Marine Biology

, Volume 144, Issue 4, pp 705–712

The effect of feeding behavior of the gastropods Batillaria zonalis and Cerithideopsilla cingulata on their ambient environment

Authors

  • Satomi Kamimura
    • Laboratory of Ecology and Systematics, Graduate School of Engineering and ScienceUniversity of the Ryukyus
    • Laboratory of Ecology and Systematics, Department of Chemistry, Biology and Marine ScienceUniversity of the Ryukyus
Research Article

DOI: 10.1007/s00227-003-1238-x

Cite this article as:
Kamimura, S. & Tsuchiya, M. Marine Biology (2004) 144: 705. doi:10.1007/s00227-003-1238-x

Abstract

Feeding behaviors of the gastropods Batillaria zonalis, a suspension and deposit feeder, and Cerithideopsilla cingulata, an obligate deposit feeder, were studied to examine their effect on dynamics of suspended materials, total nitrogen (TN) and total organic carbon (TOC) in sediments. Suspension feeding in B. zonalis was observed in detail visually, as it had been previously unreported. An experimental system where B. zonalis and C. cingulata were cultured for 10 weeks, using previously frozen microalgae Nannochloropsis oculata as food, was then constructed. During feeding observations, the suspension-feeding B. zonalis formed a mucus “food cord” to entangle particulate materials, which were subsequently ingested. The feeding mode of B. zonalis is hence categorized as ctenidial filter feeding. For the culture experiments, decreases in suspended materials were seen only in the B. zonalis cultures, while the control (no gastropods) and C. cingulata cultures remained nearly unchanged. Sediment TN and TOC showed no significant differences between B. zonalis (with mean TN at 0.0345% and mean TOC at 0.261%) and control cultures (with TN at 0.0389% and TOC at 0.331%), but the sediments in C. cingulata cultures had lower levels (with TN at 0.0204% and TOC at 0.156%). The C/N ratios were similar for both B. zonalis (7.55) and C. cingulata (7.68) cultures, and both were lower than the control cultures (8.55). The filtration rate for B. zonalis was lower than that previously observed in bivalves inhabiting the same intertidal flat (e.g. Cyclina sinensis, Grafrarium tumidum and Barbatia virescens). However, Batillaria zonalis occurs at higher abundances than these bivalves. Therefore, it is expected that this species has a large affect upon the transport of suspended materials to the sediments. The addition of TN and TOC to sediments in B. zonalis cultures was probably caused by biodeposition, but deposit feeding by B. zonalis may have restrained the accumulation of those components. The impact of deposit feeding in Cerithideopsilla cingulata cultures was most probably stronger than sedimentation and biodeposition, because of the lower sediment TN and TOC. Bioturbation by both B. zonalis and C. cingulata yields the same effect on sediment quality, as indicated by the low C/N in the culture sediment of both treatments, despite difference in feeding modes. This paper demonstrates, for the first time, the importance of gastropods in bioturbation and removal of suspended materials in subtropical tidal flat habitats.

Introduction

Many materials in the intertidal-flat ecosystem are supplied from the adjacent terrestrial and marine habitats (Knox 1986; Raffaelli and Hawkins 1996; Levinton 2001). Invertebrate feeding patterns in intertidal flats play an important role in the transportation and degradation of organic materials (Knox 1986; Raffaelli and Hawkins 1996; Lenihan and Micheli 2001). To understand the process of energy flux in the intertidal flat, it is therefore important to understand the feeding mechanisms of dominant species and their roles in nutrient transport via ingestion and fecal discharge.

The gastropods Batillaria zonalis (Family Batillariidae) and Cerithidea (Cerithideopsilla) cingulata (Family Potamididae) are dominant species in the Tomigusuku tidal flat located in the southern part of Okinawa Island, Japan (Kamimura 2000). These species probably have an important role in the energy flux of their habitat. In a study of their fatty acids, their food sources were reported to be macroalgae, bacteria and diatoms (Meziane and Tsuchiya 2000). Feeding selectivity was compared in a study of the congeneric species, B. attramentaria and C. california occurring in salt marshes located in central California (Whitlatch and Obrebski 1980). Here, B. attramentaria generally ingested larger diatoms than did C. california. Thus, there are some qualitative studies about food sources for Batillaria and Cerithidea, but no studies that quantitatively assess their roles in energy flux. B. multiformis (Morton and Morton 1983), B. cumingi (Koike et al. 1989) and B. flectosiphonata (Ozawa 1996) are reported to be deposit feeders and grazers of benthic diatoms. There are indications, however, that B. zonalis can do both suspension and deposit feeding (Kamimura 2000), but it has yet to be demonstrated in detail.

Dauer et al. (1981) proposed the term “interface feeder” for species that utilize particles from the sediment surface, in suspension and resuspended sediment. For example, the polychaetes Boccardia pugettensis, Pseudopolydora kempi japonica (Taghon and Greene 1992), Nereis diversicolor (Nielsen et al. 1995), Spiochaetopterus oculatus and Spio setosa (Bock and Miller 1997) are reported as “interface feeders”. The bivalve, Macoma balthica, also seems to feed on suspended and sediment-surface particles (Brafield and Newell 1961; Olafsson 1986) like the polychaetes. Batillaria zonalis is also expected to be an “interface feeder”.

The functional difference between suspension feeders and deposit feeders indicates that they should have different roles in energy flux (Tsuchiya and Kurihara 1980; Christensen et al. 2000). Obligate suspension feeders take particles from the water column and eject their feces and/or pseudofeces onto the sediments (for mussel see Bjork et al. 2000; for oyster see Hayakawa et al. 2001; and for clam see Jie et al. 2001). It has been reported that deposit feeders eat bacteria and organic contents in decaying cordgrass [Littoraria irrorata (Newell and Barlocher 1993)], decrease chlorophyll a [Yoldia limatula (Ingalls et al. 2000)] and decrease organic matter [Nereis virens (Kristensen and Blackburn 1987); Nucula proxima (Cheng and Lopez 1991)] in the sediment. For interface feeders like B. zonalis, it is important to determine whether they show biodeposition like Nereis diversicolor (Christensen et al. 2000), or have a role that is particularly different from suspension and deposit feeders.

The purpose of this study is to describe the mechanism of suspension feeding of B. zonalis and to estimate the effect of B. zonalis as an interface feeder in comparison to C. cingulata, an obligate deposit feeder in the same habitat. The suspension feeding mechanism in B. zonalis was observed in detail and compared with other suspension feeders. To quantify the effect, changes in concentrations of suspended materials, total nitrogen (TN), total organic carbon (TOC) and the C/N ratio in sediments were measured.

Materials and methods

Collection of gastropods

B. zonalis and C. cingulata were collected from the Tomigusuku intertidal flat located in the southern part of Okinawa Island, Japan [26.5° N, 128° E (Fig. 1)]. This intertidal flat has gravelly, sandy and muddy areas. Both species mainly inhabited the sandy and muddy areas. Snails were collected at random and subsequently kept in filtered seawater for more than 24 h before the experiments, in order to acclimate them. Each snail was marked with a number, and wet weights were recorded prior to the experiment.
Fig. 1

Location of the sampling site

Observation on the suspension feeding of B. zonalis

The suspension-feeding mechanism of B. zonalis in a petri dish (with seawater) was observed under a stereomicroscope (Olympus SZ1145TR) and recorded using an attached camera (Olympus PM-C35DX). Also, the feeding behaviors of B. zonalis placed in a 500-ml beaker were recorded with a video camera (Sony DCR-VX1000) and single lens reflex camera (Minolta AB800). During these observations, B. zonalis was provided with previously frozen planktonic algae (Nannochloropsis oculata) at 1×106 cells/ml).

Experimental systems

Feeding experiments were conducted in the laboratory adjacent to the Tomigusuku intertidal flat from 11 February 2001 to 22 April 2001 (10 weeks). The experimental setup is shown in Fig. 2. Culture chambers were made from 3-l PVC beakers perforated with many holes on the bottom, which was covered with a thin layer of cotton. A layer 5–7 cm thick of “low-organic matter” sediment (sieved through a 1 mm mesh) was added over the cotton. The experiment had three replicates of the following three treatments: (1) without animals (control); (2) with 20 individuals of B. zonalis; and (3) with 20 individuals of C. cingulata. An example of one replicate of the experimental system is indicated in Fig.  2. All the chambers were placed in a single 200-l holding tank that simulated flow tide and ebb tide every 6 h, by automatically filling and draining filtered seawater (by Advantec cartridge filter TCPD-03A-SIME). This means that control and treatments chambers were exposed to the same simulated cycle, although the placements of each chamber in an aquarium were not controlled. Draining and filling time for the 200-l holding tank was 15 min, every 6 h. The filtered seawater ebbed and flowed through each chamber passing through the holes and the sediments within the 200-l holding tank (Fig. 2). The water depth in each chamber ranged from 11 cm to zero over the simulated cycle. The seawater temperature was kept at 20°C using a submersible heater, maintaining a temperature similar to that of seawater during this season. Each chamber received an independent supply of algal food for 15 min every hour during the 6-h high tide period, except on 8 days when remaining suspended algae and filtration rates were measured. Algal food was prepared by dissolving frozen N. oculata [2–3 μm in diameter at 1.00×1010 cells/ml (Marine Bio)] to a concentration of 1.00×108 cells/ml with filtered seawater in a holding tank (15  l). The resulting solution was then automatically delivered into the chambers as above. In 15 min, 25 ml algal solution was added to each chamber, resulting in an estimated initial concentration of 1.67×106 cells/ml. These cell concentrations, which corresponded to 20–45 mg/l, were decided based on the amount of suspended material in the study area (unpublished data).
Fig. 2

Design of the experimental ebb-and-flow system in the laboratory, showing one replicate of the experimental system. All the 3-l chambers were placed in one 200-l holding tank that recreates maximum and minimum tides every 6 h, by using two pumps for filling and draining seawater. The seawater temperature was kept at 20°C, and each chamber was supplied with an algal suspension (frozen Nannochloropsis oculata of 2–3 μm diameter) for 15 min every 1 h

Estimation of remaining suspended materials and filtration rate

In order to determine the effect of each treatment on removal of the algal suspension, we estimated the concentration of remaining suspended materials in the chambers and the filtration rate during simulated “high tide”. Samplings were conducted on 5 days for remaining suspended materials and on 3 days for filtration rate. Automatic supply of N. oculata was stopped, and 5 ml N. oculata diluted to 6.67×108 cells/ml was added to all chambers after the seawater temperature stabilized (20°C). Initial concentrations in each chamber were estimated to be 2.22×106 cells/ml. After 10 min of aeration for uniform algae concentration, 50-ml seawater samples from all chambers were taken in order to measure initial algal concentrations. After 3 h, additional 50-ml samples were taken to estimate the concentration of remaining suspended algae in each chamber. All seawater samples were filtered onto Whatman GF/C filter, and chlorophyll was measured after extraction in 90% acetone overnight at 4°C in a refrigerator [using a modified method (Tachibana and Nasu 1994)]. The concentration of remaining suspended materials (Re) was calculated using the formula:
$$\operatorname{Re} \; = \;{\left( {{\text{C}}_{3} /{\text{C}}_{0} } \right)} \times 100,$$
where C0 is the concentration 10 min after N. oculata was added, and C3 is the concentration 3 h later.
A similar, but more complex procedure was used to estimate filtration rates on 3 days. Here, samples were taken 1 and 2 h (time 1 and time 2, respectively) after initial sampling (time 0). Then, N. oculata was added again to all chambers with aeration, and additional samples were taken at initial time (time 0′) and 1 h later (time 1′). All samples were filtered and chlorophyll content estimated as above. Filtration rate was calculated using a modified Coughlan (1969) formula:
$${\text{FR}}\; = \;{\left( {{\text{M}}/{\text{W}}} \right)}{\left\{ {{\left[ {\log \;_{{\text{e}}} {\left( {{\text{C}}_{{{\text{t}} - 1}} /{\text{C}}_{{\text{t}}} } \right)}} \right]} - {\left[ {\log \;_{{\text{e}}} {\left( {{\text{C}}_{{{\text{t}} - 1}} '/{\text{C}}_{{\text{t}}} '} \right)}} \right]}} \right\}}$$
where FR is the filtration rate (ml g wet weight−1 h−1), M is the volume of seawater in the chamber, W is the wet weight of snails per test chamber, Ct-1 and Ct are the concentrations at time t and after 1 h in the experimental seawater with snails, and Ct-1′ and Ct′ are the same for the control culture. The control replicates from the corresponding sampling times were calculated as the mean of [loge (Ct-1′/Ct′)] and subtracted from [loge (Ct-1/Ct)].

Snail biomass and estimation of biodeposition

The experiment was run for 10 weeks. Then, B. zonalis and C. cingulata were transferred to filtered seawater more than 24 h to collect their feces, and the wet weights of snail were recorded after this incubation. Feces in filtered seawater were filtered onto pre-combusted and pre-weighed Whatman GF/C filters. The feces on the filters were dried (80°C for 48 h) and weighed, then exposed to fumes of 1 N HCl for 24 h to remove inorganic carbon. Surface sediments from test chambers were collected using slide glass in place of a spatula. Sediment samples were dried (80°C for 48 h) and treated with 2 N or 6 N HCl. After this treatment, the samples were rinsed twice with deionized distilled water and dried again. TN and TOC in the feces and surface sediments samples were measured using a Sumigraph NC-80 NC analyzer (Sumitomo Chemical Company) connected to a GC-8A gas chromatograph and Chromatopac C-R6A recorder (Shimadzu). The C/N ratio was calculated using TN and TOC.

Statistical analysis

Wet weight of individuals and feces data were analyzed by t-test for comparison between treatments (P<0.05). The Bonferroni-Dunn procedure, a multiple comparison procedure, is attributed to Dunn (1961). The procedure was used for comparison of remaining suspended materials or sediments between control and two treatments.

Results

Suspension feeding behavior of B. zonalis

As soon as B. zonalis were placed in seawater, they began to suck suspended particulate materials through their siphonal canal (Fig. 3A). The emerging particulate materials became entangled in a mucus “food cord” extending from the gills along the food groove (Fig. 3A, B). The food cord was green, remarkably the same color as N. oculata. The food cord pooled on the foot (Fig. 3B) and then the snail moved its snout (possibly propodium) over the food cord and ingested it with its mouth (Fig. 3C). Upon ingestion, the food cord was removed from the foot (Fig. 3D), but continued to extend from the gills along the food groove (Fig. 3E). Occasionally, B. zonalis showed no interest in the food cord, and it was extruded from the foot and deserted as pseudofeces.
Fig. 3A–E

Batillaria zonalis. A B. zonalis sucks particulate materials through its siphonal canal. The emerging particulate materials become entangled in a mucus “food cord” extending from the gills along the food groove. The arrow indicates the direction of suspended material and food cord movement. B Thirty seconds after placing B. zonalis in the N. oculata suspension in seawater, the food cord is found along the food groove (in the circle). The food cord pools on the foot. C Almost 1 min later, B. zonalis eats the food cord by moving its head and using its mouth. D After eating, the pooled food cord is lost from the foot. E The food cord continues to extend from the gill

Remaining suspended materials and filtration rate

The algal suspension in the B. zonalis treatment showed dramatic decreases, and its concentration was 37.9±11.2% (mean±SD) of the initial concentration after 3 h. In the control and C. cingulata chambers, however, there were only slight changes over 3 h, the concentration for the control treatment being 82.3±7.3% while that for the C. cingulata treatment was 91.9±15.2%. There are significant differences between treatments after 3 h (P<0.0167 by the Bonferroni-Dunn test).

Mean filtration rates (±SD) for each time period were as follows: FRt=1=8.52±4.12, FRt=2=11.61±5.62, FRt=1′=12.28±7.25 ml g wet weight−1 h−1. There was high variation in filtration rates, and considerable variation in initial concentrations between chambers at all times, perhaps because of resuspension and the effect of suspension feeding. Therefore, we tried to show whether initial concentration affected the filtration rate in Fig. 4. There was no trend indicating a filtration rate that depended on initial concentrations where the initial concentrations ranged from 0.104 μg chlorophyll l−1 to 0.282 μg chlorophyll l−1. Therefore, a mean overall filtration rate for B. zonalis (10.56 ml g wet weight−1 h−1) was calculated for comparative purposes.
Fig. 4

Initial concentration of N. oculata and filtration rate (ml g wet weight−1 h−1) for B. zonalis. Diamonds at 1 h, triangles at 2 h, squares at 1 h from when N. oculata was added again (1′). Different colors (black, gray and white) indicate each chamber; these are combined data for all 3 sampling days (mean±SD)

Snail biomass and estimation of biodeposition

There was no significant difference in the wet weights of B. zonalis before and after 10 weeks [P>0.05 by the t-test (Table 1)]. Also C. cingulata did not show significant differences as between before and after culturing [P>0.05 by the t-test (Table 1)].
Table 1

Batillaria zonalis and Cerithideopsilla cingulata. Comparison of wet weight including shell (g), before and after experiments, using a t-test (P<0.05). Data shown are means±SD (n=20) for all replicates

Before

After

P

B. zonalis

2.00±0.26

2.00±0.26

>0.999

1.95±0.34

1.94±0.34

0.824

1.96±0.37

1.95±0.37

0.658

C. cingulata

1.33±0.21

1.33±0.20

0.923

1.35±0.21

1.34±0.21

0.776

1.31±0.25

1.32±0.25

0.908

TN in the surface sediments increased in all samples after 10 weeks (Fig. 5A). Sediments in the B. zonalis and the control treatments had similar TN (0.0346±0.0051% and 0.0389±0.0068%, respectively) and TOC levels (0.261±0.040% and 0.331±0.044%, respectively), and both were higher than in the C. cingulata treatment [for which TN was 0.0204±0.0009% and TOC was 0.156±0.009% (P<0.0167 by the Bonferroni-Dunn test) (Fig. 5A, B)]. The C/N ratios in the sediments for B. zonalis and C. cingulata treatment were similar (7.55±0.14 and 7.68±0.11, respectively), and they were lower than in the control treatment [8.55±0.43 (P<0.0167 by the Bonferroni-Dunn test) (Fig. 5C)].
Fig. 5A–C

Sediment components for culture experiment for B. zonalis (BZ), Cerithideopsilla cingulata (CC) and control. A Total nitrogen (TN). B Total organic carbon (TOC). C C/N ratio. Blank bar Sediments at the start; gray bar after 10-weeks culture; * indicates significant difference (P<0.0167 by the Bonferroni-Dunn test). Values are mean±SD, n=3

For the dry weight of feces collected after the experiment, there was no significant difference between B. zonalis (0.743±0.201 mg g wet weight−1 day−1) and C. cingulata (0.825±0.045 mg g wet weight−1 day−1) [P>0.05 by the t-test (Fig. 6A)]. Fecal TN (2.42±0.59% for B. zonalis feces and 2.60±0.45% for C. cingulata feces) and TOC levels (17.2±0.59% for B. zonalis feces and 16.0±2.8% for C. cingulata feces) also showed no significant differences [P>0.05 by the t-test (Fig. 6B, C)]. However, the C/N ratio for B. zonalis feces (7.18±0.08) was higher than for C. cingulata feces (6.14±0.16) [P<0.05 by the t-test (Fig. 6D)].
Fig. 6A–C

Values for feces collected after 10 weeks’ culture. A Dry weights. B Total nitrogen (TN). C Total organic carbon (TOC). D C/N ratio. BZ B. zonalis; CC C. cingulata; * indicates significant difference (P<0.05 by the t-test). Values are means±SD, n=3

Discussion

In this study, we show the suspension feeding process of B. zonalis. This is the first report describing the mechanism of suspension-feeding activity in Batillaria. Suspension feeding in gastropods species is relatively uncommon in comparison to bivalves, and is reported to occur in 37 species (Fenchel et al. 1975; Declerck 1995; Chaparro et al. 2002). A review of such species distinguished three feeding styles (Declerck 1995): (1) ctenidial filter feeding, where particles are filtered from the respiratory current passing through the mantle cavity; (2) mucus-net feeding, where suspended mucus nets are used to take particles from ambient currents; and (2) ciliary-tract feeding, where particles are trapped in ciliated fields on lobes of the foot. B. zonalis is categorized as a ctenidial filter feeder because its ctenidium is used for trapping particles in the respiratory current.

It is hypothesized that suspension feeding developed from gill cleaning mechanisms (Jørgensen 1966). Suspension feedings for B. zonalis and for bivalves [e.g. clams and cockles (Ruppert and Barnes 1994); mussels (Beninger and St-Jean 1997); and pearl oysters (Prouvreau et al. 2000)] are similar in that their gills (ctenidia) are important organs for catching and entangling particles within mucus. However, the feeding and disposing processes are different, as mentioned below. The mouth of a bivalve lies near the gill, and labial palps sort ingested particles, whether into the mouth or out through the out-flow siphon (Brusca and Brusca 1990; Ruppert and Barnes 1994; Morse and Zardus 1997). For disposing of particles, bivalves transport them to the out-flow siphon after sorting. On the other hand, the mouth of B. zonalis lies under part of its head and they have to turn their heads to eat the food cord, including particles. The food cord is elongated from the gills, pooled on the foot and, if uneaten, would be pseudofeces, as observed in this study. It appears that bivalves have a more functional feeding process than B. zonalis, and the feeding process of B. zonalis is secondarily derived from ctenidial cleaning. Their uptake of materials in seawater is so definite that they may be called suspension feeders. B. zonalis might also be called a “modified suspension feeder” because of the involvement of the of food cord in their feeding, in order to distinguish between this feeding process and bivalves’ methods.

Several bivalve suspension feeders have been observed at our collection site (Bachok 2000). A report by the Okinawa Prefectural Government (2000) provides filtration rates for Cyclina sinensis, G. tumidum and Barbatia virescens. When compared with these bivalves, Batillaria zonalis has a lower filtration rate per wet weight, but the abundance of B. zonalis is higher than that of these bivalves (Table 2). Hence, it is expected that the population of B. zonalis in this tidal flat may be an important contributor to the transport of suspended materials.
Table 2

Dominant suspension feeders in Tomishiro tidal flats. The values shown are means

Organism

Experimental

temperature (°C)

Filtration rate (ml g wet weight−1 h−1)

Habitat density (g wet weight/625 cm2)

In situ filtration rate (ml/625 cm2 h−1)

Bivalvesa

  Cyclina sinensis

20

6.39

0.329b

2.10

  Grafarium tumidum

20

3.37

1.159b

3.91

  Barbatia virescens

20

23.13

0.041

0.95

  Psammotaea elongata

5.769b

  Semele camicolor

1.469b

Gastropod

  Batillaria zonalisc

20

10.56

16.8d

177.41

aData on bivalves are from Okinawa Prefecture (2000), except as noted below

bData from Bachok (2000)

cData for B. zonalis are from this study, except as noted below

dData from Kamimura (2000)

Obligate suspension feeders take particles from the water column and deposit their feces and/or pseudofeces onto the sediments. For example, in mussel (Bjork et al. 2000), clam (Jie et al. 2001) and oyster (Hayakawa et al. 2001), biodeposition increases organic material content in the sediment. Although the interface feeder, Nereis diversicolor, shows biodeposition in the manner of an obligate suspension feeder, it is also known that deposit feeding in this species occurs during phytoplankton-poor periods (Christensen et al. 2000). In contrast, B. zonalis treatments in this study show that the addition of TN and TOC in the sediments occurs at similar levels as in the control treatments. At the same time, however, the C/N ratios in the sediments of the B. zonalis treatment were lower than for the control treatment. These results indicate that the addition of TN and TOC in the B. zonalis sediments are quantitatively similar to controls, but different in quality. It is assumed that increases of TN and TOC in the control treatments were mostly due to physical sedimentation, and occurred when the algal-rich seawater drained through the sediment during the low tide simulation. In contrast, biodeposition in B. zonalis cultures is the most likely cause of increased TN and TOC in their sediments. This is supported by the higher removal of suspended algal material in the B. zonalis treatment, indicating their suspension-feeding activity. Furthermore, settled material on the sediments should be available for adventitious deposit feeding by B. zonalis and it is possible that this deposit feeding limited the accumulation of TN and TOC in sediments. Thus, the TN and TOC levels in sediments of the B. zonalis treatment may implicate both features of suspension and deposit feeding styles. In addition, other bioturbation factors such as foraging and mucus secretion must be considered, but it is likely that these factors are also present in the Cerithideopsilla cingulata treatment.

Deposit feeders, in general, take organic materials from sediments and eject feces onto those same sediments. It has been reported that deposit feeding decreases organic content in sediments (e.g. Kristensen and Blackburn 1987; Cheng and Lopez 1991). On the other hand, it is known also that some types of biodeposition by deposit feeders, such as fecal production and mucus secretion, positively affect microbial biomass (Newell 1965; Riemann and Schrage 1978; Tsuchiya and Kurihara 1979; Davies and Beckwith 1999). In this study, the sediments with C. cingulata show lower values of TN, TOC and C/N than those of the control treatments. The lower TN and TOC results are due to deposit feeding and the lower C/N suggests that the number of bacteria was increasing (Kautsky and Evans 1987). The contents of C. cingulata sediments are probably a result of the antagonistic effects of deposit feeding versus sedimentation and biodeposition. However, the impact of deposit feeding was most probably stronger, given the relatively lower TN and TOC content in C. cingulata sediments. The initial (time 0) culture sediments had a low organic content (Fig. 5). Therefore, materials deposited on the sediment were “fresh” and useful for deposit feeders, and might be one cause of the clear results in sediments inhabited by C. cingulata.

B. zonalis and C. cingulata are considered to show some similarities in their ecological niches, such as inhabiting the upper 5 cm of sediments and foraging on the surface (our observations). However, they differ greatly in feeding modes: B. zonalis is a suspension and deposit feeder and C. cingulata is an obligate deposit feeder. In this study, mode differences were evident, namely the lower algal-suspension content in the 3-h B. zonalis treatment, and the lower TN and TOC accumulation in C. cingulata sediments. The amount of feces can also positively affect the TN and TOC content of sediments. Dry weight, TN and TOC in feces of both species show no significant differences, but B. zonalis did not egest pseudofeces during the incubation time when there were no suspended materials (see “Materials and methods”). Non-suspension feeders, like C. cingulata, produce few pseudofeces, if any. So, B. zonalis sediments were affected by both feces and pseudofeces, while C. cingulata sediments only received fecal material. These feces and pseudofeces also may positively affect microbial biomass on each sediment. Bioturbation by both gastropods yields the same effect on quality of sediments, as indicated by sediments C/N, despite feeding mode differences.

Acknowledgements

We thank Okinawa Prefectural Land Development Public Corporation and Metocean Environment Inc for a support grant and for laboratory systems. We also would like to express our gratitude to Dr G.C. Fiedler and M. Black for helpful suggestion on English terms. Many grateful thanks are due to Dr. S. Ohgaki and an anonymous reviewer for their valued suggestions and comments.

Copyright information

© Springer-Verlag 2003