The influence of deep-sea-bed CO2 sequestration on small metazoan (meiofaunal) community structure and function
Kevin R. Carman and John W. Fleeger - Louisiana State University
David Thistle - Florida State University
Abstract
Background
Proposed Research
Questions
Approach
Methods
We are initiating a new project that will examine the influence of deep-sea CO2 sequestration on the meiofaunal (metazoans < 1.0 mm in body length) component of the benthic community. Meiofauna are by far the most abundant and diverse metazoan organisms in the benthos, and their relative importance to both standing stock and biological diversity increases with increasing depth in the ocean. Because of their small size and high abundance, relatively small samples can be used for analyses. Also because of their small size, meiofauna have higher metabolic rates and shorter population turnover times than do macrobenthos, and thus are particularly useful for the study of environmental disturbances over relatively short time scales.
Our research will be integrated with the DOE project of Drs. Jim Barry and Peter Brewer (MBARI), which involves the application of liquid CO2 to the seabed in experimental corrals at depths ranging from 3200-3600 m in the Monterey Canyon. Sediment cores will be collected from treatment (exposed to CO2) and control sites with an ROV operated from a research vessel. The Barry-Brewer project is focused primarily on the responses of macro- and megafaunal benthos to CO2 exposure; our research will complement theirs by examining the ecologically distinct meiofaunal invertebrates that form an important component of the benthic community.
We propose to examine both structural (abundance, taxonomic and functional-group composition, and vertical distribution in sediment) and functional (body size/biomass distributions, and animal health) effects of CO2 exposure on meiobenthos, with emphasis on the following hypotheses:
- H1: Meiofauna will emigrate from sediments and/or alter their vertical distribution within sediments.
- H2: Some species will be eliminated from the community because of exposure to sequestered CO2. Community composition will thus be simplified and biological diversity will decline.
- H3: CO2 exposure will adversely affect the health of individuals, which will be reflected in the depletion of their lipid energy reserves.
- H4: Meiofaunal fecundity and demographics will be adversely affected.
The vertical distribution of meiofauna in the top 2 cm of sediments will be used to test for effects on surface-dwelling meiofauna. General effects on community composition will be determined from species-level (harpacticoid copepods) and functional-group (nematode feeding type and tail morphology) influences of CO2 exposure. Analysis of lipid content of individual harpacticoid copepods will be used as an indicator of health. Effects of CO2 exposure on fecundity will be determined by measuring brood size in female copepods and abundances of larval and juvenile harpacticoid copepod developmental stages.
Since the industrial revolution, the atmospheric concentration of CO2 has increased by approximately 30% as a consequence of accelerated use of fossil fuels and deforestation (Keeling and Whorf 1998). This rate of increase is unprecedented in the past 400,000 years and has raised concerns that it may lead to accelerated global warming (Carbon and Climate Working Group (CCWG) 1999). During the 1980's, the combustion of fossil fuels created approximately 5.5 ± 0.5 Gt C y-1 (Gt = gigaton = billion metric tons) of CO2. Approximately 2.2 Gt C y-1 was removed from the atmosphere. The consensus is that most (approximately 2 Gt y-1) of this carbon was absorbed in the ocean (Reichele et al. 1999) and that the rest was removed by the terrestrial biosphere. In spite of these large CO2 sinks, atmospheric CO2 increased at a rate of approximately 3.3 Gt C y-1 (Schimel et al. 1995), and thus atmospheric CO2 concentrations continue to rise (Reichele et al. 1999). Reichele et al. (1999) estimate that atmospheric CO2 carbon will need to be decreased by about 1 Gt y-1 by the year 2025 and by up to 4 Gt y-1 by 2050 in order to achieve atmospheric stabilization. Attempts to reduce atmospheric CO2 concentrations by curtailing global CO2 emissions have thus far been relatively unsuccessful, and fall well short of the goals identified above. Alternative strategies are therefore being considered.
The world ocean contains approximately 50 times as much CO2 as does the atmosphere and represents a tremendous potential sink for excess atmospheric CO2 (CCWG 1999). For example, the deep ocean has the capacity to absorb an amount of CO2 equivalent to all of the known fossil-fuel reserves (Reichle et al. 1999). Thus, it is reasonable to consider the ocean as reservoir in which excess CO2 can be sequestered. However, we lack an understanding of how increased CO2 concentrations will influence marine ecosystems.
Two general strategies have been proposed to accelerate the transfer of CO2 from the atmosphere to the ocean. One strategy involves the stimulation of phytoplankton primary production via the addition of a limiting nutrient, such as iron (Chisholm and Morel 1991). The second strategy involves concentrating CO2 from the atmosphere and injecting it onto the seabed of the deep ocean (Ormerod 1996). The ecological consequences of neither of these strategies are known. Our proposal focuses on the second strategy, and will examine the influence of elevated CO2 concentrations, and associated changes in pH, on the structure and function of meiofaunal (animals > 0.062 mm and < 0.50 mm) benthic invertebrates in the deep sea.
As gaseous CO2 is moved from the sea surface to depth, the increased pressure dramatically changes its physical properties. At approximately 1000 m, CO2 becomes liquid but is less dense than seawater. Below 3000 m, liquid CO2 changes to a gas hydrate, which is denser than seawater (Brewer et al. 2000). It is therefore likely that mitigation strategies involving direct injection of CO2 would necessitate that the CO2 be delivered to depths of > 3000 m. The injection of CO2 at these depths may have direct and indirect effects. In particular, because of the participation of CO2 in the natural buffering of the ocean, the addition of CO2 will shift the pH toward acidity.
Relatively little is known about the effects of reduced pH and/or elevated CO2 concentrations on marine benthic invertebrates, probably because both pH and especially CO2 concentrations are relatively stable in most marine settings. A few correlative field studies have considered pH or CO2 as physicochemical variables that might influence the abundance of (primarily shallow-water) benthic organisms: results are either equivocal (e.g., pH covaries with a variety of other environmental variables such as O2 or H2S) or apparent effects are minimal (Siemens et al. 2001; Paula et al. 2001; Vopel et al. 1996; Meadows et al. 1994). However, Thiermann et al. (1997) observed reduced abundances of benthic fauna in the immediate vicinity of a shallow-water hydrothermal vent where pH was as low as 6. Ellis et al. (2001) observed that variation in pH had a more adverse effect on soil nematodes than did metal contaminants. Around deep-sea hydrothermal vents, low pH (5.2) has adverse effects on calcareous foraminifera (carbonate dissolution is enhanced; Jonasson et al. 1995).
Nevertheless, very little is known about the potential effects of CO2 sequestration on deep-sea organisms (Shirayama 1995; Omori et al. 1996). Toxicity models are based on the assumption that reductions in pH associated with elevated CO2 concentrations will be the primary mechanism by which organisms are adversely affected (e.g., Auerbach et al. 1997). Exposure to reduced pH can cause sublethal effects, such as altered behavior (Knutzen 1981) and loss of chemosensory ability (Hara 1976), and continued exposure can be lethal. As summarized by Seibel and Walsh (2001) deep-sea organisms are highly adapted to stable pH and CO2 conditions and have greatly reduced buffering capacities relative to shallow-water species; therefore,even slight changes in pH can have important influences on metabolic activities. Tamburri et al. (2000), however, showed that mobile, deep-sea scavengers suffer from respiratory distress when exposed to hypercapnic conditions (and that this effect was not associated with decreased pH). Thus, it appears likely that injection of CO2 into the deep sea may adversely affect organisms both directly (via elevated CO2 concentrations) and indirectly (via reduced pH).
Shirayama (1997) outlined a strategy for evaluating the potential influences of CO2 sequestration and identified the major features of deep-sea organisms that should be considered in such planning.
First, deep-sea animals live in an energy-limited environment and have very low respiration rates. Thus, any perturbation that reduces their ability to acquire resources or stresses that elevate their respiration rate could significantly influence their ability to survive and reproduce.
Second, because deep-sea organisms have long life spans, low reproductive rates, low abundance and are highly diverse, any major source of mortality could have devastating consequences; centuries could be required for recovery of the community (Young and Richardson 1998). The combination of high species diversity and low abundance means that most species are rare. Thus, large-scale mortality could result in the extinction of large number of species. Low reproductive rates would result in slow recoveries of surviving species. Major negative impacts on deep-sea organisms may therefore have profound and long-lasting consequences for deep-sea ecology, biodiversity, and biogeochemistry.
Third, deep-sea organisms are very sensitive to environmental disturbance. Deep-sea species are highly adapted to specific physiological conditions that exist in the deep ocean. Of particular relevance to this proposal, pH is very stable in the deep sea, varying by only 0.1 or 0.2 pH units. Therefore major changes in pH could have dramatic implications for deep-sea organisms.
Our research will be integrated with the DOE-funded research currently being conducted by Drs. Jim Barry and Peter Brewer, which is focused on (1) the long-term fate of CO2 hydrate, (2) geochemical influences of CO2 disposal on sediments and pore waters, (3) modeling the fate of released CO2, and (4) responses of benthic organisms, as well as bottom-dwelling fish, to CO2 exposure. The research that we propose would complement the work of Drs. Barry and Brewer, and result in a uniquely thorough analysis of the responses of benthic biota to CO2 exposure.
As part of his analysis of the benthos, Dr. Barry's group is measuring abundances and viability of benthic flagellates, ciliates, and nematodes. Our research will focus on structural and functional responses of the meiofaunal copepods and nematodes, and thus will complement, but not be redundant with his research. There are several compelling reasons for focusing on meiofauna. Based on predictive regressions (Vincx et al. 1994), total meiofaunal density at depths comparable to the Monterey Canyon (3200-3600 m) is expected to be about 300-500 x 103 individuals m-2. Thus, small samples can accurately characterize meiofaunal abundance and biodiversity, which makes meiofauna better-suited for experimental manipulations than larger organisms. Meiofauna are well-suited for studies of environmental disturbances (anthropogenic or natural) because: (1) They are small and have higher metabolic rates and faster turnover times than do macrobenthos. Meiofauna therefore respond to disturbances over relatively short time scales (Coull and Chandler 1992; Carman et al. 1997). (2) They typically lack larval dispersing stages and spend almost all of their life cycle in the sediment. Thus emigration/immigration events do not ordinarily contribute substantially to their community structure, and causality of changes in abundance and community structure over time can be more reliably linked to a disturbance event. In addition to these practical advantages, the meiofauna are worth studying in their own right. They are the most abundant deep-sea metazoans, and their representation in the deep-sea fauna increases with depth (Thiel 1979). Further, their diversity in the deep sea far exceeds that of the megafauna and macrofauna (Lambshead 1993).
Taxonomic composition of the deep-sea metazoan meiofauna is typically dominated by nematodes, which often comprise > 90% of the total meiofauna. All known nematode feeding types are found in the deep sea, including selective and non-selective deposit feeders, epistrate feeders and omnivores/predators (Thistle et al., 1995). Nematode functional groups in the deep sea can also be distinguished on the basis of tail morphology (Thistle et al. 1995a; Vanreusel et al. 1997). Tail types are diverse and variable and tail and buccal morphology together have proven to be an effective method to discriminate nematode communities (Thistle et al. 1995a). Harpacticoid copepods are typically second in abundance in the deep sea; they are ubiquitous and become proportionately more abundant relative to macrofauna in the deep sea (Thistle 2001). Species diversity of harpacticoids in the deep sea is very high (Thistle 1983), and thus harpacticoids are well suited to studies of biodiversity. Other abundant meiofaunal groups in the deep sea include kinorhynchs and ostracodes.
Our research will examine structural (abundance, taxonomic and functional-group composition, and surface-dwelling meiobenthos) and functional (fecundity, body size/biomass distributions, lipid-energy reserves, and stable isotopic analysis of food resources) influences of CO2 exposure (and related changes in pH) on meiobenthos.
Goal 5 of the Carbon and Climate Working Group (1999) identifies the need Òto develop a scientific basis for evaluating potential management strategies to enhance carbon sequestration in the environment and capture/disposal strategiesÓ, and more specifically to Òdetermine the feasibility, environmental impacts, stability, and effective time scale for capture and disposal of industrial CO2 in the ocean and geological reservoirsÓ. At issue in this proposal is the degree to which CO2 sequestration disrupts deep-sea benthic communities. If CO2 effects are important to metazoans, the meiofauna can be expected to react. To assess this possibility, we will specifically test the following hypotheses concerning the influence of CO2 sequestration on the deep-sea meiofaunal community.
Some animals, such as some species of harpacticoid copepods (Bell et al. 1988) and some species of amphipods (Mees & Jones 1997) that live on or in the seabed can swim into the overlying water. This ability gives them the capacity to escape from deteriorating environmental conditions by entering the water column and being transported by near-bottom currents (Service and Bell 1987). This escape response provides a tool with which to determine if sequestered CO2 (directly or indirectly) stresses the benthic community. If CO2 exposure creates stressful conditions, species with the ability to swim should leave sediments and enter the overlying water. This behavioral response should occur on a time scale of minutes to hours and therefore be readily detectable in an experiment of several-weeks duration.
We propose to test for swimming-escape behavior by placing emergence traps approximately 1, 2, 5, and 40 m away from a CO2 source (ÒgradientÓ experiment described below), and thus sample a gradient of CO2 exposure. If CO2 is stressful, we expect that the number of animals caught in the traps will decrease with distance from the source. The influence of CO2 exposure on meiofaunal emergence will be determined by regression analysis relating meiofaunal abundance in traps to CO2 exposure (as determined by water pH values).
Meiofaunal abundances generally decrease with increasing depth in the seabed. In the deep sea, the decrease occurs on a scale of centimeters (Thistle and Sherman 1985; Tietjen et al. 1989). Meiofaunal species differ in their use of this aspect of their habitat. Some are surface specialists (living near the sediment-water interface) and can be identified by morphological traits (Coull 1972; Thistle 1982); others live well down in the inhabited layer (Joint et al. 1982; Foy and Thistle 1991). Almost all meiofaunal animals are motile and adjust their vertical positions in response to changing environmental conditions. Even in the deep sea, differences in factors such as sediment type (Carman et al. 1987) and hydrodynamic conditions (Thistle and Levin 1998) can alter vertical distributions. If the layer of CO2 interferes with the exchange of materials between the overlying water and the seabed, or alters pH in ways that affect metazoans, then the vertical profiles of the meiofauna will change. Meiofaunal movement rates are millimeters per minute, so a several-week experiment would detect effects of vertical environmental gradients created by CO2.
We hypothesize that surface specialists will be disproportionately impacted by exposure to CO2. This effect should be manifested as reduced abundances and/or displacement to greater depths within the sediment of species that are surface specialists. We will specifically examine the question of whether surface specialists move deeper into the sediment (as an alternative to swimming out of the sediment) when exposed to CO2. The influence of CO2 on vertical distribution will be examined by determining meiofaunal abundances in the 0-1 and 1-2 cm layers of sediment.
As noted under H1, some meiofaunal species may respond to elevated CO2 levels by leaving the sediment. Such an exodus would clearly reduce biodiversity in the community, especially if emigrating animals are unable to successfully reenter the seabed because CO2 deposition affected a large area. Animals that are unable to leave the sediment will also be at risk because of their inability to avoid adverse conditions by moving away from them. Thus biodiversity could be adversely affected by both emigration and toxic effects to obligately benthic taxa. We will test for changes in biodiversity of meiofauna by analyzing harpacticoid copepod species and nematode feeding and tail-morphology functional groups.
Elevated CO2 levels and/or associated changes in pH may create stressful conditions such as decreased respiratory efficiency. Neutral lipids (triglycerides) are the principle energy-storage material of harpacticoid copepods and can be depleted over a period of several days under stressful conditions (Carman et al. 1991). Neutral-lipid energy reserves will be measured in individual harpacticoid copepods, which will subsequently be identified to species. From these observations we will determine if the lipid-energy reserves of copepod species are differentially influenced by exposure to CO2.
H4: Meiofaunal fecundity and demographics will be adversely affected.
Many environmental stressors cause reduced rates of reproduction and/or disproportionately high mortality to early developmental stages. We will examine fecundity by measuring brood size in female copepods and examine effects on early developmental stages by determining the abundance of larval (nauplius) and juvenile (copepodite) harpacticoid copepods. We will also examine meiofaunal biomass of nematodes, which is correlated with fecundity, by measuring biovolume with image-analysis techniques to estimate body-size distributions. Nematode biovolume will be converted to biomass by the methods of Feller and Warwick (1988).
The Barry-Brewer DOE research project involves the application of liquid CO2 to the seabed in experimental corrals at depths ranging from 3200-3600 m in the Monterey Canyon. Experiments are conducted with the aid of an ROV (Tiburon) operated from the R/V Western Flyer. Technical details of experimental design and application of CO2 to the seabed are described in the Barry-Brewer proposal. We will obtain samples for analysis of meiofauna by participating in their ongoing experimental work. Their experimental design is evolving as they discover better methods for applying CO2. At present, they are doing two types of experiments, both of which involve the application of CO2 to circular PVC corrals (50 cm dia) that are pushed into the seabed. In the first type of experiment (= fixed-treatment experiment), corrals extend approximately 15 cm above the sea floor. Corrals are filled with CO2, and the CO2 is replenished after approximately 2 weeks. Liquid CO2 gradually moves out of corrals and forms a cap on the surrounding seabed. After approximately seven weeks of treatment, 7.6-cm dia sediment cores are collected from around the periphery of the corrals. In each experiment, 3 replicate CO2 corrals, 3 control (no CO2) corrals, and 3 ambient (no corrals) sites are sampled (the latter as a control for the effect of corrals). The second design (= gradient experiment) is similar to the first, except that corrals are inserted so that they extend approximately 33 cm above the seabed. This approach allows for the containment of a larger volume of CO2, which subsequently spreads over a larger area. Samples (7.6-cm dia sediment cores) are collected at intervals of 1, 2, 5, and 40 m along a transect that extends away from the corral; a second transect is sampled on the opposite side of the corral. CO2 exposure is determined from pH, and observations by Drs. Barry and Brewer indicate that pH increases at a rate of 0.2-0.3 pH units per 5 m away from the CO2 source, thus allowing for good resolution of CO2 effects that can be realistically sampled using the ROV. Two replicates of the gradient design are setup for an experiment.
We plan to participate in 2 experiments per year over a 2-year period (one of each type described above per year). We currently plan to deploy emergence traps only during gradient experiments, but, if sufficient ROV time is available, emergence traps will also be deployed during the fixed-treatment experiments.
Three replicate 7.6-cm dia cores will be collected at each fixed-treatment or gradient sampling point. We will collect a 2.5-cm dia subcore from each core, which will be vertically sectioned (0-1 and 1-2 cm) and preserved in buffered 4% formaldehyde for quantitative determination of meiofaunal community structure (abundances and biomass, harpacticoid species identifications, nematode functional groups, and harpacticoid copepod fecundity (described below)). Meiofauna will be sorted and quantified from each vertical layer of each subcore. In fixed-treatment experiments, data from the three subsamples will be pooled for statistical comparisons of CO2 treatments versus controls. In the gradient experiment, data from the three subcores will be used as replicates for each point on the exposure gradient, and used in a regression analysis of the effects of CO2 exposure.
Analyses of lipid-energy reserves are best determined on specimens frozen in liquid nitrogen. Because of limitations to ROV time, we will obtain samples to freeze from Jim Barry's macrofauna cores by collecting the material that passes through a 0.3-mm sieve, on a 62-µm sieve. This smaller fraction contains the meiofauna, which we will freeze in liquid nitrogen. Animals from these frozen samples will be used for analyses of lipid-energy reserves.
Meiofaunal Community Structure
Meiofauna will be stained with Rose Bengal, sorted to major taxon, and enumerated using the method of Sherman et al. (1984), which allows for subsampling of an abundant taxon (nematodes), and complete sampling of other taxa (e.g., harpacticoid copepods, kinorhynchs, ostracods, etc.). More detailed analysis will be performed on harpacticoid copepods and nematodes. Harpacticoids will initially be characterized by determining size structure, fecundity (number of egg-bearing females and number of eggs per ovigerous female), and demographic structure (juvenile: adult ratio). Adult harpacticoids will be identified to species, and the abundance of surface- and subsurface-dwelling species will be quantified. Nematodes will be analyzed by the use of image analysis to estimate mean body size (biovolume) and biomass, and identified to major functional group using: (1) the trophic-group approach (Wieser 1953; Moens & Vincx 1997), which is based on buccal morphology and provides insight into the types of food ingested, and (2) the tail-morphology approach (Thistle et al. 1995a), which indicates life styles (e.g., differences in mobility). The effect of CO2 on nematode body-size distributions will also be examined.
We estimate that there will be a combined total of approximately 50 harpacticoid copepods in the three samples collected at each site (most of which will be in the top 1 cm of sediment; Thistle 2001). Thus, from the 2 years of sampling, approximately 4200 copepods will be sorted and analyzed in two stages: (1) size structure, fecundity, etc, and (2) species identification. One hundred nematodes from both the 0-1 and 1-2 cm layers (total 200 per core sample) will be analyzed in two stages: (1) buccal morphology, and (2) tail-morphology. Thus, from the 2 years of sampling, approximately 16,800 nematodes will be analyzed.
Copepod Lipid Reserves
Sediment samples frozen at sea in liquid nitrogen will be shipped to the laboratory on dry ice, and stored at Ð80¡C. In the laboratory, a sample will be thawed and washed on sieves of 500, 250, 125, and 62 µm to create aliquots of uniform size for sorting. The unstained harpacticoids are removed from each aliquot under a dissecting microscope. The harpacticoids are then stained with Nile red (Carman et al. 1991). After staining, each individual is mounted in a standard position in a depression slide and placed under epifluorescent illumination. With the aid of an image-analysis system, the operator measures the area of the lipid droplets, which glow bright yellow on a dark background. Switching to bright-field illumination, the operator traces the outline of the animal's body. The animal is then preserved in formaldehyde on a spot plate, transferred to a glycerin drop on a microscope slide, and identified with the aid of published and working keys (Thistle et al. 1995b). The procedure allows the assessment of changes in neutral lipid (an indicator of stress) on a species- and biovolume-specific basis among experimental treatments.
We anticipate analyzing the lipid reserves of approximately 50 individual copepods per sample location. Thus, from the 2 years of sampling, approximately 4200 copepods will be analyzed in two stages: (1) lipid content, and (2) species identification.
Emergence Traps
In shallow water, inverted-funnel traps have been used to catch benthic taxa that swim from sediments into the overlying water (Hicks 1986). A trap consists of a base and a collecting chamber (an inverted funnel and a closed space above it). For deployment, the trap is inserted into the seabed and the collecting chamber opened. Any animals that swim out of the enclosed sediment are directed by the inverted funnel into the catch chamber. At the end of the experiment, the collecting chamber is closed, and the trap is recovered. Collected animals are preserved in buffered 4% formaldehyde quantified. We have successfully tested a prototype, deep-sea trap with DSRV Alvin. Although traps will need to be fabricated for this project, the design can be used essentially as is with the ROV. Animals recovered in emergence traps will be sorted to major taxon. Harpacticoid copepods will be identified to species, and nematode feeding and tail-morphology groups will be determined as described above.
Because essentially nothing is known about emergence of meiofauna in the deep sea, we cannot predict the number of animals that will collected and sorted from emergence-trap samples. If abundances are low, all individuals will be sorted and enumerated. If the abundance of one or more taxon is high, the subsampling method of Sherman et al. (1984; described above) will be used.
Responsibilities
As noted above, several thousand meiofaunal harpacticoids and nematodes will be individually analyzed. Sorting and analysis of meiofauna is labor intensive, and each of the three PI's will directly participate and contribute significantly toward the project.
As PI, Carman will be responsible for project coordination. He will work with Dr. Barry to coordinate sampling and will travel to California to participate in field sampling. Carman will be responsible for bulk sorting of meiofaunal samples from sediment samples and emergence traps. Carman will also be responsible for stable-isotope analyses.
Fleeger will be responsible for demographic/fecundity analysis of harpacticoid copepods. He will also analyze the biomass and size-structure of nematodes in addition to determining functional groups based on tail morphology.
Thistle will be responsible for identification of harpacticoid species and analysis of their community structure, including functional-group analysis. He will also take the lead in the field implementation of emergence-trap studies, and will participate in the yearly cruise in which emergence traps are deployed and retrieved. Carman and Thistle will work together on the Nile Red analysis of copepod lipids.