Transport and breakdown of fecal pellets: Biological and sedimentological consequences

Limnol. Oceanogr., 29(l), 0 1984, by the American

1984, 64-72 Society of Limnology

and Oceanography,

Inc.

Transport and breakdown of fecal pellets: Biological and sedimentological consequencesl Gary L. Taghon,= Arthur R. M. Nowell, and Peter A. Jumars School of Oceanography

WB- 10, University

of Washington,

Seattle 98 195

Abstract Fecal pellets of benthic animals are important in sediment transport processes, yet few quantitative data are available on their salient physical characteristics. We measured, directly and independently, the densities (specific gravities), sizes, and settling velocities of pellets produced by Amphicteis scuphobrunchiata, a deposit-feeding polychaete worm. Pellet density was measured by an isosmotic density gradient technique. Densities ranged from 1.086 to 1.282 gacm-3 and settling velocities from 3.03 to 5.94 cmes-I. Pellets transported as bedload for variable distances; the oldest pellets tested (6 h after production) traveled a median distance of 3.1 m, while freshly egested pellets traveled 9.5 m before disintegrating. Worms would reingest disaggregated pellets, but feeding rates correlated positively with pellet age, consistent with previous findings that this species feeds at a faster rate on energetically more profitable sediment. These results suggest substantial interactions among benthic animals, fecal pellets, and sediment transport processes.

Fecal pellets produced by aquatic organisms affect sedimentary processes in two ways. Because fecal pellets are aggregatesof particles, pellet sinking rates can be much greater than the rates of their smaller constituents (Haven and Morales-Alamo 1968; McCall 1979). Thus, pellets enhance material fluxes through the water column by increasing sedimentation rates; this can also lead to deposition of particles that, because of hydrodynamic or chemical characteristics, otherwise might not be deposited in a given environment (Haven and MoralesAlamo 1968; Smayda 1969; Small et al. 1979; Robison and Bailey 198 1; Silver and Bruland 198 1). Most interest in such vertical sedimentation processes has focused on pellets of planktonic origin. Fecal pellets also play a role in sediment transport processes operating within the benthic boundary layer. Deposit feeders comprise the dominant trophic group in softbottom environments and ingest and package ambient sediment particles during their feeding activities. Although it has been recognized for some time that pellets can affect sediment transport via both bedload and suspended load transport modes (Rhoads I This research was supported by the Office of Naval Research and is a contribution from the School of Oceanography, University of Washington. * Present address: School of Oceanography, Oregon State University, Marine Science Center, Newport

97365.

and Young 1970; Young 197 1; Rhoads 1974; Risk and Moffat 1977; McCall 1979), most of this recognition has been qualitative in nature. Pellets have been portrayed generally as exacerbating sediment erodibility (e.g. Schafer 1972). Nowell et al. (198 1) provided quantitative data on transport of deposit-feeder fecal pellets, showing that pellets of different morphologies, produced by different species, have increased or decreased critical shear stressesfor movement relative to unpelletized sediment. Still less is known about dynamic animalsediment-fluid interactions. Diagramatic representations of sediment cycling among benthic animals, fecal pellets, and pellet breakdown and supply of potential food to animals by sediment transport (e.g. Haven and Morales-Alamo 1966; Young 197 1; Risk and Moffat 1977) unfortunately have remained iconic. We have only the most basic quantitative data on responses of a few deposit-feeding species to fluid flows and to sediment transport (e.g. Taghon et al. 1980). Feeding and, consequently, sediment pelletization rates have long been known to be functions of environmental variables such as temperature, and some deposit feeders also vary their feeding rates in response to sediment nutritional value (Taghon 198 1; Taghon and Jumars in press). Again, only a few species have been studied in this context. Our understanding of pellet transport processes has not been oriented within a 64

65

Fecal pellet transport

sound fluid dynamics framework, in part due to methodological difficulties. Size, density, and settling velocity are crucial in determining whether and in what mode a particle will be transported by a given flow (Smith 1977), yet we know of no study that has determined directly these three parameters for pellets of either benthic or planktonic origin. Density (specific gravity) has proven especially difficult to measure directly. Techniques to calculate pellet density indirectly by using Stokes’ law and pellet size and settling velocity (e.g. Komar et al. 198 1) are not applicable beyond such theoretical approaches; Reynolds numbers of deposit-feeder pellets are generally > 1 and thus outside the region of viscosity-dominated fluid forces. Effects of particle shape on settling velocity also hinder such indirect calculations of particle density (Janke 1966; Komar pers. comm.). We report here a series of experiments examining some biological and sedimentological consequences of fecal pellet production by a deposit-feeding polychaete We worm, Amphicteis scaphobranchiata. describe a direct method for determining pellet density, which allows comparison of directly measured pellet settling velocities with those calculated from previous, empirically derived, particle-settling equations. The distances pellets transported as bedload and pellet reingestion rates by A. scaphobranchiata were examined as a function of pellet age. The results suggest a mechanism wherein nutritionally valuable sediments (older pellets) are retained within the animal’s immediate environment and less valuable sediments (fresh pellets) are removed by fluid forces. Finally, we present a model of interactions among benthic animals, fecal pellets, and sediment transport processes. We thank P. Komar, E. Gallagher, and A. McElroy for their comments, J. Wolcott for help with data analysis, and the Director of the Friday Harbor Laboratories for continued use of these facilities. Materials and methods Experimental animals-Amphicteis scaphobranchiata is a tube-dwelling, amphar-

etid polychaete found subtidally in

Puget

Sound, Washington. This species is also found at abyssal depths (2,000+ m) off the coast of southern California (Hartman 1969). The feeding and defecation behavior of A. scaphobranchiata is described by Nowell et al. (in prep.). Briefly, this species uses a “mucous sling” mechanism to eject its fecal pellets beyond its normal feeding area. We dredged worms and sediment from N 18-m depth in West Sound, Orcas Island, Washington. Animals were separated from sediment and kept under constantly running seawater at Friday Harbor Laboratories (FHL) and appeared to remain healthy and to feed normally throughout the experiments. Animals rebuilt their tubes in containers of sediment previously sieved through a 61-pm mesh; the median diameter of this sediment is 18.2 pm (Nowell et al. 1981). Amphicteis scaphobranchiata produces elongate, gradually tapering fecal pellets (figs. 4C, D: Nowell et al. 198 1) intermediate between a cylinder and a cone. We measured densities, sizes, and settling velocities of individual pellets from 17 animals. In one series of experiments, we used pellets produced by worms feeding on the 0.05, Wilcoxon matched-pairs signed-ranks test: Siegel 1956), while all worms fed at slower rates on freshly egested pellets (P < 0.005). Animal-fecal interactions-A

pellet-sediment

transport

priori, deposit feeders should avoid ingesting their intact fecal pellets, simply because the material is less nutritious following gut passage.In fact, packaging digested sediment into discrete pellets may be the dominant mechanism that deposit feeders use to avoid reingesting this material. Pellets are usually much larger than the particles a deposit feeder can or will ingest (Levinton and Lopez 1977; pers. obs.). Discrete particles are relatively easily manipulated and can be placed outside the foraging area by the animal itself (Nowell et al. 198 1: fig. 4C; Nowell et al. in prep.) or removed by water currents (Schafer 1972). Our experiments reveal subtleties in these processes. Levinton and Lopez (1977) reported that snails of the genus Hydrobia 2 4 6 8 IO 12 14 16 18 immediately reingest their freshly egested Distance to disintegration (m) pellets when they are mechanically disaggregated, but did not report feeding rates on Fig. 2. Histograms of distances pellets transported this material. Amphicteis scaphobranchiata as bedload before disintegrating. A. Freshly egested will also reingest fresh pellets, but at a sigpellets from worms feeding on 61-250qm sediment fraction; median = 7.9 m. B. Freshly egested pellets nificantly reduced rate (Table 3). Although from integration can depend on its age (Fig. 2). Such dynamic processes suggest feedback 0.10, U-test). Pellet production rates- Amphicteis sca- interactions among benthic deposit feeders, phobranchiata increases its rates of feeding fecal pellets, and sediment transport. There and fecal pellet production as sediment food are theoretical reasons, based on energetic value increases (Taghon and Jumars in considerations, why deposit feeders should press). Deposit feeders digest sediment-as- vary feeding rate on fresh and aged pellets sociated microorganisms, and in a classic in the observed fashion (Taghon 198 1). In series of experiments Newell (1965) dem- a local (to the animal) collection of pellets, onstrated that following gut passagethe sed- recently egested pellets of lower nutritional iment could be recolonized rapidly by such value can be removed intact from an indimicroorganisms. Thus, we expected A. sca- vidual’s feeding area by fluid forces, while phobranchiata to feed on its own freshly older pellets which have higher nutritional egested pellets at a slower rate than on aged value will be more likely to disintegrate lopellets or on field-collected sediment. This cally and thus become available for reingesexpectation was borne out by the eight in- tion. We suggest, as a working hypothesis, that the mechanical and chemical properdividuals tested (Table 3). Pellet production rates were not significantly different when ties of A. scaphobranchiata mucus are “ap-

71

Fecal pellet transport Table 3. Pellet production rates, in mm3.pellet-‘, by A. scaphobranchiata feeding on three sediment sources.

Worm

1 2 3 4 5 6 7 8

Worm size (mm’)

Freshly egested pellets

Pellets aged 6h

Fieldcollected sediment

130 170 190 170 200 200 180 250

2.8 8.8 5.4 0.8 16.0 11.3 6.4 4.6

13.2 10.6 16.5 5.6 23.5 22.0 16.1 27.5

5.5 12.9 17.7 7.8 19.7 16.6’ 14.6 18.8

(sensu Gould and Vrba 1982) associated with the role of mucus in this species’ novel defecation pattern (Nowell et al. in prep.) and with the advantage to a deposit feeder of a mucus easily degraded by microbial processes (cf. Calow 1979). Clearly, A. scaphobranchiata fecal pellets vary in physical characteristics. These variations can arise from differences among pellets produced by the same individual, heterogeneity among animals (e.g. pellet size being a function of animal size), as well as differences in sediment particle size composing the pellets (Table 1 vs. Table 2). The variability in pellet density and settling velocity arising from these sources is assessed in Table 4. Pellets produced by the same individual were less variable in density than in settling velocity. This is to be expected, as variability in both density and size of a worm’s pellets will affect the variability in settling velocity. Interestingly, interindividual variability in pellet density is no larger than intraindividual variability. Because all worms fed on the same sediment, this result implies that the physical characteristics of the mucous binder and the degree of particle compaction during gut passagedo not vary appreciably among A. scaphobranchiata individuals. On the other hand, interindividual variability in settling velocity is higher than intraindividual variability. Again, this is as expected because the variations in body size among the experimental animals (Table 3) lead to wider variations in pellet sizes and thus increased variability in pellet settling velocities. Finally, although the sediment size fed to the worms significantly aftations”

Table 4. Standard deviations (SD, g-cm-‘) and coefficients of variation (C.V., %) for fecal pellet density and settling velocity, based on intraindividual and interindividual differences. Intraindividual values are the averages for the 14 individuals in Tables 1 and 2 producing 12 pellets. Averages for worms producing 22 pellets were used in calculating interindividual variability. Density

Intraindividual Interindividual, feeding on ~6 1-pm sediment fraction Interindividual, feeding on 6 l-250~pm sediment fraction

Settling

velocity

SD

C.V.

SD

C.V.

0.028 0.020

2.4 1.7

0.38 0.74

9.1 17.2

0.023

2.0

0.78

19.4

fected pellet densities and settling velocities (Tables 1,2), the variabilities of these properties about their means were not affected (Table 4). These variations in pellet characteristics emphasize the probabilistic nature of sediment transport problems. Grass (1970) has shown that the instantaneous critical bottom shear stresses associated with instabilities of individual sediment grains have distinct probability distributions. For any given bottom shear stress, what proportion of and in what mode pellets are transported will be functions of the probability distributions of .pellet size and density. How far pellets transport, and the amount of material transported by pellets in bedload, will be functions of the age distribution of a collection of pellets. For natural communities composed of several species, the cumulative effects of deposit-feeder fecal pellets on sediment transport mode (suspended vs. bedload) and rate might be expressible simply as the summation of individual species’ effects. However, it is more likely that synergisms will occur based on different hydrodynamic characteristics among pellets per se,on shielding or armoring effectswhere small pellets are “hidden” from the flow by large pellets (cf. Einstein 1950), and on different nutritional requirements among the deposit feeders producing the fecal pellets (Levinton and Lopez 1977). The net effects of these interactions on pellet production rates, transport modes, and transport rates are not yet readily predictable.

72

Taghon

References CALOW, P. 1979. Why some metazoan mucus secretions are more susceptible to microbial attack than others. Am. Nat. 114: 149-152. CHASE, R. R. 1979. Settling behavior of natural aquatic particulates. Limnol. Oceanogr. 24: 4 17-426. EINSTEIN, H. A. 1950. The bed-load function for sediment transportation in open channel flows. U.S. Dep. Agr. Tech. Bull. 1026. GIBBS, R. J., M. D. MATTHEWS, AND D. A. LINK. 197 1. The relationship between sphere size and settling behavior. J. Sediment. Petrol. 41: 7-18. GOULD, S. J., AND E. S. VRBA. 1982. Exaptation-a missing term in the science of form. Paleobiology 8: 4-15. GRASS, A. J. 1970. Initial instability of fine bed sand. Proc. Am. Sot. Civ. Eng. 96(HY3): 619-632. HALLERMEIBR, R. J. 198 1. Terminal settling velocity of commonly occurring sands. Sedimentology 28: 859-865. HARTMAN, 0. 1969. Atlas of the sedentariate polychaetous annelids from California. Alan Hancock Found. Univ. S. Calif. HAVEN, D. S., AND R. MORALES-ALAMO. 1966. Aspects of biodeposition by oysters and other invertebrate filter feeders. Limnol. Oceanogr. 11: 487498. -, AND -. 1968. Occurrence and transport of fecal pellets in suspension in a tidal estuary. Sediment. Geol. 2: 14 l-l 5 1. HAWLEY, N. 1982. Settling velocity distribution of natural aggregates. J. Geophys. Res. 87: 9489-9498. JANKE, N. C. 1966. Effect of shape upon the settling velocity of regular convex geometric particles. J. Sediment. Petrol. 36: 370-376. JONCKHEERE,A. R. 1954. A distribution-free K-sample test against ordered alternatives. Biometrika 41: 133-143. KOMAR, P. D. 198 1. The applicability of the Gibbs equation for grain settling velocities to conditions other than quartz grains in water. J. Sediment. Petrol. 51: 1125-l 132. -, A. 1’. MORSE, L. F. SMALL, AND S. W. FOWLER. 198 1. An analysis of sinking rates of natural copepod and euphausiid fecal pellets. Limnol. Oceanogr. 26: 172-l 80. LEVINTON, J. S., AND G. R. LOPEZ. 1977. A model of renewable resources and limitation of depositfeeding benthic populations. Oecologia 31: 177190. MCCALL, P. L. 1979. The effects of deposit feeding oligochaetes on particle size and settling velocity of Lake Erie sediments. J. Sediment. Petrol. 49: 813-818. NEWELL, R. 1965. The role of detritus in the nutrition of two marine deposit feeders, the prosobranch Hydrobia ulvae and the bivalve Macoma balthica. Proc. Zool. Sot. Lond. 144: 25-45. NOWELL, A. R., P. A. JUMARS, AND J. E. ECKMAN.

et al. 198 1. Effects of biological activity on the entrainment of marine sediments. Mar. Geol. 42: 133153. RHOADS, D. C. 1974. Organism-sediment relations on the muddy sea floor. Oceanogr. Mar. Biol. Annu. Rev. 12: 263-300. -, AND D. K. YOUNG. 1970. The influence of deposit-feeding organisms on sediment stability and community trophic structure. J. Mar. Res. 28: 150-178. RICKWOOD, D., AND G. D. BIRNIE. 1975. Metrizamide, a new density gradient medium. Fed. Eur. Biochem. Sot. Lett. 50: 102-l 10. RISK, M. J., AND J. S. MOFFAT. 1977. Sedimentological significance of fecal pellets of Macoma balthica in the Minas Basin, Bay of Fundy. J. Sediment. Petrol. 47: 1425-1436. ROBISON, B. H., AND T. G. BAILEY. 198 1. Sinking rates and dissolution of midwater fish fecal matter. Mar. Biol. 65: 135-142. RUBEY, W. W. 1933. Settling velocities of gravel, sand and silt particles. Am. J. Sci. 25: 325-338. SCHAFER,W. 1972. Ecology and palaeoecology of marine environments. Oliver and Boyd. SIEGEL, S. 1956. Nonparametric statistics for the behavioral sciences. McGraw-Hill. SILVER, M. W., AND K. W. BRULAND. 198 1. Differential feeding and fecal pellet composition of salps and pteropods, and the possible origin of the deepwater flora and olive-green “cells.” Mar. Biol. 62: 263-273. SMALL, L. F., S. W. FOWLER, AND M. Y. UNLU. 1979. Sinking rates of natural fecal pellets. Mar. Biol. 51: 233-241. SMAYDA, T. J. 1969. Some measurements of the sinking rate of fecal pellets. Limnol. Oceanogr. 14: 621-625. SMITH, J. D. 1977. Modeling of sediment transport on continental shelves, p. 539-577. In E. D. Goldberg et al. [eds.], The sea, v. 6. Wiley-Interscience. TAGHON, G. L. 198 1. Beyond selection: Optimal ingestion rate as a function of food value. Am. Nat. 118: 202-214. AND P. A. JUMARS. In press. Variable ingestion rate and its role in optimal foraging behavior of marine deposit feeders. Ecology. -, A. R. NOWELL, AND P. A. JUMARS. 1980. Induction of suspension feeding in spionid polychaetes by high particulate fluxes. Science 210: 562564. WANLESS, H. R., E. A. BURTON, AND J. DRAVIS. 198 1. Hydrodynamics of carbonate fecal pellets. J. Sediment. Petrol. 51: 27-36. YOUNG, D. K. 197 1. Effects of infauna on the sediment and seston of a subtidal environment. Vie Milieu Suppl. 22, p. 557-571.

Submitted: 18 June 1982 Accepted: 6 June 1983