Reefs’ backlight: Identifying coral reef bright spots with the fossil record
Lepore, others, O’Dea
2017-01-07
- 1 Preliminary
- 2 Abstract
- 3 Introduction
- 4 Methods
- 5 Results
- 6 Discussion and Conclusion
- 6.1 Diversity patterns
- 6.2 Abundance patterns
- 6.3 Something else
- 6.4 Conclusion
- 6.5 Discussion and Conclusion
- 6.6 Conclusion
- 6.6.1
The first paragraph tells readers the overall messages they should take away. With each sentence, it makes a clear, direct statement. - 6.6.2
The second paragraph broadens the above conclusion by putting the achievement into perspective. It compares the work presented to related work in other domanins, indicates the work's significance and anticipates on future results.
- 6.6.1
- 7 Figures
- 8 Tables
- 9 Supplemental material
- References
1 Preliminary
1.1 Note to myself
To remove table tags that would not knit to word use replace “\(???)[^\)]+)” by “”.
2 Abstract
Why did you do it? (50 words)Bright spots are, among coral reefs, “substantially better than expected given the environmental conditions and socioeconomic drivers they are exposed to” (xxxref Cinner). A recently developed model can provide insight about what caused bright spots to be different if the such causes are detectable at present; but if the causes acted in times past, bright spots may still be identifyed but what drove their difference remains uncertain. What did you do? (50 words)Here we present a case study from Bocas del Toro, Caribbean Panama. We examined time-averaged communities of reef coral from two time intervals: the mid-Holocene (around 7,000 years ago) and the subrecent (last several years-decades). We tested for statistically significant differences in community structure through time, over millennial timescales, and among sites within each time interval–which were exposed to natural and anthropogenic pressures that are practically difficult or impossible to indistinguish. What happened [when you did that]? (50 words)We found that the difference in community structure of reef corals from Bocas del Toro was statistically significant between subrecent and mid-Holocene reefs and also among sites within each time intervalmz05. Most importantly, we identified Punta Caracol reef as a bright spot; its community structure was similar to pristine whereas all other subrecent reefs were significantly degraded. What do the results mean in theory? (50 words) This key finding means that bright spots may result from environmental drivers that modern ecology cannot reveal. If some driver acted upon a number of communities of reef coral in times past only, and a bright spot emerged, the ecological difference among the reefs may persist even if all reefs experience identical conditions at present. Bright spots that emerge and persist in such particular scenario may be thought about as having an advantage. What do the results mean in practice? (50 words)In practice, if we understand that the occurrence of bright spots does not necesarily mean that we can identify and replicate what caused them to be so, stakeholders will expect outputs from management plans that are more realistic. What is the key benefit for readers? (25 words)This study provides stakeholders working in Bocas del Toro a robust reference against they can measure how far from a pristine coral reefs are in this region. And it should help stakeholders elsewhere to understand better why it is important to incorporate the fossil record into their ecological studies. What remains unresolved? (no word limit)An imortant next step is to incorporate the lessons learned from the fosisl record into bright spotes models. This models should substantially improve if the expected scenario against which they identify bright spots recognises that some of the variability that coral reefs display today may have past causes.
3 Introduction
…
4 Methods
4.1 Coral reefs in Bocas del Toro, humans and impact
4.1.1 Introduction
The distribution, shape and life of coral reef ecosystems worldwide result from the interplay of multiple processes–including geological, climatic, oceanographic and ecological processes. After coral reefs had existed for millions of years, humans evolved and, only recently, human impact became a significant driver of coral reefs’ decline. This worldwide issue is perhaps most urgent in the Caribbean, and most notable from its latest consequences.
Since the 1970s, coral reefs in the Caribbean changed dramatically. The cover of live corals decreased and that of macroalgae increased in unprecedented manner (Jackson et al. 2014, xxxref Aronson and Precht 1997, Greenstein et al. 1998, cited in Fredston-Hermann et al. (2013)). This is clearly anomalous (xxxref Silvia Earle, 1972) and indicates that reefs have severely declined (Jackson et al. 2014). This degradation followed the rocketing growth of human population, tourism and shipping, via the overexplotation of natural resources and the spread of disease (Jackson et al. 2014).
The most likely mechanisms and most evident proximate causes have been recently reviewed and are summarized next (Jackson et al. 2014 and references thereafter). Between mid 1970s and early 1980s, enormous shipping through the Panama Canal facilitated alien species, including disease, to invade the Caribbean sea. White band disease infected and massively killed Acropora corals. An unidentified disease also infected and massively killed Diadema sea urchins. Diadema and parrot fish are both herbivores that critically regulate the balance between macroalgae and corals in reefs. Because humans had overexploited parrot fish, when Diadema died-off, macroalgae bloomed and reefs–originally dominated by corals–became dominated by macroalgae. This phase-shift peaked by mid 1990s and continues today. Recently, greater overfishing, pollution, numbers of tourists and of extreme warming events have worsen coral reefs.
To better understand the consequences of human impact on coral reefs now and in the future we need more historical context. A good place to focus is Bocas del Toro, in Caribbean Panama.
4.1.2 Setting
Isla Colon is the North-East limit of Bahia Almirante and one of seven islands in Bocas del Toro Archipelago is locaed at the North-West of Panama, in the Caribbean sea (8°30-9°40 N, 82°56-81°18 W, xxxfig. map). Here, Coral reefs are patchy or fringing, and ocurr in environments that range from off-shore (typically, with sandy beaches or reef flats) to lagoonal (typically, with fringing mangroves and extensive seagrass beds)(xxxref check Guzman et al. 2005).
The geomorphology of the archipelago and its biodiversity originated over the past several million years. About 20 Ma ago (before the isthmus of Panama formed) the Bocas del Toro Basin was a deep tropical marine bed (xxxref Coates and Jackson 1998, cited in Guzman et al. 2005). Volcanoes were active between 16-10 Ma and deposited igneous and sedimentary rocks in many parts of the archipelago. When volcanic activity ceased, marine transgressions deposited a sedimentary sequence near the coast, adequate for (a) diverse and abundant marine fauna to settle, (b) new species of reef coral and benthic foraminifers to originate, and (c) reefs to form (xxxref Coates and Jackson 1998, Collins et al. 1996, cited in Guzman et al. 2005). Sea level changed substantially through time and affected the separation between islands and the mainland, and the bathymetry and topography of the archipelago. In the mid-Holocene, sea level stabilized. As a result, the archipelago´s shape settled and coral reefs developed (Summers et al. 1997, cited in Guzman et al. 2005, Fredston-Hermann et al. (2013)).
The continental shelf of Bocas del Toro is narrow and the maximum depth along the coast ranges 20-50 m (Guzman et al. 2005). Surf and tides are more important outside than inside the archipelago. Tides may be mixed, semidiurnal or diurnal and their amplitude ranges below 0.5 m (xxxref Glynn 1972 cited in Guzman et al. (2005)). Winds that more strongly influence the archipelago blow from the North and Northeast; Their effect on surf and tides is reduced by the northern islands and reefs (DMA 1988, cited in Guzman et al. 2005).
The climate is typical of the wet Caribbean. Precipitation is 330 cm annually (Cubit et al. 1989, Kaufmann and Thompson 2005, cited in Cramer 2013). The dry and rainy seasons are unclearly defined. Generally, rainfall is lower in March and between September-October and higher in July and December (Kaufman and Thompson, cited in Guzman et al. 2005).
The vegetation on the mainland is mainly banana plantations, clear-felled cattle lands and rainforest; On the islands, it is mainly lowland wet tropical-forest (xxxref Carruthers et al. 2005, xxxCramer et al. 2012, xxxref Cramer 2013 cited in Schlöder, O’Dea, and Guzman 2013).
The lagoon of Bahia Almirante receives little oceanic flushing and high runoff. Although it is open to the ocean, this lagoon is semi-enclosed and receives creeks that drain the floodplain of the Changuinola River, through extensive banana plantations; This resuls in high input of sediments and pollutants (Guzmán and Jimenez 1992, Guzmán and Garcia 2002, Guzmán 2003, cited in Cramer 2013). Additionally, a mountain chain runs paralell to the mainland coast of the lagoon (xxx IGNTG 1988 cited in Guzman et al. 2005); And because the distance between the mountains and coast is short (> 3.5 km), river flow after the rains increases immediately (Guzman et al. 2005). Consequently, the hyposaline runoff forms a lens of superficial water that is around 0.5 m thick and rich in suspended organic material.
4.1.3 Human ocuppation and impact
Although human impact in Bocas del Toro has rocketed only recently, humans have likely affected land and sea most of the time since coral reefs here initiated develpoment (Cramer 2013). Since around 10000 years ago, humans have cleared the land and fished marine coasts in Central America (Cramer 2013); By then, they had already occupied the Pacific coast of Panama (Fredston-Hermann et al. 2013). In Bocas del Toro, however, the earliest hunters and gatherers were recorded around 5000 years ago (Ranere and Cooke 1991, Cooke 2005, cited in Fredston-Hermann et al. 2013) and the earliest human settlements and fishing activities, around 1500 years ago (Wake et al. 2004, 2013, cited in Fredston-Hermann et al. 2013).
When the Spanish arrived to America, approximately 500 years ago, human population in Bocas del Toro had depleted the largest animals from mangroves, seagrass meadows, and coral reefs (Wing and Wing 2001, Pandolfi et al. 2003, cited in Cramer 2013). European contact resulted in a catastrophic mortality of indigenous peoples(Cramer 2013); Their population recovered pre-contact levels in the 1800s, but it was distributed sparsely until 1900s, because the orography and climate of Bocas del Toro was challenging for human settlement and agriculture (Cramer 2013).
What followed was rapid environmental change and had the greatest impact on ecosystems in Bocas del Toro. Since 1915, mainland and islands in Bocas del Toro have been deforested to grow and export bananas. Channels have been dredged and ships have been departing from Almirante port, crossing Bahia Almirante and exiting into the Caribbean sea through Bocas del Drago (xxxref Greb et al. 1996, cited in Seemann et al. 2014). Cumulatively, deforestation, dredging and shipping has increased erosion, sediments, nutrients and pollution (xxxref Berry et al. 2013, xxxref Burke et al. 2004, , cited in Seemann et al. 2014). Since 1993, population and tourism have strikingly increased (xxxref Guerrón-Montero 2005, cited in Seemann et al. 2014), coasts have extensively developed, destructive fishing methods have been used and marine resources have been overfished. Consequently, over the past few decades ecosystems have substantially changed (xxxref Saric 2005, cited in Seemann et al. 2014).
4.1.4 Striking ecological shifts
Since around 2 Ma until recently, fringing and lagoonal reefs in Bocas del Toro were dominated by Acropora cervicornis (Lamarck, 1816) and Acropora palmata (Lamarck, 1816). Thus, these species dominated most of the time since the modern reefs initiated development, in the mid-Holocene (Aronson et al. 2004, Klaus et al. 2011, Cramer et al. 2012, Cramer 2013, Fredston-Hermann et al. 2013, cited in Schlöder, O’Dea, and Guzman 2013). For example, A. cervicornis, and also Montastraea “annularis” [Ellis and Solander, 1786] and Porites furcata, dominated Lennond reef (in Isla Colon, Bahia Almirante) during the mid-Holocene (Fredston-Hermann et al. 2013). Acropora spp. and Porites spp. corals also dominated reefs in Bocas del Toro during the late-Holocene (~4000-0 years ago, xxxref) (Aronson et al. 2004, Cramer et al. 2012, cited in Fredston-Hermann et al. 2013). Today, Porites remains abundant and Agaricia tenuifolia (Dana, 1846) became important too [xxxref Guzman and Guevara 1998]. Acropora corals, however, have declined catastrophically (Cramer et al. 2012, cited in Fredston-Hermann et al. 2013).
The recent ecological shift from Acropora corals, which love light, to Porites and Agaricia corals, which resist turbid-water better, suggests that changes in water clarity induced by human impacts (mainly land clearing and dredging) were implicated (Rogers 1990, Cramer et al. 2012).
4.1.5 Ecology
Most of what we know about the ecology of coral reefs in Bahia Almirante comes from studies around the beginning of the 21st century (Héctor M. Guzman and Guevara 1998, Guzman et al. (2005); For other reefs in Bocas del Toro see Hector M. Guzman and Guevara 1998, Guzman and Guevara (1999), Guzman and Guevara (2001)). Then, coral reefs developed to a maximum depth of 23 m; Lower, light is insufficient for corals to grow (Héctor M. Guzman and Guevara 1998). Of all the species of reef coral found in Panama (33), 53% occurred here; Coral cover ranged 20-50% and averaged 35% (Guzman and Guevara 1999, Héctor M. Guzman and Guevara (1998)). Porites furcata (90% cover) dominated down to 2 m depth and Agaricia tenuifolia (40% cover) from 2-6 m. The depth range from 6-15 m was composed by A. tenuifolia, Madracis mirabilis, and Siderastrea siderea (15%, 7% and 5%, respectively). S. siderea and coralline algae dominated deeper waters. Sponges where the second most important sessile organism and were found in all reefs. Facing Bahia Almirante and along the leeward coast of Isla Colon, reefs form an almost continuous barrier (Héctor M. Guzman and Guevara 1998). On a reef around the center of the island, coral and algae cover was 26.9% and 21.4%; Richness of reef coral species was 17 (Guzman et al. 2005).
More recently, between 2010-2011, cover and richness of reef corals in Bahia Almirante was <10% and 7 species (Seemann et al. 2014). Porites furcata was the dominant reef coral; Adult and carnivorous fish species were absent (Seemann et al. 2014).
Multiple environmental factors cummulatively impact coral reefs’ ecology in Bahia Almirante. Here, runoff from the Changuinola river is particularly important; Its plumes enter the bay through an inlet at Boca del Drago (Seemann et al. 2014). Worldwide, coral bleaching in September 2010 killed reef corals massively. Bocas del Toro was no exception: Mortality and degradation of reef ecosystems was particularly extensive (NOAA 2010, cited in Seemann et al. (2014)). Here, not only sea temperatures increased but also salinity lowered; Seasonal rain fall was high and oxygen depleted because water exchange was limited (Kaufmann and Thompson 2005, cited in Seemann et al. (2014)).
4.1.6 Resilience
Once they degrade severely, reefs ecosystems can struggle to recover. Four reefs from Bocas del Toro were experimentally degraded; After two years, their fate was unclear. Some recovered to pre-disturbance condition; others did not (Schlöder, O’Dea, and Guzman 2013). The impact on biodiversity may not be immediate. Although dead dead coral rubble in Bocas del Toro can host sessile infauna, biodiversity in a degrading reef will eventually lower because given enough time the three-dimensional structure of the reef should flatten (Nelson, Kuempel, and Altieri 2016). Moreover, simple (flat) reef sructure is unfavors reefs’ recovery after coral bleaching (Graham et al. 2015). Because Acropora corals have historically been the greatest builders of structurally complex reefs worldwide, and because they are almost extinct in the Caribbean, reefs’ resilience in Bocas del Toro is seriously compromised.
4.2 Sampling
The sampled area was southwest coast of Isla Colon, in Bocas del Toro, Caribbean Panama (figure nf12). Here, multiple subrecent reefs (with living corals) and a large track of a mid-Holocene reef occurr close to one another (xxx-xxx m appart?). This proximity ensured that the hydrological conditions accross the sampled reefs were comparable—the reefs shared similar zonation, sedimentological conditions and topography (Fredston-Hermann et al. 2013).
4.2.1 Subrecent reefs
Coral communities from subrecent reefs were sampled from the reef benthos between 1.7-12 m underwater on SCUBA. Approximately 10 cm of the rubble that makes up the reef matrix was manually excavated from around living corals (figure yz14). Three to five bulk samples (~10 kg each) were collected at random along one transect laid parallell to the reef margin from one site at each of five reefs (22 subrecent samples in total; figure bs04). The separation between reefs and samples within reefs ranged xxx-xxx m and xxx-xxx m, respectively.
Because they covered only the surface of the reefs, subrecent communiteis represented reef corals that lived recently—relaive to the accretion history of the reef. Yet, subrecent communities captured more than a snapshot of the latest assemblages of reef corals; they captured ecological patterns that are pervasive over a time interval over which the reef built a substantial part of its three-dimenisonal structure.
4.2.2 Fossil site
Coral communities from the mid-Holocene were collected from a fossil reef near Lennond Village, in Isla Colon, Bocas del Toro. This fossil reef was accessible for sampling during construction work in 2006-2008 and was inundated after that. We manually collected 2-5 bulk samples (~10 kg each) in-situ at random at xxx-xxx m below current sea level (~xxx-xxx m below sea level during the mid-Holocene; xxxref) from the walls of trenches excavated with an excavating machine at each of four sites (16 mid-Holocene samples in totalanalysis np01). Samples within trenches and sites within the mid-Holocene reef track were separated by xxx-xxx m and xxx-xxx m, respectively (figure yz14).
4.2.3 Age of fossil site
…
4.3 Samples Processing
The samples were processed in the same way, regardless they came from mid-Holocene or the subrecent reefs. They were transported after collection to the the marine lab at Smithsonian Tropical Research Institure (STRI) and kept at room temperature. Each sample was then washed and sieved and the coral skeletons greater than 2 mm were identified to the lowest possible taxonomic level–generally to species or genus level–and weighted. Scleractinian corals were the focus of this study but Millepora spp. were also included because in Bocas del Toro (xxxdiscusswithAaron and in Lake Enriquillo?) they notably build reefs. A reference collection was built and is available at STRI to help this and future studies identify corals consistently (xxxref abstractt to APANAC conference).
To ensure that all samples were identified with a similar level of expertise, we identified corals in two rounds. In round one we identified most corals but we were conservative and classified any coral which identity was unclear as unknonw. As we identified more samples our ability to identify corals improved. In round two we reviewed all unknown corals from round one to standarize how we identified corals to our maximum level of expertise. This revision was make by a single observer–the member of our team who had identified the largest number of the samples.
…
4.4 Analysis
4.4.1 Diversity
To test if patterns of diversity were significantly different between communities we analysed the variance in taxonomic richness (S) and Shannon–Wiener diversity (H’) between time intervals. H’ approximated the normal distribytion, but S did not, except during the mid-Holocene.
`xxx S/H' remained not normal after log-transformation. Brown and Forsythe (xxx 1974) tests of homogeneity of variance indicated heteroscedasticity for xxx S/H' xxx and/but not H'/S.` Based on these preliminary results we used parametric (t test or ANOVA) or non-parametric (Wilcoxon or Kruskal-Wallis rank sum) tests to analyse S and H’ between time intervals and among sites within each time interval (table rg11).
To perform multiple comparisons with adjusted P values we used the "TukeyHSD" and "kruskalmc" functions ["stats" and "pgirmess" software packages, respectively (xxx Giraudoux, 2014; R Core Team, 2014)].4.4.2 Variation in community structure based on taxonomic data
…
4.4.2.1 Test differences in taxonomic composition
To test for differences in community structure of reef-building corals between subrecent and mid-Holocene reefs we proceded as follows.
First, to limit our analysis to corals that we confidently identified, we excluded from each sample the corals that could not identify (because they were poorly preserved).
Second, to reduce the influence of dominana taxa, the skeletal mass of each coral species in each sample was first square root transformed. Second, to allow comparison between samples of different total weight, we converted the abundance of each taxa in each sample to relative abundance (xxx check that the order is right). Relative abundances were then used to calculate the Bray-Curtis (BC) dissimilarity among any two samples.
From my thesis (needs rewording):A permutational multivariate analysis of variance (xxxref Anderson, 2001a) was used to test for significant differences in dissimilarities among subrecent and mid-Holocene samples (function “adonis”, package “vegan”). This is a non-parametric analysis that partitions dissimilarities among sources of variation, fits a linear model and outputs an “ANOVA-like” summary based on some number of random permutations of the data, which we set to 10,000.
4.4.2.2 Is dispersion confounding differences in community structure?
When significant differences in community composition among time intervals occurred, we tested whether such differences were confounded with dispersion within time intervals (xxxref Anderson, 2001b). To measure the dispersion within time intervals, we calculated the average distance of community dissimilarities within each time interval to the centroid of that time interval in multivariate space. We then used ANOVA to test if the dispersions of one or more time intervals were different (implemented utilising the “betadisper” function).
4.4.2.3 Is water depth confounding differences in community structure?
…
4.4.2.4 Correlation between MDS and relative abundance of taxa
To visualise the general pattern of the Bray-Curtis dissimilarities among samples we used non-metric multidimensional scaling (MDS) ordination [“isoMDS” function, “stats” software package (R Core Team, 2014)]. To identify the coral taxa that principally discriminated communities through time, the correlation between their relative abundance and the MDS ordination of Bray-Curtis dissimilarities was tested for significance (“envfit” function); significant correlations were represented as vectors overlaid on the MDS ordination–for simplicity, represented were only the species that made up to xxx95%? of the total relative abundance.
4.4.3 Variation in community structure based on functional data
…
5 Results
5.1 General patterns through time
The difference in Shannon-Wiener diversity and richness was statistically not-significant between time intervals (table dv00) and also among mid-Holocene sites but was statistically not-significant among subrecent sites (table dl08). Among subrecent sites, pairwise contrasts revealed that, after adjusting for multiple p values, differences in richness but not in Shannon-Wiener diversity were statistically significant between: (a) Airport point 2 vs. Airport point 1, (b) Casa Blanca vs. Airport point 2 and (c) Airport point 2 vs. Punta Caracol (table wh08).
The most common taxa, those that occurred in 50% or more mid-Holocene and subrecent communities, totalled six and eight; Respectively, they represented 30% and 50% of all found taxa (figure ef22; analyses ef23) and comprised 85% and 96% of the coral abundance (figure ef22; analyses ef23 and bt01). Mean richness was 7.938 in mid-Holocene communities and 7.455 in subrecent communities, thus, the most common taxa contributed to around 75% and 100% of the local richness (figure ao00). Most other taxa occurred in substantially fewer communities: 79% and 62% of the other taxa were present in less than 25% of the communities from mid-Holocene and subrecent reefs(analysis kn02).
Acropora cervicornis was the most abundant taxa across mid-Holocene reefs (analysis bt00), and in each of 81% (13/16) of the mid-Holocene communities but in only 23% (5/22) of the subrecent communities(analysis bt02). Branching Porites was the most abundant taxa across subrecent reefs, closely followed by Agaricia spp.(analysis bt00), and in each of 50% (11/22) of the subrecent communities but in only 6% (1/16) of the mid-Holocene communities. And Agaricia spp. was the most abundant taxa in each of 27% (6/22) of the subrecent communities but in none of the mid-Holocene communities (figure ef22; analysis bt01). The greatest changes through time in relative abundance were that Acropora cervicornis decreased by 26% and Agaricia spp. and branching Porites increased by 21% and 17% (table nk07). All other taxa changed through time 5% or less.
Map _A. cervicornis_ in the subrecent reefs
Note that _A. cervicornis_ was common but little abundant, which is surprising given that it is not commonly found in living reefs.
- Is it in a many samples of a few sites?
- Is it in a few samples of all sites?The MDS ordination of BC dissimilarities showed that overall separation was clear between mid-Holocene and subrecent communities (figure td11). The difference in community structure through time was statistically significant (table ez03) and a betadispersion test indicated that this result was not an atrifact of differences in disperison between time intervals (table eu02). The result was the same excluding Acropora cervicornis (before calculating relative abundance) (tables sj08, tu01; figure td11_yx18).
5.2 Punta Caracol: An exceptional subrecent site
Communities from the subrecent site Punta Caracol contrasted with all other subrecent communities and generally likened mid-Holocene ones. Here, the abundance of (a) Acropora cervicornis was highest, (b) Millepore spp. was considerably high, (c) Agaricia spp. and branching Porites were lowest (figures pe12 and dy02; analysis rq12). (In mid-Holocene sites, the mean abundance of Agaricia spp. and branching Porites was about 6% and 15%, i.e. close to the relatively low abundance of these taxa in Punta Caracol(analysis rq14).). These communities grouped separately from subrecent and within mid-Holocene ones (figure td11); When Acropora cervicornis was excluded, this pattern was less striking (figure yx18).
(This section may be excluded from the paper.)
## Environmental drivers
### Depth versus community structure
When the relationship is (visually) examined between depth and the MDS ordination (betadiversity) of modern and fossil samples, the correlation seems significant for subrecent but not for mid-Holocene communities (figure `xo15` and `xo16`).
In contrast, when that relationship is (visually) examined not form the MDS ordination but from a linear model relating depth and each component of the MDS ordination, taken independently, the relationship seems significant for both subrecent and mid-Holocene communities, although for only component 2--not component 1--of the MDS ordination (fig. 1.2 and fig. 2.2 at https://bookdown.org/maurolepore/mds_depth/xo17.html). The residuals of the model seem to be not-correlated with depth; This means that if a pattern exists, the model captured it; there is no significant pattern left.
Therefore, examination of the MDS or the linear model sugests opposite results: the relationship between depth and community structure changed through time when interpreted from the MDS but remained unchanged when interpreted from linear models.
The interpretation from linear models, suggests that the effect of depth in affecting community structure predates the recend advent of substantial human impact; i.e. it is natural.
The interpretation from the MDS suggests that the effect of depth in affecting community structure became important since later than the mid-Holocene. Likely, human impact caused the biggest change by darkening the water so the variation of community structure along the depth gradient changed recently. But this could in part have started in the late-Holocene, prior to human impact. ENSO increased frequency and severity of storms, resulting in more runoff and water more turbid, which in turn changed the depth range of coral communities.
Future research should focus on capturing more communities from the late-Holocene. This was partly done by Aronson, but their description was probabbly too coarse for ecological analysis. And they sampled destructively, so there is no way to back to the samples and re-do the analyses. Cores should run perperndicular to the coast to capture variation in community structure along the depth profile, similarly to what this study did with subrecent samples. This will also describe the geomorphology of the reefs in more detail. When did it hit sea level? How fast did it prograde? Can we use Katie Cramer's cores? What do they look like in terms of accretion rates? When did they hit sea level? if ever? We need to core the reef flat too.6 Discussion and Conclusion
6.1 Diversity patterns
[By Aaron; moved from results]Mid-Holocene reef assemblages had greater number of taxa than modern communities on the whole (20 vs. 16), but mid-Holocene assemblages were not more diverse than modern (Table dv00). Variation in richness and diversity was high amongst sites of both ages (Figure bs04).
[By Aaron; moved from results]Reef coral communities from Punta Caracol differed from other subrecent communities and were more similar to mid-Holocene communities with the most abundant coral being Acropora cervicornis, high abundances of Millepora spp and low abundances of Agaricia spp. and branching Porites (figure pe12, also analysis rq12).
6.2 Abundance patterns
The commonest taxa in mid-Holocene communities were Acropora cervicornis and Agaricia spp. while in subrecent communities Agaricia spp. and branching Porites, dominated, occurring in 100% of the samples (analysis ef24).
6.3 Something else
Water depth significantly correlated with betadiversity in subrecent coral communities xo15.
6.4 Conclusion
6.5 Discussion and Conclusion
6.6 Conclusion
6.6.1 The first paragraph tells readers the overall messages they should take away. With each sentence, it makes a clear, direct statement.
Over the past few decades, almost all leeward reefs from Isla Colon have been structured differently compared with an ecological baseline from the mid-Holocene (approximately 7000 years ago). Baseline communities of reef corals were dominated by Acropora cervicornis; subrecent communities have been dominated by Porites spp. and Aaricia spp. except at Punta Caracol; It has remained indistinguishible from the baseline.
7 Figures
Figure nf12: Map of Bocas del Toro, Caribbean Panama, showing the location of mid-Holocene (~7 000 years old) and subrecent reef sites.
Figure nf12: Sites sampled in mid-Holocene (~7 000 years old) and subrecent reef sites, Isla Colon, Bocas del Toro region, Caribbean Panama.
Figure gc22: Draft stratigraphic and biotic descriptions of trenches dug into the mid-Holocene Sweet Bocas reef, Bocas del Toro.
Figure yz14: Methods used to collect bulk samples from a mid-Holocene (a-e) and subrecent (f) reefs. Construction work (a) exposed a fossil reef in Bocas del Toro, Panama, where bulk samples (b, c) were collected in situ. Many coral colonies were in life position (d, e). Subrecent reefs near the fossil reef were sampled on SCUBA (f).
Figure bs04: Accumulation curves of reef corals in mid-Holocene and subrecent reef sites from Bocas del Toro.
Figure kt06: Reef corals from Bocas del Toro, ranked by their relative abundance within two time intervals: the mid-Holocene and subrecent.
Figure td11_yx18: Ordination of Bray-Curtis dissimilarities of subrecent and mid-Holocene communities of reef corals from Bocas del Toro, Panama. Named vectors indicate taxa that make up to 95% of the total relative abundance. The lenght of each vector represents the strengh of the correlation between the relative abundance of each taxa and the ordination of Bray-Curtis dissimilarities. The data is excludes unknown taxa (toptd11 and bottomyx18) and Acropora cervicornis (bottomyx18). (Unknown taxa and Acropora cervicornis were removed before the relative abundance of each taxa was calculatedsee definition of function wrangle.)
8 Tables
xxxinsert eq07
Figure eq07: Ages of fossil samples from Bocas del Toro.
| metric | method | alternative | estimate | statistic | p.value |
|---|---|---|---|---|---|
| shannon | Welch Two Sample t-test | two.sided | -0.038 | -0.356 | 0.724 |
| richness | Wilcoxon rank sum test with continuity correction | two.sided | NA | 201.500 | 0.455 |
| metric | time_interval | method | term | parameter | df | statistic | sumsq | meansq | p.value |
|---|---|---|---|---|---|---|---|---|---|
| richness | subrecent | Kruskal-Wallis rank sum test | NA | 4 | NA | 9.663 | NA | NA | 0.047 |
| richness | mid-Holocene | Anova | g | NA | 3 | 2.295 | 27.321 | 9.107 | 0.130 |
| richness | mid-Holocene | Anova | Residuals | NA | 12 | NA | 47.617 | 3.968 | NA |
| shannon | subrecent | Anova | g | NA | 4 | 3.357 | 0.952 | 0.238 | 0.034 |
| shannon | subrecent | Anova | Residuals | NA | 17 | NA | 1.206 | 0.071 | NA |
| shannon | mid-Holocene | Anova | g | NA | 3 | 0.182 | 0.068 | 0.023 | 0.907 |
| shannon | mid-Holocene | Anova | Residuals | NA | 12 | NA | 1.501 | 0.125 | NA |
| metric | comparison | estimate | conf.low | conf.high | adj.p.value |
|---|---|---|---|---|---|
| shannon | Punta Caracol-Casa Blanca | 0.525 | 0.013 | 1.038 | 0.043 |
| richness | Airport point 2-Airport point 1 | 4.050 | 0.146 | 7.954 | 0.040 |
| richness | Casa Blanca-Airport point 2 | -4.400 | -8.081 | -0.719 | 0.015 |
| richness | Punta Caracol-Airport point 2 | -4.600 | -8.281 | -0.919 | 0.011 |
| term | df | SumsOfSqs | MeanSqs | F.Model | R2 | p.value |
|---|---|---|---|---|---|---|
| time_interval | 1 | 1.5648 | 1.5648 | 12.0337 | 0.2505 | 1e-04 |
| Residuals | 36 | 4.6812 | 0.1300 | NA | 0.7495 | NA |
| Total | 37 | 6.2459 | NA | NA | 1.0000 | NA |
| term | df | sumsq | meansq | statistic | p.value |
|---|---|---|---|---|---|
| Groups | 1 | 0.0398 | 0.0398 | 2.6224 | 0.1141 |
| Residuals | 36 | 0.5458 | 0.0152 | NA | NA |
| term | df | SumsOfSqs | MeanSqs | F.Model | R2 | p.value |
|---|---|---|---|---|---|---|
| time_interval | 1 | 1.2853 | 1.2853 | 2.7805 | 0.0107 | 2e-04 |
| Residuals | 257 | 118.7976 | 0.4622 | NA | 0.9893 | NA |
| Total | 258 | 120.0829 | NA | NA | 1.0000 | NA |
| term | df | sumsq | meansq | statistic | p.value |
|---|---|---|---|---|---|
| Groups | 1 | 0.0000 | 0e+00 | 3e-04 | 0.9864 |
| Residuals | 257 | 0.2193 | 9e-04 | NA | NA |
9 Supplemental material
9.1 Supplemental figures
Figure kt04: Reef corals from Bocas del Toro, ranked by their relative abundance in each sample from mid-Holocene and subrecent sites, after aplying transformation \(\log(1 + x)\)."
Figure pe12: Reef corals from Bocas del Toro, ranked by their relative abundance in each sample from mid-Holocene and subrecent sites. Bars without errorbars corresond to taxa that occurred on only one sample.
Figure ef22: Count of communities (samples) in which different taxa occurred in mid-Holocene and subrecent reefs from Bocas del Toro. Taxa above the dotted line occurred in 50% or more communities.
Figure dy02: Taxa in subrecent communities from Bocas del Toro, in three invervals of water-depth (cut points at 5 m and 8.5 m), ranked by percent relative abundance. Genrally, different taxa occurr in about the same abundance accross sites within each time interval. Notable exceptions are that the substantially increased abundance of branching Porites at the mid-Holocene site Sweet Bocas-9 and of Millepora spp. at the subrecent site Punta Caracol.
Figure nq01: Relative abundance of each taxa in mid-Holocene and subrecent communities compared accross sites. The
Figure ao00: Richness in each of subrecent and mid-Holocene sites from Bocas del Toro, Panama.
Figure az42: Colonies of Acropora palmata from a mid-Holocene reef in Bocas del Toro.
Figure xo15_xo16: Water depth and community structure are related in subrecent (topxo15) but not in mid-Holocene (bottomxo15) reefs.
9.2 Supplemental tables
| metric | time_interval | statistic | p.value |
|---|---|---|---|
| shannon | subrecent, mid-Holocene | 0.968 | 0.334 |
| shannon | subrecent | 0.955 | 0.387 |
| shannon | mid-Holocene | 0.913 | 0.131 |
| richness | subrecent, mid-Holocene | 0.915 | 0.007 |
| richness | subrecent | 0.900 | 0.029 |
| richness | mid-Holocene | 0.929 | 0.237 |
| species | cumsum_pcnt |
|---|---|
| Acropora_cervicornis | 30.461 |
| Porites_branching | 53.992 |
| Agaricia_spp | 70.517 |
| Millepora_spp | 80.880 |
| Favia_fragum | 83.929 |
| Orbicella_spp | 86.881 |
| Porites_non_branching | 89.656 |
| Madracis_spp | 91.839 |
| Acropora_palmata | 93.904 |
| Manicina_areolata | 95.062 |
| Eusmilia_fastigiata | 96.209 |
| Leptoseris_cucullata | 97.258 |
| Oculina_spp | 98.257 |
| Mussa_angulosa | 98.795 |
| Cladocora_arbuscula | 99.089 |
| Colpophyllia_natans | 99.350 |
| Pseudodiploria_spp | 99.579 |
| Montastraea_cavernosa | 99.744 |
| Solenastrea_bournoni | 99.905 |
| Scolymia_cubensis | 99.967 |
| Acropora_prolifera | 100.000 |
| species | mid_holocene | subrecent | diff_pcnt | change |
|---|---|---|---|---|
| Acropora_cervicornis | 43.441 | 17.481 | 25.960 | decrease |
| Agaricia_spp | 5.960 | 27.091 | 21.130 | increase |
| Porites_branching | 15.220 | 31.841 | 16.621 | increase |
| Orbicella_spp | 5.474 | 0.429 | 5.044 | decrease |
| Madracis_spp | 0.281 | 4.085 | 3.804 | increase |
| Millepora_spp | 12.230 | 8.497 | 3.733 | decrease |
| Porites_non_branching | 4.555 | 0.995 | 3.559 | decrease |
| Leptoseris_cucullata | 0.081 | 2.016 | 1.935 | increase |
| Oculina_spp | 0.061 | 1.937 | 1.876 | increase |
| Mussa_angulosa | 1.044 | 0.032 | 1.012 | decrease |
| Manicina_areolata | 1.517 | 0.799 | 0.718 | decrease |
| Favia_fragum | 3.380 | 2.719 | 0.661 | decrease |
| Eusmilia_fastigiata | 0.850 | 1.443 | 0.593 | increase |
| Cladocora_arbuscula | 0.325 | 0.263 | 0.062 | decrease |
| Scolymia_cubensis | 0.076 | 0.049 | 0.027 | decrease |
| Acropora_palmata | 4.131 | NA | NA | NA |
| Acropora_prolifera | 0.065 | NA | NA | NA |
| Colpophyllia_natans | 0.521 | NA | NA | NA |
| Montastraea_cavernosa | 0.329 | NA | NA | NA |
| Pseudodiploria_spp | 0.459 | NA | NA | NA |
| time_interval | acrocervi_agarspp | acrocervi_poribran | agarspp_poribran |
|---|---|---|---|
| mid_holocene | 7.288 | 2.854 | 0.392 |
| subrecent | 0.645 | 0.549 | 0.851 |
| trench | collected_from | description |
|---|---|---|
| 1 | Sweet Bocas-2 | Used. |
| 2 | Sweet Bocas-3 | Used. |
| 3 | Sweet Bocas-4 | No corals in samples. |
| 4 | Sweet Bocas-5 | Used. |
| 5 | Sweet Bocas-6 | Used. |
| 6 | Sweet Bocas-7 | No corals in samples. |
| 7 | Sweet Bocas-8 | Not comparable. Not from the core of the reef. |
| 8 | Sweet Bocas-9 | Not comparable. Not from the core of the reef. |
References
Cramer, Katie L. 2013. “History of Human Occupation and Environmental Change in Western and Central Caribbean Panama.” Bulletin of Marine Science 89 (4). University of Miami-Rosenstiel School of Marine; Atmospheric Science: 955–82.
Cramer, Katie L., Jeremy B. C. Jackson, Christopher V. Angioletti, Jill Leonard-Pingel, and Thomas P. Guilderson. 2012. “Anthropogenic Mortality on Coral Reefs in Caribbean Panama Predates Coral Disease and Bleaching.” Journal Article. Ecology Letters 15 (6): 561–67. doi:10.1111/j.1461-0248.2012.01768.x.
Fredston-Hermann, A. L., A. O’Dea, F. Rodriguez, W. G. Thompson, and J. A. Todd. 2013. “Marked Ecological Shifts in Seagrass and Reef Molluscan Communities Since the Mid-Holocene in the Southwestern Caribbean.” Journal Article. Bulletin of Marine Science 89 (4): 983–1002. doi:10.5343/bms.2012.1077.
Graham, Nicholas A. J., Simon Jennings, M. Aaron MacNeil, David Mouillot, and Shaun K. Wilson. 2015. “Predicting Climate-Driven Regime Shifts Versus Rebound Potential in Coral Reefs.” Journal Article. Nature 518 (7537): 94–97. doi:10.1038/nature14140.
Guzman, Hector M, and Carlos A Guevara. 1999. “Arrecifes Coralinos de Bocas Del Toro, Panamá: III. Distribución, Estructura, Diversidad Y Estado de Conservación de Los Arrecifes de Las Islas Pastores, Cristóbal, Popa Y Cayo Agua.” Journal Article. Revista de Biología Tropical 47: 659–76. http://www.scielo.sa.cr/scielo.php?script=sci_arttext&pid=S0034-77441999000400002&nrm=iso.
———. 2001. “Arrecifes Coralinos de Bocas Del Toro, Panamá: IV. Distribución, Estructura Y Estado de Conservación de Los Arrecifes Continentales de Península Valiente.” Journal Article. Revista de Biología Tropical 49 (1): 53–66.
Guzman, Hector M, Penelope AG Barnes, Catherine E Lovelock, and Ilka C Feller. 2005. “A Site Description of the Caricomp Mangrove, Seagrass and Coral Reef Sites in Bocas Del Toro, Panama.” Journal Article. Caribbean Journal of Science 41 (3): 430–40.
Guzman, Hector M., and Carlos A. Guevara. 1998. “Arrecifes Coralinos de Bocas Del Toro, Panamá: II. Distribución, Estructura Y Estado de Conservación de Los Arrecifes de Las Islas Bastimentos, Solarte, Carenero Y Colón.” Journal Article. Revista de Biología Tropical 46: 889–912. http://www.scielo.sa.cr/scielo.php?script=sci_arttext&pid=S0034-77441998000400002&nrm=iso.
Guzman, Héctor M., and Carlos A. Guevara. 1998. “Arrecifes Coralinos de Bocas Del Toro, Panamá: Distribución, Estructura Y Estado de Conservación de Los Arrecifes Continentales de La Laguna de Chiriquí Y La Bahía Almirante.” Journal Article. Revista de Biología Tropical 46: 601–23. http://www.scielo.sa.cr/scielo.php?script=sci_arttext&pid=S0034-77441998000300014&nrm=iso.
Jackson, Jeremy, Mary Donovan, Katie Cramer, and Vivian Lam. 2014. “Status and Trends of Caribbean Coral Reefs: 1970-2012.” Report. Global Coral Reef Monitoring Network.
Nelson, Hannah R., Caitlin D. Kuempel, and Andrew H. Altieri. 2016. “The Resilience of Reef Invertebrate Biodiversity to Coral Mortality.” Journal Article. Ecosphere 7 (7): n/a–n/a. doi:10.1002/ecs2.1399.
Schlöder, Carmen, Aaron O’Dea, and Hector M Guzman. 2013. “Benthic Community Recovery from Small-Scale Damage on Marginal CariBBean Reefs: An Example from Panama.” Journal Article. Bulletin of Marine Science 89 (4): 1003–13.
Seemann, Janina, Cindy T González, Rodrigo Carballo-Bolaños, Kathryn Berry, Georg A Heiss, Ulrich Struck, and Reinhold R Leinfelder. 2014. “Assessing the Ecological Effects of Human Impacts on Coral Reefs in Bocas Del Toro, Panama.” Journal Article. Environmental Monitoring and Assessment 186 (3): 1747–63.