Ecosystems (2010) 13: 328–337 DOI: 10.1007/s10021-010-9321-x
2010 Springer Science+Business Media, LLC

Watershed Effects on Chemical Properties of Sediment and Primary Consumption in Estuarine Tidal Flats: Importance of Watershed Size and Food Selectivity by Macrobenthos Takashi Sakamaki,1,2* Jennifer Y. T. Shum,1 and John S. Richardson1 1

Department of Forest Sciences, University of British Columbia, Vancouver, British Columbia V6T 1Z4, Canada; 2Transdisciplinary Research Organization for Subtropics and Island Studies, University of the Ryukyus, Okinawa 903-0213, Japan

ABSTRACT Particulate organic matter transported from rivers to estuaries (POMR) varies quantitatively and qualitatively across estuaries; however, a lack of comparative studies poses a challenge in general understanding of responses of estuarine food webs to POMR input. We studied 20 estuarine tidal flats of the Pacific Northwest coast of North America, with watershed areas ranging from 7 to 8000 km2. We used carbon-stable isotope (d13C) to test the hypothesis that the nutritional contribution of POMR to macrobenthos is proportional to relative abundances of POMR in tidal flat sediments. The predominant origin of total POM (TPOM) in tidal flat sediments generally shifted from marine-origin POM (POMM) to POMR as watershed area increased; however, terrestrial-origin POMR with high C/N predominated sediment TPOM even in estuaries with small watershed areas. Some

macrobenthos species assimilated POM sources in proportion to sediment TPOM composition, and incorporated POMR in POMR-predominant sediments. These species were considered to have low food selectivity; however, the relative nutritional contribution of POMR to these macrobenthos was still lower than the fraction of POMR in sediment TPOM. Other species disproportionately utilized POMM and/or benthic microalgae regardless of the relative abundance of POMR, indicating their high food selectivity. The species-specific, low- or highfood selectivity was likely linked with depositfeeding and filter-feeding, respectively. Hence, our hypothesis was supported conditionally. Our findings indicate that watershed area, relative abundance of POMR in an estuary, and food selectivity of estuarine species are key factors controlling the tightness of linkage between watersheds and estuarine food webs.

Received 20 July 2009; accepted 1 February 2010; published online 2 March 2010

Key words: intertidal sediments; river discharge; basal resources; subsidy; autochthonous and allochthonous organic matter; primary consumers; benthic invertebrates; feeding modes; land–sea connectivity.

Author Contributions: TS conceived of the study, performed field research, conducted data analyses and wrote the manuscript. JYTS performed field research, helped with sample processing, and contributed to the writing of manuscript. JSR contributed to study design, and assisted with data analyses and writing of the manuscript. *Corresponding author; e-mail: [email protected]

328

Watershed Effects on Estuarine Tidal Flats

INTRODUCTION Particulate organic matter (POM) is an important component of materials transported from rivers to estuaries (Howarth and others 1991; Gomi and others 2002; Wipfli and others 2007; Cole and others 2007). The degree of contribution of POM transported from river (POMR) to POM pools in estuarine water and sediment depends on watershed and estuarine characteristics, and varies across estuaries (Abril and others 2002; Middelburg and Herman 2007; Sakamaki and Richardson 2008a, b). Furthermore, human-related disturbances to watersheds and hydrological changes due to climate change have a potential to dramatically alter the quantity and quality of POMR transported from watershed to estuary, among other materials (Howarth and others 1991; Correll and others 1992; Thrush and others 2004; Darnaude 2005, 2005). To date, however, the response of estuarine food webs to spatial or temporal changes in POMR input is not predictable. Previous studies have derived different conclusions on the contribution of POMR to food webs between estuaries. Some studies have demonstrated the importance of POMR for estuarine food webs (Darnaude 2005; Kasai and Nakata 2005; Schlacher and Connolly 2009). It was also shown that stable isotopic signatures of some estuarine macrobenthos species varied along an upstream– downstream gradient within a single estuary (Kasai and Nakata 2005; Doi and others 2005). This suggests that the nutritional incorporation of POMR by those macrobenthos was proportional to the relative abundance of POMR in the POM pool of estuaries. On the contrary, many other studies have demonstrated a greater importance of other sources, such as marine-origin POM (POMM) and benthic microalgae, rather than POMR (for example, Deegan and Garritt 1997; Kang and others 2003; Yokoyama and others 2005; Yokoyama and Ishihi 2007). Some of these studies concluded that diets of estuarine organisms were not associated with the relative abundance of POMR (Sobczak and others 2002; Yokoyama and Ishihi 2003; Martineau and others 2004). Overall, due to the inconsistent conclusions from previous studies, it is still unclear whether estuarine food webs really incorporate POMR as a basal resource. Furthermore, it is unknown how sensitively the incorporation of POMR into estuarine food webs is related to the relative abundance of POMR in estuaries. This questions how selective estuarine primary consumers are in their assimilation of trophic resources. One can expect that if primary consumers are non-selective,

329

their diet will consist of different POM sources in proportion with the composition of ambient POM pools; in this case, nutritional incorporation of POMR is proportional to the relative abundance of POMR. On the other hand, if primary consumers are highly selective, then their nutritional incorporation of POMR is predicted to be unrelated to the relative abundance of POMR, whichever POMR or another POM source is selectively utilized. This uncertainty in the basal resource selectivity of estuarine food webs and the dietary incorporation of POMR stems from the lack of comparative studies across a broad gradient of the relative abundance of POMR in estuaries. POMR abundance in an estuary is expected to be highly dependent on physical processes, such as efflux of POMR and its retention in the estuary. A geological study, which considered watershed areas of from 100 to 6.1 9 106 km2 has shown that the sediment load from rivers to the ocean increases in proportion to watershed area (Milliman and Syvitski 1992). Based on this finding, it can be expected that the degree of POMR contribution to estuarine POM pool is also positively related to the watershed area. However, the prediction of the relative abundance of POMR in estuarine POM pools could be more complex, because the dispersion of POMR and its dilution by marine-origin POM (POMM) occur in estuaries. Specifically, a relationship between the relative abundance of POMR in estuaries and watershed area has not been quantified. We studied 20 estuarine tidal flats of the Pacific Northwest coast of North America, and tested our hypothesis that across these estuaries the nutritional contribution of POMR to tidal flat macrobenthos is proportional to the relative abundance of POMR in sediment. We studied tidal flats for the following three reasons: estuarine tidal flats (1) are important and highly productive ecosystems that contribute to the high biodiversity of coastal areas, (2) are depositional, long-term repositories of mixtures of materials transported from rivers and marine systems, and (3) enable a standardized design of sampling location across estuaries, due to their ease of access and relatively spatially homogenized physical and chemical conditions. The watershed areas connecting to the studied estuaries ranged from 7 to 8000 km2, providing a broad gradient in quantity and quality of POMR transported into the estuaries. In this study, we also quantified the relationship between watershed area and the relative abundance of POMR in tidal flat sediments. To distinguish POM of different origins, we applied biogeochemical indicators: carbon-stable isotope

330

T. Sakamaki and others

(d13C) distinguishes marine or watershed origins, and the ratio of carbon to nitrogen (C/N) helps to discern aquatically or terrestrially produced POM.

METHODS Study Sites We studied 20 estuarine tidal flats encompassing a wide range of watershed areas (Figure 1). In particular, we collected all our samples from a total of 17 estuaries in July and August 2006, 12 of which flow into the Strait of Georgia (10 from the southeast coast of Vancouver Island and 2 from the Sunshine Coast); the other 5 estuaries are located in Puget Sound, northwestern Washington State (4793¢–4971¢N, 12220¢–12499¢W). We also included data obtained from three other estuaries in the Vancouver Lower Mainland from summer 2004 (Sakamaki and Richardson 2008b) into data analyses of this study.

Field Sampling Because comparing multiple estuaries is the cornerstone of this study, the sampling locations of all estuaries were standardized based on the following conditions. In cases where there are multiple channels within one estuarine tidal flat, we collected samples along the largest main channel. In each of the estuarine tidal flats, three sampling stations (upper, middle, and lower intertidal) were established along the main channel to span the

Courtenay R.(861) Anderson C.(39) Trent R.(82) Wilfred C.(28) Rosewall C.(44) Roberts C.(29) Nile C.(20) French C.(69) Englishman R.(323) Noons C.(7) Little Qualicum R.(248) unnamed irrigation c.(27) Nanaimo R.(826)

Vancouver Island Cowichan R.(1235)

Nicomekl R.(179) California C.(74)

49 N

Nooksack R.(3237) Samish R.(228)

Skagit R.(8011)

50 km 125 W

Stillaguamish R.(1443)

124 W

123 W

Figure 1. Studied estuaries in the Pacific Northwest coast region of North America. Names of rivers or creeks entering the studied estuaries are shown in the map. The numbers in the parentheses indicate watershed areas (km2).

length of the river in the intertidal zone. All sampling stations were set 5–15 m away from the river channel. At each of the three stations, we marked three points 10–15-m apart along the river channel to serve as replicate sampling points. For most of the estuarine tidal flats, the sampling stations were located within a wide, open intertidal zone bordered by long, clear tide marks. However, in the intertidal flat of two estuaries, a narrow strip of intertidal zone stretches far above the river mouth due to the low river gradient and small river discharge. In these two tidal flats, the upper station was located far above the river mouth, but the sediment total POM (TPOM) was dominated by POMM with high d13C, which proved that the influence of POMR was not overestimated. To standardize a consistent representation of sizes of all estuaries, we conducted all our sampling in the tidal flats at approximately the lowest of spring tides. In any given estuary, the distance between two consecutive sampling stations was approximately the same, but this distance varied (approximately 0.1–1 km) between estuaries depending on the size (for example, length, gradient) of the tidal flat. All the sampling points were set in non-vegetated areas with gravel, sand, or mud sediment. In most of the sampling stations, no vegetation was found at all, although saltmarsh plant vegetation was found near the upper tidal stations in some estuaries (approximately farther than 30 m). Coarse detritus of saltmarsh plants, seagrass and macroalgae were hardly observed around the sampling stations. To determine the chemical signatures (that is, carbon-stable isotopes, C/N) of sediment TPOM, one surface sediment core was collected using a plastic tube with a diameter of 5 cm at each replicate sampling point (a total of 9 in each estuary). The 0–1-cm layer of the sediment cores was then sliced and stored in a plastic bag. Macrobenthos were collected at each replicate sampling point by excavating a 15 9 15 cm2 patch of sediment to a depth of 25 cm and screening the sediment through a 1-mm sieve in water. The collected macrobenthos were identified and sorted into six species commonly found in the estuaries studied, which are an amphipod (Corophium spinicorne), a polychaete (Nereis limnicola), and four bivalves (Macoma balthica, Mya arenaria, Venerupis philippinarum, and Nuttallia obscurata). To determine the chemical signatures of POM from different origins (POMR, POMM, and benthic microalgae), water and superficial tidal flat sediment samples were also collected. For each estuary, downstream freshwater of the river was collected once at base flow conditions under fair weather

Watershed Effects on Estuarine Tidal Flats after several consecutive days of fair weather. A total of six samples of offshore seawater (<50 cm in depth) were collected by boats, kayaks, or by walking out on a pier (five from Straight of Georgia, one from Puget Sound). For benthic microalgae, the surface sediment (approximately 3 mm) was scraped by a scoop at every replicate sampling point, and combined to form a representative sample for each station (three samples in total in each estuary). The water and sediment samples were later processed to obtain concentrated POM and benthic microalgae. In addition, saltmarsh plants growing above intertidal flats, and macroalgae drifted onto the intertidal flat sediments were sampled in some estuaries. In the rivers without government monitoring for river discharge, we measured the discharge at the river water sampling point on the sampling day using a cross-sectional measurement of depth and current velocity by a current velocity meter (Swoffer, Model 2100). For the other monitored rivers, we obtained the discharge data for the river water sampling day or closest day from government websites (Environment Canada, Washington State Department of Ecology, USGS).

Sample Processing and Analyses The macrobenthos except for Corophium spinicorne were dissected to obtain a specific body part for carbon-stable isotope analyses. For bivalves, the foot was cut and analyzed; for polychaetes, we used the cuticle, circular, and longitudinal muscles near the head of each individual and avoided the intestinal lumen. The water samples for POMR and POMM were screened through a 1-mm sieve to remove coarse particles and filtered through pre-combusted Whatman GF/F glass-fiber filters. Processing benthic microalgae samples followed a slightly modified version of Riera and Richard’s method (1996): the tidal flat sediment scraped for benthic microalgae analysis was spread to a thickness of 1–2 cm on a plastic plate, left for 1 h under light, covered with pre-combusted (2 h at 550C) sand (>250 um) of approximately 5-mm thickness, and left again under light for 2 h. Throughout these processes, the sediment surface was kept moist by sometimes spraying distilled water. Then the topmost layer of sand (about 3 mm) was scraped, dispersed in filtered seawater, screened through a 63-lm sieve, and filtered through pre-combusted Whatman GF/F filters. All samples for carbon-stable isotope and C/N (that is, sediment, macrobenthos tissue, POMR,

331

POMM, benthic microalgae, macroalgae, and saltmarsh plant) were acidified with 10% HCl, rinsed with deionized water, and dried at 60C. The carbon-stable isotopic compositions and C/N were determined by a continuous flow, isotopic ratio, mass spectrometer system (Europa, Hydra 20/20) at the University of California Davis Stable Isotope Facility. For most macrobenthos species, each individual was analyzed separately with the exception of Corophium spinicorne in which we pooled 3–5 individuals for one analysis in order for the mass to fall within range of detection of the analysis. Isotopic compositions were reported as parts per thousand deviations (&) from Pee Dee Belemnite for d13C. Carbon and nitrogen contents of the samples were also reported at the same time.

Data Analyses Our discrimination of organic matter sources had possible confusions between POMR and C3 saltmarsh plants (–25.7 ± 1.6&, n = 4) and between benthic microalgae and macroalgae (–15.9 ± 1.5&, n = 9) due to the overlap of d13C signatures. Even if we confound those sources to some extent, it does not significantly affect our interpretation of the study results due to their similarities in spatial origins; saltmarsh plant can be considered a part of POMR transported from a watershed, and macroalgae can be considered in the same category with benthic microalgae. However, we assumed that in our study, POMM, POMR, and benthic microalgae had significant effects on sediment TPOM and/or macrobenthos, whereas C3 saltmarsh plant or macroalgae had negligible contributions. This was based on the following reasons: (1) around most of the sampling stations we did not find vegetation or their coarse detritus and (2) in many bare tidal flats, substantial contributions have been reported for POMM, POMR, and/or benthic microalgae (for example, Yokoyama and others 2005; Sakamaki and Richardson 2008b). For d13C and C/N of tidal flat sediment TPOM, variations in those variables were tested by twoway ANOVAs with estuaries and stations as sources of variation. Because the effect of estuaries predominated the variations of the variables, the variables obtained from the three stations of each tidal flat were averaged and used for further statistical analyses. Multiple linear regression analyses were used to test the relationship of tidal flat sediment TPOM d13C with sediment TPOM C/N and watershed area. Linear regression analyses were used to test the relationships for river discharge versus watershed area (dependent vs. independent),

332

T. Sakamaki and others

tidal flat sediment TPOM C/N versus watershed area, sediment organic carbon content versus watershed area, sediment organic carbon content versus sediment TPOM d13C, benthic microalgal d13C versus sediment TPOM d13C, macrobenthos d13C (each species) versus benthic microalgal d13C, and macrobenthos d13C (each species) versus sediment TPOM d13C. To evaluate intra-specific variability of dietary intake, standard deviations (SDs) of macrobenthos d13C were calculated for each of the macrobenthos species within each station and within each estuary. The difference of those SDs among the species was tested by a Kruskal–Wallis test. T tests and paired t tests were used to test whether d13C of each macrobenthos species significantly differed from d13C of POMM and benthic microalgae, respectively. We used SAS version 9.1 (SAS Inc, Cary, NC) for all the statistical analyses in this study.

RESULTS For d13C and C/N of sediment TPOM, the variation between the estuaries (explained variance = 74–86%, F19,38 = 6.5–14.9, P < 0.0001) far exceeded the variation between the three stations within each estuary (explained variance = ca. 2.7–3.3%, F2,38 = 2.7-4.4, P = 0.02–0.08). Sediment TPOM with relatively high d13C had low C/N, whereas that with relatively low d13C showed a wide range of C/N (Figure 2). The means of d13C for POM sources potentially involved in the POM pool of the studied estuarine tidal flats were as follows: POMR; -25.9 ± 2.0& (mean ± SD, n = 20), POMM; -19.4 ±

Figure 2. Relationship between chemical properties of tidal flat sediment TPOM (d13C and C/N) and watershed area. Each data point represents one estuary.

2.2& (n = 6), benthic microalgae; -17.9 ± 2.3& (n = 20). Sediment TPOM d13C (&, Y) was negatively related to both sediment TPOM C/N (X1) and watershed area (km2, X2) (R2 = 0.72, Y = -0.42X1 - 1.03 Log10X2 - 15.2, X1: P < 0.0001, X2: P = 0.0009). Sediment TPOM C/N was not related to watershed area (P = 0.78). The discharge (m3 s-1, Y) was positively related to watershed area (km2, X) (R2 = 0.92, Y = 0.025X - 6.32, P < 0.0001). The organic carbon content of tidal flat sediments was not significantly related to either watershed area (P = 0.10) or sediment TPOM d13C (P = 0.34). The relationship between macrobenthos d13C and sediment TPOM d13C was significant and positive for three of six species analyzed, that is, Nereis limnicola, Macoma balthica, and Corophium spinicorne (Figure 3; Table 1). The variance of macrobenthos d13C within each station and within each estuary significantly differed between the macrobenthos species (Kruskal–Wallis test: v2 = 11.4, DF = 5, and P = 0.044 for within each station; v2 = 29.5, DF = 5, and P < 0.0001 for within each estuary) (Table 1). Nereis limnicola, Macoma balthica, and Corophium spinicorne showed relatively larger variances within each station as well as within each estuary, compared with the other three species. For these three species with larger variance in d13C, d13C of most individuals were higher than that of sediment TPOM (after 1& was subtracted from macrobenthos d13C taking account of trophic fractionation; Deniro and Epstein 1978). Some individuals of these species had much higher d13C than d13C of any potential sources considered (that is, POMM, POMR, benthic microalgae); this was most evident in Macoma balthica as nearly all individuals had d13C signatures higher than that of potential sources. Some other individuals of these three species had d13C relatively closer to d13C of POMR particularly in the relatively lower d13C of sediment TPOM. For the other three species analyzed, Venerupis philippinarum, Mya arenaria, and Nuttallia obscurata, their d13C showed no significant relationship with sediment TPOM d13C. We found exceptionally low d13C for Mya arenaria and Nuttallia obscurata in a single estuary (the estuary of unnamed irrigation creek (Figure 1), indicated by arrows in Figure 3). However, even after excluding the data from this estuary, the relationships between their d13C and sediment TPOM d13C were still not significant. For these three species, d13C of most individuals clustered tightly near d13C of POMM and benthic microalgae regardless of sediment TPOM d13C. d13C of these three species as well as Corophium spinicorne

Watershed Effects on Estuarine Tidal Flats Sediment TPOM -30

-25

-20

-15

A

-10 -30 -10

-25

-20

13

C (‰)

-15

-10

-30 -10

B

1:1

333

-25

-20

-15

-10 -10

C

-15

-15

-15

POMM

-20

-20

-20

POMR

-25

-25

-25

-30

-30

-30

Benthic algae

-25

-20

-15

-10 -30 -10

-20

-15

-30 -10 -10

-20

-15

-10 -10

F

-15

-15

-15

-20

-20

-20

-25

-25

-25

-30

-30

-30

C (‰)

E

-25

13

D

-25

Macrobenthos

-30

Fig. 3. Relationship between macrobenthos d13C and sediment TPOM d13C for six macrobenthos species. Each white dot represents one individual macrobenthos sample. Each black dot represents a mean in one estuary. The dotted lines indicate the d13C values for benthic algae, POMM, and POMR; they serve to compare with macrobenthos d13C as well as sediment TPOM d13C. Plotted macrobenthos diet d13C have been corrected for trophic fractionation (1 &). A point would lie on the 1:1 line if d13C of macrobenthos diet was equivalent to sediment TPOM d13C. The regression line is for the black dots (estuary representatives), and the detailed results of the regression analyses are shown in Table 1. In the regression analysis for Mya arenaria and Nuttallia obscurata, both cases, where a point from one estuary indicated by the arrow was included and excluded, were considered. a Nereis limnicola, b Macoma balthica, c Corophium spinicorne, d Venerupis philippinarum, e Mya arenaria, and f Nuttallia obscurata.

Table 1.

Regression Analysis for Macrobenthos d13C (Y) Versus Sediment d13C (X)

Species

Nereis limnicola Macoma balthica Corophium spinicorne Venerupis philippinarum Mya arenaria Nuttallia obscurata

Median SD of d13C

Regression n

R2

P

a

(95% CI)

b

Station

Estuary

16 16 17 7 9 8 16 15

0.39 0.25 0.29 0.52 0.05 0.16 0.001 0.18

0.010 0.047 0.027 0.069 0.567 0.329 0.920 0.111

0.65 0.63 0.45 0.30 -0.18 0.26 -0.02 0.19

(0.18–1.12) (0.01–1.24) (0.06–0.84) (-0.03–0.64) (-0.90–0.54) (-0.34–0.86) (-0.48–0.44) (-0.05–0.43)

-3.14 -1.26 -11.42 -12.16 -23.56 -13.06 -19.18 -14.26

0.60 0.47 0.39 0.32 0.12

0.85 0.64 0.87 0.32 0.35

0.26

0.32

For both macrobenthos d13C and sediment d13C, the means of each estuary were used for the analyses. The medians of standard deviation of macrobenthos d13C within a single station and within a single estuary are also shown. The results for regression include the number of estuaries where the species was analyzed (n), coefficient of determination (R2), probability for a = 0 (P), parameters estimated for the model, Y = aX + b, and 95% confidence interval for a (in round parenthesis). For the results of Mya arenaria and Nuttallia obscurata in italics, the estuary of the unnamed irrigation creek, which had macrobenthos d13C deviating from other estuaries (indicated by the arrow in Figure 3), was excluded from the linear regression analyses.

did not significantly differ from d13C of POMM (t test: P > 0.28) and benthic microalgae (paired t test: P > 0.077), whereas d13C of Nereis limnicola and Macoma balthica significantly deviated from

d13C of POMM (P < 0.004) and benthic microalgae (P < 0.030). Across estuaries there was no significant relationship between macrobenthos d13C and benthic

334

T. Sakamaki and others

microalgae d13C (P > 0.15 for all the six species), or between benthic microalgae d13C and sediment TPOM d13C (P = 0.72).

DISCUSSION Relative Abundance of POMR in Tidal Flat Sediments Generally, benthic microalgae are a minor fraction of sediment TPOM composition (Cook and others 2004; Sakamaki and Nishimura 2006). Thus, based on comparison of sediment TPOM d13C for POMM and POMR, relatively higher sediment TPOM d13C suggests POMM predominance in the sediment, whereas lower sediment TPOM d13C suggests POMR predominance. As predicted, the predominant origin of sediment TPOM generally shifted from POMM to POMR across the estuarine tidal flats with increasing watershed area. Larger river discharge is expected to increase POMR flux into the estuary, and enhance fluvial force to distribute POMR to a wider area around the estuary. Our results provide a quantitative context for predicting the POMR predominance in tidal flat sediment TPOM based on watershed area. However, we need to note that in our study, the POMR contribution to the estuarine tidal flat POM pool is possibly higher than other parts of the estuary and/or other regions due to the following two reasons in regards to our study design. First, our sampling stations are limited to areas of the intertidal flat near the river channel where they may be affected by POMR more significantly than other parts of an estuary. Second, many of the watersheds we studied contained relatively steep mountainous areas, so such a geological feature of the studied region may enhance POMR inputs to the estuaries (Milliman and Syvitski 1992). POMR has diverse origins, for example, anthropogenic, terrestrial plant, freshwater algae, and its composition depends in part on watershed characteristics (for example, discharge, water chemistry) (Hopkinson and others 1998; Onstad and others 2000), whereas POMM is in general represented by phytoplankton. This explains our finding that the POMR-predominant estuarine tidal flats had a broader range of C/N, compared with POMMpredominant estuaries. Importantly, our results indicate that when sediment TPOM C/N is high, POMR can predominate sediment TPOM even with a small watershed area. The relatively higher C/N is probably due to high contribution of terrestrialorigin organic matter. In general, terrestrial-origin organic matter with high C/N tends to be less

degradable than aquatically produced organic matter with low C/N (for example, algae) (Enrı´quez and others 1993). Therefore, if such low-degradable terrestrial POM enters tidal flat sediments, and if the hydrodynamic conditions are sufficiently calm to retain it within the sediment, it can remain for a long time. Even sporadic inputs of terrestrial POM due to human-caused disturbance (for example, forestry, agricultural activities) and/or natural events (for example, landslide, erosion) in watersheds could cause substantial and durable effects on the chemical properties of tidal flat sediments. Such characteristics of terrestrial POM input and retention may be responsible for the difficulty in predicting its predominance in tidal flat sediments.

Nutritional Contribution of POMR to Macrobenthos The significant relationships between sediment TPOM d13C and macrobenthos d13C, which were shown for Nereis limnicola, Macoma balthica, and Corophium spinicorne, indicate that they utilized sediment TPOM less selectively and that their diets were linked tightly with the POM source compositions of sediment TPOM. It should be noted that the lack of significant relationship of benthic microalgae d13C with either sediment TPOM d13C or macrobenthos d13C indicates that benthic microalgae d13C was independent of sediment TPOM d13C, and benthic-microalgae-feeding by macrobenthos was not responsible for the significant relationships between sediment TPOM d13C and macrobenthos d13C. For these three macrobenthos species with low-food selectivity, the individuals with lower d13C than d13C of POMM and benthic microalgae, which were found in relatively lower sediment TPOM d13C, likely assimilated POMR as a food source. These findings basically support our hypothesis that across multiple estuaries, the nutritional contribution of POMR to macrobenthos is proportional to the relative abundance of POMR in sediment. Although the POMR used by the above three macrobenthos species was proportional to the POMR abundance, the slope in the regression between sediment TPOM d13C and macrobenthos d13C was between 0 and 1. This indicates that the relative nutritional contribution of POMR to the macrobenthos was lower than the fraction of POMR in sediment TPOM. Therefore, the three macrobenthos species with less food selectivity (that is, Nereis limnicola, Macoma balthica, and Corophium spinicorne) still had some degree of bias toward particular sources with relatively higher d13C than

Watershed Effects on Estuarine Tidal Flats POMR (for example, POMM, benthic microalgae) through their foraging and/or assimilation processes. POMR was one of the substantial food sources for these species in the POMR-abundant estuaries, but its contribution was not overwhelmingly predominant. In contrast to the abovementioned three species, Venerupis philippinarum, Mya arenaria, and Nuttallia obscurata likely utilized benthic microalgae and/or POMM selectively, being less affected by the predominant origin of sediment TPOM. Venerupis philippinarum, Mya arenaria, and a subspecies of Nuttallia obscurata are known to practice filterfeeding (Nakamura 2001; Sakamaki and others 2002; Forster and Zettler 2004). Thus, filter-feeding may be responsible for their tight nutritional linkage to POMM and/or benthic microalgae both of which can be major components of algal POM in estuarine water (de Jonge and van Beusekom 1992). Although POMR was not a substantial food source for these species in most of the estuaries, we found one single estuary with exceptionally low d13C for Mya arenaria and Nuttallia obscurata. This estuary has a small watershed and no significant contribution of POMR to the sediment POM pool, but receives the input of algal-predominant POMR from the river with by far the highest algal concentration when compared with rivers of the other studied estuaries (Sakamaki and Richardson 2008b). This idiosyncratic case suggests that these species utilize algal-origin POM regardless of its spatial origin, and also can incorporate POMR substantially through filter-feeding even despite the lack of POMR predominance in the sediment. For Corophium spinicorne, Macoma balthica, and Nereis limnicola, it is plausible that their deposit feeding is responsible for their lower selectivity in food source incorporation as well as their proportional use of POMR to the relative abundance of POMR in sediment. Meanwhile, for these three species, the dietary sources and pathways are probably diversified within each species, as their d13C showed wide intra-specific variability. Many macrobenthos species, including Corophium spinicorne and Macoma balthica, have been reported to switch their feeding behavior between filter- and deposit-feeding, depending on environmental conditions such as food availability, water current, and predation risk (Riisga˚rd and Kamermans 2001). Thus, the intra-specific diversity in food source utilization may be explained partially by the difference in switching feeding behavior between individuals and small spatial scale differences in food sources.

335

The dietary pathways for the less selective species likely includes not only direct feeding on POMM, POMR, and/or benthic microalgae but also some other indirect feeding on such basal resources, because some individuals showed higher d13C than these potential sources taken into account. There are some possible pathways which could cause such high macrobenthos d13C. These macrobenthos may have practiced opportunistically omnivorous feeding, which can include consumption of meiobenthos and small macrobenthos, particularly due to their unselective deposit feeding (Tita and others 2000). This potentially increases their d13C from that of the original basal organic matter sources by trophic fractionation (Deniro and Epstein 1978). Bacterial production can also cause isotopic fractionation; in some cases, bacterial d13C increases (Macko and Estep 1984). Therefore, it is possible that the macrobenthos fed on bacteria growing on organic/inorganic particles of sediment (Findlay and Tenore 1982; Crosby and others 1990; van Oevelen and others 2006; Alfaro and others 2006) and caused the higher macrobenthos d13C than sediment TPOM d13C. However, direction and degree of isotopic fractionation through bacterial production greatly depend on the biochemical component of the substrate (Hullar and others 1996; Bouillon and Boschker 2006), so the effects of bacteria-feeding on macrobenthos d13C remain unclear. Our results suggest that macrobenthos species showing nutritional dependence on POMR have characteristics that make it difficult to, through the use of stable isotopes, correctly evaluate the POMR contribution to those macrobenthos: that is, indirect incorporation of POM with isotopic fractionation can occur and there can be high intra-specific variability in dietary sources and pathways. These are likely distinct for deposit-feeding macrobenthos. Although indirect pathways (for example, bacterivory, omnivory) and the degree of its effect on macrobenthos d13C have not been specified, they could potentially mislead the estimation of POMR contribution to macrobenthos diets. Caraco and others (1998) have also pointed out a similar issue due to potential bacterial mediation in evaluation of POMR importance for estuarine food webs using stable isotopic signatures. Meanwhile, the high intra-specific variability in dietary intake potentially confounds representativeness of the estimation of POMR contribution to a targeted species. Hence, in future studies as well as reviews of past studies, POMR contribution to estuarine food webs needs to be carefully evaluated taking

336

T. Sakamaki and others

into account these factors that potentially affect the evaluation.

CONCLUSIONS Watershed area was a significant predictor for the strength of the POMR effect on tidal flat sediments in most of the estuaries, although the predominance of terrestrial-origin POMR with high C/N in the tidal flat sediment was not predictable by watershed area. For some macrobenthos species, the nutritional contribution of POMR was proportional to the relative abundance of POMR in the tidal flat sediments. Such non-selectivity of their diet is considered to be due to their depositfeeding. However, it should be noted that even less selective species still could have some degree of bias toward other sources such as POMM and/or benthic microalgae. In the tidal flat with abundant POMR, these less-selective species substantially utilized POMR, but the nutritional contribution of POMR was not overwhelming. On the contrary, some other species were not sensitive to the relative abundance of POMR in the tidal flat sediments, and disproportionately utilized POMM and benthic microalgae in most estuaries. These highly selective species probably depended greatly on filter-feeding. Overall, these results indicate that POMR plays a nutritionally important role in estuarine food webs when the relative abundance of POMR is high and estuarine organisms are less selective in their food utilization. Our findings highlight the importance of watershed area, relative abundance of POMR in estuaries, and food selectivity of estuarine species as factors controlling the tightness of the linkage between watersheds and estuarine food webs.

ACKNOWLEDGMENTS We thank G. Hood for assistance with field sampling. This study was financially supported by Research Foundation for the Electro-technology of Chubu, Japan, the Natural Sciences and Engineering Research Council of Canada, and the Forest Science Program (British Columbia, Canada).

REFERENCES Abril G, Nogueira M, Etcheber H, Cabec¸adas G, Lemaire E, Brogueira MJ. 2002. Behaviour of organic carbon in nine contrasting European estuaries. Estuarine Coast Shelf Sci 54:241–62. Alfaro AC, Thomas F, Sergent L, Duxbury M. 2006. Identification of trophic interactions within an estuarine food web (northern

New Zealand) using fatty acid biomakers and stable isotopes. Estuarine Coast Shelf Sci 70:271–86. Bouillon S, Boschker HTS. 2006. Bacterial carbon sources in coastal sediments: a cross-system analysis based on stable isotope data of biomarkers. Biogeosciences 3:175–85. Caraco NF, Lampman G, Cole JJ, Limburg KE, Pace ML, Fischer D. 1998. Microbial assimilation of DIN in a nitrogen rich estuary: implications for food quality and isotope studies. Marine Ecol Prog Ser 167:59–71. Cole JJ, Prairie YT, Caraco NF, McDowell WH, Tranvik LJ, Striegl RG, Duarte CM, Kortelainen P, Downing JA, Middleburg JJ, Melack J. 2007. Plumbing the global carbon cycle: integrating inland waters into the terrestrial carbon budget. Ecosystems 10:171–84. Cook PLM, Revill AT, Clementson LA, Volkman JK. 2004. Carbon and nitrogen cycling on intertidal mudflats of a template Australian estuary. III. Sources of organic matter. Marine Ecol Prog Ser 280:55–72. Correll DL, Jordan TE, Weller DE. 1992. Nutrient flux in a landscape: effects of coastal land use and terrestrial community mosaic on nutrient transport to coastal waters. Estuaries 15:431–42. Crosby MP, Newell RIE, Langdon CJ. 1990. Bacterial mediation in the utilization of carbon and nitrogen from detrital complexes by Crassostrea virginica. Limnol Oceanogr 35:625–39. Darnaude AM. 2005. Fish ecology and terrestrial carbon use in coastal areas: implications for marine fish production. J Anim Ecol 74:864–76. Deegan LA, Garritt RH. 1997. Evidence for spatial variability in estuarine food webs. Marine Ecol Prog Ser 147:31–47. de Jonge VN, van Beusekom JEE. 1992. Contribution of resuspended microphytobenthos to total phytoplankton in the Ems estuary and its possible role for grazers. Neth J Sea Res 30:91–105. Deniro MJ, Epstein S. 1978. Influence of diet on the distribution of carbon isotopes in animals. Geochim Cosmochim Acta 42:495–506. Doi H, Matsumasa M, Toya T, Satoh N, Mizota C, Maki Y, Kikuchi E. 2005. Spatial shifts in food sources for macrozoobenthos in an estuarine ecosystem: Carbon and nitrogen stable isotope analyses. Estuarine Coast Shelf Sci 64:316–22. Enrı´quez S, Duarte CM, Sand-Jensen K. 1993. Patterns in decomposition rates among photosynthetic organisms: the importance of detritus C:N:P content. Oecologia 94:457–71. Findlay S, Tenore K. 1982. Nitrogen source for a detritivore: detritus substrate versus associated microbes. Science 218:371–3. Forster S, Zettler ML. 2004. The capacity of the filter-feeding bivalve Mya arenaria L. to affect water transport in sandy beds. Mar Biol 144:1183–9. Gomi T, Sidle RC, Richardson JS. 2002. Understanding processes and downstream linkages of headwater systems. BioScience 52:905–16. Hopkinson CS, Buffam I, Hobbie J, Vallino J, Perdue M, Eversmeyer B, Prahl F, Covert J, Hodson R, Moran MA, Smith E, Baross J, Crump B, Findlay S, Foreman K. 1998. Terrestrial inputs of organic matter to coastal ecosystems: An intercomparison of chemical characteristics and bioavailability. Biogeochemistry 43:211–34. Howarth RW, Fruci JR, Sherman D. 1991. Input of sediment and carbon to an estuarine ecosystem: influence of land use. Ecol Appl 1:27–39. Hullar MAJ, Fry B, Peterson BJ, Wright RT. 1996. Microbial utilization of estuarine dissolved organic carbon: a stable iso-

Watershed Effects on Estuarine Tidal Flats tope trace approach tested by mass balance. Appl Environ Microbiol 62:2489–93. Kang CK, Kim JB, Lee KS, Kim JB, Lee PY, Hong JS. 2003. Trophic importance of benthic microalgae to macrozoobenthos in coastal bay system in Korea: dual stable C and N isotope analysis. Mar Ecol Prog Ser 259:79–92. Kasai A, Nakata A. 2005. Utilization of terrestrial organic matter by the bivalve Corbicula japonica estimated from stable isotope analysis. Fish Sci 71:151–8. Macko SA, Estep MLF. 1984. Microbial alteration of stable nitrogen and carbon isotopic compositions of organic matter. Org Geochem 6:787–90. Martineau C, Vincent WF, Frenette J, Dodson JJ. 2004. Primary consumers and particulate organic matter: Isotopic evidence of strong selectivity in the estuarine transition zone. Limnol Oceanogr 49:1679–86. Middelburg JJ, Herman PMJ. 2007. Organic matter processing in tidal estuaries. Mar Chem 106:127–47. Milliman JD, Syvitski JPM. 1992. Geomorphic/tectonic control of sediment discharge to the ocean: The importance of small mountainous rivers. J Geol 100:525–44. Nakamura Y. 2001. Filtration rates of the Manila clam, Ruditapes philippinarum: dependence on prey items including bacteria and picocyanobacteria. J Exp Mar Biol Ecol 266: 181–92.

337

Sakamaki T, Richardson JS. 2008a. Retention, breakdown and biological utilisation of deciduous tree leaves in an estuarine tidal flat of southwestern British Columbia, Canada. Can J Fish Aquat Sci 65:38–46. Sakamaki T, Richardson JS. 2008b. Effects of small rivers on chemical properties of sediment and diets for primary consumers in estuarine tidal flats: a comparison between forested and agricultural watersheds. Mar Ecol Prog Ser 360:13–24. Schlacher TA, Connolly RM. 2009. Land-ocean coupling of carbon and nitrogen fluxes on sandy beaches. Ecosystems 12:311–21. Sobczak WV, Cloern JE, Jassby AD, Mu¨ller-Solger AB. 2002. Bioavailability of organic matter in a highly disturbed estuary: The role of detrital and algal resources. Proc Natl Acad Sci USA 99:8101–5. Syvitski JPM, Vorosmarty CJ, Kettner AJ, Green P. 2005. Impact of humans on the flux of terrestrial sediment to the global coastal ocean. Science 308:376–80. Tita G, Desrosiers G, Vincx M, Nozais C. 2000. Predation and sediment disturbance effects of the intertidal polychaete Nereis virens (Sars) on associated meiofaunal assemblages. J Exp Mar Biol Ecol 243:261–82. Thrush SF, Hewitt JE, Cummings VJ, Ellis JI, Hatton C, Lohrer A, Norkko A. 2004. Muddy waters: elevating sediment input to coastal and estuarine habitats. Front Ecol Environ 2:299–306.

Onstad GD, Canfield DE, Quay PD, Hedges JI. 2000. Sources of particulate organic matter in rivers from the continental USA: Lignin phenol and stable carbon isotope compositions. Geochim Cosmochim Acta 64:3539–46.

van Oevelen D, Moodley L, Soetaert K, Middleburg JJ. 2006. The trophic significance of bacterial carbon in a marine intertidal sediment: Results of an in situ stable isotope labeling study. Limnol Oceanogr 51:2349–59.

Riera P, Richard P. 1996. Isotopic determination of food source of Crassostea gigas along a trophic gradient in the estuarine bay of Marennes-Oleron. Estuarine Coast Shelf Sci 42:347–60. Riisga˚rd HU, Kamermans P. 2001. Switching between deposit and suspension feeding in coastal zoobenthos. In: Reise K, Ed. Ecological comparisons of sedimentary shores. Berlin: Springer. p 73–101.

Wipfli MS, Richardson JS, Naiman RJ. 2007. Ecological linkages between headwaters and downstream ecosystems: transport of organic matter, invertebrates, and wood down headwater channels. J Am Water Resour Assoc 43:72–85.

Sakamaki T, Nishimura O, Sudo R. 2002. Comparison of organic matter dynamics between tidal flat ecosystems with different biota by an experimental system. Environ Eng Res (JSCE) 39:209–18 (in Japanese).

Yokoyama H, Ishihi Y. 2007. Variation in food sources of the macrobenthos along a land-sea transect: a stable isotope study. Mar Ecol Prog Ser 346:127–41.

Sakamaki T, Nishimura O. 2006. Dynamic equilibrium of sediment carbon content in an estuarine tidal flat: characterization and mechanisms. Mar Ecol Prog Ser 328:29–40.

Yokoyama H, Ishihi Y. 2003. Feeding of the bivalve Theora lubrica on benthic microalgae: isotopic evidence. Mar Ecol Prog Ser 255:303–9.

Yokoyama H, Tamaki A, Koyama K, Ishihi Y, Shimada K, Harada K. 2005. Isotopic evidence for phytoplankton as a major food source for macrobenthos on an intertidal sandflat in Ariake Sound, Japan. Mar Ecol Prog Ser 304:101–16.