AN IN SITU COMPARISON OF BENTHIC
PRIMARY PRODUCERS ON A CARIBBEAN
CORAL REEF
THE IMPORTANCE OF GORGONIANS AND CALCIFYING ALGAE
ByAmber K.T. Riley
Bachelor of Biology University of Amsterdam
August 2020
Examinator: dhr. dr. ir. J.M. de Goeij Supervisor: mr. N.A. Kornder MSc
2
ABSTRACT
Coral reefs are highly valuable ecosystems, yet these ecosystems face worldwide degradation through global climate change and other (local) anthropogenic stressors. Understanding how these ecosystems will change in the future is therefore crucial. Efforts to understand reef cycling using ecological modeling are key, since altered cycling could explain how different dominant benthic primary producers affect the surrounding reefs. One of the ways benthic primary producers differ is in their primary production. Net photosynthesis (Pn) rates give indications of primary productions on reefs and are often used in modelling studies. Pn rates change throughout the day in response to light availability, peaking during midday. Researchers tend to use the maximum Pn rate in models, excluding daily light variation and the effect this has on the Pn rate. But models without a parameter for light variation could overestimate primary production. Modelling studies have also tended to focus on hard corals and fleshy algae, although other benthic primary producers, such as gorgonians and calcifying algae, are also very abundant on reefs. This study aimed to quantify the relationship of Pn rates and light for common dominant benthic primary producers on a Caribbean coral reef ecosystem (Curaçao, Lesser Antilles) while also comparing Pn rates of less researched functional groups to more commonly researched functional groups. In situ incubation chamber were used, where dissolved oxygen (DO) and photosynthetically active radiation were measured. Regression relationships between Pn and light were linear for all species. Gorgonians were found to photosynthesize at similar rates compared to hard corals, while only one calcifying algae species (crustose coralline algae (CCA)) was found to have similar Pn rates as one of the fleshy algae species (Lobophora spp.). This finding in combination with the high abundance of gorgonians and CCA on reefs suggests that these species notably contribute to carbon assimilation on Curaçao’s reefs.
3
Table of Contents
INTRODUCTION ... 4
MATERIALS AND METHODS ... 6
Study site ... 6
Organisms ... 6
In situ incubations ... 7
DATA PROCESSING AND DERIVED VARIABLES ... 9
Regression analysis ... 9 ImageJ ... 9 Pn comparison ... 10 RESULTS ... 11 Regression analysis ... 11 Pn comparison ... 12 DISCUSSION ... 14 Regression analysis ... 14 Pn comparisons ... 14 CONCLUSION ... 15 ACKNOWLEDGEMENTS ... 15 REFERENCES ... 16 APPENDIX 1 ... 19 APPENDIX 2 ... 28 APPENDIX 3 ... 32 APPENDIX 4 ... 33 APPENDIX 5 ... 35 DATA REPOSITORY ... 36
4
INTRODUCTION
Coral reefs face worldwide degradation through the impact of global climate change and other (local) anthropogenic stressors. Degradation of coral reefs is often associated with changes in reef
community composition (Norstrom, et al., 2009). Phase shifts, defined as changes from one
ecological state to another, have been documented worldwide (Done, 1992; Norstrom, et al., 2009). Phase shifts on coral reefs are usually typified by reefs changing from a coral-dominated state to an algae-dominated state (McManus & Polsenberg, 2004). Because coral reef ecosystems have a very high cultural and economic value, there has been a lot of effort to understand how these ecosystems will change in the future (Cesar, et al., 2003; Woodhead, et al., 2019). One of these efforts involves the assessments of nutrient and carbon cycling on coral reefs, since altered nutrient and carbon cycling may explain how changes in the dominant benthic primary producers influence the surrounding reef (de Goeij, et al., 2013; Haas, et al., 2016; Pawlik, et al., 2019).
Benthic primary producers influence reef cycling in different ways. An example is how corals and algae differ greatly in the amount of bioavailable dissolved organic matter (DOM) they produce, with algae releasing higher amounts of DOM than coral (Haas, et al., 2010; Haas, et al., 2011; Nelson, et al., 2013; Haas, et al., 2016). Shifts in the dominant benthic primary producer can therefore change the bioavailable DOM, which supports the growth of heterotrophic bacteria. Since heterotrophic bacteria respire more than autotrophic bacteria, increased growth of heterotrophs can kill corals by creating prolonged hypoxia zones (Haas et al., 2016). Heterotrophic bacteria have also been linked to pathogenic disease outbreaks on corals. The excess DOM exuded by benthic algae has also been shown to enrich virulence factors in microbial communities, harming and/or killing corals (Nelson et al., 2013; Cardenas et al., 2017). As corals die, more space is created for algae, resulting in a positive feedback loop called DDAM (DOC, disease, algae, microorganisms) (Haas et al., 2016). These shifts in microbial communities (‘microbial phase shift’) can mediate coral reef phase shifts (Silveira et al 2017).
Another way benthic primary producers influence reef cycling is through differences in (primary) productivity (Hatcher, 1988). Net photosynthetic (Pn) rates of benthic primary producers indicate overall reef productivity. The transformation of light into energy determines reef biomass and growth, since most dominant benthic organisms are primary producers (coral and algae) (de Bakker, et al., 2017; Hatcher, 1988).
Modelling studies usually only use Pn rates obtained around midday, when Pn rates peak due to high irradiance. However, it is often the case that coral reef ecosystem models exclude the effect of light on Pn rates, usually using maximum Pn rates and excluding daily light variation (Grigg, et al., 1984; Polovina, 1984; Arias-Gonzalez, et al., 1997; Mongin & Baird, 2014; Alva-Basurto & Arias-Gonzalez, 2014). But research shows there are significant differences in Pn rates depending on irradiance (Anthony & Hoegh-Guldberg, 2003; Langdon & Atkinson, 2005; Ramsby, et al., 2014; Sawall &
5 Hochberg, 2018), which could cause overestimation of the productivity of coral reefs. Studies
showing the relationship between photosynthesis and light could be used to calibrate ecosystem models by including a parameter for irradiance.
A way to better understand the influence of (anthropogenic) stressors on the reef ecosystem is ecological modelling, specifically mass-balanced models, in which models assume that the ecosystem modeled is in a steady state for each of the living groups, implying that inputs equal outputs (Walters, et al., 1997; Holmes & Johnstone, 2010; Han, et al., 2017). However, modelling studies have tended to focus on carbon and nutrient cycling of hard corals and fleshy algae (i.e. macroalgae and turf algae) (Tanaka, et al., 2008; Haas, et al., 2010; Haas, et al., 2011; Haas, et al., 2016; Mueller, et al., 2016), although other benthic primary producers, such as gorgonians and calcifying algae, are among the most abundant organisms on coral reefs (Inoue, et al., 2013; Quinlan, et al., 2019). There are studies that show that crustose coralline algae (CCA), a group of calcifying algae, have a higher contribution to organic production than previously thought (Chisholm. 2003), indicating that it is likely that these understudied primary producers are underrepresented in studies and models about coral reef cycling.
Further, few studies make direct comparisons between Pn rates of commonly researched benthic groups (hard corals and fleshy algae) and less researched benthic groups (gorgonians and calcifying algae) within the same coral reef (Wanders, 1976). Therefore, the aim of this study was to quantify the relationship of Pn rates and light for common dominant benthic primary producers on a Caribbean coral reef ecosystem. Additionally, this study also aimed to compare Pn rates of gorgonians and calcifying algae to hard corals and fleshy algae respectively. To do this, we measured dissolved oxygen (DO) and photosynthetically active radiation (PAR) during in situ incubations. In situ incubations were chosen because they allow for more accurate Pn rate measurements, by causing less stress to organisms being measured and mimicking environmental conditions better than conventional methods. The inclusion of the effect irradiance has on Pn rates and understudied functional groups can make models of coral reef ecosystems more accurate.
6
MATERIALS AND METHODS
Study site
Research was conducted on the Caribbean island of Curacao, in November and December of 2019. Incubations were done on the reef slope in front of the CARMABI Research station, situated at the SW side of the island at the entrance of Piscadera Bay. The incubation site was at a depth of 10 meters and consisted of an old sunken fishing boat where the incubations were set up through the duration of the whole experiment (Figure 4). Organisms used in the incubations were either taken from the reef slope in front of the research station (hard corals, fleshy algae, calcifying algae) or Boei 1 (gorgonians) (Figure 1B).
Figure 1. Geographical locations. (A) Location of Curacao in the Caribbean. Source: Mueller (2014). (B)
Curacao and an overview of CARMABI research station, Piscadera Bay, incubation site and Boei 1. Adapted from: Wanders (1976).
Organisms
In situ incubations consisted of hard corals (Colphophyllia natans (n=3) and Madracis mirabilis (n=4)), gorgonians (Eunicea spp. (n=3) and Plexaura spp. (n=3)), fleshy algae (Dictyota spp. (n=3) and
Lobophora spp. (n=3)), calcifying algae (CCA (n=3) and Peyssonnelia spp. (n=3)) and controls (n=4). Hard
corals were collected from the CARMABI reef slope with a hammer and chisel and stored on a ‘coral tree’ platform (Figure 2A) directly next to our incubation site for acclimation. Gorgonians were collected at Boei 1 by chipping the organisms off the reef framework with a hammer and chisel. Gorgonians were lifted to the boat using incubation chambers, loaded in a cooler and transported to CARMABI research station. The gorgonians were then immediately outplanted next to the incubation site in reef crevices by using marine epoxy to make a base for the gorgonians and subsequently lodging this base into small holes of the reef framework (Figure 2B). Substrata not covered by coral was
7 scrubbed with a toothbrush to remove other organisms. Specimen were acclimated for approximately 1 week before incubations. After incubations, coral species were outplanted back onto the reef.
Figure 2. (A) coral tree. (B) gorgonian garden.
Fleshy algae were carefully collected by hand around the incubation site immediately before incubation and calcifying algae were collected with a knife from the reef slope. Algae were gently shaken in gloved hands or scrubbed with a toothbrush to remove other organisms preceding incubations. After use in incubations, algae species (except CCA) were kept in a -20°C freezer until freeze dried (Coolsafe, Scanvac; LaboGene). These samples were then later milled, combusted at 550°C and weighed at the University of Amsterdam for ash free dry weight (AFWD) estimations. AFDW of coral species and CCA was determined by taking scaled photographs and analyzing them in ImageJ.
In situ incubations
Organisms were incubated for approximately 4 hours during peak daylight hours in incubation chambers of either 3L or 6L (Figure 3). All incubation chamber instruments (except the stirrers and loggers) were kept in acid baths between incubations. Rubber parts were taken out of the acid baths and dried after 1 hour.
8
Figure 3. In situ incubation chamber set-up.
For accurate DO measurements, all air bubbles were removed from the incubation chambers by fanning water onto the equipment with gloved hands (cylinders, plates, DO logger and stirrer) and sucking water through the tubing and syringes once the incubation site was reached. Measurements of DO and PAR were performed continuously throughout the incubations, at 1-minute intervals. DO was measured by a HOBO U26 Dissolved Oxygen Data Logger (Onset Computer Corporation, US) which was attached to the lids of the incubation chambers before each incubation. PAR was assessed by an Odyssey Integrating PAR sensor (Dataflow Systems PTY Limited, NZ) lodged in the middle of the incubation chambers on the platform (Figure 4B).
Figure 4. Incubations site and set-up. (A) Incubation site: an old sunken fishing boat. Incubation
equipment was carried to and from the incubation site using a weighted Albert Heijn basket. (B) Incubation chambers were set up on the bottom of the fishing boat.
9
DATA PROCESSING AND DERIVED VARIABLES
Regression analysis
DO data were extracted as DO concentrations in mg/L from the logger with HOBOware Pro software, using a U-DTW-1 HOBO waterproof shuttle for offloading. PAR data were extracted as Integrating light (IL) in µmol photons m-2 s-1 from the logger with Odyssey Data Logging Software. All DO and PAR data were analyzed for each incubation in Rstudio (v1.3.959). DO and PAR data were plotted over time (minutes) for each incubation. Each graph was analyzed for slope breakpoints, which is when a change in slope occurs, within the DO data using the segmented package (v1.2-0) (Muggeo, 2008). These slope breakpoints divided the DO and PAR data into segments with differing DO slopes, for which the corresponding PAR means were calculated. Segments that were 10 minutes or shorter were not analyzed. The trendline and PAR mean per segment were plotted in a graph, together with the raw DO and PAR data.
In all incubation experiments, rates of change per segment were calculated by dividing the difference between start and end DO concentrations (∆[O2]) by the segment duration (∆t). The average control (∆[O2]/∆t) was subtracted from each segment and this was subsequently corrected for by the number of liters in the incubation chamber to get absolute fluxes in µmol O2 gAFDW-1 h-1. These fluxes were then divided by the AFDW in grams of the incubated organism and subsequently converted to hours and µmol O2 to get the final Pn rates in µmol O2 g AFDW-1 h-1 (Appendix 2). The Pn rates were plotted against IL per incubation organism. Either a linear model or a polynomial model was fitted to the Pn data, depending on (in order) which model had the lowest adjusted R2, Akaike Information Criterion and number of factors explaining the relationship.
ImageJ
For hard coral species and CCA the surface area was determined with freehand selection in ImageJ (v1.52a). For gorgonians, the length and diameter of each tube was determined with straight lines in ImageJ and from this the surface area per tube was calculated. The surface area of all tubes for one gorgonian were added up to get the total surface area.
surface area = 2𝜋rh + 2𝜋r2
AFDW for all organisms (except fleshy algae species and Peyssonnelia spp.) were subsequently determined by using previously generated allometric conversions (Appendix 3).
10
Pn comparison
Assumptions of the data were tested (Residuals VS Fitted, Normal Q-Q, Scale-Location, Residuals VS Leverage, Variance Inflation Factors, Overdispersion test) and a log transformation was needed for Pn data to get a normal distribution. A two-way ANOVA test was done to compare Pn between species. A post-hoc analysis was done with Tukey’s test, with the p-value adjusted for multiple comparisons. A p-value lower than 0.05 was considered significant.
11
RESULTS
Regression analysis
Per species several Pn rates were extracted in relation to light: Colphophyllia natans (n = 12), Madracis
mirabilis (n = 18), Plexaura spp. (n = 13), Eunicea spp. (n = 15), Dictyota spp. (n = 11), Lobophora spp.
(n = 12), Crustose Coralline Algae (n = 5) and Peyssonnelia spp. (n = 8) (Appendix 1 and 2).
Figure 4. Dissolved oxygen (DO) and photosynthetically active radiation (PAR) measured during a ~4 hour incubation. This example graph is the incubation of the first Madracis mirabilis individual. This
graph was made for each incubation (n=29).
A linear model for Pn-IL graphs was chosen for all species (Appendix 4). Pn linear regression formulas and R2 are displayed in Table 1, graphs with the regression analysis are shown in Appendix 4. Coral species had data points between the 50-450 µmol photons m-2 s-1. In contrast Dictyota spp. had no data points lower than 250 µmol photons m-2 s-1, Lobophora spp. had no data points between the 50-300 µmol photons 2 s-1 and calcifying algae had no data points lower than 200 µmol photons m-2 s-1 and higher than 400 µmol photons m-m-2 s-1.
12
Table 1. Summary of regression formulas and R2 per species. Y stands for net photosynthesis (Pn) in µmol O2 gAFDW-1 h-1 and x stands for integrating light (IL) in µmol photons m-2 s-1.
Group Species Formula R2
Corals Colpophyllia natans 𝑦 = −5.8 + 0.094𝑥 0.58
Madracis mirabilis 𝑦 = −6 + 0.11𝑥 0.56
Eunicea spp. 𝑦 = −1.5 + 0.069𝑥 0.18
Plexaura spp. 𝑦 = −2.5 + 0.093𝑥 0.09
Algae Dictyota spp. 𝑦 = 92 + 0.68𝑥 0.49
Lobophora spp. 𝑦 = −21 + 0.68𝑥 0.85
Crustose Coralline Algae 𝑦 = −210 + 1.1𝑥 0.91
Peyssonnelia spp. 𝑦 = 0.34 + 0.17𝑥 0.45
P
ncomparison
Pn and IL means are displayed in Table 2. Dictyota spp. had the highest Pn rates, followed (in order) by CCA, Peyssonnelia spp., Lobophora spp., Madracis mirabilis, Colpophyllia natans and Eunicea spp.
Table 2. Summary of the average net photosynthesis (Pn) and integrating light (IL) per species. Standard error included.
Group Species Mean net photosynthesis (µmol O2 gAFDW-1 h-1)
Mean integrating light (µmol photons m-2 s-1)
Corals Colpophyllia natans 20.23 ± 4.05 277 ± 33
Madracis mirabilis 22.86 ± 4.59 358 ± 31
Eunicea spp. 17.75 ± 4.30 278 ± 26
Plexaura spp. 21.46 ± 4.93 261 ± 16
Algae Dictyota spp. 352.69 ± 31.01 385 ± 32
Lobophora spp. 168.57 ± 50.87 278 ± 69
Crustose Coralline Algae 102.32 ± 21.19 279 ± 18
Peyssonnelia spp. 49.02 ± 4.93 285 ± 19
There were significant differences between Pn rates of species (ANOVA, p < 0.05) (Figure 5). P-values for specific species-species comparisons can be found in Appendix 5. Algal photosynthesis was significantly higher than coral photosynthesis (Tukey post-hoc, p<0.05). Dictyota spp. had the highest Pn rates, followed by Lobophora spp. and CCA. Pn was higher in fleshy algae compared to calcifying algae (Tukey post-hoc, p<0.05), with the only exception being CCA in comparison to Lobophora spp.
13 (Tukey post-hoc, p>0.05) (Figure 5B). In contrast, no significant difference was found between hard corals and gorgonians (Tukey post-hoc, p>0.05) (Figure 5A).
Figure 5. Boxplot of net photosynthetic (Pn) rates for all species. Algal photosynthesis is 10x higher than coral photosynthesis. Significance is indicated with the compact letter display. A different letter means species have significantly different Pn rates (p<0.05). The same letter means species do not have significantly different Pn rates. (A) Hard corals (Colpophyllia natans and Madracis mirabilis) and gorgonians (Eunicea spp. and Plexaura spp.). (B) Fleshy algae (Dictyota spp. and Lobophora spp.) and calcifying algae (crustose coralline algae (CCA) and Peyssonnelia spp.).
A
14
DISCUSSION
This research focused on finding the relationship between natural light variations and Pn rates of different coral and algae species. In addition, it aimed to shed light on less researched dominant benthic primary producers, such as gorgonians and calcifying algae, and how they compare to more commonly researched primary producers such as hard corals and fleshy algae respectively. Linear regressions were found among species to explain the relationship between Pn and light. Gorgonians were found to be in the same range of Pn rates as hard corals. In contrast calcifying algae were not found to be in the same range as fleshy algae, with the only exception being that CCA was found to be in the same Pn range as Lobophora spp.
Regression analysis
It is common to find a linear fit for Pn-IL curves within natural light ranges (Wanders, 1976; Anthony & Hoegh-Guldberg, 2003). Under 400 to 500 µmol photons m-2 s-1, Pn-IL curves tend to stay linear (Anthony & Hoegh-Guldberg, 2003; Langdon & Atkinson, 2005; Ramsby, et al., 2014; Sawall & Hochberg, 2018). Pn-IL curves had varying ranges of IL due to either (absence of) rain or cloud cover. Especially the regression fit for CCA may be inaccurate compared to previous studies due to the lack of Pn data points for lower IL values in this study (Wanders, 1976).
Later incubations of gorgonians show very low Pn rates. Pn rates of the third incubation are visibly lower than the first incubation. This is also the case for Plexaura spp., where in the second incubation Pn rates were also visibly lower, causing the regression slope to be less steep. This might have been caused by environmental factors, as the coral reef in front of the CARMABI research station is relatively degraded compared to Boei 1 and other reefs around the island, which could have caused stress to soft corals that were acclimatizing longer due to unsuitable environmental conditions (de Bakker, et al., 2017). Added to this is that incubations were done during the islands rainy season and that the reef slope used for incubations is at the entrance of a bay with major (nutrient) runoff, which could have also caused stress (Mueller et al., 2014).
P
ncomparisons
The Pn rates found for corals and algae in this study were comparable to previous studies (Wanders, 1976; Ramsby, et al., 2014; Sawall & Hochberg, 2018). CCA Pn rates were in the same range as Pn rates of Lobophora spp. Previous studies have shown that CCA is a common dominant benthic primary producer on Caribbean coral reefs (Chisholm, 2003). High Pn rates and high abundance could indicate that CCA notably contributes to carbon assimilation on reefs, also indicated by previous research (van der Heijden & Kamenos, 2015).
15
Peysonnellia spp. had relatively low Pn rates compared to other algae. Peyssonnelia spp. collected was usually growing in crevices of the reef framework and/or shaded areas. Taking
Peyssonnelia spp. substrata from these shaded areas to full light in situ incubation chambers probably
induced stress, which could have in turn caused lower Pn rates.
Pn rates of gorgonians were comparable to hard coral Pn rates. Gorgonians are also one of the most abundant non-reef building species on Caribbean coral reefs (Johnson & Hallock, 2020), which indicates that they are an important part of overall reef cycling. Recent research has also suggested that gorgonians have an increased resistance to elevated temperatures, ocean acidification, and nutrient enrichment compared to hard corals (Goulet, et al., 2017; Tsounis, et al., 2018). It is predicted that gorgonians will likely become more common throughout the Caribbean with further hard coral declines (Inoue, et al., 2013; Tsounis & Edmunds, 2017). This indicates the importance of including gorgonians in (future) modelling efforts.
CONCLUSION
This research shows how light affects Pn rates of common coral and algae species. With evidence for a linear relationship to irradiance, failure to take daily light variation into account may lead to overestimation of primary production due to peak daylight hours having (on average) a higher amount of irradiance compared to the rest of the day. It also gave an indication on how much gorgonians and calcifying algae contribute to overall reef carbon assimilation. Data collected in this study can be used in future modelling studies.
Future research should focus on apprehending Pn rates (Pn-IL curves), respiration rates (to approximate gross photosynthesis) and DOM release of less researched functional groups and species to include in nutrient and carbon cycling of coral reef ecosystems. Including less researched functional groups, like gorgonians and calcifying algae, will give a better representation of overall reef cycling.
Curacao’s coral reefs are valued at more than $445 million per year through their support to the tourism and fishing industry alone (WAITT Institute, 2017). With this economic and cultural importance in mind, coral reef ecosystem modelling is essential in the way forward to better mitigate mass coral ecosystem die off (Pandolfi, et al., 2003).
ACKNOWLEDGEMENTS
I would like to thank Niklas Kornder and Jasper de Goeij for their guidance, insights and knowledge. I would also like to acknowledge Peter Roessingh and Florian van der Steen for their help with my data processing. Lastly, I would like to thank CARMABI research station for their hospitality and the use of their labs and offices.
16
REFERENCES
Alva-Basurto, J. C. & Arias-Gonzalez, J. E., 2014. Modelling the effects of climate change on a Caribbean coral reef food web. Ecological Modelling, Volume 289, pp. 1-14.
Anthony, K. R. N. & Hoegh-Guldberg, O., 2003. Variation in coral photosynthesis, respiration and growth characteristics in contrasting light microhabitats: an analogue to plants in forest gaps and understoreys?.
Functional Ecology, Volume 17, pp. 246-259.
Arias-Gonzalez, J. E., Delesalle, B. & Galzin, R., 1997. Trophic functioning of the Tiahura reef sector, Moorea Island, French Polynesia. Coral Reefs, Volume 16, pp. 231-246.
Cesar, H., Burke, L. & Pet-Soede, L., 2003. The economics of worldwide reef degradation, Arnhem: Cesar Environmental Economics Consulting.
Chisholm, J. R., 2003. Primary productivity of reef-building crustose coralline algae. Limnology and
Oceanography, 48(4), pp. 1376-1387.
de Bakker, D. M. et al., 2017. 40 years of benthic community change on the Caribbean reefs of Curacao and Bonaire: the rise of slimy cyanobacterial mats. Coral Reefs, Volume 36, pp. 355-367.
de Bakker, D. M. et al., 2017. 40 Years of benthic community change on the Caribbean reefs of Curacao and Bonaire: the rise of the slimy cyanobacterial mats. Coral Reefs, Volume 36, pp. 355-367.
de Goeij, J. M. et al., 2013. Surviving in a marine desert: The sponge loop retains resources within coral reefs. Science, pp. 108-110.
den Haan, J. et al., 2016. Nitrogen and phosphorus uptake rates of different species from a coral reef community after a nutrient pulse. Scientific Reports, Volume 6.
Done, T. J., 1992. Phase shifts in coral reef communities and their ecological significance. Hydrobiologia, pp. 121-132.
Goulet, T. L., Shirur, K. P., Ramsby, B. D. & Iglesias-Prieto, R., 2017. The effects of elevated seawater temperatures on Caribbean gorgonian corals and their algal symbionts, Symbiodinium spp.. Plos One, 12(2).
Grigg, R. W., Polovina, J. J. & Atkinson, M. J., 1984. Model of a coral reef ecosystem. Coral Reefs, Volume 3, pp. 23-27.
Haas, A. F. et al., 2016. Global microbialization of coral reefs. Nature Microbiology, Volume 1.
Haas, A. F. et al., 2010. Organic matter release by the dominant primary producers in a Caribbean reef lagoon: implication for in situ O2 availability. Marine Ecology Progress Series, Volume 409, pp. 27-39. Haas, A. F. et al., 2011. Effects of coral reef benthic primary producers on dissolved organic carbon and microbial acitivity. Plos One, 6(11).
Han, D., Xue, Y., Zhang, C. & Ren, Y., 2017. A mass balanced model of trophic structure and energy flows of a semi-closed marine ecosystem. Acta Oceanologica Sinica, Volume 36, pp. 60-69.
17 Hatcher, B. G., 1988. Coral reef primary productivity: A beggar's banquet. Trends in Ecology & Evolution, 3(5), pp. 106-111.
Holmes, G. & Johnstone, R. W., 2010. Modelling coral reef ecosystems with limited observational data.
Ecological Modelling, Volume 221, pp. 1173-1183.
Inoue, S., Kayanne, H., Yamamoto, S. & Kurihara, H., 2013. Spatial community shift from hard to soft corals in acidified water. PANGAEA, 3(7), pp. 638-687.
Institute, W., 2017. Marine scientific assessment: The state of Curacao's coral reefs, s.l.: s.n.
Johnson, S. K. & Hallock, P., 2020. A review of symbiotic gorgonian research in the western Atlantic and Caribbean with recommendations for future work. Coral Reefs, Volume 39, pp. 239-258.
Kornder, N., n.d. in prep.
Langdon, C. & Atkinson, M. J., 2005. Effects of elevated pCO2 on photosynthesis and calcification of corals and interactions with seasonal change in temperature/irradiance and nutrient enrichment. Journal of
Geophysical Research, Volume 110.
McManus, J. W. & Polsenberg, J. F., 2004. Coral-algal phase shifts on coral reefs: Ecological and environmental aspects. Progress in Oceanography, 60(2-4), pp. 263-279.
Mongin, M. & Baird, M., 2014. The interacting effects of photosynthesis, calcification and water circulation on carbon chemistry variability on a coral reef flat: A modelling study. Ecological Modelling, Volume 284, pp. 19-34.
Mueller, B. et al., 2016. Effect of light and nutrient availability on the release of dissolved organic carbon (DOC) by Caribbean turf algae. Scientific Reports, Volume 6.
Mueller, B. et al., 2014. Effect of light availability on dissolved organic carbon release by Caribbean reef algae and corals. Bulletin of Marine Science, 90(3), pp. 875-893.
Muggeo, V. M. R., 2008. Segmented: an R package to fit regression models with broken-line relationships.
R news, 8(1), pp. 20-25.
Nelson, C. E. et al., 2013. Coral and macroalgal exudates vary in neutral sugar composition and
differentially enrich reef bacterioplankton lineages. International Society for Microbial Ecology, Volume 7, pp. 962-979.
Norstrom, A. V., Nystrom, M., Lokrantz, J. & Folke, C., 2009. Alternative states on coral reefs, beyond coral-macroalgal phase shifts. MARINE ECOLOGY PROGRESS SERIES, Volume 376, pp. 295-306.
Pandolfi, J. M. et al., 2003. Global Trajectories of the long-term decline of coral reef ecosystems. Science, Volume 301, pp. 955-958.
Pawlik, J. R., Burkepile, D. E. & Thurber, R. V., 2019. A vicious circle? Altered carbon and nutrient cycling may explain the low resilience of Caribbean coral reefs. BioScience, pp. 470-476.
Polovina, J. J., 1984. Model of a coral reef ecosystem. Coral Reefs, Volume 3, pp. 1-11.
Quinlan, Z. A. et al., 2019. Species-specific differences in the microbiomes and organic exudates of crustose coralline algae influence bacterioplankton communities. Frontiers in Mircobiology, Volume 10.
18 Ramsby, B. D., Shirur, K. P., Iglesias-Prieto, R. & Goulet, T. L., 2014. Symbiodinium Photosynthesis in Caribbean Ocotocorals. Plos One, 9(9).
Sawall, Y. & Hochberg, E. J., 2018. Diel versus time-integrated (daily) photosynthesis and irradiance relationships of coral reef organisms and communities. Plos One, 13(12).
Tanaka, Y. et al., 2008. Production of dissolved and particulate organic matter by the reef-building corals Porites cylindrica and Acropora pulchra. Bulletin of Marine Science, 82(2), pp. 237-245.
Tsounis, G. & Edmunds, P. J., 2017. Three decades of coral reef community dynamics in St. John, USVI: a contrast of scleractinians and octocorals. Ecosphere, 8(1).
Tsounis, G. et al., 2018. Variability of size structure and species composition in Caribbean octocoral communities under contrasting environmental conditions. Marine Biology, Volume 165.
van der Heijden, L. H. & Kamenos, N. A., 2015. Reviews and syntheses: Calculating the global contribution of coralline algae to total carbon burial. Biogeosciences, Volume 12, pp. 6429-6441.
Walters, C., Christensen, V. & Pauly, D., 1997. Structuring dynamic models of eploited ecosystems from trophic mass-balance assessments. Reviews in Fish Biology and Fisheries, Volume 7, pp. 139-172.
Wanders, J. B. W., 1976. The role of benthic algae in the shallow reef of Curacao (Netherlands Antilles). I: Primary productivity in the coral reef. Aquatic Botany, Volume 2, pp. 235-270.
Woodhead, A. J. et al., 2019. Coral reef ecosystem services in the Antropocene. Functional Ecology, Volume 33, pp. 1023-1034.
19
APPENDIX 1
Figure S1-S3. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all Colpophyllia natans species incubations. Legend of figure
S1 is applicable to all graphs.
S1 S2
20
Figure S4-S7. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all Madracis mirabilis species incubations. Legend of figure
S4 is applicable to all graphs.
S4 S5
S6 6=
21
Figure S8-S10. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all Eunicea spp. species incubations. Legend of figure S8 is
applicable to all graphs.
S8 S9
22
Figure S11-S13. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all Plexaura spp. species incubations. Legend of figure S11 is
applicable to all graphs.
S11 S12
23
Figure S14-S16. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all Dictyota spp. species incubations. Legend of figure S14 is
applicable to all graphs.
S14 S15
24
Figure S17-S19. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all Lobophora spp. species incubations. Legend of figure S17 is
applicable to all graphs.
S17 S18
25
Figure S20-S22. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all crustose coralline algae species incubations. Legend of figure S20 is applicable to all graphs.
S20 S21
26
Figure S23-S25. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all Peyssonnelia spp. species incubations. Legend of figure S23 is
applicable to all graphs.
S23 S24
27
Figure S26-S29. Dissolved Oxygen (DO) and Photosynthetically Active Radiation (PAR) measured plotted against the time (minutes) for all control incubations. Legend of figure S26 is applicable to all graphs.
S26 S27
28
APPENDIX 2
Table S1. Calculations from raw dissolved oxygen (DO) data to final absolute fluxes. N = number of individuals (or replications). ∆[O2] = DO concentration per segment. Final Pn fluxes were obtained by correcting for controls, the volume of the incubation chamber and the weight of the individual.
Group Species N Segment Time (min) per segment
Slope (mg/L)
Intercept (mg/L)
∆[O2] (mg/L) Final Pn Flux (µmol O2
g AFDW-1 h-1)
Hard corals Colpophyllia natans 1 1 49 0,0279 7,1285 1,3681 24,85
2 17 0,0134 7,8364 0,2281 11,94 3 61 0,0252 7,0594 1,5357 22,40 4 59 0,0132 8,5749 0,7796 11,76 5 33 0,0021 10,6510 0,0684 1,84 2 1 54 0,0099 7,1784 0,5330 10,02 2 182 0,0192 6,6767 3,5019 19,53 3 14 0,0104 8,7665 0,1455 10,55 3 1 153 0,0134 7,2976 2,0504 32,18 2 14 0,0193 6,3901 0,2705 46,39 3 17 0,0036 9,0169 0,0618 8,72 4 71 0,0177 6,4279 1,2585 42,56 Madracis mirabilis 1 1 73 0,0550 7,1675 4,0161 69,00 2 42 0,0250 9,3482 1,0497 31,34 3 50 0,0024 12,4890 0,1218 3,05 4 34 0,0020 11,7630 0,0668 2,46 5 61 0,0176 8,6611 1,0711 22,02 2 1 13 0,0121 7,1067 0,1571 13,63 2 132 0,0428 6,7173 5,6558 48,31 3 95 0,0288 8,7516 2,7374 32,49 3 1 37 0,0108 7,7463 0,3990 29,38 2 67 0,0155 7,5684 1,0414 42,36
29 3 131 0,0084 8,3054 1,1049 22,98 4 15 -0,0050 11,4690 -0,0754 -13,71 4 1 81 0,0233 7,2969 1,8882 28,63 2 64 0,0184 7,6920 1,1798 22,64 3 22 -0,0005 10,4410 -0,0106 -0,59 4 35 0,0143 7,9806 0,4989 17,51 5 27 0,0247 5,8755 0,6656 30,28 6 26 0,0079 9,7228 0,2044 9,65 Gorgonians Eunicea spp. 1 1 69 0,0078 7,8616 0,5394 47,59 2 30 0,0058 7,9998 0,1744 35,39 3 75 0,0037 8,2086 0,2775 22,51 4 11 0,0006 8,7545 0,0062 3,40 5 70 0,0039 8,1379 0,2722 23,66 2 1 124 0,0129 7,7464 1,6032 31,96 2 17 0,0169 7,7464 0,2865 41,66 3 30 0,0015 9,4300 0,0440 3,62 4 64 0,0135 7,3818 0,8625 33,32 5 20 0,0038 9,6410 0,0769 9,50 3 1 48 0,0004 7,4108 0,0191 1,71 2 11 0,0007 7,6014 0,0076 3,00 3 75 0,0007 7,1602 0,0561 3,23 4 77 0,0009 6,9645 0,0689 3,86 5 49 0,0004 7,9436 0,0211 1,86 Plexaura spp. 1 1 43 0,0008 7,9340 0,0365 6,42 2 52 0,0050 7,7543 0,2597 37,87 3 58 0,0063 7,6277 0,3670 47,98 4 12 0,0009 8,4627 0,0103 6,51 5 26 0,0072 7,4204 0,1869 54,51 6 64 0,0059 7,6757 0,3746 44,39 2 1 40 0,0096 7,8010 0,3826 6,89
30 2 123 0,0211 7,3457 2,5939 15,19 3 14 0,0118 8,8663 0,1646 8,47 4 73 0,0183 7,7042 1,3381 13,20 3 1 114 0,0106 7,8489 1,2135 15,52 2 55 0,0064 8,3310 0,3523 9,34 3 96 0,0087 7,9416 0,8361 12,70
Fleshy algae Dictyota spp. 1 1 28 0,0553 7,0990 1,5483 269,97
2 79 0,0875 6,1905 6,9138 427,28 3 53 0,0781 7,1914 4,1406 381,43 4 68 0,0584 10,3410 3,9709 285,11 5 50 0,0480 12,7080 2,4017 234,51 2 1 24 0,0510 7,3039 1,2234 346,06 2 117 0,0885 6,4000 10,3516 600,62 3 87 0,0611 10,2580 5,3120 414,49 4 47 0,0435 14,2740 2,0426 295,02 3 1 185 0,0581 7,3226 10,7528 345,16 2 45 0,0471 9,3527 2,1210 279,90 Lobophora spp. 1 1 95 0,0262 7,3018 2,4901 204,85 2 78 0,0210 7,8019 1,6356 163,87 3 77 -0,0012 11,6280 -0,0908 -9,23 2 1 92 0,0413 7,1360 3,7967 338,28 2 79 0,0320 7,9843 2,5302 262,53 3 79 -0,0018 13,7890 -0,1453 -15,09 3 1 230 0,0295 7,2727 6,7938 234,82
Calcifying algae Crustose coralline algae 1 1 201 0,0092 7,7231 1,8473 61,40
2 54 0,0075 8,0713 0,4028 49,83
2 1 250 0,0147 7,6411 3,6759 101,16
3 1 168 0,0200 7,5417 3,3598 143,34
2 97 0,0217 7,2484 2,1096 155,87
31 2 213 0,0155 7,5904 3,3000 40,90 2 1 137 0,0072 7,5160 0,9799 62,65 2 113 0,0035 8,0215 0,3917 30,35 3 1 47 0,0099 7,5054 0,4641 43,97 2 135 0,0133 7,3412 1,8013 59,41 3 38 0,0153 6,9872 0,5809 68,07 4 46 0,0120 7,7213 0,5497 53,21 Controls - 1 1 260 0,0000 7,1383 0,0011 0,02 - 2 1 240 0,0000 7,0615 -0,0013 -0,04 - 3 1 250 0,0000 6,9990 0,0020 0,03 - 4 1 265 0,0000 7,4191 0,0004 0,00
32
APPENDIX 3
Table S2. Ash free dry weight (AFDW) conversions. AFDW conversions used for Dictyota spp., Lobophora spp. and Peyssonnelia spp. n = number of
replicates. Species tissue thickness [cm] CN ratio dry mass
AFDW Organic carbon dry mass
Organic nitrogen dry mass
C/AFDW unit n
Massive coral (used for
Colpophyllia natans) 0,1358 7,0072 1,1098 0,0843 0,0277 0,0042 0,3181 g cm-2 15 Madracis mirabilis 0,2207 4,9284 1,4185 0,0995 0,0143 0,0029 0,1442 g cm-2 3 Eunicea spp. 5,2409 0,1351 0,0393 0,0142 0,0027 0,3664 g cm-3 3 Plexaura spp. 5,2793 0,2103 0,0392 0,0151 0,0029 0,3881 g cm-3 4 Dictyota spp. 1,3553 16,2827 0,0045 0,0026 0,0008 0,0001 0,3218 g cm-2 3 Lobophora spp. 1,3553 30,9700 0,0102 0,0067 0,0028 0,0001 0,4392 g cm-2 3
Crustose Coralline Algae 0,0258 7,2950 0,1259 0,0129 0,0022 0,0003 0,1851 g cm-2 4
APPENDIX 4
S30 S31 S32 S33 S34 S3534
Figure S30-S37. Net photosynthesis (Pn) plotted against integrating light (IL) for each species. Y stands for net photosynthesis (Pn) in µmol O2 gAFDW-1 h-1 and x stands for integrating light (IL) in µmol photons m-2 s-1. Legend of figure S30 is applicable to all graphs.
S36 S37
35
APPENDIX 5
Table S3. Post-Hoc Analysis with Tukey’s Test. Abbreviations: cnat = Colpophyillia natans, mad =
Madracis mirabilis, eun = Eunicea spp., plex = Plexaura spp., dict = Dictyota spp., lobo = Lobophora spp., cca = Crustose Coralline Algae, pey = Peyssonnelia.
Species comparison
Difference Lower limit Upper limit Adjusted p value Significance
(‘***’0.001, ‘**’ 0.01, ‘*’0.05) cca-cont 0,097 0,003 0,191 0,039 ** cnat-cont 0,020 -0,061 0,101 0,997 dict-cont 0,300 0,218 0,381 0,000 *** eun-cont 0,017 -0,061 0,096 0,999 lobo-cont 0,150 0,062 0,238 0,000 *** mad-cont 0,022 -0,055 0,100 0,991 pey-cont 0,048 -0,038 0,133 0,698 plex-cont 0,021 -0,059 0,101 0,995 cnat-cca -0,077 -0,151 -0,002 0,038 ** dict-cca 0,203 0,127 0,278 0,000 *** eun-cca -0,079 -0,151 -0,007 0,021 ** lobo-cca 0,053 -0,029 0,135 0,503 mad-cca -0,074 -0,145 -0,004 0,032 ** pey-cca -0,049 -0,129 0,031 0,581 plex-cca -0,076 -0,149 -0,002 0,040 ** dict-cnat 0,280 0,221 0,338 0,000 *** eun-cnat -0,002 -0,057 0,052 1,000 lobo-cnat 0,130 0,063 0,196 0,000 *** mad-cnat 0,002 -0,050 0,055 1,000 pey-cnat 0,028 -0,036 0,092 0,899 plex-cnat 0,001 -0,055 0,057 1,000 eun-dict -0,282 -0,338 -0,227 0,000 *** lobo-dict -0,150 -0,217 -0,082 0,000 *** mad-dict -0,277 -0,331 -0,224 0,000 *** pey-dict -0,252 -0,317 -0,187 0,000 *** plex-dict -0,279 -0,336 -0,221 0,000 *** lobo-eun 0,132 0,068 0,196 0,000 *** mad-eun 0,005 -0,044 0,054 1,000 pey-eun 0,030 -0,031 0,092 0,815 plex-eun 0,004 -0,049 0,057 1,000 mad-lobo -0,127 -0,190 -0,065 0,000 *** pey-lobo -0,102 -0,175 -0,030 0,001 *** plex-lobo -0,129 -0,194 -0,063 0,000 *** pey-mad 0,025 -0,034 0,085 0,910 plex-mad -0,001 -0,052 0,050 1,000 plex-pey -0,027 -0,090 0,036 0,912
36
