«Benthos: Trophic modelling of the Ross Sea M.H. Pinkerton, J. Bradford-Grieve, D.A. Bowden National Institute of Water and Atmospheric Research Ltd ...»
Benthos: Trophic modelling of the Ross Sea
M.H. Pinkerton, J. Bradford-Grieve, D.A. Bowden
National Institute of Water and Atmospheric Research Ltd (NIWA), Private Bag 14901,
Wellington 6021, New Zealand.
Email: firstname.lastname@example.org; Tel.: +64 4 386 0369, Fax: +64 4 386 2153
Although the Ross Sea benthos has been extensively studied from the late 19th century to present-
day (e.g., Borchgrevink 1901; Littlepage & Pearse 1967; Bullivant & Dearborn 1967; Bullivant 1967a, b; Dayton & Oliver 1977; Lipps et al. 1979; Battershill 1989; review by Starmans et al.
1999; Gambi & Bussotti 1999; Barry et al. 2003; Rehm et al. 2006), large-scale estimates of benthic biomass in the Ross Sea is limited. Sampling methods and lack of calibration of numbers to biomass are at the base of difficulties in arriving at consolidated mega- and macrobenthos biomass data.
Some key features of the Antarctic benthos are given below (e.g., Arntz et al. 1997).
There are distinct differences between various benthic subsystems (30m, shelf, slope, deep water).
Distribution of benthic macrofauna biomass is very patchy. There are areas of the Ross Sea that contain extraordinarily high benthic faunal abundances and others that have relatively low biomass.
Some benthic production is linked to surface production, at other times it is decoupled.
Two types of method have been used commonly to study the benthic ecosystem of the Ross Sea:
“remote” and “direct” measurements. Diver-swum transects give an indication of the spatial distribution of macro and mega-benthic fauna in terms of abundance (numbers of individuals) and diversity (number and types of species) but are restricted to shallow waters (50 m). In deeper water, sampling using “remote” methods such as camera (video or still) systems have been used to obtain data on what is living on, or extending from, the sediment surface. Camera resolution can be an impediment to observing smaller individuals in the frames. Remote measurements do not directly measure biomass or production of different organisms and are usefully combined with direct sampling (e.g., Cummings et al. 2003; Mitchell & Clark 2004; Hanchet et al. 2008).
If box or other corers are used, then a much smaller area is sampled and smaller organisms enumerated. However, it is generally not possible to sample densely enough to elucidate the patchiness in the distribution of benthos biomass. Also, it is common for workers to use different mesh sizes to separate benthic organisms from sediment making combining different studies difficult or impossible.
1.1 The benthic ecosystem sub-model The conceptual structure of the benthic ecosystem used in this study is based on a widely-used but simple energetic model of benthic communities (e.g., Smith 1987, 1989; Christiansen et al.
2001; Gage 2003; Piepenburg et al. 1995; Nodder et al. 2003; Bradford-Grieve et al. 2003). The conceptual model of the Ross Sea benthos used in the present study has three components: (1) megabenthos; (2) macrobenthos; (3) meiobenthos. Benthic bacteria and benthic detritus are
It is unclear whether a more detailed subdivision of the benthic community would benefit the overall trophic model. Jarre-Teichmann et al. (1997) developed a trophic model of the benthic shelf community of the eastern Weddell Sea, dividing the benthic macrofauna into the following compartments: Crinoidea, Holothuroidea, Ophiuroidea, Mollusca, Bryozoa, Polychaeta, Asteroidea, Porifera, Echinoidea and Tunicata/Hemichordata. Such a detailed subdivision of the benthic community is not common in trophic models, because the benthos tends to be relatively poorly characterised in terms of biomass, and spatial and/or temporal variability. Data also becomes progressively scarcer as the water depth increases, especially in Southern Ocean regions.
The structure and function of the benthic ecosystem and the characteristics of the bentho-pelagic coupling are typically unevenly distributed at a range of spatial scales, depending on factors including substrate, water depth, ice-cover, and proximity to primary producers (macroalgae, phytoplankton, epontic algae). Some of the environmental factors that are likely to exert some control on benthic faunal biomass also have a temporal variation (especially ice cover and primary production). The complexity of the relationship between benthic faunal density and environment, and the patchiness of the distribution, makes it difficult to estimate a “characteristic” biomass, structure and function of the Ross Sea benthic fauna. As a starting point, we estimate megafaunal and macrofaunal biomass for the coastal and offshore regions of the Ross Sea separately as explained below. For meiobenthos, we use a relationship between meiofaunal biomass and depth.
2.1 Weights and carbon content conversions We require knowledge of weights of individual megabenthic organisms to convert measured abundances (ind/m2) to biomass density. The wet mass of several taxa are greatly biased by water content, massive inorganic outer shells and/or inorganic carbon-rich (CaCO3) skeletal material, and variable amounts of organic carbon as a percentage of wet and dry weights (Rowe 1983).
Some of the megabenthos is not appropriately enumerated in terms of abundance of individuals, including structure forming “massive” organisms and colonial species. For two groups (porifera, ectoprocta), measurements of abundances are in terms of area cover, and these are converted to biomass using an estimate of wet weight per m2 for that organism.
As yet, the weight relationships of Ross Sea benthos has been worked out only for some shallow water hard bottom organisms in Terra Nova Bay (Gambi et al. 1994), soft bottom shallow water polychaetes (Gambi et al. 1997), the shallow water nemertean Parborlasia corrugatus (Heine et al. 1991), the echinoid Sterechinus neumayeri (Brey et al. 1995) and benthic littoral communities (Cattaneo-Vietti et al. 2000). In addition, because of the limited resolution of underwater imagery, many benthic taxa measured remotely cannot be identified to species level, making estimation of “typical” sizes from the literature uncertain. Here, typical individual weights were obtained by weighing specimens collected from the Ross Sea on the recent New Zealand IPYCAMLR voyage (Hanchet et al. 2008) and shown in Table 1.
Table 1 also shows typical individual weight, area weights, and conversion factors between wetweight and carbon, as wet weights must be converted to organic carbon content for modelling.
Wet-weight to carbon conversion factors were taken from a number of publications including 2 Vinogradov (1953) (various groups), Galeron et al. (2000) (various groups), Dayton et al. (1974) for Porifera, and Brey (2005) for Holothurians. Proportions of carbon associated with living material rather than inorganic skeletal material were estimated as by Lundquist & Pinkerton (2008).
Biomass of megabenthos in the Ross Sea is estimated from three sets of data.
First, there are a number of studies of the near-shore Ross Sea megabenthos in waters shallower than 30 m depth. The majority of studies of megafauna have been conducted in the McMurdo Sound and Terra Nova Bay regions (Dayton et al. 1969, 1970, 1974, 1994; Dayton & Oliver 1977; Oliver & Slattery 1985; Battershill 1989; Dayton 1990; Lenihan 1992; Lenihan & Oliver 1995; Brey et al. 1995; Chiantore et al. 1998; Cattaneo-Vietti et al. 2000; Gambi et al. 2000;
Heilmayer et al. 2003). Even though waters less than 100 m deep make up 1% of the total Ross Sea study region, these are considered separately because shallow waters may contribute a disproportionate amount to the total megabenthic biomass of the Ross Sea. Biomasses and densities of the molluscan species in Terra Nova Bay are reported in Cattaneo-Vietti et al. 2000 (see Table 4 in that paper). Their study found the Antarctic scallop (Adamussium colbecki) to be the most common species of mollusc in Terra Nova Bay, with densities up to 59 ind m-2 (25 average). The study showed that the bivalve assemblage was diverse, with A. colbecki making up only 12% of the mollusc individuals (by number) on average. A. colbecki was also found to be common further south in McMurdo Sound, where abundances up to 85 ind m-2 occur at between 4–15 m (Stockton 1984). Typical dry weights of tissue of individuals are 0.2–4 gDW ind-1 for shell heights between 20 and 80 mm (Heilmayer & Brey 2003). A population median individual size may be c.50 mm (Heilmayer et al. 2003) and an average individual weight may be of the 3 order of 1.3 gDW ind-1. This gives an average biomass density of A. colbecki of 30 gDW m-2 in Terra Nova Bay and McMurdo Sound. This range of density is low compared to estimates of biomass of A. colbecki at New Harbour, McMurdo Sound, where biomass measurements range from 59–66 gDW m-2 (Brey & Clarke 1993), and to up to 120 gDW m-2 for a 20–40 m population with densities of around 60 ind m-2 (Road Bay, Terra Nova Bay; Chiantore et al. 1998). Carbon is assumed to make up about 34% of dry weight of molluscs (Brey 2005). Hence, a lower bound on the density of A. colbecki alone is estimated to be of the order of 11 gC m-2 in parts of Terra Nova Bay and McMurdo Sound. Benthic megafauna in Terra Nova Bay and McMurdo Sound regions also include the regular urchin Sterechinus neumayeri in addition to exceedingly high densities of the infaunal bivalve Laternula elliptica that have been found at Faraglione (Terra Nova Bay) at depths below 25 m (S. Thrush, N. Andrew and G. Funnell, unpublished data). These studies would hence suggest megafauna biomass densities in some coastal areas may be substantially greater than 11 gC m-2, though this includes inorganic carbon in the shell.
Second, we consider data from the ROAVERRS (Research on Ocean/Atmosphere Variability and Ecosystem Response in the Ross Sea) research cruise studying megabenthos of deeper Ross Sea (Figure 1a). This voyage sampled two areas: (1) along the coast from Cape Adare and Terra Nova Bay, out to 500 m depth; and (2) in the Ross Sea from 300 to 1200 m depth (Barry et al. 2003).
Barry et al. (2003, see Table 10) gives data on the abundance of benthic megafauna over large areas of the Ross Sea from this program. Dr Jim Barry has kindly provided these data from 55 stations in the Ross Sea to this study. Data were gathered using a towed camera system but organism size or biomass were not measured. We assume that the abundances given by Barry et al. (2003) include the major contributors to the biomass of the megabenthos, though note that smaller organisms may be under-represented.
Third, data were obtained from the NIWA Deep-water Towed Imaging System (DTIS) on the New Zealand IPY-CAML voyage to the Ross Sea (Hanchet et al. 2008). This voyage completed tows of the video imaging system on the Ross Sea shelf (8 tows), slope (8 tows), deep water within the study area (2 tows), and deep water north of the study area (8 tows): Figure 1b. “Shelf” is all areas landward of the 600 m depth contour; “slope” is depths 600–1800 m in the shelf region; and “deep” is all areas 1800 m in depth in the study area, and deeper than 1000 m to the north of the study area. Data on major megabenthic groups were obtained in “real-time” onboard the vessel. More extensively processed data from the voyage will be available in due course, but these preliminary data are the best available results at present (April 2009). The data were merged onto the common set of megabenthic groups given in Table 1. ROAVERRS data are likely to be more quantitative, as the optical resolution of the images are higher and the still images have been subject to more detailed processing than the underway IPY-CAML video data. To reconcile the IPY-CAML video data with the ROAVERRS data, we calculated log-average values for each benthic group of biota from the region of overlap in the Ross Sea shelf (73–77°S, 167°E–180°).
Log-averages were used to reduce biasing of the average by occasional high values. This overlap consisted of 9 IPY-CAML stations and 31 ROAVERRS stations. Where the ratio of the logaverages for a particular group between the two surveys was between 0.1 and 10, we adjusted the IPY-CAMLR data by this value. This was the case for Asteroid, Ophiuroid, Echinoid, Holothurian, Crinoid, Mollusc, Annelida, Pycnogonid, Hydrocoral, Ascidian, Alcyonacea, Pennatulacea, Gorgonacea and Hexacoral groups. The abundances of Arthropod_shrimp and Hydroid groups measured by IPY-CAMLR on the shelf were very much lower than those measured by ROAVERRS (factor of 190 for Arthropod shrimp and 26 for Hydroids). This is probably because the resolution of the video data from IPY-CAMLR is sufficient to see animals greater than about 5 cm in size whereas the still images used on ROAVERRS data mean that individuals 2 cm are likely to be counted. Both these groups include many small individuals in the 2–5 cm size range. We used only data from ROAVERRS on the shelf, and unadjusted data 4 from IPY-CAMLR on the slope and shelf. We acknowledge that biomass values for Arthropod_shrimp and Hydroid may consequently be underestimated on the slope and deep water. No area coverage measurements of Porifera and Ectoprocta are currently available from the IPY-CAML voyage so we used individual counts along the transects as an indicator of abundances of these groups adjusted to match the log-average percentage cover values from the ROAVERRS voyages. Final estimates of biomass for all groups are given in Table 2.
Figure 1. Location of stations from the a: ROAVERRS and b: IPY-CAML benthic surveys of the Ross Sea region.
Combining these data in the appropriate proportions for the study area of the trophic model allows us to estimate megabenthic faunal abundance for the whole study region (Table 2). The average abundance of benthic individuals in the present trophic model study region was 0.15 individuals/m2. Our data show that the benthic megafauna of the deeper Ross Sea was dominated in terms of carbon biomass by anenomes (22.0%), holothurians (16.1%), ophiuroids (12.1%), and porifera (10.6%). Combining these components gives an average megafaunal biomass density for the non-coastal waters of the Ross Sea of 1.4 gC m-2.