Robert F. L. Self

A SIMPLE MODEL OF FLOW-SEDIMENT-ORGANISM INTERACTION

Abstract Previous models of organism-sediment interaction have been limited primarily to eddy diffusion-type mixing. While such models are useful in interpreting some stratigraphic records, they cannot be expected to provide mechanistic insights into the diversity of interactions among sedimentary processes, organism activities, and boundary layer flows. To provide a more general formulation capable of dealing with all three types of processes in concert, a simple, discrete-time Markov model is generated. This version of the model deals explicitly only with particulate and not liquid phases. The initial formulation is a two-compartment, ergodic model of deposit feeding. Particle selection by deposit feeders results in incorporation of the selected particles into fecal pellets, and fecal pellet disaggregation completes the cycle. Particles which are more strongly selected or which are incorporated into more robust pellets thus spend more time and reach higher relative concentration in the fecal-pellet compartment. The next level of complexity is achieved by adding burial from the “free” sediment (non-fecal-pellet) compartment. This model predicts that particles which are more strongly selected or incorporated into more robust fecal pellets will have greater residence times in surficial sediments (both pellets and “free” sediments) before they become part of the stratigraphic record, and will therefore achieve higher concentrations in surficial sediments than will less preferred particles. The full flow-sediment-organism system is reached by adding the potential for lateral advection of particles. The probability of such advection is allowed to differ between pellets and “free” sediments. Increasing selection and robustness of fecal pellets then can either increase or decrease relative concentration and residence time of a given particle type in surficial sediments. The model thus provides a means of explaining differences in stratigraphic patterns among materials deposited simultaneously. It also suggests that deposit-feeder selection may increase the average age (i.e., time since initial deposition) of surficial sediments by maintaining preferred particle types near the surface. Selection may likewise control the texture of surficial sediments-in a manner which is easily confused with purely physical sorting. Parameters which require careful measurement in testing these simple predictions are selectivity by deposit feeders (probability of selection), rate of fecal-pellet breakdown under realistically simulated conditions (probability of pelletal disaggregation), differential erodibilities of pellets versus “free” sediments (relative probabilities of lateral advection), and rates of burial of “free” sediments versus pelletized sediments (probabilities of burial).

The use of radon-222 as a tracer of sediment irrigation and mixing on the Washington continental shelf

Abstract Observations of the natural radioisotope Rn-222 were used in conjunction with radiochemical, biological, and sedimentological techniques to investigate the irrigation and mixing of sediment at three sites (about 90 m water depth) along the Washington continental shelf. The Rn-222 deficit (difference between observed and predicted concentrations) within the seabed was greatest near the Columbia River during the summer. Molecular diffusion at most locations and times could account for only a portion of the deficit; irrigation and mixing of the seabed explain the remainder. Biological mixing (bioturbation) and physical irrigation (wave and tidal pumping) have relatively little effect on the Rn-222 deficit. Physical mixing (erosion and deposition) has a significant effect soon after a major erosion event (storm); however, no such event occurred prior to the observations. The conclusion is that biological irrigation and molecular diffusion are the primary causes of the Rn-222 deficit, and that biological irrigation is responsible for the observed spatial and temporal variations in the deficit. Therefore, biological irrigation is greatest near the Columbia River during the summer. The spatial increase of irrigation near the Columbia is probably due to a relatively large population of tube-dwelling, filter-feeding polychaetes. The temporal increases of irrigation during the summer is probably due to annual growth cycles of the benthic community, and is not due to water temperature changes.

Gut-marker and gut-fullness methods for estimating field and laboratory effects of sediment transport on ingestion rates of deposit-feeders

Abstract By developing a method for quantifying feeding rate in the presence of sediment transport, we determined whether sediment transport or one of its correlates influenced feeding rate of deposit-feeders in the field. The method entailed application and subsequent removal of magnet-susceptible marker particles, so that no marker was present in the ambient sediments during the period over which feeding rate actually was measured. Error variance was reduced by a paired statistical design in which the technique was used twice in the same individuals, to assess feeding rate immediately prior to the sediment transport event as well as during it. Low levels of sediment transport significantly affected feeding rates of Pseudopolydora kempi japonica Imajima & Hartman (a spionid polychaete) both in a small, imprecise, laboratory flume and in an intertidal environment characterized from late spring through early fall by a short pulse of sediment transport coincident with the incoming tide. Deposit-feeding rates immediately after the latter pulse were enhanced by a (geometric) mean of 50% in field trials from May through October. Both in the laboratory and in the field, however, the magnitude of the (albeit statistically significant) sediment transport effect was overshadowed by variation due to nutritional history of individuals and to other unidentified factors. The form of the field data for P. kempi japonica (feeding rates estimated for two successive intervals in the same individuals) allowed us, in addition, to analyze the response of gut fullness to mean feeding rate (averaged over the two feeding bouts), previous feeding rate (first bout), most recent feeding rate (second bout) or recent change in feeding rate (difference between first and second bouts). The strongest correlation was found with the last of these variables, suggesting that gut fullness may provide a tracer-free method of gaining some kinetic information on continuously feeding animals. Contrary to prevailing impressions, gut fullness in such continous feeders may provide more information on recent change in feeding rate than on feeding rate itself. We gathered additional data on P. kempi japonica that differed in emersion history (in tide-pools versus outside them prior to the pulse of sediment transport with the incoming tide) but were not exposed to markers. Animals both inside and outside tide-pools showed fuller guts immediately after the sediment transport pulse, consistent with an interpretation of accelerated feeding caused by some aspect of the sediment transport event. Incidentally collected individuals of Hobsonia florida (Hartman) (an intertidal ampharetid polychaete), on the other hand, showed no obvious response to the same conditions that triggered accelerated feeding and fuller guts in Pseudopolydora kempi japonica.

Adhesive-based selection by a tentacle-feeding polychaete for particle size, shape and bacterial coating in silt and sand

We tested particle selection by a surface deposit-feeding, tentaculate spionid polychaete, Pseudopolydora kempi japonica Imajima and Hartman. In experiments with peroxide-cleaned, sizegraded but otherwise natural silts and sands, individual worms showed peak preference for particles 80‐99 mm in diameter, compared with previous (Self and Jumars, 1988) documentation of peak preference for particles of 7 mm in experiments conducted with plastic and glass beads. These results imply that microtektites will not in general be good tracers of mixing of mineral grains of comparable size. Animals exhibited statistically signie cant but not marked differences in size selectivity for subrounded versus subangular grain shapes; the size preference peak was broader in subangular grains, for which orientation of the grain can alter probabilities of both contact and retention. When one size class of grains was coated with the bacterium Halomonas halodurans (ATCC 29686), animals in general showed enhanced selection (relative to controls with no food value on any size class) of that size and smaller grains, even though these smaller grains lacked food value. Greatest selection, however, generally occurred for the coated size class. Results frominclusion of glass beads in some of the experiments and from separate experiments with tentacle analogs imply that this selective capability may be largely passive and mechanical. Natural grains, due to surface texture, have more surface area for adhesive contact than do smooth glass beads so that larger grains than beads are retained. Bacteria-coated grains, in turn, appear to be picked up preferentially due to adhesion with the bacterial coating; petroleum jelly-coated microscope slides also succeed in selective retention of the size class that is bacterially coated. Substantial selection by adhesiveutilizing deposit feeders apparently can be achieved without investment in complex, time- and energy-consuming sensory systems and behaviors. Experiments with natural grains showed notably more scatter than prior experiments with glass beads, but this difference is consistent with the mechanism. For nonspherical particles, both contact and retention depend on orientation as well as size.

In vitro amino acid and glucose uptake rates across the gut wall of a surface deposit feeder

Abstract We used the everted-sleeve method (Karasov & Diamond, 1983) and dual radiolabelled substrates to measure carrier-mediated and diffusive fluxes of glucose and four amino acids (AA) across the gut wall of the mobile, surface deposit-feeding holothuroid Parastichopus californicus Stimpson. The relationship between substrate concentration and carrier-mediated uptake rate for d -glucose, l -aspartic acid, l -glutamic acid, l -methionine, and l -tyrosine was adequately described by the classic Michaelis-Menten equation. Passive uptake rates were controlled by substrate interaction with the hydrophobic gut membrane. For the hydrophilic compounds (glucose, methionine and aspartic and glutamic acids) passive uptake contributed 11% or less to total uptake but accounted for 21% of the total for the rare but hydrophobic AA l -tyrosine. Carrier-mediated uptake rates of hydrophilic (and lipophobic) AA were inversely proportional to the relative availability of substrate in sedimentary food. Balance in absorptive delivery was achieved by compensating for rarity in sedimentary food (low luminal concentration) and retarded cross-membrane diffusion with added carriers. Thus the relatively rare (in the food) and lipophobic AA l -methionine possessed the highest carrier-mediated maximal flux ( J max = 57 pM (mg wet mass of gut tissue) −1 min −1 ) as well as the highest apparent half-saturation concentration. Glucose, aspartic and glutamic acid, all relatively common sediment organic constituents with intermediate passive permeabilities, had intermediate J max values ranging from 6 to 26 pM · mg −1 · min −1 . The rare but lipophilic and thus highly permeant AA l -tyrosine had the lowest J max of 2.6 pM · mg −1 · min −1 , with passive diffusion accounting for substantial uptake. The distribution of nutrient uptake rates along the gut depends upon substrate carrier densities and luminal concentrations of the products of digestion. Rates interpolated at regional (fore-, mid- and hindgut) concentrations were high in the foregut and midgut but decreased in the hindgut for the rare AA methionine and tyrosine; conversely they were low in the foregut and high in both midgut and hindgut for the common AA glutamic and aspartic acid. Despite these varying patterns of AA flux along the gut, when integrated over the whole gut length to estimate total amount of AA absorbed, just as in the sediment food and animal tissue the acidic AA dominated while tyrosine was rare. The total amount of methionine absorbed far surpassed tissue requirements, however, suggesting that it is hyperessential. Our results indicate that AA compositional equality of uptake ratios with food and tissue composition ratios serves as a useful null hypothesis for identifying hyperessential nutrients. Karasov has used AA: sugar uptake ratios to classify vertebrates as herbivores, omnivores or carnivores. By this criterion Parastichopus californicus is clearly a herbivore, showing a very low ratio.