Effect of the Western Pearlshell Mussel Margaritifera Falcata on Pacific

Effect of the western pearlshell mussel Margaritifera falcata on Paciﬁ c lamprey Lampetra tridentata and ecosystem processes

Michael P. Limm and Mary E. Power

M. P. Limm ([email protected]), 4180 Valley Life Science Building, Univ. of California, Berkeley, Berkeley, CA 94720, USA. – M. E. Power, 4184 Valley Life Science Building, Univ. of California, Berkeley, Berkeley, CA 94720, USA .

Suspension feeders concentrate organic material from the water column and enhance deposition to the surrounding benthos. On the South Fork of the Eel River (Mendocino, California) two suspension feeders, the freshwater mussel Margari- tifera falcata and Pacifi c lamprey larvae Lampetra tridentata, co-occur in areas with low fl ow velocities and boundary sheer stresses. We investigated mussel/lamprey larvae interactions, and their impacts on nutrient and organic matter cycling, in fl ow-through enclosures placed where lamprey larvae and mussels naturally occurred. Over the 80-day study, lamprey larvae grew faster in the presence of mussels and in food addition treatments. Our results suggest that lamprey larvae benefi t from native mussels, and that lamprey populations and organic matter retention in rivers may decrease with the rapid decline of native freshwater mussels.

Bivalve suspension feeders play important roles in freshwater within these aggregations may exhibit stress and starvation and marine ecosystems. Th ey fi lter large volumes of water symptoms (Baker and Hornbach 1997) and slower growth and seston, and deposit faeces and pseudofaeces locally on (Bertness and Grosholz 1985, Okamura 1986, but see the benthos (Kryger and Riisgard 1988, Norrko et al. 2001, Hanson et al. 1988). Experimental research suggests bivalve Howard and Cuff ey 2006a). Bivalves enhance the nutri- fi ltering depletes food particles in the benthic boundary layer ents available to aquatic plants (Peterson and Heck 1999, and may lead to slower bivalve growth rates (Wildish and Aquilino et al. 2009) and deposit-feeding animals (Howard Kristmanson 1984). and Cuff ey 2006b, Spooner and Vaughn 2006) by increas- Given that bivalves can reduce food availability for local ing the quantity (Graf and Rosenberg 1997, Spooner and suspension feeders, observations of dense larval Pacifi c Vaughn 2006) and quality (Kautsky and Evans 1987, Peterson lamprey Lampetra tridentata within western pearlshell mus- and Heck 1999) of deposited material, nutrient cycling rates sel Margaritifera falcata aggregations in coastal rivers of (Peterson and Heck 1999, Vaughn et al. 2008, Vaughn California (Eel River, J. Howard unpubl. and Klamath River, 2010) and organic matter retention (Wotton and Malmqvist R. Reed pers. comm.) and the co-occurrence of brook lam- 2001). Since bivalve biomass can exceed that of all other prey Lampetra planeri with pearl mussels Margaritifera mar- local macroinvertebrates (Dame 1996) they can dramatically garitifera in Europe (Geist et al. 2006), is intriguing. Larval alter ecosystem structure and function (Strayer et al. 1999, lamprey, also fi lter feeders, and mussel populations are in Vaughn et al. 2008). decline globally due to habitat loss caused by river regulation How bivalves aff ect other suspension feeders is less clear, and pollution (Renaud 1997, Moser and Close 2003, Lydeard and may depend on species composition and local physical et al. 2004, Strayer et al. 2004). Both are found along stream conditions that control the supply and delivery of food par- margins and in pools where fl ow-induced boundary shear ticles such as fl ow velocity, bed topography, channel geom- stresses are reduced (Hardisty 1944, Howard and Cuff ey etry and depth. Bivalve fi ltration reduces edible particles in 2003). Margaritifera can live 100 years and aggregations the water column by removing material faster than replen- can exceed 80 individuals per square meter (Hastie and Toy ishment by advective and turbulent processes (Kryger and 2008). Lampetra larvae rear in freshwater for 5 – 8 years before Riisgard 1988, Strayer et al. 1999). For example, since its undergoing metamorphosis and migrating to salt water habi- introduction to the Hudson River, the invasive zebra mussel, tats (Beamish and Northcote 1989). Th e larvae burrow into Dreissena polymorpha, has reduced phytoplankton biomass sediment and actively pump water and materials into their by 80– 90% (Caraco et al. 1997). On local scales, bivalves fre- oral hood, out through their gills, and into surrounding sedi- quently form dense aggregations with hundreds (e.g. Mytilus ment. If larval lamprey feed on suspended material, why are californianus, Paine 1974) to thousands (e.g. Dreissena poly- they abundant in dense aggregations of mussels, whose feeding morpha , Hunter and Bailey 1992) per square meter. Bivalves can deplete edible particle concentrations in the water column?

1076 In the summer of 2004, we studied the interaction between Margaritifera falcata and Lampetra tridentata in the South Fork of the Eel River, a coastal river in northern California. Our study addressed three questions: 1) are lamprey food limited? 2) Do mussels infl uence growth of lamprey larvae (hereafter referred to as lamprey)? 3) Do mussels and lamprey infl uence organic matter deposition and microbial activity? Th ese questions were investigated with fi eld experiments in which lamprey were reared with and without additional food, and with and without mussels present.

Methods

Site

Our study area along the South Fork of the Eel River Figure 1. Photograph of experimental enclosures in South Fork of in Mendocino County, California (39 ° 44 ” N, 123 ° 39 ” W, the Eel River, Mendocino, CA. Th e enclosures were arranged in a Fig. 1) was within the Heath and Marjorie Angelo Coast block design (two blocks are visible in the photograph). Range Reserve of the Univ. of California Natural Reserve System. Th is region has a Mediterranean climate with warm, (F5,99 0.035, p 0.99) with a mean wet mass of 1.0 g dry summers and wet, cool winters. Most rainfall occurs (SD 0.2) and a mean length of 82 mm (SD 5). between October and April. Th e drainage area at our study 2 site covers approximately 140 km . River habitat consists of Experimental enclosures shallow runs, riffl es and large pools (1– 7 m deep) during We reared lamprey in fl ow-through enclosures. We used summer low fl ow periods. Vegetation in the watershed is a 36 cm tall, 24 cm diameter mesh baskets (mesh size 4 mm mixed-evergreen forest dominated primarily by old-growth by 2 mm diamond shape) with a solid bottom. Th e enclo- Douglas fi r Pseudotsuga menziesii and redwood Sequoia sem- sures were fi lled with sediment collected from dry depos- pervirens trees. Th e major aquatic food web components its near the study site. All sediment was passed through a consist of producers (primarily diatoms and fi lamentous 2 mm mesh sieve. We homogenized the 2 mm fraction and green algae), grazing insects (midges, mayfl ies, caddisfl ies) used this to fi ll the enclosures to a depth of 0.15 m. We low- and snails, predatory insects (stonefl ies, dragonfl ies, aquatic ered enclosures carefully into the stream, and placed rocks beetles and hemiptera), fi sh (stickleback Gasterosteus aculea- around the outside of the base for stability. Depths varied tus, juvenile steelhead Oncorhynchus mykiss, California roach within the study reach, so we arranged enclosures into fi ve Lavinia symmetricus ), and fi lter feeders (lamprey larvae Lam- blocks and assigned each treatment randomly within each petra tridentata, unionoid mussels Margaritifera falcata and individual block. Anodonta californiensis ) (Power et al. 1996). We conducted experiments in a 100 m long run that Mussels becomes a slow pool (fl ow velocities 0 – 5 cm s–1 ) under Mussels were haphazardly collected from nearby aggregations summer base fl ows. Margaritifera and larval Lampetra are and pooled. We measured an individual mussel ’ s length and patchily distributed within the study area. During base wet mass (live tissue plus shell) and randomly assigned it to fl ow, the mean depth of the run is 0.4 – 0.6 m, stream a mussel treatment enclosure. Initial mussel mass and length width is 13 m, and mean velocity is 0.02 – 0.04 m s–1 . A were similar between treatments (F 2.51, p 0.12) 1,48 thin layer ( 1 m) of mixed alluvium covers the bedrock with a mean wet mass of 25 g (3.9 g) and a mean length of channel. Th e median grain size (D50 ) of our study site is 68.7 mm 3.4 mm SD (these mussels were approximately approximately 50 mm based on pebble counts (Wolman 15 – 20 years old, Howard and Cuff ey 2006b). and Union 1954). Sedges Carex nudata line the riverbank and stream margins. Experiment 1

Collection Food addition on lamprey growth Lamprey To determine if lamprey are food limited, we subjected We sampled lamprey from two large pools in the South Fork lamprey to four food treatments: three diff erent types of of the Eel River using a backpack electrofi sher in mid-July food and an ambient control. Each treatment was replicated 2004. Th e two pools were approximately 150 m apart. We fi ve times, with three lamprey randomly assigned to each used short bursts to stimulate larvae to emerge from the bed, experimental enclosure. Th ree lamprey per enclosure are collected them with hand nets, and stored them in a bucket equivalent to 58 individuals m–2 , within the range (40– 100 fi lled with aerated stream water. Mass and length of lightly individuals m–2 ) we observed in open sites with fi ne sedi- anesthetized (with MS-222) larvae were measured locally, ments. Th e three food addition treatments received aquatic then individuals were randomly assigned to a treatment and (Cladophora glomerata and epiphytes), terrestrial (leaves), or placed into assigned fl ow-through enclosures. Initial lam- artifi cial (fi sh fl akes) foods. Th e algae and leaves represent prey wet mass and length were similar between treatments two diff erent detrital sources, while the fi sh fl akes represent a

1077 higher quality food in the form of carbon and nitrogen. Each enclosure to retain re-suspended material. We measured the food type was dried and ground to a fi ne powder prior to water depth to calculate the volume from which we were addition. Cladophora was visually estimated to have 10 – 30% sampling. To re-suspend surface organic matter, we rotated of its macroalgal surface covered with epiphytes, predomi- a plastic 20 cm diameter disk 360° every second for ten nantly the diatoms Gomphonema sp. and Cocconeis sp. Leaves seconds. A 500 ml sample was immediately collected and were collected from the stream and include bay, oak, mad- fi ltered through a 47 mm, 1.2 μ m glass fi ber fi lter in the rone, maple and Douglas fi r needles. No food was added to laboratory. We dried, weighed, ashed and re-weighed the the control treatment. fi lter to quantify AFDM. Based on the amount of organic biodeposits (faeces and After 80 days, we sampled sediment in the enclosures pseudofaeces) generated per week by fi ve mussels from previ- to quantify microbial respiration and AFDM. To compare ous research (Howard and Cuff ey 2006a), we added 3 g dry organic matter and microbial activity between the surface mass of each food type to their respective enclosures once a and deeper sediments, we collected the sediment from the week. To prevent food from getting swept out of the enclosure top two cm and from the bottom eight cm. We homoge- by water currents, we used a plastic cylinder with a 0.22 m inner nized sediment from each layer and placed a subsample of diameter into the enclosure. Th e bottom of the plastic cylinder measured volume into a 500 ml plastic bottle. Each bottle rested on the sediment and the top was above the water line. was fi lled with stream water, sealed, and inverted to mix After food was added and allowed to settle, we removed the and dislodge any air bubbles. Th e bottle was then topped plastic cylinder slowly from the enclosure. We observed very off with water, resealed, and inverted again. We then took little loss of suspended material in the process. After 80 days the initial dissolved oxygen measurement using a DO probe. we processed the lamprey as described below. Th e bottle was then resealed, placed into a dark chamber in the stream, and agitated every 5 min. After 30– 40 min, Lamprey element and stable isotopes we measured dissolved oxygen again using the DO probe. To assess whether lamprey larvae were assimilating the sup- Respiration values were standardized by both sediment mass plemented food, we measured the carbon and nitrogen sta- and organic matter present. In the laboratory the sediment ble isotope ratios of the three supplemented foods and larvae was placed into a drying oven at 60° C for 72 h. Th e sediment tissue from each experimental enclosure. After 80 days, the was weighed and then ashed in a muffl e furnace at 550 ° C lamprey were removed, killed with MS-222, and placed on for four h. After removal from the furnace, the sediment was ice. In the laboratory, we measured larval length and mass dried again at 60 ° C for 72 h and weighed. and collected a tissue sample from the tail of individual larvae for stable isotope analysis. Food and larval tissue samples Analysis were rinsed and stored at – 5 ° C. In preparation for analysis, samples were freeze-dried, ground, and placed into tin cap- We analyzed the larval growth in experiment 1 (when reared sules. We measured carbon and nitrogen content and stable with and without mussels) with a mixed-model ANOVA isotopes using an elemental analyzer connected to a stable design with treatment fi xed and block random. In experi- isotope analyzer. Th e δ 13 C and δ 15 N values we report are in ment 2 (when reared with and without the three food addi- reference to Pee Dee Belemnite and atmospheric nitrogen tions) we analyzed larval lamprey growth with mixed model standards, respectively. design with treatment fi xed and block random. If treatment was signifi cant we compared means using a Tukey ’ s test Experiment 2 (alpha 0.05). We analyzed data from the 30 and 80 day measures of Lamprey growth near mussels AFDM, the 80 day measure of respiration, and the 80 day To investigate whether mussels enhance or reduce lamprey measure of larval lamprey δ 13 C and δ 15N with a mixed model growth, we randomly assigned three lamprey to a mussel and with treatment fi xed and block random. We used Tukey ’ s test control treatments. For the mussel treatment, fi ve mussels (alpha 0.05) to compare means if treatments were signifi - were placed into an enclosure. Five mussels per enclosure cantly diff erent. JMP (SAS Inst.) software was used for all (area 520 cm 2) produce a density of 96 individuals m–2 , analyses. close to ambient densities observed in nearby mussel aggregations (∼ 85 individuals m–2 , Howard and Cuff ey 2006b). No mussels were added to the control treatment. Each treat- Results ment was replicated fi ve times. Experiment 1 Organic matter and respiration To quantify lamprey and mussel impacts on organic matter Food addition accrual and respiration in the sediment, we sampled sedi- After 80 days, one lamprey in the Cladophora addition treatment from four treatment enclosures. Th e treatments were ment and two lamprey in the leaf addition treatments were three lamprey, three lamprey with fi ve mussels, fi ve mussels, not found. Th e two lamprey missing in the leaf addition and an ambient control with no lamprey or mussels present. treatments were from diff erent replicates. All mussels were We sampled organic matter in the enclosures after 30 days actively fi ltering at the end of the experiment. and 80 days. Before sampling organic matter at 30 days, we In the food addition experiment, lamprey grew nearly visually inspected enclosures and counted invertebrates on twice as fast in the leaf and fi sh-fl ake treatments as in the the surface. We then placed a plastic cylinder within each control treatment (F3,12 2.66, p 0.07, Fig. 2). Lamprey

1078 Table 1. Results of lamprey tissue composition from IRMS analysis.

% C SE % N SE δ13 C SE δ 15 N SE

Lamprey 53.6 0.9 11.1 0.4 23.2 0.2 2.7 0.1 Lamprey with 54.1 0.8 10.7 0.4 23.2 0.2 2.5 0.1 mussels Lamprey fed 52.7 1.0 11.4 0.5 23.7 0.3 2.7 0.1 cladophora Lamprey fed 53.2 0.6 11.6 0.3 23.3 0.2 2.7 0.1 leaves Lamprey fed 53.4 1.2 11.2 0.5 23.5 0.3 3.3 * 0.2 fi sh fl akes Note: * denotes signifi cance at a 0.01 Algae 21.6 0.1 2.2 0.0 21.7 0.08 0.9 0.02 Leaves 48.7 0.1 0.6 0.01 27.7 0.03 0.1 0.14 Fish fl akes 43.8 0.3 5.9 0.05 20.9 0.03 9.9 0.2

Figure 2. Larval lamprey growth after 80 days with and without mussels in the enclosures. Error bars represent 1 SE. and lamprey (mean 9.8, SE 4.5) enclosures relative to control enclosures that lacked both mussels and lamprey (mean 5.8, SE 1.4), but the diff erences were not signifi - grew the slowest on ground Cladophora , which had lower cant (F 1.70, p 0.21). carbon content (22% of total) than ground fi sh fl akes and 3,16 ground leaves (49% and 44% of total, respectively). Lam- prey growth was also low in the same block (block 2), but in Discussion experiment 2, the eff ect was not signifi cant. δ15 Lamprey tissue N in the fi sh fl ake treatment was sig- Filter-feeding bivalves concentrate and deposit organic-rich nifi cantly higher than in the other treatments (Table 1), as material, excrete nutrients into their surroundings, and mix δ15 was the N of the fi sh fl akes relative to that of the leaves and stabilize sediment. In coastal habitats, bivalves enhance and Cladophora , which suggests lamprey were assimilating leaf growth in salt marsh cordgrass Spartina alternifl ora the nitrogen from the fi sh fl akes. We observed no diff erences (Bertness 1984), the seagrass Th alassia testudinum (Peterson δ13 in %C, %N, and C of lamprey muscle tissue between and Heck 2001) and the seaweed Porphyra perforata (Aquilino food addition treatments. et al. 2009). In freshwater, higher periphyton abundance has been observed on sediment (Vaughn et al. 2007) and on the Experiment 2 shells of two unionid mussel species relative to non-feeding sham mussels (Spooner and Vaughn 2006). In our study Lamprey growth near mussels the mussel Margaritifera falcata signifi cantly enhanced Over the eighty day period, lamprey grew twice as fast Pacifi c lamprey larvae growth and increased respiration in (0.08 mm day-1 vs 0.037 mm day -1) when reared with mus- the sediment. sels (F1,4 71.66, p 0.001, Student ’ s t 2.77, Fig. 3). We Mussel biodeposits, by increasing the quality and/or also observed a signifi cant block eff ect with lamprey growing quantity of available food, may have fueled faster lamprey signifi cantly more slowly in one block (block 2). growth. In both marine and freshwater systems, mussels increase bulk deposition rates and the organic and nutrient Organic matter and respiration concentration of the deposited material (Kautsky and Evans Surface organic matter was similar amongst treatments after 30 days (F3,16 2.23, p 0.12). After 80 days, organic matter was similar amongst treatments in the top two cm (F 3,12 0.62, p 0.62) or the bottom eight cm (F3,16 2.84, p 0.39) of sediment. A block eff ect was observed after 80 days in the top two cm of sediment (F4,12 14.6, p 0.001), with signifi cantly lower organic matter in block 3 (Tukey ’ s, alpha 0.05). We observed signifi cant diff erences in respiration between treatments in the top two cm of sediment (F3,16 3.22, p 0.05, Fig. 4). Respiration was signifi cantly higher in the mussel lamprey treatment than in control or lamprey treatments (Tukey’ s, alpha 0.05). Respiration was not diff erent between treatments in the bottom eight cm (F3,16 1.11, p 0.38). Due to unequal variances we log-transformed Gumaga count data. After 30 days we observed higher numbers of the caddisfl y Gumaga nigricula in the mussel (mean 14.8, Figure 3. Larval lamprey growth after 80 days in control (no food SE 2.7), mussel lamprey (mean 12.0, SE 2.6), added) and food addition treatments. Error bars represent 1 SE.

1079 enclosures may have ingested mussel biodeposits settling into their burrow. Lamprey larvae are often described as suspension feeders (Moore and Mallatt 1980, Malmqvist and Br ö nmark 1982) and food has been kept in suspension during previous lamprey growth experiments (Hardisty 1944, Moore and Mallatt 1980). In our food addition experiment, however, lamprey grew faster even though added food was deposited. Lamprey fed ground fi sh fl akes, which are enriched in 15 N, had signifi cantly higher δ 15N isotope values than other lamprey. Th ese results, and the presence of sand grains in lamprey stomach contents (Moore and Beamish 1973), suggest lamprey may feed directly off the bed surface and/or fi lter interstitial water. Bioturbation in the enclosures may have infl uenced lamprey feeding and growth. Mussel biodeposits, coupled with animal movement and activity (e.g. water pumping, van Duren et al. 2006), could have increased food particle re-suspension and fl ux over lamprey burrows. Mussels are also bioturbators (Dame 1996), and their activity can mix material both in the water column (van Duren et al. 2006) and the sediment (Vaughn and Hagenkamp 2001). In addition to increasing food particle deposition, mussel activity while moving may have increased the food available to lamprey both in the water above their burrows and in interstitial spaces. Th e increase in lamprey growth but no diff erence in organic matter accrual is at face value, paradoxical, but bioturbation may have played a role. We did not observe organic Figure 4. Respiration measured after 80 days in the top two cm and matter accruing at higher rates near mussel aggregations, as bottom eight cm of sediment. Error bars represent 1 SE. seen in previous studies (Kryger and Riisgard 1988, Grenz et al. 1990, Graf and Rosenberg 1997, Peterson and Heck 1987). In the South Fork of the Eel River, Margaritifera fal- 1999, Norkko et al. 2001, Spooner and Vaughn 2006). Ani- cata aggregations increase both the deposition rate and the mal activity can remove fi ne particles from interstitial spaces percent organic material of deposits relative to background (Zanetell and Peckarsky 1996), and the more numerous levels (Howard and Cuff ey 2006a). Lamprey larvae feed pre- invertebrates (Gumaga , and possibly other invertebrates we dominantly on organic detritus, which typically accounts for did not quantify) in the mussel enclosures may have removed 96– 98% of their stomach contents (the remainder includes mussel biodeposits by ingesting them, or resuspending and diatoms and bacteria, Moore and Beamish 1973, Sutton and dispersing them out of the enclosures. Bowen 1994, Mundahl et al. 2005). Of the organic detri- In addition to bioturbation, rapid degredation of bio- tus ingested, lamprey assimilate 60% (Sutton and Bowen deposits may have off set the higher rate of deposition in 1994, Mundahl et al. 2005). Th e evidence of faster lamprey mussel enclosures. Respiration rates in sediment from growth in both the mussel and food addition treatments mussel and mussel lamprey enclosures were higher than suggests lamprey were food limited in the ambient control in control enclosures. Mussels and other bioturbators are treatments. known to mix organic material and oxygen into the sedi- Th e lamprey δ 13 C values (ca – 23 ‰ ) suggest lam- ment and to enhance water fl ux across the water-sediment prey feed on a mix of aquatic sources (e.g. riffl e and pool boundary (Spooner and Vaughn 2006). Th is mixing can derived algae, with relatively depleted and enriched carbon, enhance microbial processing and degradation of organic respectively as shown by Finlay et al. 1999) and/or possi- matter (Dame 1996). Bacteria abundance and exoenzy- bly terrestrial sources. Lamprey tissue values were similar to matic activity increases rapidly after mussel biodeposits Margaritifera tissue δ 13 C and δ 15 N values ( – 22.9 ‰ , 2.6 ‰ , settle, and degradation can occur over short time periods respectively) reported by Howard et al. (2005) during the (Stuart et al. 1982, Grenz et al. 1990). Stuart et al. (1982) summer period in the South Fork of the Eel River. Finlay observed a rapid increase in bacterial abundance in depos- et al. (1999) report summer δ 13 C values in the South Fork of ited mussel faeces and pseudofaeces relative to background the Eel River of – 17.9 ‰ for epilithic algae in pool habitats, levels. Maximum mineralization occurred three days after – 26.2 ‰ for epilithic algae in riffl e habitats, and – 27.5 ‰ for deposition and gradually declined in their study. Grenz terrestrial detritus. Stable isotope values of control lamprey et al. (1990) also observed similar rapid increases in ( δ13 C – 23.2 ‰ , δ 15 N 2.7 ‰ ) refl ect aquatic and ter- bacteria production on sediments enriched with mussel restrial source contributions to both suspended (seston) and biodeposits. Th ese studies suggest biodeposits degrade deposited material. with in days. Th e pathway by which mussel biodeposits and the added Lamprey, like mussels, may act as bioturbators. Lamprey food reached lamprey is unclear. Lamprey in the mussel increased respiration in the sediment, but only when mussels

1080 were present. Th eir movement and fi ltering activity may mix Acknowledgements – We thank G. Benigno and E. Limm for help in and aerate the sediment and increase microbial activity. We the fi eld and P. Steel for his tireless work on site. We thank the Univ. did not quantify lamprey movement, but we did observe new of California Natural Reserve System for the protected research site. lamprey burrows in the enclosures during the experiment. Th e project was funded by the National Science Foundation via grants to the National Center for Earth Surface Dynamics (EAR Lamprey also fl ush the surrounding sediment with water and 0120914) and to Consumer-Resource Interactions and Stoichio- introduce unassimilated particles during fi ltering. Lamprey metrically Explicit Spiraling (DEB 0543363). We thank W. Sousa selectively fi lter and ingest particles smaller than 400 μ m for reviewing the manuscript and providing helpful comments. (Moore and Mallat 1980), while larger particles are ejected out through the gills and into the surrounding sediment. Sea lamprey Petromyzon marinus and brook lamprey larvae were References found to ingest 5 mg AFDM g–1 ammocoete day–1 in July (Sutton and Bowen 1994). If approximately 60% is assimi- Aquilino, K. M. et al. 2009. Local-scale nutrient regeneration lated, the remaining 40% is released into the sediment. Th ree facilitates seaweed growth on wave-exposed rocky shores in an lamprey in each enclosure would add approximately 6 mg upwelling system. – Limnol. Oceanogr. 54: 309 – 317. of organic material to the enclosure daily and possibly Baker, S. M. and Hornbach, D. J. 1997. Acute physiological eff ects of zebra mussel (Dreissena polymorpha ) infestation on two increase microbial activity. In our experiment, however, this unionid mussels, Actinonaias ligamentina and Amblema plicata . increase in organic matter or sediment respiration by lam- – Can. J. Fish. Aquat. Sci. 54: 512 – 519. prey larvae was not detectable when mussels were absent. Beamish, R. J. and Northcote, T. G. 1989. Extinction of a popula- Interactions between mussels and lamprey larvae may tion of anadromous parasitic lamprey, Lampetra tridentata , depend on environmental conditions. Spatial and temporal upstream of an impassable dam. – Can. J. Fish. Aquat. Sci. 46: variation in temperature can alter mussel assemblage struc- 420 – 425. ture and species performance (Spooner and Vaughn 2009), Beamish, F. W. H. and Lowartz, S. 1996. Larval habitat of Amer- both of which infl uence periphyton accrual, organic mat- ican brook lamprey. – Can. J. Fish. Aquat. Sci. 53: 693 – 700. Bertness, M. D. 1984. Ribbed mussels and Spartina alternifl ora produc- ter accrual, nutrient cycling, and invertebrate abundance tion in a New England salt marsh. – Ecology 65: 1794 – 1807. and assemblage (Spooner and Vaughn 2006, Vaughn et al. Bertness, M. D. and Grosholz, E. 1985. Population dynamics of 2008). Physicochemical characteristics of the streambed can the ribbed mussel, Geukensia demissa : the costs and benefi ts of infl uence habitat suitability for larval mussels (Geist and an aggregated distribution. – Oecologia 67: 192 – 204. Auerswald 2007), and fl ow-particle size relationships may Caraco, N. F. et al. 1997. Zebra mussel invasion in a large, turbid infl uence lamprey larvae burrow suitability. Lamprey larval river: phytoplankton response to increased grazing. – Ecology densities increase with deposited organic matter and chlo- 78: 558 – 602. rophyll a (Hardisty 1944, Malmqvist 1980, Beamish and Cardinale, B. J. et al. 2002. Species diversity enhances ecosystem functioning through interspecifi c facilication. – Nature 415: Lowartz 1996), suggesting that they do track variation in 426 – 429. food availability. Where mussel aggregations enhance these Dame, R. F. 1996. Ecology of marine bivalves: an ecosystem resources, they may facilitate lamprey during their larval approach. – CRC Press. rearing stages. Finlay, J. C. et al. 1999. Eff ects of water velocity on algal carbon Between species facilitation can enhance ecosystem func- isotope ratios: implications for river food web studies. – tion in aquatic ecosystems, even when species are within Limnol. Oceanogr. 44: 1198 – 1203. similar functional groups (Cardinale et al. 2002). In our Geist, J. 2010. Strategies for the conservation of endangered fresh- study, mussels facilitated lamprey growth and microbial res- water pearl mussels (Margaritifera margaritifera L.): a synthesis of conservation genetics and ecology. – Hydrobiologia 644: piration was highest when both species were present. While 69 – 88. both species are predominantly suspension feeders, they Geist, J. and Auerswald, K. 2007. Physicochemical stream bed may partition resources and have a ‘ complementarity eff ect ’ characteristics and recruitment of the freshwater pearl (Hooper and Vitousek 1997) on ecosystem processes. Mus- mussel (Margaritifera margaritifera). – Freshwater Biol. 52: sels feed predominantly on particles less than 20 microns in 2299 – 2316. size (Vaughn et al. 2008), while lamprey can fi lter larger par- Geist, J. et al. 2006. Th e status of host fi sh populations and fi sh ticles ( 400 μ m, Moore and Mallat 1980). Lamprey may species richness in European freshwater pearl mussel (Marga- utilize larger food particles in mussel pseudofaeces, and the ritifera margaritifera ) streams. – Aquat. Conserv. Mar. Fresh- water Ecosyst. 16: 251 – 266. combined excretion and egestion from mussels and lamprey Graf, G. and Rosenberg, R. 1997. Bioresuspension and biodeposi- into surrounding sediments could explain the higher micro- tion: a review. – J. Mar. Syst. 11: 269 – 278. bial respiration we observed. Grenz, C. et al. 1990. In situ biochemical and bacterial variation of Our study provides additional evidence of the key role sediments enriched with mussel biodeposits. – Hydrobiologia suspension feeders play in freshwater ecosystems, includ- 207: 153 – 160. ing to other suspension feeders. Both mussels and lamprey, Hanson, J. M. et al. 1988. Th e eff ects of water depth and density and their fi sh hosts, face similar threats: altered fl ow regime, on the growth of a unionid clam. – Freshwater Biol. 19: habitat degradation, and habitat destruction (Renaud 1997, 345 – 355. Hardisty, M. W. 1944. Th e life history and growth of the brook Moser and Close 2003, Lydeard et al. 2004, Strayer et al. lamprey (Lampetra planeri ). – J. Anim. Ecol. 13: 110 – 122. 2004, Geist 2010). Given their apparently complementary Hastie, L. C. and Toy, K. A. 2008. Changes in density, age structure functional role, the loss of either mussels or lamprey will and age-specifi c mortality in two western pearlshell (Marga- alter stream carbon and nutrient cycling, and the loss will ritifera falcata) populations in Washington (1995– 2006). not be merely subtractive. – Aquat. Conserv. 18: 671 – 678.

1081 Hooper, D. U. and Vitousek, P. M. 1997. Th e eff ects of plant Peterson, B. J. and Heck, K. L. 2001. Positive interactions between composition and diversity on ecosystem processes. – Science suspension-feeding bivalves and seagrass-a facultative mutual- 277: 1302–1305 ism. – Mar. Ecol. Prog. Ser. 213: 143 – 155. Howard, J. K. and Cuff ey, K. M. 2003. Freshwater mussels in a Power M. E. et al. 1996. Disturbance and food chain length in California North Coast Range river: occurrence, distribution, rivers. – In: Polis, G. A. and Winemiller, K. O. (eds), Food and controls. – J. N. Am. Benthol. Soc. 22: 63 – 77. webs: integration of patterns and dynamics. Chapman and Howard, J. K. and Cuff ey, K. M. 2006a. Th e functional role of Hall, pp. 286 – 297. native freshwater mussels in the fl uvial benthic environment. Renaud, C. B. 1997. Conservation status of Northern Hemisphere – Freshwater Biol. 51: 460 – 474. lampreys (Petromyzontidae). – J. Appl. Ichthyol. 13: Howard, J. K. and Cuff ey, K. M. 2006b. Factors controlling the 143 – 148. age structure of Margaritifera falcata in 2 northern California Spooner, D. E. and Vaughn, C. C. 2006. Context-dependent streams. – J. N. Am. Benthol. Soc. 25: 677 – 690. eff ects of freshwater mussels on stream benthic communities. Howard, J. K. et al. 2005. Toward using Margaritifera falcata as – Freshwater Biol. 51: 1016 – 1024. an indicator of base level nitrogen and carbon isotope ratios: Spooner, D. E. and Vaughn, C. C. 2009. Species richness and insights from two California Coast Range rivers. – Hydrobiologia temperature infl uence mussel biomass: a partitioning 541: 229 – 236. approach applied to natural communities. – Ecology 90: Hunter, R. D. and Bailey, J. F. 1992. Dreissena polymorpha (zebra 781– 790. mussel): colonization of soft substrata and some eff ects on Strayer, D. L. et al. 1999. Transformation of freshwater ecosystems unionid bivalves. – Nautilus 106: 60 – 67. by bivalves – a case study of zebra mussels in the Hudson River. Kautsky, N. and Evans, S. 1987. Role of biodeposition by Mytilus- – Bioscience 49: 19 – 27. edulis in the circulation of matter and nutrients in a baltic Strayer, D. L. et al. 2004. Changing perspectives on pearly mussels, coastal ecosystem. – Mar. Ecol. Prog. Ser. 38: 201 – 212. North America ’ s most imperiled animals. – Bioscience 54: Kryger, J. and Riisgå rd, H. U. 1988. Filtration rate capacities in 6 429 – 439. species of European freshwater bivalves. – Oecologia 77: 34 – 38. Stuart, V. et al. 1982. Conversion of kelp debris and fecal material Lydeard, C. et al. 2004. Th e global decline of nonmarine molluscs. from the mussel Aulacomya ater by marine microorganisms. – Bioscience 54: 321 – 329. – Mar. Ecol. Prog. Ser. 7: 47 – 57. Malmqvist, B. 1980. Habitat selection of larval brook lamprey (Lam- Sutton, T. and Bowen, S. 1994. Signifi cance of organic detritus in petra planeri, Bloch) in a south Swedish stream. – Oecologia the diet of larval lampreys in the Great Lakes Basin. – Can. J. 45: 35 – 38. Fish. Aquat. Sci. 51: 2380 – 2387. Malmqvist, B. and Brö nmark, C. 1982. Filter feeding in larval van Duren, L. A. et al. 2006. Eff ects of mussel fi ltering activity on Lampetra planeri – eff ects of size, temperature and particle boundary layer structure. – J. Sea Res. 55: 3 – 14. concentration. – Oikos 38: 40 – 46. Vaughn, C. C. 2010. Biodiversity losses and ecosystem function in Moore, J. W. and Beamish, F. W. H. 1973. Food of larval sea freshwaters: emerging conclusions and research directions. lamprey (Petromyzon marinus ) and American brook lamprey – Bioscience 60: 25 – 35. (Lampetra lamottei ). – J. Fish. Res. Board Can. 30: 7 – 15. Vaughn, C. C. and Hakenkamp, C. C. 2001. Th e functional role Moore, J. W. and Mallatt, J. M. 1980. Feeding of larval lamprey. of burrowing bivalves in freshwater ecosystems. – Freshwater – Can. J. Fish. Aquat. Sci. 37: 1658 – 1664. Biol. 46: 1431 – 1446. Moser, M. L. and Close, D. A. 2003. Assessing Pacifi c lamprey status Vaughn, C. C. et al. 2007. Context-dependent species identity in the Columbia River basin. – Northwest Sci. 77: 116 – 125. eff ects within a functional group of fi lter-feeding bivalves. Mundahl, N. D. et al. 2005. Diet, feeding rate, and assimilation – Ecology 88: 1654 – 1662. effi ciency of American brook lamprey. – Environ. Biol. Fish. Vaughn, C. C. et al. 2008. Community and foodweb ecology 72: 67 – 72. of freshwater mussels. – J. N. Am. Benthol. Soc. 27: Norkko, A. et al. 2001. Benthic-pelagic coupling and suspension- 409– 423 feeding bivalves: linking site-specifi c sediment fl ux and Wildish, D. J. and Kristmanson, D. D. 1984. Importance to mus- biodeposition to benthic community structure. – Limnol. sels of the benthic boundary layer. – Can. J. Fish. Aquat. Sci. Oceanogr. 46: 2067 – 2072. 41: 1618 – 1625. Okamura, B. 1986. Group living and the eff ects of spatial position Wolman, M. G. and Union, A. 1954. A method of sampling coarse in aggregations of Mytilus edulis . – Oecologia 69: 341 – 347. river-bed material. – Trans. Am. Geophys. Union 35: 951– 956. Paine, R. T. 1974. Intertidal community structure. – Oecologia 15: Wotton, R. S. and Malmqvist, B. 2001. Feces in aquatic ecosys- 93 – 120. tems. – Bioscience 51: 537 – 544. Peterson, B. J. and Heck, K. L. 1999. Th e potential for suspension Zanetell, B. A. and Peckarsky, B. L. 1996. Stonefl ies as ecological feeding bivalves to increase seagrass productivity. – J. Exp. Mar. engineers – hungry predators reduce fi ne sediments in stream Biol. Ecol. 240: 37 – 52. beds. – Freshwater Biol. 36: 569 – 577.

1082