoption of the Paris Agreement FCCC/CP/2015/L. 9/Rev. 1. 1. 2015. https://doi.org/10.1007/BF02327128
- 6. Pörtner HO, Roberts D, Masson-Delmotte V, Zhai P, Tignor M, Poloczanska E, et al. IPCC Special Report on the Ocean and Cryosphere in a Changing Climate. IPCC Intergovernmental Panel on Climate Change: Geneva, Switzerland. 2019.
- 7. De’ath G, Fabricius KE, Sweatman H, Puotinen M. The 27–year decline of coral cover on the Great Barrier Reef and its causes. Proceedings of the National Academy of Sciences. 2012; 109(44):17995–9.
- 8. De Bakker DM, Van Duyl FC, Bak RP, Nugues MM, Nieuwland G, Meesters EH. 40 Years of benthic community change on the Caribbean reefs of Curaçao and Bonaire: the rise of slimy cyanobacterial mats. Coral Reefs. 2017; 36(2):355–67.
- 9. Dudgeon SR, Aronson RB, Bruno JF, Precht WF. Phase shifts and stable states on coral reefs. Marine Ecology Progress Series. 2010; 413:201–16.
- 10. Eggertsen M, Chacin DH, Åkerlund C, Halling C, Berkström C. Contrasting distribution and foraging patterns of herbivorous and detritivorous fishes across multiple habitats in a tropical seascape. Marine biology. 2019; 166(4):51.
- 11. Heron SF, Maynard JA, Van Hooidonk R, Eakin CM. Warming trends and bleaching stress of the world’s coral reefs 1985–2012. Scientific reports. 2016; 6:38402. pmid:27922080
- 12. Obura DO. Kenya. Marine pollution bulletin. 2001; 42(12):1264–1278. pmid:11827110
- 13. Spalding M, Spalding MD, Ravilious C, Green EP. World atlas of coral reefs. Univ of California Press; 2001.
- 14. Obura DO. An Indian Ocean centre of origin revisited: Palaeogene and Neogene influences defining a biogeographic realm. Journal of Biogeography. 2016; 43(2):229–42.
- 15. Carbone F, Accordi G. The Indian Ocean coast of Somalia. Marine Pollution Bulletin. 2000; 41(1–6):141–59.
- 16. Obura D. The diversity and biogeography of Western Indian Ocean reef-building corals. PloS one. 2012; 7(9).
- 17. Gamoyo M, Obura D, Reason CJ. Estimating connectivity through larval dispersal in the Western Indian Ocean. Journal of Geophysical Research: Biogeosciences. 2019; 124(8):2446–59.
- 18. McClanahan TR, Obura DO. Status of Kenyan Coral Reefs. Coastal Management. 1995; 23:57–76.
- 19. McClanahan TR, Graham NA, Calnan JM, MacNeil MA. Toward pristine biomass: reef fish recovery in coral reef marine protected areas in Kenya. Ecological applications. 2007; 17(4):1055–67. pmid:17555218
- 20. Osuka K, Samoilys M, Mbugua J, de Leeuw J, Obura D. Marine habitats of the Lamu-Kiunga coast: an assessment of biodiversity value, threats and opportunities. ICRAF Working paper number no. 248 World Agroforestry Centre. https://doi.org/10.5716/WP16167
- 21. Church J, Obura DO. Sustaining Coral Reef Ecosystems and Their Fisheries in the Kiunga Marine National Reserve, Lamu, Kenya. International Coral Reef Symposium Okinawa Japan 890882. 2004; pg 22.
- 22. Darling ES, McClanahan TR, Côté IM. Combined effects of two stressors on Kenyan coral reefs are additive or antagonistic, not synergistic. Conservation Letters. 2010; 3(2):122–30.
- 23. McClanahan TR, Ateweberhan M, Omukoto J. Long-term changes in coral colony size distributions on Kenyan reefs under different management regimes and across the 1998 bleaching event. Marine Biology. 2008; 153(5):755–68.
- 24. McClanahan TR, Baird AH, Marshall PA, Toscano MA. Comparing bleaching and mortality responses of hard corals between southern Kenya and the Great Barrier Reef, Australia. Marine Pollution Bulletin. 2004; 48(3–4):327–35. pmid:14972585
- 25. McClanahan T, Muthiga N, Mangi S. Coral and algal changes after the 1998 coral bleaching: interaction with reef management and herbivores on Kenyan reefs. Coral reefs. 2001; 19(4):380–91.
- 26. McClanahan TR. Decadal coral community reassembly on an African fringing reef. Coral Reefs. 2014 Dec 1;33(4):939–50.
- 27. Samoilys MA, Osuka K, Maina GW, Obura DO. Artisanal fisheries on Kenya’s coral reefs: Decadal trends reveal management needs. Fisheries Research. 2017; 186:177–91.
- 28. Okemwa E. The Intergovernmental Oceanographic Commission of UNESCO and regional capacity building. Marine Policy. 1998; 22(3):197–207.
- 29. Obura D, Grimsditch G. Resilience assessment of coral reefs: assessment protocol for coral reefs, focusing on coral bleaching and thermal stress. Gland, Switzerland: IUCN; 2009.
- 30. Freile D, Milliman JD, Hillis L. Leeward bank margin Halimeda meadows and draperies and their sedimentary importance on the western Great Bahama Bank slope. Coral Reefs. 1995; 14(1):27–33.
- 31. Rees SA, Opdyke BN, Wilson PA, Henstock TJ. Significance of Halimeda bioherms to the global carbonate budget based on a geological sediment budget for the Northern Great Barrier Reef, Australia. Coral Reefs. 2007; 26(1):177–88.
- 32. Kohler KE, Gill SM. Coral Point Count with Excel extensions (CPCe): A Visual Basic program for the determination of coral and substrate coverage using random point count methodology. Computers & Geosciences. 2006; 32(9):1259–69.
- 33. Pante E, Dustan P. Getting to the point: Accuracy of point count in monitoring ecosystem change. Journal of Marine Biology. 2012; 2012.
- 34. Clarke KR, Gorley RN. Getting started with PRIMER v7. PRIMER-E: Plymouth, Plymouth Marine Laboratory. 2015:20.
- 35. Anderson M, Gorley R, Clarke KP. for PRIMER: guide to software and statistical methods. PRIMER-E, Plymouth, UK. 2008.
- 36. R Development Core Team RF. R: A language and environment for statistical computing. 2011; 409.
- 37. McClanahan TR, Ateweberhan M, Darling ES, Graham NA, Muthiga NA. Biogeography and change among regional coral communities across the Western Indian Ocean. PloS one. 2014; 9(4).
- 38. Gibson SY, Van der Marel RC, Starzomski BM. Climate change and conservation of leading‐edge peripheral populations. Conservation Biology. 2009; 23(6):1369–73. pmid:20078636
- 39. Riegl B. Corals in a non-reef setting in the southern Arabian Gulf (Dubai, UAE): fauna and community structure in response to recurring mass mortality. Coral Reefs. 1999; 18(1):63–73.
- 40. Harriott VJ, Harrison PL, Banks SA. The coral communities of Lord Howe Island. Marine and Freshwater Research. 1995; 46(2):457–65.
- 41. Adjeroud M, Wallace CC, Bosserelle P, Payri C, Menou JL, Pichon M. Reefs at the edge: coral community structure around Rapa, southernmost French Polynesia. Marine Ecology. 2016; 37(3):565–75.
- 42. Beger M, Sommer B, Harrison PL, Smith SD, Pandolfi JM. Conserving potential coral reef refuges at high latitudes. Diversity and distributions. 2014; 20(3):245–57.
- 43. Ateweberhan M, McClanahan TR. Partitioning scleractinian coral diversity across reef sites and regions in the Western Indian Ocean. Ecosphere. 2016; 7(5):e01243.
- 44. Jouffray JB, Nyström M, Norström AV, Williams ID, Wedding LM, Kittinger JN, et al. Identifying multiple coral reef regimes and their drivers across the Hawaiian archipelago. Philosophical Transactions of the Royal Society B: Biological Sciences. 2015; 370(1659):20130268.
- 45. Schutte VG, Selig ER, Bruno JF. Regional spatio-temporal trends in Caribbean coral reef benthic communities. Marine Ecology Progress Series. 2010; 402:115–22.
- 46. Ribas-Deulofeu L, Denis V, De Palmas S, Kuo CY, Hsieh HJ, Chen CA. Structure of benthic communities along the Taiwan latitudinal gradient. PloS one. 2016; 11(8).
- 47. Jokiel PL, Brown EK, Friedlander A, Rodgers SK, Smith WR. Hawai'i coral reef assessment and monitoring program: spatial patterns and temporal dynamics in reef coral communities. Pacific Science. 2004; 58(2):159–74.
- 48. Vroom PS, Braun CL. Benthic composition of a healthy subtropical reef: baseline species-level cover, with an emphasis on algae, in the Northwestern Hawaiian Islands. PloS one. 2010; 5(3).
- 49. Williams GJ, Smith JE, Conklin EJ, Gove JM, Sala E, Sandin SA. Benthic communities at two remote Pacific coral reefs: effects of reef habitat, depth, and wave energy gradients on spatial patterns. PeerJ. 2013;1:e81. pmid:23734341
- 50. Adjeroud M. Factors influencing spatial patterns on coral reefs around Moorea, French Polynesia. Marine Ecology Progress Series. 1997; 159:105–19.
- 51. Adjeroud M, Poisson E, Peignon C, Penin L, Kayal M. Spatial patterns and short-term changes of coral assemblages along a cross-shelf gradient in the Southwestern Lagoon of New Caledonia. Diversity. 2019; 11(2):21.
- 52. Fajemila OT, Langer MR, Lipps JH. Spatial patterns in the distribution, diversity and abundance of benthic foraminifera around Moorea (Society Archipelago, French Polynesia). PloS one. 2015; 10(12).
- 53. Pante E, Adjeroud M, Dustan P, Penin L, Schrimm M. Spatial patterns of benthic invertebrate assemblages within atoll lagoons: importance of habitat heterogeneity and considerations for marine protected area design in French Polynesia. Aquatic Living Resources. 2006; 19(3):207–17.
- 54. Penin L, Adjeroud M, Pratchett MS, Hughes TP. Spatial distribution of juvenile and adult corals around Moorea (French Polynesia): implications for population regulation. Bulletin of Marine Science. 2007; 80(2):379–89.
- 55. Hsieh HJ, Wen CK, Huang YC, Chen KS, Dai CF, Chen CA. Spatial Patterns and Environmental Settings of Non-reefal Coral Communities Across the Tropic of Cancer in the Penghu Archipelago (Pescadores), Taiwan. Zoological Studies. 2016; 55(45).
- 56. Teichberg M, Wild C, Bednarz VN, Kegler HF, Lukman M, Gärdes AA, et al. Spatio-temporal patterns in coral reef communities of the Spermonde Archipelago, 2012–2014, I: comprehensive reef monitoring of water and benthic indicators reflect changes in reef health. Frontiers in Marine Science. 2018; 5:33.
- 57. Francini-Filho RB, Coni EO, Meirelles PM, Amado-Filho GM, Thompson FL, Pereira-Filho GH, et al. Dynamics of coral reef benthic assemblages of the Abrolhos Bank, eastern Brazil: inferences on natural and anthropogenic drivers. PloS one. 2013; 8(1).
- 58. Obura DO. Status of coral reefs in Kiunga Marine reserve, Kenya. Coral reef degradation in the Indian ocean. 2002:47.
- 59. O'Leary JK, McClanahan TR. Trophic cascades result in large‐scale coralline algae loss through differential grazer effects. Ecology. 2010; 91(12):3584–97. pmid:21302830
- 60. Bak RP. The growth of coral colonies and the importance of crustose coralline algae and burrowing sponges in relation with carbonate accumulation. Netherlands Journal of Sea Research. 1976; 10(3):285–337.
- 61. O’Leary JK, Potts DC, Braga JC, McClanahan TR. Indirect consequences of fishing: reduction of coralline algae suppresses juvenile coral abundance. Coral Reefs. 2012; 31(2):547–59.
- 62. Smith TB, Maté JL, Gyory J. Thermal refuges and refugia for stony corals in the eastern tropical Pacific. InCoral Reefs of the Eastern Tropical Pacific 2017 (pp. 501–515). Springer, Dordrecht. https://doi.org/10.1007/978-94-017-7499-4
- 63. Abrego D, Willis BL, van Oppen MJ. Impact of light and temperature on the uptake of algal symbionts by coral juveniles. PLoS One. 2012; 7(11).
- 64. Grottoli AG, Wellington GM. Effect of light and zooplankton on skeletal δ13C values in the eastern Pacific corals Pavona clavus and Pavona gigantea. Coral Reefs. 1999;18(1): 29–41.
- 65. Wethey DS PORTER JW. Sun and shade differences in productivity of reef corals. Nature. 1976; 262(5566):281–2.
- 66. Kojis BL, Quinn NJ. Seasonal and depth variation in fecundity of Acropora palifera at two reefs in Papua New Guinea. Coral reefs. 1984; 3(3):165–72.
- 67. Obura D, Gudka M, Rabi FA, Gian SB, Bijoux J, Freed S, et al. Coral reef status report for the Western Indian Ocean (2017). In Nairobi Convention 2017. Global Coral Reef Monitoring Network (GCRMN)/International Coral Reef Initiative (ICRI).
- 68. McClanahan T, Maina J, Pet-Soede L. Effects of the 1998 coral morality event on Kenyan coral reefs and fisheries. AMBIO: A Journal of the Human Environment. 2002; 31(7):543–50.
- 69. Lambo AL, Ormond RF. Continued post-bleaching decline and changed benthic community of a Kenyan coral reef. Marine Pollution Bulletin. 2006; 52(12):1617–24. pmid:16962145
- 70. Barott K, Smith J, Dinsdale E, Hatay M, Sandin S, Rohwer F. Hyperspectral and physiological analyses of coral-algal interactions. PloS one. 2009; 4(11).
- 71. Littler MM, Littler DS, Brooks BL. Harmful algae on tropical coral reefs: bottom-up eutrophication and top-down herbivory. Harmful algae. 2006; 5(5):565–85.
- 72. McCook L, Jompa J, Diaz-Pulido G. Competition between corals and algae on coral reefs: a review of evidence and mechanisms. Coral reefs. 2001; 19(4):400–17.
- 73. Steneck RS, Dethier MN. A functional group approach to the structure of algal-dominated communities. Oikos. 1994: 476–98.
- 74. Smith JE, Brainard R, Carter A, Grillo S, Edwards C, Harris J, et al. Re-evaluating the health of coral reef communities: baselines and evidence for human impacts across the central Pacific. Proceedings of the Royal Society B: Biological Sciences. 2016; 283(1822):20151985. pmid:26740615
- 75. Jackson JB. Adaptation and diversity of reef corals. BioScience. 1991: 475–82.
- 76. Lirman D, Orlando B, Maciá S, Manzello D, Kaufman L, Biber P, et al. Coral communities of Biscayne Bay, Florida and adjacent offshore areas: diversity, abundance, distribution, and environmental correlates. Aquatic Conservation: marine and freshwater ecosystems. 2003; 13(2):121–35.
- 77. Pandolfi J. Coral community dynamics at multiple scales. Coral reefs. 2002; 21(1):13–23.
- 78. Done TJ. Patterns in the distribution of coral communities across the central Great Barrier Reef. Coral reefs. 1982; 1(2):95–107.
- 79. Coles SL, Brown BE. Coral bleaching—capacity for acclimatization and adaptation. Advances in marine biology. 2003; 46:183. pmid:14601413
- 80. Potts DC, Done TJ, Isdale PJ, Fisk DA. Dominance of a coral community by the genus Porites (Scleractinia). Marine Ecology Progress Series. 1985; 25:79–84.
- 81. Darling ES, Alvarez‐Filip L, Oliver TA, McClanahan TR, Côté IM. Evaluating life‐history strategies of reef corals from species traits. Ecology Letters. 2012; 15(12):1378–86. pmid:22938190
- 82. Obura DO. Environmental stress and life history strategies: A case study of corals and river sediments from Malindi, Kenya (Doctoral dissertation, ph. D. Thesis). University of Miami. 1995.
- 83. Obura DO. Can differential bleaching and mortality among coral species offer useful indicators for assessment and management of reefs under stress? Bulletin of Marine Science. 2001; 69(2):421–42.
- 84. Intergovernmental Panel on Climate Change. Global Warming of 1.5° C: An IPCC Special Report on the Impacts of Global Warming of 1.5° C Above Pre-industrial Levels and Related Global Greenhouse Gas Emission Pathways, in the Context of Strengthening the Global Response to the Threat of Climate Change, Sustainable Development, and Efforts to Eradicate Poverty. Intergovernmental Panel on Climate Change; 2018.
- 85. Van Woesik R, Houk P, Isechal AL, Idechong JW, Victor S, Golbuu Y. Climate‐change refugia in the sheltered bays of Palau: analogs of future reefs. Ecology and evolution. 2012; 2(10):2474–84. pmid:23145333
- 86. Beyer HL, Kennedy EV, Beger M, Chen CA, Cinner JE, Darling ES, et al. Risk‐sensitive planning for conserving coral reefs under rapid climate change. Conservation Letters. 2018; 11(6):e12587.
- 87. McClanahan TR, Obura DO. Status of Kenyan coral reefs, Coastal Management, 2015; 23(1):57–76.
- 88. Obura DO. Resilience and climate change: lessons from coral reefs and bleaching in the Western Indian Ocean. Estuarine, Coastal and Shelf Science. 2005; 63(3):353–72.
- 89. West JM, Salm RV. Resistance and resilience to coral bleaching: implications for coral reef conservation and management. Conservation Biology. 2003; 17(4):956–67.
- 90. Guzman HM, Benfield S, Breedy O, Mair JM. Broadening reef protection across the Marine Conservation Corridor of the Eastern Tropical Pacific: distribution and diversity of reefs in Las Perlas Archipelago, Panama. Environmental Conservation. 2008; 35(1):46–54.
- 91. Ortiz-Lozano L, Pérez-España H, Granados-Barba A, González-Gándara C, Gutiérrez-Velázquez A, Martos J. The Reef Corridor of the Southwest Gulf of Mexico: Challenges for its management and conservation. Ocean & Coastal Management. 2013; 86:22–32.
- 92. Nyström M, Graham NA, Lokrantz J, Norström AV. Capturing the cornerstones of coral reef resilience: linking theory to practice. Coral Reefs. 2008; 1;27(4):795–809.
- 93. van Nes EH, Scheffer M. Implications of spatial heterogeneity for catastrophic regime shifts in ecosystems. Ecology. 2005; 86(7):1797–807.