Intraspecific variations in delta13C indicate ontogenetic diet changes in deposit-feeding polychaetes.
Identification of the food resources assimilated by detritivores in general and deposit feeders in particular is a difficult problem for ecologists (Lopez et al. 1989). To improve analyses of food webs Cohen et al. (1993) urged that consumers be grouped by species, but such distinctions are currently impossible among species that have vague diets dominated by "amorphous detritus" (e.g., Benke and Jacobi 1994) or "the organic fraction of ingested sediment" (Lopez and Levinton 1987). Cohen et al. (1993) added that when metaphoetesis (an ontogenetic change in diet) occurs, descriptions of food webs should distinguish among the size classes or life-cycle stages of that species. In this study, I present evidence of ontogenetic changes in diet within the free-living benthic forms of four species of deposit-feeding worms.
Many organisms undergo ontogenetic changes in niche; the best-known examples are amphibians, holometabolous insects, and many fishes (Werner and Gilliam 1984, Bergman and Greenberg 1994, Persson and Eklov 1995, Olson 1996). Many deposit feeders (and other benthic invertebrates) also have complex life cycles, beginning as planktonic larvae before shifting to a benthic existence (Thorson 1950). Ontogenetic niche changes also can occur continuously during growth, without a discrete metamorphosis or change in habitat (Polis 1984).
Accumulating evidence suggests that benthic juveniles of deposit-feeding species might change diet as they grow. By definition, deposit feeders ingest nutritionally poor material that is diluted with completely nonnutritive mineral grains (Lopez and Levinton 1987). Larvae of deposit feeders, on the other hand, ingest relatively high-quality diets (plankton or yolk). After settlement, juveniles face a transition from the rich larval diet to a poor adult one, and a digestive constraint associated with small body size (Penry and Jumars 1990) might make an abrupt shift to deposit feeding difficult (Jumars et al. 1990). Because gut volume scales as body volume to a power of 1.0 (Penry and Jumars 1990), while gut throughput rate scales to a power of 0.77 (Cammen 1980), gut residence time (volume/throughput rate) - and the extent to which material can be digested in the gut - will increase with body size (Penry and Jumars 1990). This size-dependent digestive constraint could be eased by either (1) an ontogenetic decrease in gut residence time or (2) ingestion of a higher quality diet during juvenile stages. Forbes (1989) measured a break in the allometry of throughput rate ([approximately equal to] egestion rate) for juveniles of Capitella sp. I, but the data show a shift from an allometric exponent of 1.21 to 0.78 at a body volume of 1.7 [mm.sup.3]. This implies, and direct measurements of gut residence time confirm, that the gut residence times of juveniles will be even shorter than would be predicted from the adult's allometric exponent (Forbes 1989). Given this scaling, ingesting a diet that can be digested more quickly than that of conspecific adults is the most likely means for juveniles to overcome the size-dependent digestive constraint (Jumars et al. 1990, Penry and Jumars 1990). Similar arguments that smaller individuals require diets that can be more easily digested have been made and confirmed for many other animals that ingest low-quality forage (Sibly 1981, Demment and Van Soest 1985, Caceres et al. 1994, Moir 1994).
One potential means for deposit-feeding juveniles to increase diet quality is to selectively ingest smaller particles that will increase the amount of surface-associated organic matter per unit volume of ingested sediment (Taghon et al. 1978). Intuition and interspecific relationships between body size and particle-size selection suggest a positive relationship between body size and the size of ingested particles, but at least some deposit-feeding worms show the opposite trend: smaller juveniles have a bias toward contacting and ingesting relatively larger particles more frequently than do conspecific adults (Hentschel 1996). Given the negative relationship between particle size and the quantity of surface-associated organic matter per unit volume (DeFlaun and Mayer 1983), this ontogenetic change in particle selection imposes an additional constraint on juveniles - not a means to overcome the digestive constraint associated with small gut size.
Scaling of gut size and feeding structures probably forces benthic juveniles to forage differently from the microphagous deposit- or suspension-feeding adults (Jumars et al. 1990). Because the morphology of juvenile and adult feeding structures can be quite similar (often only a change in the size of feeding structures occurs), any ontogenetic changes in foraging behaviors are probably subtle and likely involve gradations between general feeding guilds (e.g., Fauchald and Jumars 1979). Hentschel (1996), for example, speculates that an ontogenetic shift in the foraging behaviors of spionid polychaetes would increase diet quality if (1) juveniles spend more time than adults suspension feeding rather than deposit feeding (supported by unpublished data from G. Taghon) or (2) juveniles forage in a more macrophagous mode where only the higher quality organic components are ingested (i.e., a feeding mode that is more similar to that of many permanent meiofauna than to that of deposit-feeding macrofauna).
Because deposit-feeding species play dominant roles in benthic food webs (Lopez et al. 1989), ontogenetic shifts in their feeding behaviors and diets can have dramatic effects on the ecology of benthic communities. Size-specific competition for high-quality juvenile food resources may, for example, limit population densities of deposit-feeding species (Gallagher et al. 1990, Hentschel and Jumars 1994). More generally, recruitment may be limited seasonally by the supply of labile foods like benthic diatoms (e.g., Bianchi 1988, Hentschel and Jumars 1994) or fresh phytodetritus (Marsh and Tenore 1990, Snelgrove et al. 1992, Gage 1994). In addition, Jumars et al. (1990) discussed how size-dependent diet shifts might influence deep-sea species diversity.
The principal aim of this study is to test the null hypothesis that the diets of deposit-feeding species are constant as benthic juveniles grow to adults. Despite several arguments suggesting that the diets of conspecific juveniles and adults may differ, their shared habitat, similar feeding structures, and the general lack of knowledge concerning the composition of deposit-feeder diets make the null hypothesis difficult to dismiss out of hand.
The simplest means to reject the null hypothesis is to quantify the assimilation of food resources with a method that is both unbiased by body size and sensitive enough to detect even gradual shifts in diet as juveniles grow. Stable carbon isotopes as a trophic tracer provide such an approach (Fry and Sherr 1984, Petersen and Fry 1987). The premise is simple: heterotrophs incorporate the ratio of 13C/12C of their food into their body tissues with little fractionation. If two consumers assimilate foods that have different isotopic ratios, the consumers' tissues will reflect them. If diet changes as juveniles grow to adults and the unknown juvenile and adult resources have distinguishable isotopic ratios, diet change will be detected as a body-size-dependent change in the consumers' isotopic ratios. Furthermore, isotopic analyses of suspected foods can offer clues to the dominant items in the diets of juveniles and adults.
Although many studies have used stable isotopes to identify trophic links in aquatic and terrestrial communities (e.g., Tieszen et al. 1979, Martin et al. 1992, Forsberg et al. 1993, Ruckelshaus et al. 1993, Bootsma et al. 1996), few have examined intraspecific variations with body size (Rau et al. 1981, 1991, Hughes and Sherr 1983, Gearing et al. 1984, Minagawa and Wada 1984). The information lost by ignoring size or life-stage effects in isotopic analyses of food webs may be especially critical to accurate interpretations of seasonal dynamics (e.g., Yoshioka et al. 1994) because ontogeny and seasonality are often confounded. In this study, all four species that I examined showed significant ontogenetic changes in their ratios of carbon isotopes, implying gradual metaphoetesis as benthic juveniles grow.
Animals were collected from two intertidal sandflats: Skagit Bay and False Bay, Washington. Characteristics of Skagit Bay are reported in Eckman (1979) and Gallagher et al. (1983, 1990); those of False Bay are reported in Pamatmat (1968) and Miller and Sternberg (1988). Collections at Skagit Bay were made [approximately]100 m seaward of the bulrush marsh at an elevation of [approximately]1.2 m above MLLW. Collections at False Bay were made in a shallow tidepool ([approximately]10-20 cm deep and 3 x 10 m in area) [approximately]100 m from shore at an elevation of [approximately]1.4 m above MLLW.
Target species in Skagit Bay were Hobsonia florida (Ampharetidae), Pseudopolydora kempi japonica (Spionidae), and Pygospio elegans (Spionidae); those in False bay were Polydora cornuta (Spionidae) and P. kempi japonica. In addition to these primary target species, the meiofaunal oligochaete Amphichaeta leydigii (Naididae), which competes with juvenile H. florida for benthic diatoms (Gallagher et al. 1990, Hentschel and Jumars 1994), was collected from Skagit Bay so that stable isotopes could be used as an additional means to assess their shared diet.
All animals were collected within 5 m of a marker stake. Collections were attempted on several dates, but sufficient numbers of live and whole individuals of each species were not always recovered. Successful collection dates for H. florida were 8 June, 4 July, and 23 July 1993. P. kempi japonica and P. elegans were collected from Skagit Bay on 10 and 28 June 1994. A. leydigii were collected on 4 July 1993. In False Bay, P. kempi japonica were collected on 30 May, 25 July, 19 September, and 6 October 1994: P. cornuta were collected on 30 May, 25 July, 10 August, 24 August, and 19 September 1994.
Worm tubes were collected by sieving sediments in the field through either a 0.5- or 1.0-mm mesh. To collect smaller H. florida juveniles and A. leydigii from Skagit Bay, I suspended surficial sediments in a slurry and captured suspended material on a 63-[[micro]meter] mesh sieve.
P. cornuta and P. kempi japonica in False Bay have lecithotrophic larvae brooded within capsules inside the maternal, sediment tube (personal observation). Blake and Woodwick (1975) described the development of P. kempi japonica. Blake (1969) described adelphophagic development (feeding on nurse eggs) of some P. ligni (renamed cornuta; Blake and Maciolek 1987) as anomalous, but this type of lecithotrophic development is the only developmental mode observed in P. cornuta from False Bay (personal observation). When sorting tubes, any that contained encapsulated larvae ([less than]1% of all collected tubes) were set aside so that larvae could be removed and collected for isotopic analysis. In addition to brooded larvae, P. cornuta eggs were collected from capsules in two maternal tubes on 30 May 1994. H. florida also broods lecithotrophic larvae within the maternal tube (Zottoli 1974), but I could not collect sufficient mass of H. florida and P. kempi japonica larvae from Skagit Bay sediments.
In the laboratory, live individuals were sorted and relaxed with 10% Mg[Cl.sup.2]. Body size was determined by counting the number of setigerous segments and measuring body length with an ocular micrometer at 6.4 or 16 x. Worms were sorted into petri dishes containing 0.2-[[micro]meter] mesh filtered seawater (FSW) and kept at 13 [degrees] C for 24 h to allow gut clearance. Adults were large enough to be analyzed individually, but juveniles and larvae had to be pooled into size classes to obtain sufficient tissue. In general, larvae were pooled into groups of 20-85 individuals, while the number of juveniles in a pooled sample ranged from 2 to 70 individuals, depending upon body size. P. cornuta eggs were pooled in groups of 100 eggs. Individual or pooled samples were filtered onto precombusted (500 [degrees] C) quartz filters and rinsed with Milli-Q reagent grade water.
To permit inferences about the food resources from which the worms assimilate their carbon, primary producers likely to be consumed by the target species were collected. In Skagit Bay they included the two dominant marsh grasses Scirpus americanus and S. maritimus (Ewing 1982), the marine macrophyte Enteromorpha sp., and benthic diatoms. Samples of Enteromorpha and sediment from which diatoms were extracted were collected on 23 July 1993 from the same location as the animals; marsh grasses were collected [approximately]100-300 m landward. False Bay has a lower diversity of plants, and collections on 10 August 1994 were limited to benthic diatoms and the macrophyte Ulva sp., which covers most of the high intertidal during summer.
TO collect benthic diatoms, I placed a layer of Kim Wipes (Kimberly Clark, Roswell, Georgia) atop a surficial-sediment slurry in freezer trays. Trays were positioned under fluorescent lights for 10-12 h to cause benthic diatoms to migrate into the Kim Wipes (Eaton and Moss 1966). Kim Wipes were then removed and vigorously rinsed in a flask of FSW. This water was repeatedly centrifuged (5000 rpm) to concentrate particles. Solids were disaggregated in a petri dish, and pieces of Kim Wipe were removed under 40x magnification. Although labor intensive, this procedure resulted in a nearly pure sample of benthic diatoms (as viewed at 100x). The diatom sample was collected on a precombusted quartz filter.
Stable isotope analysis
Samples on filters were placed in precombusted (900 [degrees] C) Vycor tubes (9 mm diameter, Corning, Corning, New York). Macrophyte and marsh-grass samples were placed directly in Vycor tubes. To remove any carbonates from samples, 0.5-1.0 mL of 10% HCl was added to each tube. Samples were dried at 80 [degrees] C for 35 d. Precombusted (500 [degrees] C) CuO flakes (0.2 mg) were added to tubes, which then were evacuated and sealed. These ampules were combusted at 900 [degrees] C for 1 h.
Combusted ampules were placed in a vacuum line, and C[O.sub.2] that resulted during combustion was purified cryogenically. [[Delta].sup.13]C of each sample was measured relative to PDB (Pee Dee Belemnite Standard) standard in a Finnigan MAT 251 mass spectrometer (Finnigan Corporation, San Jose, California), where
[Mathematical Expression Omitted]. (1)
The standard deviation of [[Delta].sup.13]C measurements from a single sample was always less than 0.03[per thousands].
Interpreting changes in [[Delta].sup.13]C with body size as an indication of ontogenetic changes in diet requires two assumptions: (1) rates of carbon turnover are sufficiently rapid to reflect any seasonal changes in diet that may be confounded with growth, and (2) isotopic fractionation does not vary significantly with body size. The first assumption is especially important because turnover rates of body carbon depend on growth rates (Fry and Arnold 1982); rapidly growing juveniles will incorporate new carbon (relative to that of existing tissue) more rapidly than slowly growing adults. The total carbon pool of a juvenile's tissue, therefore, will be coupled much more tightly to temporal changes in the [[Delta].sup.13]C of available food resources than will that of adults. To conclude that size-dependent changes in [[Delta].sup.13]C are caused by ontogenetic changes in diet rather than an artifact due to size-dependent rates of carbon turnover, juveniles and adults must be sampled for a period in excess of the time required to reflect significant changes in the [[Delta].sup.13]C of their food.
I conducted several laboratory experiments with P. kempi japonica to estimate the rates of carbon turnover by postlarvae, juveniles, and adults. Animals from False Bay were placed in individual wells of tissue culture dishes. Each well was filled with FSW and a thin layer ([approximately]0.5 mm) of clean glass beads (1-100 [[micro]meter] diameter; Ferro Corporation, Cleveland, Ohio) so worms could construct tubes. Every 24-48 h, worms were removed from tubes and transferred individually to new wells containing new beads and FSW. During each transfer, Gerber Rice Cereal (GRC, [[Delta].sup.13]C = -26.4[per thousands]) was added so worms could feed ad libitum on a diet of known and constant [[Delta].sup.13]C.
The GRC was added by first making a GRC flour (particle diameter [approximately]50-90 [[micro]meter]) with a household coffee grinder (Starbucks Corporation, Seattle, Washington) and then mixing 0.1 g of the GRC flour in 10 mL FSW. A 0.1-mL portion of the resulting broth was pipetted into each well. At each transfer, excess food was seen in every dish. The frequent transfers ensured that worms were feeding on fresh GRC with [[Delta].sup.13]C = -26.4[per thousands], rather than on GRC that had been allowed to age and undergo significant bacterially mediated isotopic fractionation during the experiment.
Experiments were designed to note size-dependent differences in carbon turnover rates and to determine whether worms that fed on GRC would have similar [[Delta].sup.13]C values regardless of their body size (cf. assumption 2). Experiments began with three size classes of P. kempi japonica: lecithotrophic larvae (0.5-0.75 mm body length) that will begin feeding soon after removal from brood capsules, large juveniles (6-8 mm), and adults (10-12 mm). Larvae and small postlarvae had to be pooled for [[Delta].sup.13]C analysis. Juveniles were chosen to be large enough to measure the [[Delta].sup.13]C of individuals. Because of time constraints imposed by the labor-intensive sorting and transfers of individual worms, the larval experiment was conducted separately (beginning 1 August 1994) from the experiment involving the juveniles and adults (beginning 6 October 1994). Field-collected specimens of each class were analyzed to provide values at the start of the experiment. Two pooled samples of postlarvae (15 individuals/sample) were collected after 15 d of feeding on GRC. Juveniles and adults were sampled after 10, 20, 40, and 60 d of feeding on GRC. High mortality and time constraints limited me to 2-4 replicates for a given size class and sampling date. In addition to the GRC experiment that began with larvae, a second group of larvae collected on 10 August 1994 was fed Gerber Mixed Cereal (GMC, [[Delta].sup.13]C = -21.6[per thousands]), and two replicate pooled samples (10 postlarvae/sample) were analyzed after 20 d of feeding on GMC.
I used body length as the measure of body size (and its mean for pooled samples). For each species, I planned a priori to analyze data pooled among different collection dates to test for body-size variations in [[Delta].sup.13]C. Data from eggs and lecithotrophic larvae were excluded a priori from all statistical analyses of size-dependent changes in the [[Delta].sup.13]C of feeding worms. Statistical analyses were computed with SYSTAT 5.2.1 (Macintosh version).
For each population, Spearman's rank correlations between [[Delta].sup.13]C and body length were computed on pooled data and data from individual collection dates. Significance of each correlation was tested using the Hotelling-Pabst test (Conover 1980). When body-size trends were significant, LOWESS smoothing (Cleveland 1979, Trexler and Travis 1993) was conducted to suggest functional relationships. Because LOWESS fits suggested significant departure from a simple linear function, two nonlinear regression models were fitted to the pooled data using quasi-Newton iterations (SYSTAT 1992): (1) a negative exponential function (3 df) and (2) a piecewise model composed of two joined lines with a breakpoint (4 df). The exponential and piecewise models have different mathematical formulations, but both lead to a similar interpretation: the size-dependent change in [[Delta].sup.13]C diminishes as worms approach adulthood (H. florida, P. cornuta, and P. kempi japonica become sexually mature at [approximately]10 mm; personal observation). The negative exponential model allows calculation of the body length at which an arbitrary degree of attenuation is reached (e.g., when 90% of the change is complete). Locating the breakpoint between the two lines of the piecewise model provides a less arbitrary means to identify the size at which growing worms near the [[Delta].sup.13]C of adults, and a 95% confidence interval around the location of the breakpoint can be calculated. F tests were used to compare fits statistically (Davis 1986).
Temporal effects within pooled data were tested by analysis of covariance (ANCOVA) with body length as the covariate and collection date as the treatment factor. Prior to performing the ANCOVA for each population, two-way ANOVA was conducted to verify that no significant interaction between body size and collection date existed ([Alpha] = 0.05). When a significant date-size interaction existed, results of the ANCOVA would be suspect, so I applied a Kruskal-Wallis test on similarly sized large worms ([greater than]9 mm) collected on different dates.
The dominant macroalga in False Bay, Ulva sp., was isotopically heavier than the dominant macroalga in Skagit Bay, Enteromorpha sp. (Table 1). Similarly, samples of benthic diatoms were more enriched with 13C in False Bay (Table 1). The two species of marsh grass in Skagit Bay had large differences in [[Delta].sup.13]C (Table 1).
Pseudopolydora kempi japonica data pooled among collection dates from Skagit Bay showed significant correlation between [[Delta].sup.13]C and body size, as did data from within each of the two collection dates [ILLUSTRATION FOR FIGURE 1 OMITTED]. F tests comparing the negative exponential and piece wise-linear models indicate that the former is the statistically better description ([ILLUSTRATION FOR FIGURE 1 OMITTED]; [F.sub.1, 19] = 5.00, P = 0.040). Both models, however, show that most of the size-dependent change in [[Delta].sup.13]C occurs when P. kempi japonica are smaller than [approximately]6 mm [ILLUSTRATION FOR FIGURE 1 OMITTED]. The exponential model indicates that [[Delta].sup.13]C values reach 90% of the asymptote (-13.38[per thousands]; [ILLUSTRATION FOR FIGURE 1 OMITTED]) at a body length of 6.5 mm ([[Delta].sup.13]C = -13.56[per thousands]). The breakpoint between the two adjoining lines of the piecewise model was determined to occur at 5.5 mm [ILLUSTRATION FOR FIGURE 1 OMITTED].
TABLE 1. The [[Delta].sup.13]C of some potential foods at each field site. Each [[Delta].sup.13]C value is that of a single sample. Primary producer [[Delta].sup.13]C ([per thousands]) False Bay Ulva sp. -7.95 (macroalga) -8.86 Benthic diatoms -19.21 -20.38 Skagit Bay Enteromorpha sp. -10.22 (macroalga) -10.73 Scirpus maritimus(*) -14.55 (marsh grass) -14.90 Benthic diatoms -20.91 -21.29 S. americanus(*) -27.10 (marsh grass) -27.31 * The genus Scirpus is reported to contain both [C.sub.3] and [C.sub.4] species (Downton 1975). Notes: Samples from Skagit Bay were collected on 23 July 1993. Samples from False Bay were collected on 10 August 1994.
Temporal effects were not detected in the [[Delta].sup.13]C data for Skagit Bay P. kempi japonica. ANCOVA did not indicate a significant effect due to collection date ([F.sub.1, 20] = 1.451, P = 0.242), but interpretation is problematic because of significant date-size interaction (two-way ANOVA: [F.sub.1, 19] = 6.165, P = 0.023). In accord with the ANCOVA, however, the Kruskal-Wallis test comparing only large worms among collection dates was not significant (H = 5.018, P = 0.285).
P. kempi japonica collected from False Bay also showed significant size-dependent variations in [[Delta].sup.13]C [ILLUSTRATION FOR FIGURE 2 OMITTED]. Significant correlations between [[Delta].sup.13]C and body size resulted for pooled data and for data from each collection date [ILLUSTRATION FOR FIGURE 2 OMITTED]. There was no significant difference between the negative exponential and piecewise-linear models ([ILLUSTRATION FOR FIGURE 2 OMITTED]; [F.sub.1, 42] = 1.28, P = 0.272), indicating that the model with fewer degrees of freedom (the negative exponential) is the more parsimonious descriptor of the data. The exponential model indicates that [[Delta].sup.13]C values reach 90% of the asymptotic value (-10.74[per thousands]; [ILLUSTRATION FOR FIGURE 2 OMITTED]) at a body length of 15.5 mm ([[Delta].sup.13]C = -10.93[per thousands]). The breakpoint between the two intersecting lines of the piecewise model occurs at 6.0 mm [ILLUSTRATION FOR FIGURE 2 OMITTED].
Pooled data for False Bay P. kempi japonica were more variable than those from Skagit Bay because the former included data from four collection dates and significant temporal trends were apparent. Visually, the data show that [[Delta].sup.13]C values for all sizes of False Bay P. kempi japonica became heavier from May to October [ILLUSTRATION FOR FIGURE 2 OMITTED]. The assumption of no interaction between collection date and body size held ([F.sub.3, 38] = 1.939, P = 0.140), and the ANCOVA revealed a highly significant effect of collection date with body size as a covariate ([F.sub.3, 41] = 12.278, P [less than] 0.001).
In general, both populations of P. kempi japonica showed similar ontogenetic trends [ILLUSTRATION FOR FIGURE 3 OMITTED]. False Bay P. kempi japonica were isotopically heavier than Skagit Bay individuals of comparable size [ILLUSTRATION FOR FIGURE 3 OMITTED], consistent with the differences in [[Delta].sup.13]C of primary producers at each sandflat (Table 1). Piecewise-linear models show a breakpoint at [approximately]6 mm for each population [ILLUSTRATION FOR FIGURE 3 OMITTED].
Polydora cornuta also showed significant size-dependent variations in [[Delta].sup.13]C [ILLUSTRATION FOR FIGURE 4 OMITTED]. Regressions and F tests indicated no significant difference between the negative exponential and the piecewise-linear models ([ILLUSTRATION FOR FIGURE 4 OMITTED]; [F.sub.1, 41] = 1.02, P = 0.345). The exponential function indicates that [[Delta].sup.13]C values reach 90% of the asymptote (-9.69[per thousands]; [ILLUSTRATION FOR FIGURE 4 OMITTED]) at a body length of 13.0 mm ([[Delta].sup.13]C = -9.89[per thousands]). The breakpoint in the piecewise model occurred at a body length of 5.6 mm [ILLUSTRATION FOR FIGURE 4 OMITTED].
Temporal effects were significant in the P. cornuta data. [[Delta].sup.13]C values became isotopically heavier from 30 May to 10 August and then shifted back to lighter values by 19 September [ILLUSTRATION FOR FIGURE 4 OMITTED]. For data from the later four collection dates, there was a highly significant difference in [[Delta].sup.13]C among dates (ANCOVA: [F.sub.3, 38] = 15.222, P [less than] 0.001).
Hobsonia florida also showed significant ontogenetic variation in [[Delta].sup.13]C [ILLUSTRATION FOR FIGURE 5 OMITTED]. Body length and [[Delta].sup.13]C were significantly correlated when data were pooled among collection dates and when data from each of the latter two collection dates were analyzed individually [ILLUSTRATION FOR FIGURE 5 OMITTED]; lack of a significant correlation in the 8 June data probably occurred. because data consisted of only five adults and a single sample of meiofaunal juveniles. F tests comparing the two models indicated that the negative exponential is the statistically superior description of the relationship ([F.sub.1, 39] = 9.25, P = 0.005). Both regression models, however, show that most of the size-dependent change in [[Delta].sup.13]C occurs when H. florida are smaller than [approximately]8 mm [ILLUSTRATION FOR FIGURE 5 OMITTED]. The exponential model indicates that [[Delta].sup.13]C values reach 90% of the asymptote (-13.94[per thousands]; [ILLUSTRATION FOR FIGURE 5 OMITTED]) at a body length of 8.6 mm ([[Delta].sup.13]C = -14.27[per thousands]), while the breakpoint in the piecewise model occurred at 8.0 mm [ILLUSTRATION FOR FIGURE 5 OMITTED].
Temporal trends in the H. florida [[Delta].sup.13]C data were not evident. ANCOVA for the effects of collection date also was not significant ([F.sub.2, 39] = 2.279, P = 0.115), but the ANOVA to test for date-size interaction was significant ([F.sub.2, 37] = 8.28, P = 0.001). The Kruskal-Wallis test comparing worms [greater than]9 mm among collection dates was not significant (H = 0.184, P = 0.912).
In contrast to the other polychaetes, Pygospio elegans had a narrow range of body sizes and [[Delta].sup.13]C [ILLUSTRATION FOR FIGURE 6 OMITTED]. P. elegans did have significant correlation between [[Delta].sup.13]C and body length when data were pooled, but separate analyses of the two collection dates yielded nonsignificant correlations at [Alpha] = 0.05 [ILLUSTRATION FOR FIGURE 6 OMITTED]. Although the [r.sup.2] of the piecewise regression model is greater than that of the exponential model [ILLUSTRATION FOR FIGURE 6 OMITTED], the former does not explain significantly more of the data's variability when their differing degrees of freedom are considered ([F.sub.1, 11] = 0.053, P [greater than] 0.750). Collection date had a significant effect in the P. elegans data (ANCOVA: [F.sub.1, 12] = 5.900, P = 0.032).
Amphichaeta leydigii were collected from Skagit Bay in sufficient numbers only on 4 July 1993. Two pooled samples (90 and 94 individuals/sample) had [[Delta].sup.13]C = -15.73 and -15.36[per thousands]; average body lengths were 0.75 mm.
Rates of carbon turnover clearly differed between P. kempi japonica that were 6-8 mm and those that were 10-12 mm in length [ILLUSTRATION FOR FIGURE 7 OMITTED]. The smaller worms approached a final [[Delta].sup.13]C [approximately equal to] -25[per thousands] in 40 d, while larger individuals required 60 d to reach the same [[Delta].sup.13]C as the smaller worms. Both size classes, however, showed significant changes in [[Delta].sup.13]C in as little as 10 d after the switch from their natural diets to the controlled diet of GRC [ILLUSTRATION FOR FIGURE 7 OMITTED]. At 60 d, both small and large worms had a fractionation factor of approximately +1.3[per thousands] [ILLUSTRATION FOR FIGURE 7 OMITTED].
The two samples of P. kempi japonica postlarvae that were fed GRC ([[Delta].sup.13]C = -26.4[per thousands]) for 15 d after removal from brood capsules reached 0.6-0.8 mm and [[Delta].sup.13]C = -24.3 and -24.7[per thousands]. Duplicate samples of postlarvae fed GMC ([[Delta].sup.13]C = -21.6[per thousands]) for 20 d reached [[Delta].sup.13]C = -19.7 and -20.0[per thousands]. Both GRC and GMC postlarvae showed similar fractionation, ranging from +1.6 to +2.1[per thousands].
Ontogenetic changes in [[Delta].sup.13]C and diet
Obvious changes in habitat, such as amphibian metamorphosis or meroplanktonic larval settlement, certainly will cause ontogenetic diet changes that probably will alter an animal's [[Delta].sup.13]C. In addition, general relationships between the size of an animal and the mean size of its food items (e.g., Hughes 1980, Vezina 1985, Fenchel 1987, Fisher and Dickman 1993) also suggest that diet changes are likely to occur during growth within a single habitat. This intuitive prediction, however, is derived from studying macrophages, i.e., animals that evaluate food items individually and are only 1-2 orders of magnitude larger than their food resources. Deposit- and suspension-feeding benthos are microphages: they process particles in bulk and obtain their nutrition from items that are several orders of magnitude smaller than themselves (Yonge 1928, Fauchald and Jumars 1979, Jumars et al. 1990). While isotopic evidence of an ontogenetic diet change in a macrophage (e.g., shrimp and crabs: Hughes and Sherr 1983) is not surprising; any evidence of intraspecific changes in the diets of microphagous species after settlement is notable because their diet components are so poorly understood (Lopez and Levinton 1987, Jumars 1993). Merely identifying the existence of ontogenetic diet changes opens many new avenues of research into the ecology of such species.
Because stable isotopic tracers have proven to be an effective tool to identify trophic links in systems dominated by detritivores and other species for which visual analysis of gut contents is ineffective, the data showing ontogenetic changes in [[Delta].sup.13]C indicate that isotopic measurements can be used to resolve possible metaphoetesis within food webs that previously have not included such detailed structure. The four polychaetes studied are not the first deposit-feeding species for which variation in [[Delta].sup.13]C has been related to body size. Gearing et al. (1984) included data showing positive, linear relationships between [[Delta].sup.13]C and body size of the deposit-feeding bivalves Pitar morrhuanna, Nucula annulata, and Yoldia limatula (their [ILLUSTRATION FOR FIGURE 5 OMITTED]). Detailed pursuit of the functional relationship, however, was beyond the scope of Gearing et al. (1984).
Correlations between [[Delta].sup.13]C and body length clearly demonstrate intraspecific changes in [[Delta].sup.13]C during growth of the deposit feeders that I studied, and nonlinear regressions indicate that most of the size-dependent change in [[Delta].sup.13]C occurs in juveniles. In addition, the nearly identical location of breakpoints in the piecewise models for P. kempi japonica from Skagit Bay and False Bay shows that both populations undergo similar ontogenetic changes, despite differences in the [[Delta].sup.13]C of available foods at each site ([ILLUSTRATION FOR FIGURE 3 OMITTED], Table 1).
Interpreting changes in [[Delta].sup.13]C as changes in diet
To interpret intraspecific variations in [[Delta].sup.13]C with body size as evidence of ontogenetic diet changes, several alternative explanations must be ruled out (see Fry and Sherr 1984). First, biochemical composition affects [[Delta].sup.13]C, the largest effect being the 2-4[per thousands] depletion of 13C in lipids and fatty tissue relative to protein and muscle (DeNiro and Epstein 1978, Tieszen et al. 1983, Kling et al. 1992). The lighter [[Delta].sup.13]C of lecithotrophic larvae and eggs relative to adults [ILLUSTRATION FOR FIGURES 2 AND 4 OMITTED], for example, probably results from very high lipid concentrations. Although the magnitude of this biochemical effect could explain the range of [[Delta].sup.13]C observed between juveniles and adults, such an explanation would require that juveniles were almost completely composed of lipids and adults entirely lacked lipids. In fact, adults have greater lipid reserves than juveniles (Hentschel, in press). If changes in biochemical composition are considered, therefore, the biochemically corrected differences in [[Delta].sup.13]C between juveniles and adults would be even greater than those measured.
Various physiological changes during growth also could alter isotopic fractionation between an animal and its food, potentially causing size-dependent changes in the animal's [[Delta].sup.13]C even when diet remains constant during ontogeny. The laboratory experiments with GRC and GMC, however, do not indicate any such physiological effects during growth [ILLUSTRATION FOR FIGURE 7 OMITTED].
Variable rates of carbon turnover with growth (Fry and Arnold 1982) also must be considered when interpreting the size-dependent variations in [[Delta].sup.13]C of field-collected animals as evidence of ontogenetic changes in diet. The laboratory experiment with P. kempi japonica illustrates the difference in carbon turnover rates between individuals that are 6-8 mm in length and those that are 10-12 mm [ILLUSTRATION FOR FIGURE 7 OMITTED]. While these two size classes appeared to require [approximately]40 and 60 d, respectively, to approach a final [[Delta].sup.13]C that closely reflects their diet of GRC [ILLUSTRATION FOR FIGURE 7 OMITTED], larvae removed from brood capsules and fed GRC as postlarvae reached similar [[Delta].sup.13]C in only 15 d. Under some circumstances, juveniles and adults could have the same diet at the time of analysis but still show differences in [[Delta].sup.13]C due to their different carbon turnover rates and seasonal changes in available foods. Adults collected in summer could, for example, be reflecting the [[Delta].sup.13]C of their winter diet, while juveniles born in the summer could have the [[Delta].sup.13]C of the adults' summer diet. Observed seasonal changes in the [[Delta].sup.13]C of worms collected from False Bay, however, dismiss this possibility because all body sizes showed seasonal shifts', date-specific relationships were essentially parallel [ILLUSTRATION FOR FIGURES 2 AND 4 OMITTED]. If size-dependent rates of carbon turnover are the cause of the observed intraspecific variations in [[Delta].sup.13]C, the magnitude of the difference between juveniles and adults should decline over time as adults gradually reflect more of the [[Delta].sup.13]C signal of their summer food. The temporal shifts in all sizes of the False Bay P. kempi japonica and P. cornuta indicate that the carbon turnover rates of both adults and juveniles are sufficiently high to couple their [[Delta].sup.13]C to their recent diet. In fact, data from the GRC experiment show that significant changes in [[Delta].sup.13]C occur as a result of diet changes in as little as 10 d for even slowly growing adults [ILLUSTRATION FOR FIGURE 7 OMITTED].
Although changes in biochemical composition, physiology, and rates of carbon turnover occur during growth, the accumulated evidence argues that these factors cannot account for the observed field data. The most parsimonious explanation of the positive relationships between [[Delta].sup.13]C and body size of the species I studied is an ontogenetic change in diet.
Inferring diet components from [[Delta].sup.13]C values
The [[Delta].sup.13]C values of likely food resources in each habitat (Table 1) suggest possible components of juvenile, adult, and intermediate diets. Generally the fractionation between an animal and its diet ranges from -2 to +3[per thousands], with a mean of +0.2[per thousands] (Peterson and Fry 1987). P. kempi japonica fed GRC for 60 d had a fractionation factor of approximately +1.3[per thousands] [ILLUSTRATION FOR FIGURE 7 OMITTED]. Together these data and the [[Delta].sup.13]C of the measured foods (Table 1) suggest that juvenile diets in False Bay are a mixture of benthic diatoms and Ulva, while adult diets are dominated more by Ulva. Analysis of photosynthetic pigments also has implicated Ulva as a dominant diet item for deposit-feeding annelids in False Bay (Levinton and McCartney 1991). In Skagit Bay, where the saltmarsh provides a more diverse mixture of potential foods, adult diets isotopically resemble Enteromorpha and S. maritimus. Skagit Bay juveniles, on the other hand, have an isotopic signature closer to benthic diatoms.
Seasonal trends in False Bay further support the interpretation that lighter [[Delta].sup.13]C of juveniles is due to a greater amount of diatoms in their diet relative to the Ulva-dominated diet of conspecific adults [ILLUSTRATION FOR FIGURES 2 AND 4 OMITTED]. In False Bay, Ulva gradually accumulates and covers most of the sandflat by late summer (personal observation). In August, the abundance of Ulva declines. Benthic diatom abundances, however, peak in mid-June and decline during summer. If both juveniles and adults assimilate a mixture of benthic diatoms and Ulva, with juvenile diets having a relatively greater proportion of diatoms, the summer increase in Ulva and decrease in diatoms should drive the [[Delta].sup.13]C of all worms toward that of Ulva in August. As the abundance of Ulva declines, the [[Delta].sup.13]C of all worms should shift back to lighter values. The data for P. kempi japonica and P. cornuta show this pattern [ILLUSTRATION FOR FIGURES 2 AND 4 OMITTED].
Isotopic composition of a single food also might vary seasonally. Schwinghamer et al. (1983), for example, found that the benthic diatom Gyrosigma sp. had [[Delta].sup.13]C [approximately equal to] -18[per thousands] in May but then shifted to [[Delta].sup.13]C [approximately equal to] -13.6[per thousands] from June through September. Several studies have measured the [[Delta].sup.13]C of benthic diatoms (see Table 3 in Currin et al. 1995), but only Schwinghamer et al. (1983) measured a seasonal change. The cause of this seasonality is not known, but C[O.sub.2] limitation is a likely explanation (Gould and Gallagher 1990, Zimba et al. 1990, Fry 1996). Although I did not make seasonal collections of benthic diatoms, samples from both Skagit Bay and False Bay had depleted [[Delta].sup.13]C values (Table 1) that are inconsistent with Schwinghamer et al.'s (1983) observation that benthic diatoms became enriched with 13C during the summer.
Unfortunately, a stable-isotope approach to diet analysis, particularly when only a single element is examined, cannot provide definitive identification of diet components beyond the isotopic similarity of measured foods and consumers. Many other potential food sources (e.g., phytoplankton, potential meiofaunal prey, sediment bacteria) could have been sampled and analyzed for [[Delta].sup.13]C. Those that are isotopically similar to benthic diatoms, macroalgae, and [C.sub.4] saltmarsh plants must also be considered potential diet items. Phytoplankton, which often have [[Delta].sup.13]C [approximately equal to] -20[per thousands] (Fry 1996), deserve special attention because adults of the spionid species used in this study are known to switch between deposit feeding and suspension feeding in response to variations in flow speed and particle flux (Taghon et al. 1980, Dauer et al. 1981). If juveniles tend to suspension feed more than adults, as Hentschel (1996) and Shimeta (1996) speculate, phytoplankton probably would be more common in juvenile diets than in adult diets, and such a diet shift could produce the ontogenetic variations in [[Delta].sup.13]C measured in this study. In general, stable-isotope analysis of food webs is quite difficult unless additional foraging limitations constrain a given consumer to feed on only a subset of the isotopically characterized foods. Such limitations cannot be made objectively for deposit feeders (and detritivores in general) because they have access to virtually all sources of organic matter that eventually deposit in sediments.
In addition, quantitative analysis of the contributions by only two end-member diet items (e.g., benthic diatoms vs. Ulva) assumes that isotopically intermediate foods do not exist. In trophically complex sediments, isotopic intermediates cannot be ruled out. For example, protists that feed on benthic diatoms and have an isotopic fractionation of +2[per thousands] would have [[Delta].sup.13]C [approximately equal to] -18[per thousands] and, thus would be a potentially dominant item in the juvenile diets. On the other side of the spectrum, fresh Ulva undergoing aerobic microbial degradation would probably shift from [[Delta].sup.13]C [approximately equal to] -8[per thousands] to detritus with [[Delta].sup.13]C [approximately equal to] -10[per thousands] (see Spies et al. 1989). Of course, if heterotrophic bacteria are a major food resource for deposit feeders, a deposit feeder's [[Delta].sup.13]C will be difficult to distinguish from that of the bacteria's food.
Independent evidence regarding the species I studied, however, supports the simplest conclusion that there is an ontogenetic shift from a diet dominated by benthic diatoms to one dominated by macroalgae (or marsh grass). Hentschel and Jumars (1994) reduced natural densities of microphytobenthos in Skagit Bay by perfusing pore waters with the herbicide DCMU [3-(3, 4-dichlorophenol)-1, 1-dimethylurea] to test the significance of benthic diatoms as a resource for juvenile H. florida. H. florida juveniles and competing A. leydigii responded with significant declines in abundances within DCMU patches relative to control patches (Hentschel and Jumars 1994). P. kempi japonica juveniles and P. elegans were not abundant enough during that field season (1991) to discern effects due to diatom densities. In the following year when P. kempi japonica abundances were high, however, B. T. Hentschel (unpublished data) transplanted sediment cores incubated in the laboratory under light or dark conditions to produce diatom-rich and diatom-poor patches; A. leydigii and juveniles of both H. florida and P. kempi japonica showed significantly higher abundances in the diatom-rich patches.
Interspecific differences in [[Delta].sup.13]C are apparent in the polychaetes collected from Skagit Bay [ILLUSTRATION FOR FIGURE 8 OMITTED] and those from False Bay [ILLUSTRATION FOR FIGURE 9 OMITTED]. In general, the interspecific differences in [[Delta].sup.13]C are less than the intraspecific bodysize effect [ILLUSTRATION FOR FIGURES 8 AND 9 OMITTED], suggesting that size or age classes of a species might represent different "ecological species" (see Polis 1984). Furthermore, the intraspecific effect of body length on [[Delta].sup.13]C does not necessarily apply across species: the smallest of the polychaete species, P. elegans, had the heaviest [[Delta].sup.13]C values [ILLUSTRATION FOR FIGURE 8 OMITTED].
The [[Delta].sup.13]C values suggest resource partitioning among species at each sandflat [ILLUSTRATION FOR FIGURES 8 AND 9 OMITTED]. These interspecific differences in diet may explain the lack of competition observed between P. elegans and P. kempi japonica (Wilson 1983, 1984). Such inferences about interspecific resource partitioning, however, should include the caveat that differences in [[Delta].sup.13]C among similarly sized individuals of different species are [approximately]1-2[per thousands] [ILLUSTRATION FOR FIGURES 8 AND 9 OMITTED] - well within the range of measured isotopic fractionation factors (Peterson and Fry 1987). Until fractionation factors for each species are determined with accuracy on the order of 0.1[per thousands], the possibility that each species assimilates the same diet but isotopically fractionates food differently should not be overlooked.
One notable characteristic of P. elegans is its ability to reproduce asexually (Rasmussen 1953, Wilson 1983), giving a narrow size range [ILLUSTRATION FOR FIGURES 6 AND 8 OMITTED] and obviating the need to provide a larva with lipids. Because [[Delta].sup.13]C values suggest that P. elegans may not have the same diet as similarly sized individuals of the other two Skagit Bay polychaetes [ILLUSTRATION FOR FIGURE 8 OMITTED], asexual reproduction and exploitation of the larger species' adult food resources could provide a means for P. elegans to bypass the high-quality diet needed by the other species' juveniles.
In contrast, comparisons between the [[Delta].sup.13]C values of the meiofaunal oligochaete A. leydigii ([[Delta].sup.13]C [approximately equal to] -15.5[per thousands]) and juveniles of H. florida [ILLUSTRATION FOR FIGURE 5 OMITTED] agree with results of field manipulations that suggest A. leydigii and H. florida of similar size compete for benthic diatoms (Gallagher et al. 1990, Hentschel and Jumars 1994). As juvenile H. florida grow and their diet changes, the degree of competition with A. leydigii should decrease. Size-dependent competition also may exist in Skagit Bay with juvenile P. kempi japonica ([ILLUSTRATION FOR FIGURE 8 OMITTED]; B. T. Hentschel, unpublished data).
Implications of the ontogenetic diet change
Although the [[Delta].sup.13]C data showed a gradual change in diet during growth, the co-occurrence of breaks in the body-size dependence of particle-size selection (Hentschel 1996), lipid reserves (Hentschel, in press), and [[Delta].sup.13]C at approximately the same body length (5-7 mm) suggests that this body size has special foraging and nutritional significance in the ontogeny of P. kempi japonica and similar species. Hentschel's (1996) model of particle contact and glass-bead experiments with both palp mimics and P. kempi japonica show that small worms with thinner feeding palps are biased toward contacting larger particles more frequently than will large worms with thicker palps. In sediments like those of Skagit Bay and False Bay, the palp-size effect will diminish when palps are [approximately]125-150 [[micro]meter] wide; this transition corresponds to a body length of [approximately]6-7 mm (Hentschel 1996). Similarly, the lipid reserves of P. kempi japonica increase during growth to a length of [approximately]6-7 mm, but the amount of lipid reserves is independent of body size for larger worms (Hentschel, in press). These ontogenetic changes in particle-size selection, lipid reserves, and diet generally stop before juveniles of these deposit-feeding polychaetes reach sexual maturity. This result contrasts with those of Forbes (1989), who found a break in the allometry of egestion rate of Capitella sp. I at a body size that corresponds closely to the onset of sexual maturity (body volume = 1.7 [mm.sup.3]).
Gradual ontogenetic changes in [[Delta].sup.13]C during the juvenile stage may result from subtle shifts in particle-size selection (Hentschel 1996), gradual increases in the growing worms' digestive capabilities (Penry and Jumars 1990), or more radical changes in foraging mode. Jumars et al. (1990) noted an absence of deposit feeding in species with body volumes less than [approximately]1 [mm.sup.3] and implicated the limit that small gut volume places on gut residence time (Penry and Jumars 1990) as the cause. In an ontogenetic context, individuals of "deposit-feeding" species are likely to feed macrophagously on higher quality food (e.g., herbivory or carnivory) until they grow to a size that permits microphagy on lower quality, bulk sediments. Volume estimates indicate that P. kempi japonica and P. cornuta that are 5-6 mm long (0.5-0.7 mm wide) have body volumes [approximately equal to]1-2 [mm.sup.3], consistent with Jumars et al.'s (1990) prediction of a shift from macrophagy to microphagy at this size.
The [[Delta].sup.13]C evidence of gradual ontogenetic diet changes in H. florida, P. kempi japonica, and P. cornuta, especially before reaching a length of [approximately]5-7 mm, and the distinct shift that seems to occur at that size suggest that the causes of any food limitation of these polychaete populations are likely to differ between individuals that are smaller or larger than [approximately]5-7 mm. If the food resources that dominate the diets of the smaller worms (e.g., benthic diatoms) limit their abundance, while foods assimilated by deposit-feeding adults are in ample supply, juvenile recruitment bottlenecks will occur (Gallagher et al. 1990, Hentschel and Jumars 1994). Substantial attention has been paid to the niche changes that many benthos undergo at larval settlement and metamorphosis (Thorson 1950, Strathmann 1985, McEdward 1995). For species that deposit feed as adults, the data and arguments presented here suggest that a second ontogenetic change in niche occurs prior to adulthood. The apparent need for a richer juvenile diet almost certainly has significant effects on the population dynamics of such species and on other members of their community.
P. Jumars provided advice and support throughout this project. P. Quay generously provided his laboratory for stable isotope analyses; I especially thank D. Wilbur for expert advice and running samples in the mass spectrometer. The Director of the Friday Harbor Labs provided access to False Bay and outstanding facilities. The Washington Dept. of Wildlife permitted access to Skagit Wildlife Area. K. Maekawa, K. Maring, J. Thomas, and A. Huston assisted in both the field and lab. During all field collections, C. Falkenhayn carried heavy supplies and then played in the mud. Comments by P. Jumars, L. A. Levin, L. Crowder, E. D. Gallagher, J. W. Deming, R. Strathmann, A. Moran, and an anonymous reviewer improved earlier drafts of this manuscript. Financial support was provided by NSF Grants OCE 92-02855 and OCE 96-17701 to P. Jumars and A. R. M. Nowell.
Benke, A. C., and D. I. Jacobi. 1994. Production dynamics and resource utilization of snag-dwelling mayflies in a blackwater riven Ecology 75: 1219-1232.
Bergman, E., and L. A. Greenberg. 1994. Competition between a planktivore, a benthivore, and a species with ontogenetic diet shifts. Ecology 75: 1233-1245.
Bianchi, T. S. 1988. Feeding ecology of subsurface deposit-feeder Leitoscoloplos fragilis Verrill. I. Mechanisms affecting particle availability on intertidal sandflat. Journal of Experimental Marine Biology and Ecology 115: 79-97.
Blake, J. A. 1969. Reproduction and larval development of Polydora from northern New England (Polychaeta: Spionidae). Ophelia 7: 1-63.
Blake, J. A., and N. J. Maciolek. 1987. A redescription of Polydora cornuta Bosc (Polychaeta: Spionidae) and designation of a neotype. Bulletin of the Biological Society of Washington 7: 11-15.
Blake, J. A., and K. H. Woodwick. 1975. Reproduction and larval development of Pseudopolydora paucibranchiata (Okuda) and Pseudopolydora kempi (Southern) (Polychaeta: Spionidae). Biological Bulletin 149: 109-127.
Bootsma, H. A., R. E. Hecky, R. H. Hesslein, and G. F. Turner 1996. Food partitioning among Lake Malawi nearshore fishes as revealed by stable isotope analysis. Ecology 77: 1286-1290.
Caceres, C. W., L. S. Fuentes, and F. P. Ojeda. 1994. Optimal feeding strategy of the temperate herbivorous fish Aplodactylus punctatus: the effects of food availability on digestive and reproductive patterns. Oecologia 99: 118-123.
Cammen, L. M. 1980. Ingestion rate: an empirical model for aquatic deposit feeders and detritivores. Oecologia (Berlin) 44: 303-310.
Cleveland, W. S. 1979. Robust locally weighted regression and smoothing scatter plots. Journal of the American Statistical Association 74: 829-836.
Cohen, J. E., et al. 1993. Improving food webs. Ecology 74: 252-258.
Conover, W. J. 1980. Practical nonparametric statistics, Second edition. John Wiley & Sons, New York, New York, USA.
Currin, C. A., S. Y. Newell, and H. W. Paerl. 1995. The role of standing dead Spartina alterniflora and benthic microalgae in salt marsh food webs: considerations based on multiple stable isotope analysis. Marine Ecology Progress Series 121: 99-116.
Dauer, D. M., C. A. Maybury, and R. M. Ewing. 1981. Feeding behavior and general ecology of several spionid polychaetes from the Chesapeake Bay. Journal of Experimental Marine Biology and Ecology 54: 21-38.
Davis, J. C. 1986. Statistics and data analysis in geology. John Wiley & Sons, New York, New York, USA.
DeFlaun, M. F., and L. M. Mayer 1983. Relationships between bacteria and grain surfaces in intertidal sediments. Limnology and Oceanography 28: 873-881.
Demment, M. W., and P. J. Van Soest. 1985. A nutritional explanation for body size patterns of ruminant and nonruminant herbivores. American Naturalist 125: 641-672.
DeNiro, M. J., and S. Epstein. 1978. Influence of diet on the distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta 42: 495-506.
Downton, W. J. S. 1975. The occurrence [C.sub.4] photosynthesis among plants. Photosynthetica 9: 96-105.
Eaton, J. W., and B. Moss. 1966. The estimation of numbers and pigment content in epipelic algal populations. Limnology and Oceanography 11: 584-595.
Eckman, J. E. 1979. Small-scale patterns and processes in a soft-substratum, intertidal community. Journal of Marine Research 37: 437-457.
Ewing, K. 1982. Plant response to environmental variation in the Skagit marsh. Dissertation. University of Washington, Seattle, Washington, USA.
Fauchald, K., and P. A. Jumars. 1979. The diet of worms: a study of polychaete feeding guilds. Oceanography and Marine Biology: an Annual Review 17: 193-284.
Fenchel, T. 1987. Ecology of protozoa. Science Tech Publishers, Madison, Wisconsin, USA.
Fisher, D. O., and C. R. Dickman. 1993. Body size-prey size relationships in insectivorous marsupials: tests of three hypotheses. Ecology 74: 1871-1883.
Forbes, T. L. 1989. The importance of size-dependent processes in the ecology of deposit-feeding benthos. Pages 171200 in G. R. Lopez, G. L. Taghon, and J. S. Levinton, editors. Ecology of marine deposit feeders. Springer-Verlag, New York, New York, USA.
Forsberg, B. R., C. A. R. M. Araujo-Lima, L. A. Martinelli, R. L. Victoria, and J. A. Bonassi. 1993. Autotrophic carbon sources for fish of the central Amazon. Ecology 74: 643-652.
Fry, B. 1996. 13C/12C fractionation by marine diatoms. Marine Ecology Progress Series 134: 283-294.
Fry, B., and C. Arnold. 1982. Rapid 13C/12C turnover during growth of brown shrimp (Penaeus aztecus). Oecologia (Berlin) 54: 200-204.
Fry, B., and E. B. Sherr. 1984. [[Delta].sup.13]C Measurements as indicators of carbon flow in marine and freshwater ecosystems. Contributions in Marine Science 27: 13-47.
Gage, J. 1994. Recruitment ecology and age structure of deep-sea invertebrate populations. Pages 223-242 in C. M. Young and K. J. Eckelbarger, editors. Reproduction, larval biology, and recruitment of deep-sea benthos. Columbia University Press, New York, New York, USA.
Gallagher, E. D., G. B. Gardner, and P. A. Jumars. 1990. Competition among the pioneers in a seasonal soft-bottom benthic succession: field experiments and analysis of the Gilpin-Ayala competition model. Oecologia 83: 427-442.
Gallagher, E. D., P. A. Jumars, and D. D. Trueblood. 1983. Facilitation of soft-bottom benthic succession by tube builders. Ecology 64: 1200-1216.
Gearing, J. N., P. J. Gearing, D. T. Rudnick, A. G. Requejo, and M. J. Hutchins. 1984. Isotopic variability of organic carbon in a phytoplankton-based, temperate estuary. Geochimica et Cosmochimica Acta 48: 1089-1098.
Gould, D. M., and E. D. Gallagher. 1990. Field measurement of specific growth rate, biomass, and primary production of benthic diatoms of Savin Hill Cove, Boston. Limnology and Oceanography 35: 1757-1770.
Hentschel, B. T. 1996. Ontogenetic changes in particle-size selection by deposit-feeding spionid polychaetes: the influence of palp size on particle contact. Journal of Experimental Marine Biology and Ecology 206: 1-24.
-----. In press. Spectrofluorometric quantification of neutral and polar lipids suggests a food-related recruitment bottleneck for juveniles of a deposit-feeding polychaete population. Limnology and Oceanography.
Hentschel, B. T., and P. A. Jumars. 1994. In situ chemical inhibition of benthic diatom growth affects recruitment of competing, permanent and temporary meiofauna. Limnology and Oceanography 39: 816-838.
Hughes, E. H., and E. B. Sherr. 1983. Subtidal food webs in a Georgia estuary: [[Delta].sup.13]C analysis. Journal of Experimental Marine Biology and Ecology 67: 227-242.
Hughes, R. N. 1980. Optimal foraging theory in the marine context. Oceanography and Marine Biology: an Annual Review 18: 423-481.
Jumars, P. A. 1993. Gourmands of mud: diet choice in deposit feeders. Pages 136-168 in R. N. Hughes, editor. Mechanisms of diet choice. Blackwell Scientific, London, UK.
Jumars, P. A., L. M. Mayer, J. W. Deming, J. A. Baross, and R. A. Wheatcroft. 1990. Deep-sea deposit-feeding strategies suggested by environmental and feeding constraints. Philosophical Transactions of the Royal Society of London Series A 331: 85-101.
Kling, G. W., B. Fry, and W. J. O'Brien. 1992. Stable isotopes and planktonic trophic structure in arctic lakes. Ecology 73: 561-566.
Levinton, J. S., and M. McCartney. 1991. Use of photosynthetic pigments in sediments as a tracer for sources and fates of macrophyte organic matter. Marine Ecology Progress Series 78: 87-96.
Lopez, G. L., and J. S. Levinton. 1987. Ecology of deposit-feeding animals in marine sediments. Quarterly Review of Biology 62: 235-260.
Lopez, G. R., G. L. Taghon, and J. S. Levinton. 1989. Ecology of marine deposit feeders. Springer-Verlag, New York, New York, USA.
Marsh, A. G., and K. L. Tenore. 1990. The role of nutrition in regulating the population dynamics of opportunistic, surface deposit feeders in a mesohaline community. Limnology and Oceanography 35: 710-724.
Martin, A., A. Mariotti, J. Balesdent, and P. Lavelle. 1992. Soil organic matter assimilation by a geophagous tropical earthworm based on [[Delta].sup.13]C measurements. Ecology 73: 118-128.
McEdward, L. R. 1995. Ecology of marine invertebrate larvae. CRC Press, Boca Raton, Florida, USA.
Miller, D. C., and R. W. Sternberg. 1988. Field measurements of the fluid and sediment-dynamic environment of a benthic deposit feeder. Journal of Marine Research 46: 771-796.
Minagawa, M., and E. Wada. 1984. Stepwise enrichment of 15N along food chains: further evidence and the relation between [[Delta].sup.15]N and animal age. Geochimica et Cosmochimica Acta 48: 1135-1140.
Moir, R. J. 1994. The 'carnivorous' herbivores. Pages 87-102 in D. J. Chivers and P. Langer, editors. The digestive system in mammals: food, form and function. Cambridge University Press, Cambridge, UK.
Olson, M. H. 1996. Ontogenetic niche shifts in largemouth bass: variability and consequences for first-year growth. Ecology 77: 179-190.
Pamatmat, M. M. 1968. Ecology and metabolism of a benthic community on an intertidal sandflat. Internationale Revue der Gesamten Hydrobiologie 53: 211-298.
Penry, D. L., and P. A. Jumars. 1990. Gut architecture, digestive constraints and feeding ecology of deposit-feeding and carnivorous polychaetes. Oecologia 82: 1-11.
Persson, L., and P. Eklov. 1995. Prey refuges affecting interactions between piscivorous perch and juvenile perch and roach. Ecology 76: 70-81.
Peterson, B. J., and B. Fry. 1987. Stable isotopes in ecosystem studies. Annual Review of Ecology and Systematics 18: 293-320.
Polis, G. A. 1984. Age structure component of niche width and intraspecific resource partitioning: can age groups function as ecological species? American Naturalist 123: 541-564.
Rasmussen, E. 1953. Asexual reproduction in Pygospio elegans Clap. (Polychaeta, Sedentaria). Nature 171: 1161-1162.
Rau, G. H., T. L. Hopkins, and J. J. Torres. 1991. 15N/14N and 13C/12C in Weddell Sea invertebrates: implications for feeding diversity. Marine Ecology Progress Series 77: 1-6.
Rau, G. H., R. E. Sweeney, I. R. Kaplan, A. J. Mearns, and D. R. Young. 1981. Differences in animal 13C, 15N, and D abundances between a polluted and an unpolluted coastal site: likely indicators of sewage uptake by a marine food web. Estuarine, Coastal and Shelf Science 13: 701.
Ruckelshaus, M. H., R. C. Wissmar, and C. A. Simenstad. 1993. The importance of autotroph distribution to mussel growth in a well-mixed, temperate estuary. Estuaries 16: 898-912.
Schwinghamer, P., F. C. Tan, and D. C. Gordon, Jr. 1983. Stable carbon isotope studies on the Pecks Cove mudflat ecosystem in the Cumberland Basin, Bay of Fundy. Canadian Journal of Fisheries and Aquatic Science 40 (Supplement 1): 262-272.
Shimeta, J. 1996. Particle-size selection by Pseudopolydora paucibranchiata (Polychaeta: Spionidae) in suspension feeding and in deposit feeding: influences of ontogeny and flow speed. Marine Biology 126: 479-488.
Sibly, R. M. 1981. Strategies of digestion and defecation. Pages 109-139 in C. R. Townsend and P. J. Calow, editors. Physiological ecology. Sinauer, Sunderland, Massachusetts, USA.
Snelgrove, P. V. R., J. F. Grassle, and R. F. Petrecca. 1992. The role of food patches in maintaining high deep-sea diversity: field experiments with hydrodynamically unbiased colonization trays. Limnology and Oceanography 37: 1543-1550.
Spies, R. B., J. E. Bauer, and D. Hardin. 1989. Stable isotope study of sedimentary carbon utilization by Capitella spp.: effects of two carbon sources and geochemical conditions during their diagenesis. Marine Biology 101: 69-74.
Strathmann, R. R. 1985. Feeding and non-feeding larval development and life-history evolution in marine invertebrates. Annual Review of Ecology and Systematics 16: 339-361.
SYSTAT. 1992. SYSTAT: Statistics. Version 5.2 edition. SYSTAT, Evanston, Illinois, USA.
Taghon, G. L., A. R. M. Nowell, and P. A. Jumars. 1980. Induction of suspension feeding in spionid polychaetes by high particulate fluxes. Science 210: 562-564.
Taghon, G. L., R. F. L. Self, and P. A. Jumars. 1978. Predicting particle selection by deposit feeders: a model and its implications. Limnology and Oceanography 23: 752-759.
Thorson, G. 1950. Reproductive and larval ecology of marine bottom invertebrates. Biological Reviews 25: 1-45.
Tieszen, L. L., T. W. Boutton, K. G. Tesdahl, and N. A. Slade. 1983. Fractionation and turnover of stable carbon isotopes in animal tissues: implications for [[Delta].sup.13]C analysis of diet. Oecologia 57: 32-37.
Tieszen, L. L., D. Hein, S. Ovortrup, J. Troughton, and S. Imbamba. 1979. Use of [[Delta].sup.13]C values to determine vegetation selectivity in East African herbivores. Oecologia (Berlin) 37: 351-359.
Trexler, J. C., and J. Travis. 1993. Nontraditional regression analyses. Ecology 74: 1629-1637.
Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic niche and species interactions in size-structured populations. Annual Review of Ecology and Systematics 15: 393-425.
Wilson, W. H., Jr. 1983. The role of density dependence in a marine infaunal community. Ecology 64: 295-306.
-----. 1984. Non-overlapping distributions of spionid polychaetes: the relative importance of habitat and competition. Journal of Experimental Marine Biology and Ecology 75: 119-127.
Vezina, A. F. 1985. Empirical relationships between predator and prey size among terrestrial vertebrate predators. Oecologia (Berlin) 67: 555-565.
Yonge, C. M. 1928. Feeding mechanisms in the invertebrates. Biological Reviews 3: 21-76.
Yoshioka, T., E. Wada, and H. Hayashi. 1994. A stable isotope study on seasonal food web dynamics in a eutrophic lake. Ecology 75: 835-846.
Zimba, P. V., M. J. Sullivan, and H. E. Glover. 1990. Carbon fixation in cultured marine benthic diatoms. Journal of Phycology 26: 306-311.
Zottoli, R. A. 1974. Reproduction and larval development of the ampharetid polychaete Amphicteis floridus. Transactions of the American Microscopical Society 93: 78-89.
|Printer friendly Cite/link Email Feedback|
|Author:||Hentschel, Brian T.|
|Date:||Jun 1, 1998|
|Previous Article:||Plankton abundance and dynamics across nutrient levels: tests of hypotheses.|
|Next Article:||Asexual diapause induced by food limitation in the rotifer Synchaeta pectinata.|