Taphonomic processes in a deep water Modiolus-brachiopod assemblage from the west coast of Scotland
Taphonomy `the study of processes of preservation and how they affect information in the fossil record' (Behrensmeyer and Kidwell, 1985) is an important geological discipline. This study, by incorporating experiments on carbonate transplanted to shallow water with benthic sampling, explored the relationship between the calcimass of living and dead (sub-fossil) constituents of a deep water brachiopod-bivalve community in the Firth of Lorn (site 1. of Curry, 1982). The community, which occurs on a gently sloping sea floor at depths between 160 and 200m, was investigated by dredging, grab sampling, coring and video transects. Biomass and calcimass of the community were determined indirectly from regressions of shell length against weight on animals collected from 10 grab samples. The large endobyssate bivalve M. modiolus, which forms large stable clumps modifies the sediment by introducing `secondary hardbottom' (sensu Surlyk, 1972) into an otherwise soft sediment. The valves of living M. modiolus serve as substrate for a diverse epifauna, the most distinctive member of which is the articulate brachiopod Terebratulina retusa. The standing calcimass (2370.5 g/m2/yr) and carbonate production (excluding polychaetes, bryozoans and barnacles) of the community is very high (330 g/m2/yr) which in the absence of carbonate destruction would result in a rate of autochthonous carbonate accumulation in excess of 1 mm/yr. M. modiolus contributes 93.5% of the standing calcimass, but accounts for only 37.8% of estimated production (mainly due to a lifespan estimated at 40 years). Three other species, the ophiuroid Ophiothrix fragilis, the bivalve Astarte sulcata and the articulate brachiopod Terebratulina retusa together contribute an additional 58.5% of carbonate production. Length-frequency histograms of 0-year class T. retusa from seasonal samples suggest that growth rate is initially slow (an increase in length from 0.212 mm to 0.539 mm over the first 260 days). Similar rates have been published from laboratory studies (Rickwood, 1977; Stricker and Read, 1985) but growth rate is well below that estimated from conventional length-frequency histograms (eg. Curry, 1982) or field studies (Thayer, 1977; Doherty, 1979). Autecological implications of a strongly sigmoidal growth curve were investigated by examining substrate related mortality. Larval T. retusa appear non-selective in their choice of substrate and although virtually all adult T. retusa are attached to mature M. modiolus, 35 alternative substrates were recorded. Patterns of substrate utilization suggest that from a length of approximately 2 mm to maturity there is an increase in the proportion of T. retusa attached to the surface of M. modiolus, indicating that alternative substrates (eg. hydroid thecae, ascidian tests and calcareous worm tubes) are more liable to fail as the Terebratulina mature. At lengths below 2 mm the surface of M. modiolus appears to be sub-optimal, the proportion of Terebratulina utilizing this substrate falling from approximately 80% of settlement to 40% at lengths of between 1.7 & 2.8 mm. It is postulated that grazing pressures, believed to be restricted to this substrate (Akpan, 1981) may account for this inflection at lengths of approximately 2.3 mm grazers being unable to disloge brachiopods above this critical size. Estimates of carbonate production were compared with the composition of biogenic carbonate from > 4 mm, > 2 mm, and > 1 mm sediment fractions of grab and box core samples. Ophiothrix fragilis and Amphiura chaijei which together were estimated to contribute 31.6% of total carbonate production accounted for only 0.5% of autochthonous carbonate in these fractions. Even accounting for the concentration of ophiuroid debris in the finer fractions, the absence of ophiuroid debris is remarkable. Excluding ophiuroid debris and the two most minor autochthonous carbonate contributors, reveals an apparent trend in the remaining autochthonous carbonate towards selective preservation bias of smaller items. These trends were not seen in field experiments of relative rates of abrasion and laboratory studies of relative dissolution rates of the major carbonate components. Furthermore, estimated rates of destruction by abrasion were very low and S.E.M. examination of grains recovered from the death assemblage did not show evidence of dissolution. This size related bias may therefore provide evidence of the consequence of selective bioerosion of non-agitated grains first proposed by Boekschoten (1968). Terebratulina retusa is an outlier to this trend, being less common than estimates of production and preservation would predict, corresponding with the semi-quantitative evidence for under-representation of articulate brachiopods in death assemblages from the Algerian shelf (Caulet, 1967) and Canadian sub-littoral (Nobel et al. 1976). T. retusa shells recovered from the sediment were extremely friable. Initially this was thought to be a localized dissolution phenomenon, however S.E.M. preparations subsequently demonstrated that the observed strength loss was a consequence of a skeletal architecture which is common to the majority of articulate brachiopods. The rate of `softening' was determined experimentally and proved to be remarkably rapid (93% strength loss 200 days after death, as measured by point loading; Collins, 1986). The friable carbonate liberates large numbers of distinctive calcite fibres into the sediment which should be recognizable as microfossils. Similar observations of structurally weakened brachiopod carbonate have subsequently been reported from the Norwegian coast, the Mediterranean, western USA and New Zealand. The reduced strength of brachiopod shells will increase the potential for mechanical and bio-mechanical fragmentation and therefore under-representation. It is probable that the magnitude of `softening' is inversely related to size and is liable to bias against smaller shells (Collins, 1986). A significant deviation from the expected 1:1 ratio of pedicle to brachial valves of T. retusa (bv/pv = 1.6) was similar to the figure for Terebratulina septentrionalis from a shallow subtidal rocky coastline on the east coast of Canada (bv/pv = 1.5: Noble & Logan 1981). The preferential preservation of brachial valves of T. retusa compares with a strongly biased preservation of pedicle valves for Macandrevia cranium from the Scottish shelf and Gryphus vitreus from Corsica. It is suggested that biased valve ratios of these four Recent examples are the consequence of selective destruction, selective transport need not be invoked. Transport was also investigated experimentally, by laying a representative sample of shells on a moderately exposed sediment surface, at a depth of 9 m. Positions were recorded in a series of photographic mosaics and final recovery included suction sampling to a depth of approximately 30 cm. The large Modiolus valves moved the greatest distances, an unusual result thought to be due to reworking of larger items by crabs with a biologically bound and bioturbated sediment and entrapping the smaller shells and valves. It is suggested that the significance of current-mediated transport has been overstated, due to over concentration on the swash zone and flume tank and the failure to recognize the potential for biologically mediated transport and sediment binding.