Southeast Ecological Science Center
Stable Isotope Analysis of Trophic Structure Reveals Dual Carbon Sources Fueling the GOM Outer Continental Shelf Ecosystem
Ken Sulak, Jim Berg, Mike Randall, George Dennis & Allen Brooks
Presented at the FlSC Strategic Review in St. Petersburg, Florida, May 9-12, 2006
Trophic relationships for deep-water fish communities of the outer continental shelf (OCS) remain poorly studied, particularly in warm temperate and tropical oceans. Obtaining sufficient numbers of specimens from deep water (>50 m) to enable robust stomach content studies of food habits and trophic relationships is logistically challenging and expensive. Moreover, many deep-water fish species eject their stomach contents when brought to the surface. Conventional stomach contents analysis has the further limitation of providing only a point-in-time 'snapshot' of food habits.
Alternatively, analysis of stable carbon, nitrogen and sulfur isotope signatures from fish tissues can provide temporally integrated information both on food sources and trophic position. Sampling of both potential carbon sources and consumer tissues can further refine understanding of ecosystem organization and trophic function. The present study was undertaken by the U.S. Geological Survey (USGS) OCS Ecosystem Program. It focused on the benthic fish fauna (all species associated with the benthic boundary zone) of the northeastern Gulf of Mexico (NEGOM) OCS, particularly the fauna of drowned shelf-edge reefs between depths of 65-120 m. Comparative taxa sampled included benthic non-reef fishes, benthic invertebrates (sessile, mobile benthic, and benthopelagic), and epipelagic fishes and invertebrates (free-living and Sargassum associates). Potential carbon sources analyzed included near-surface and near-bottom holoplankton (phytoplankton, diatoms, and zooplankton), terrestrial plant carbon, pelagic macroalgae (Sargassum and its epiphyte, Cladophora liniformis), and bottom sediment (potential carbon source via resuspension).
Sampling was conducted in two regions of complex fossil reef topography, the "Pinnacles" region west of DeSoto Canyon, and the West Florida Shelf OCS east of DeSoto Canyon (Fig. 1). Sampling operations were conducted during three cruises in 2001-2003 over 53-133 m bottom depth. Samples of bottom fauna were obtained primarily by angling and trapping from both hard bottom (deep reefs) and by trawling from adjacent soft-substrate shelf areas. A remotely operated vehicle (ROV) was used for in situ sampling. Samples of pelagic taxa were obtained from the overlying water column by angling and using midwater trawls. Suspended particles <125 μm were sampled using plankton nets. Two size categories of holoplankton (all types of plankton) were sampled simultaneously using nested 0.5 m diameter plankton nets. A 335 m mesh net was nested within a mated 125 m mesh net, such that particles >335 m (mesoplankton, composed primarily of zooplankton) were retained by the inner mesoplankton net, while only particles <335 m and >125 m (POM, containing a mix of phytoplankton and small zooplankton) were retained in the outer POM net. Particles smaller than 125 m passed through both nets, and were not sampled in this study. The dual net array was towed horizontally for five minutes at 3-5 m depth for shallow, near-surface samples, and at an altitude of 5-10 m above the bottom for deep, near-bottom samples. Bottom sediment was sampled using a custom 0.0484 m2 box grab that penetrates the bottom substrate, taking a nearly undisturbed 5-10 cm deep rectangular plug of sediment. Epipelagic fishes, Sargassum, attached epibiota, and associated fauna were obtained in dip nets and in a 1.0 m, 500 m mesh conical plankton net towed at the surface. Benthic fishes and invertebrates were obtained using an otter trawl, benthic sled trawl, benthic dredge, a rope-fiber tangle device, baited fish trap, angling with bait, and a suction-sampler deployed from a remotely operated vehicle (ROV). Demersal planktivorous fishes were obtained by angling using artificial meroplankton lures (Sabiki rigs). Tissue samples were prepared at sea, and processed for analysis in the USGS laboratory in Gainesville, FL. Target sample size per taxon was N>20 to enable robust statistical comparisons.
Isotopic analyses were accomplished in a dedicated stable isotope analysis laboratory. The ratio of heavy (13C) to light (12C) isotopes of carbon in a sample was compared to that of the standard, Pee Dee Belemnite to yield the isotopic signature, δ13C, of the sample source. Similarly, the ratio of sample 15N:14N isotopes was compared to the standard, atmospheric nitrogen, to yield the δ15N isotopic signature of the sample source. Both ratios increase (become enriched) per trophic level. Classically, δ13C is predicted to increase 0-1 ‰ per trophic level, thus tending to conserve the carbon isotopic signature of the original food source. In contrast, δ15N is predicted to increase by 3 ‰ per trophic level, thus incrementally revealing how many trophic steps separate a consumer from the original food source.
Fig. 1. Northern Gulf of Mexico OCS Sampling
Fig. 2. OCS Deep-Reef Food Web: Plankton-Driven?
114 producer and consumer entities were analyzed (N = 1,265 specimens). A 13C sample size of N>20 was achieved for 21 entities; 10-19 was achieved for 37 entities, 5-9 for 19 entities, and 1-4 for 37 entities. Most entities represented by N<5 samples were either large, mobile and/or cryptic predators difficult to sample with our field methods, or species that are typically rare in the NEGOM OCS fauna. Although 71 sediment samples were obtained, organic matter mass available proved insufficient for analysis in all but four samples. Ten tissue samples of the benthic planktivorous fish, Pronotogrammus marticensis were analyzed, returning a mean 34S value of 19.93±0.55. Across all consumer taxa sampled, individual 13C values ranged between –9.29 ‰ and –22.00 ‰, residing nearly within the signature spread found for holoplankton and macroalgae combined (–12.62 ‰ and –24.18 ‰). No 13C consumer signatures indicative of terrestrial carbon, (e.g., ≤=-26 ‰), were obtained. One-way ANOVA contrast of 11 empirically-defined consumer guilds resolved three statistically distinct (p < 0.05) major consumer groups, and two statistically distinct subgroups (Fig. 2) defined by 13C data. One-way ANOVA contrast in 13C data for six benthic invertebrate consumers revealed that the particulate-feeding gorgonian Nicella sp. is statistically distinct (mean 13C = -15.01±2.17 ‰) from the other five invertebrate consumers (overall mean 13C = -17.29±0.90 ‰). Two of the three major carbon consumer groups form a trophic chain linked to holoplankton as the fundamental carbon source (Fig. – Upper Right). This chain includes all consumer guilds except for the distinctive Benthic Miniparticulate Feeder guild (BMF). This guild appears to be linked into a distinct macroalgal carbon trophic chain (Fig. 3). One-way ANOVA also resolved five statistically distinct trophic levels defined by means in 15N across both carbon producers and consumers. In turn, these effectively define a food chain with four steps.
Fig. 3. Testing the Old Plankton Sole-Source Paradigm
Fig. 4. Dual Oceanic Carbon Sources on
Fig. 5. Surface Sargassum May Dictate OCS