Limnology and Oceanography In Press
Temporal and Spatial Variation in the δ15N and δ13C of coral tissue and zooxanthellae in Montastraea faveolata collected from the Florida reef tract
Peter K. Swart, Amel Saied and Kathryn Lamb Marine Geology and Geophysics Rosenstiel School of Marine and Atmospheric Sciences University of Miami Miami Fl 33149 [email protected]
Abstract
Small pieces of coral skeletons with their associated tissues were collected at monthly intervals, between January 1995 and December 1996, from specimens of Montastraea faveolata at five locations on the Florida Reef tract. The locations (Triangles, Pickles Reef, Crocker Reef, Hen and Chickens, and The Rocks) represent both near shore and off shore environments. During each sampling trip between one to four pieces of coral skeleton and associated tissues were collected from the sides of different colonies living in water depths between 3 to 4 m. At one site (Pickles) samples were also collected from 8 m water depth. The tissue and zooxanthellae were removed from the skeletons, separated, and subsequently analyzed for δ15N and δ13C. The mean δ15N value in the coral tissue of all samples was +6.6 (+/- 0.6 ‰) while the δ13C was -13.3 (+/- 0.5 ‰) (n=197). The δ15N and δ13C of the zooxanthellae were +4.7 (+/-1.1‰) and -12.2 (+/- 1.0 ‰) respectively (n=147). The differences in the δ15N and δ13C between the zooxanthellae and the coral tissue were statistically significant. No statistically significant differences were observed between near shore and offshore stations in either δ15N or δ13C. The absence of a difference between the inshore and offshore stations casts doubt on both whether the δ15N of the coral tissues is related to anthropogenic influences and whether the δ15N value itself can be used as an indicator of sewage contamination in corals. Between 1995 and 1997 there was a long term increase of 1 ‰ in the δ13C of the coral tissue and zooxanthellae and a long term decrease of approximately 0.8 ‰ in the δ15N of the coral tissue and the zooxanthellae. The increase in the δ13C of the organic material (OM) was mimicked in the δ13C of the skeletal material from corals from two reefs in the area suggesting a strong connection between the δ13C of the coral tissues and the skeletal material. There appears to be clear seasonal variations in the δ13C of the coral tissue at certain locations such as Pickles Reef with δ13C of both the coral tissues and the zooxanthellae becoming more positive between July and August. The difference between the δ13C of the zooxanthellae and the coral tissue also varies seasonal with the maximum difference occurring in July of each year. In contrast the maximum δ13C in the skeleton appears to occur later in the year, between September and November. positive influences on the growth of the Introduction coral. Under shallow water conditions the coral-zooxanthellae system is autotrophic It is well established that certain (Muscatine and Cernichiari, 1969). scleractinian corals have symbiotic Evidence of the autotrophic nature of associations with dinoflagellate algae zooxanthellate corals is found in the (zooxanthellae) which are beneficial to the difference in the δ13C of the zooxanthellae host (Goreau, 1959; Muscatine and and coral tissue at various water depths. At Cernichiari, 1969; Wilbur and Simkiss, shallow depths where light intensity is high 1979). The zooxanthellae are able to pass the δ13C of the zooxanthellae and the coral organic compounds to the coral resulting in tissue are relatively similar (Land et al.,
Page -1- In Press Limnology and Oceanography
1975; Muscatine et al., 1989), and the δ13C 2000a), although Muscatine and Kaplan of the coral tissue is significantly more (1994) also investigated δ15N as an positive (-10 to -14 ‰) than the supposed indicator of autotrophic and heterotrophic food source of the coral, zooplankton (~ -20 responses. The study by Muscatine and ‰). This indicates that sufficient Kaplan (1994) showed a slight decrease in photosynthate is being translocated so that δ15N with increasing depth, although this the δ13C values of the coral tissue and pattern was not always consistent. In contrast to δ13C, the δ15N value was generally enriched in the coral tissue compared to the zooxanthellae. Common to all previous studies on the δ15N and δ13C of coral tissues is the fact that they have ignored any temporal variation in the isotopic composition of the soft tissues of the coral. Usually such 80o 25 samples are taken during the summer 25o 05 #Y 1 months when weather conditions are more favorable. However, in the study of Swart 2 #Y et al. (1996) it was noticed that the δ13C of #Y A o 3 B 25 00 #Y the coral tissues which were collected
7 4 N during the summer months (June -July 1990) #Y 7 #Y 6 W E were isotopically more positive (-15 ‰) C #Y 5 S than those measured in September 1990 (-17 13 10 0 10 Km ‰). The difference between the δ C of the zooxanthellae and coral tissue also changed Figure 1: Location of the Reefs studied in the Florida Keys. Site from about +3 ‰ in June 1990 to +7 ‰ in 2(Triangles), Site 4 (Pickles), Site 5 (Crocker), Site 6 (Hen & September. Based on these data the authors Chickens), and Site 7 (The Rocks). In addition water samples were collected from Marker 2 (Site 1), and Molasses Channel (Site speculated that the changes might be 3). The location of sites from which water samples are collected induced by changes in the partitioning of the for the FKNMS water quality network are shown in the squares (Molasses Channel (A) and Molasses Reef (B)). The coral internal C pool. These data prompted the skeletons analyzed were collected from Crocker Reef (Site 5) and initiation of this study whose goal was to Cheeca Reef (C). The shading denote 20' depth contours. Land and semi-emergent mud-banks are shown in the black shading. investigate temporal changes in tissues and zooxanthellae of corals from the near shore zooxanthellae are similar. With increasing and offshore reefs over an extended period. depth the δ13C of the coral tissues become In particular we were interested in whether more negative and the δ13C approaches that there were any seasonal changes in the δ13C of the zooplankton. Such variations are as previously observed by Swart et al. taken as indicating a change from (1996) and whether there were differences autotrophy to heterotrophy. Studies of the in the δ13C or δ15N relative to the position of δ15N of coral tissue have mainly the reef. concentrated on their potential as indicators of anthropogenic waste (Heikoop et al.,
Page -2- In Press Limnology and Oceanography
Study Site were placed on ice until removal of the tissues (within 24 hours) by air-brushing. The study sites chosen were five patch reefs The zooxanthellae and tissue samples were off Key Largo in the Florida Keys (Figure separated using the methods of Szmant et 1). At each site, small pieces of skeletons al., (1989). Previous work using this with living coral tissue (~2 cm in diameter) method has shown that there is only a small were chipped off the sides of the heads. A amount of cross contamination by total of 197 samples were analyzed for the zooxanthellae in the coral tissue (less than δ15N and δ13C of the coral tissue and 147 for 5%; Fitzgerald and Szmant, 1997). In the δ15N and δ13C of the zooxanthellae. contrast, there can be significant With the exception of Pickles Reef the contamination of the zooxanthellae by the corals were collected at only one depth. At coral tissue during separation (up to 50%; Pickles corals were collected from water Fitzgerald and Szmant, 1997). depths of 8.5, 3.5, and 3 m. For the Isotopic analyses: The isotopic composition purposes of this study the samples from 3.5 (δ15N and δ13C) of organic coral samples and 3 m have been grouped together. The were determined using a CN analyzer corals were never collected from the same interfaced with a continuous-flow individuals which had been sampled during isotope-ratio mass spectrometer (CFIRMS) previous visits. The initial rationale was to (Europa Scientific). External precision protect the corals, but it was later realized determined through the analysis of replicate that continued sampling would not only standard material is 0.1 ‰ for C and 0.2 ‰ eventually destroy the colony, but also for N. severely stress the individual and perhaps Analyses of the dissolved inorganic carbon 13 lead to changes in the δ C induced by (DIC): The CO2 was removed from the stress. Water samples were also collected in sample by acidification in a stream of He order to check for seasonal variation in the gas and analyzed using a Europa 20-20 mass δ13C of the dissolved inorganic carbon (DIC) spectrometer by comparison with a pulse of in the water from the same sites at which the injected reference gas. External precision corals were collected well as two additional for this method as determined by measuring sites (Site 1; Maker 2) and Site 3 (Molasses replicate samples is ~0.08 ‰. Channel) (Figure 1). In order to compare Analysis of the coral skeleton: Samples of changes in the δ13C of OM to changes in the skeletal material from colonies of δ13C of coral skeletons, two corals were Montastraea faveolata at Crocker Reef (8 cored, a specimen of Montastraea faveolata m) and Siderastrea siderea at Cheeca Rocks at Crocker Reef (10m) and a specimen of (3 m) were drilled from the slab using a Siderastrea siderea from Cheeca Rocks hand held drill at a resolution of (Figure 1). approximately 20 samples pre year. Material was analyzed using a Kiel III attached to a Finnigan Delta plus mass Methods
Tissue removal: After collection, samples
Page -3- In Press Limnology and Oceanography spectrometer. Nutrient data: For tissues has a mean value of +6.6 (+/- 0.6 comparison with changes in the ‰). In comparison, the δ15N of the concentration of inorganic N and salinity, zooxanthellae has a mean value of +4.8 (+/-
-10.0 9.0 a 8.0 -11.0 7.0 )
oo -12.0 / o
/oo) 6.0 o C ( -13.0 13 N ( δ 5.0 15 -14.0 Tissue 4.0 Zooxanthellae Tissues -15.0 3.0 Jan-95 Jul-95 Dec-95 Jun-96 Dec-96 Zooxanthellae 0.0 b Date 2.0 -0.5 Jan 95 Apr 95 Jul 95 Oct 95 Feb 96 May 96 Aug 96 Dec 96 -1.0 Date )
oo -1.5 / o 15 -2.0 Figure 3: Changes in the δ N of the coral tissues (diamonds) and C ( 13
δ -2.5 the zooxanthellae (squares) over the time period January 1995 to Crocker December 1996. There is a statistically significant decrease in -3.0 Cheeca the δ15N of both the zooxanthellae and the coral tissue. The r2 of -3.5 1995 1995.5 1996 1996.5 1997 the coral tissue and the zooxanthellae are 0.25 (n=16) and 0.34 Date (n=14) respectively. Outliers from the data are not shown (See Table 1). Figure 2: (a) Changes in the δ13C of the coral tissues (diamonds) and the zooxanthellae (squares) over the time period January 1995 to December 1996. There is an increase in the δ13C of both 1.0 ‰). The zooxanthellae are statistically the zooxanthellae (r2=0.10) and the coral tissue (r2=0.29) over significantly more negative than the coral this time period. The change in the δ13C of the coral tissues is statistically significant at the 95% level (p<0.05, n=16) ; (b) tissues at the 95 % confidence limits. Changes in the δ13C of the coral skeleton of Montastraea Carbon: The mean δ13C of all the coral faveolata collected from Crocker Reef and Siderastrea siderea collected from Cheeca Rocks over the same time period as tissue, -13.3 (+/- 0.5‰), is statistically more measurement of the δ13C of the coral tissue and zooxanthellae. negative than that in the zooxanthellae - These two records both exhibit a small increase during the experimental period superimposed on an overall decrease in the 12.2 (+/- 1.0 ‰) at the 95 % confidence δ13C of the skeleton as a result of the 13C Suess effect (See Figure limits. 7). Outliers from the data are not shown (See Table 1). Spatial variation: Average δ15N and δ13C we have used the data collected and values of the coral tissues and zooxanthellae analyzed by FIU/SERC. The methods used from the inshore and offshore sites are and the errors on these analyses are shown in Figures 4 and Table 2-5. At the discussed in Boyer et al. (1999) and Boyer 95% confidence limits there are no and Jones (2002). statistical differences in either the δ15N or Results δ13C of the coral tissue or zooxanthellae 15 13 The δ N and δ C data are presented in between different reefs. Table 1 and Figures 2 and 3 as the mean Depth variation: Changes in δ13C and δ15N values of the separated zooxanthellate and relative to depth were studied only at Site 4 coral fractions measured at each site during (Pickles Reef). Although the δ15N of the each month of the study. The values coral tissue of the deeper samples was represent either the mean of several samples slightly enriched in both the tissue and the or the values of individual samples. zooxanthellae fraction, there were no 15 Nitrogen: The overall δ N of the coral statistically significant differences between
Page -4- In Press Limnology and Oceanography the samples collected from 8.5 and 3 m. other sites. The δ13C values of the tissues and the Dissolved inorganic carbon: The δ13C of the DIC showed no long term variation -9 Tissue (inner) throughout the length of the experiment. -10 Tissue (outer)
) However, significant seasonal variation was Zooxanthllae (inner) oo
/ -11 Zooxanthellae (outer) present (Table 6 and Figure 6). Although o -12 there was considerable overlap in the δ13C C (
13 -13
of the DIC between sites, the inner locations ÷ -14 such as Marker 2 and The Rocks possessed 13 -15 lower mean δ C values than the more 345678 15 o ocean sites. These differences were not ÷⎜ N ( /oo) statisitically different at the 95 % confidence Figure 4: Mean δ13C and δ15N of coral tissue from all sites. The inner sites are shown in the lighter symbols and the outer sites limits. the darker symbols. Zooxanthellae are indicated by the triangles Carbon isotopic composition of the coral and coral tissues by the squares. skeleton: The δ13C was measured in the zooxanthellae were slightly more negative in skeleton of one corals collected from the deeper corals, although the differences Crocker Reef as well as an additional coral were not statistically significant at the 95% from a site outside the study area (Cheeca confidence limits (Figure 4). Rocks; Figure 1). Over the period of Temporal variation: There were no measurements of the OM (January 1995- significant seasonal variations in the δ15N of December 1996), the δ13C of both the coral either the coral tissues or zooxanthellae at skeletons showed an increase, similar in any of the sites. There was a decrease in the magnitude to that observed in the OM δ15N from the start of the measurements (Figure 2b). throughout 1995 and 1996, which was statistically significant at the 95% Discussion confidence limits. This decrease was greater in the zooxanthellae than in the coral tissue Antropogenic influences: The δ15N of OM (Figure 3). In contrast the δ13C of the coral in the marine environment has been used to tissue show clear seasonal trends at some distinguish N derived from fertilizers (0 ‰) sites, but not at others. At the Pickles site (Shearer et al. 1974; Kreitler, 1979; Heaton, 13 - there are intra-annual variations in the δ C 1986), NO3 produced from the oxidation of of the tissues collected from both the waste (+10 to +22‰) (Kreitler, 1979), and - shallow and deep site (Figure 5). The values NO3 produced from the oxidation of organic of the tissues become more positive during N in the soil (+4 to +9‰) (Gormley and the summer months and the differences Spalding, 1979, Mariotti, 1974). Numerous between the δ13C of the tissue and the papers have been published which report the zooxanthellae (∆t-z) are larger during this use of δ15N in benthic organisms in order to period. At the end of the summer the δ13C distinguish sewage (Heikoop et al. 2000a,b; of the tissue becomes more negative and the Sammarco et al., 1999; Risk and Erdmann, ∆t-z becomes smaller. Such clear seasonal patterns were not observed at any of the
Page -5- In Press Limnology and Oceanography
2000; Costanzo et al., 2001, 2004). 1989) and appear not to be related to Simplistically these authors have interpreted anthropogenic sources, but rather a result of 15 - positive δ N values (> + 7 to +10‰) as the uptake of isotopically positive NO3 reflecting input of sewage. However, there which is in turn a result of fractionation are other workers which have shown during the process of assimilation or convincingly that sewage derived N has denitrification. Hence based on the
-10
a 2.50 a -11 2.00 ) ) oo -12 / o oo 1.50 / o C ( C ( 13 δ -13 13 1.00 δ
-14 0.50 Tissues Zooxanthellae 0.00 -15 Jan 95 Apr 95 Jul 95 Oct 95 Jan 96 Apr 96 Jul 96 Oct 96 Jan 97 Apr 97 Jan 95 Apr 95 Jul 95 Oct 95 Feb 96 May 96 Aug 96 Dec 96 Mar 97 Date 2 3.00 b 1.8 1.6 b u 2.00 1.4 1.2 1.00 1 ) 0.8 oo / zoox-tiss o 0.6