Nutrient recycling by consumers in streams can influence ecosystem nutrient availability and the assemblage and growth of photoautotrophs. Stream fishes can play a large role in nutrient recycling, but contributions by other vertebrates to overall recycling rates remain poorly studied. In tributaries of the Pacific Northwest, coastal giant salamanders (Dicamptodon tenebrosus) occur at high densities alongside steelhead trout (Oncorhynchus mykiss) and are top aquatic predators. We surveyed the density and body size distributions of D. tenebrosus and O. mykiss in a California tributary stream, combined with a field study to determine mass-specific excretion rates of ammonium (N) and total dissolved phosphorus (P) for D. tenebrosus. We estimated O. mykiss excretion rates (N, P) by bioenergetics using field-collected data on the nutrient composition of O. mykiss diets from the same system. Despite lower abundance, D. tenebrosus biomass was 2.5 times higher than O. mykiss. Mass-specific excretion summed over 170 m of stream revealed that O. mykiss recycle 1.7 times more N, and 1.2 times more P than D. tenebrosus, and had a higher N:P ratio (8.7) than that of D. tenebrosus (6.0), or the two species combined (7.5). Through simulated trade-offs in biomass, we estimate that shifts from salamander biomass toward fish biomass have the potential to ease nutrient limitation in forested tributary streams. These results suggest that natural and anthropogenic heterogeneity in the relative abundance of these vertebrates and variation in the uptake rates across river networks can affect broad-scale patterns of nutrient limitation.
Citation: Munshaw RG, Palen WJ, Courcelles DM, Finlay JC (2013) Predator-Driven Nutrient Recycling in California Stream Ecosystems. PLoS ONE 8(3): e58542. https://doi.org/10.1371/journal.pone.0058542
Editor: Wayne M. Getz, University of California, Berkeley, United States of America
Received: October 13, 2012; Accepted: February 5, 2013; Published: March 8, 2013
Copyright: © 2013 Munshaw et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the National Science and Engineering Research Council of Canada’s USRA and Discovery Grant program (www.nserc-crsng.gc.ca), and the Canada Research Chairs program (www.chairs-chaires.gc.ca). No individuals employed or contracted by the funders played any role in study design, data collection and analysis, decision to publish, preparation of the manuscript.
Competing interests: This work has previously been reviewed by Dr. Mary Power, Dr. John Schade, Dr. Jill Welter, and Dr. James Hood. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and materials.
The productivity of primary producers in aquatic ecosystems is limited by a range of biotic and abiotic factors, with light and nutrient availability among the primary drivers –. Nutrient limitation can over-ride the importance of light, and even in low-light environments, increasing the availability of nutrients can increase primary producer biomass . Nitrogen (N) and phosphorus (P) are key macronutrients that can individually or simultaneously limit growth depending on their availability and the requirements of the demand assemblage , . External ambient nutrient fluxes of N and P are driven by weathering, fixation, and runoff from terrestrial ecosystems  as organic and inorganic components of soils release dissolved mobile ions in surface waters . In addition, internal biotic processes can subsidize fluxes into the system, increasing nutrient availability.
In aquatic environments, animals process organic compounds through consumption, metabolism, and excretion into labile dissolved forms that can be taken up by autotrophs and other microbial organisms . This process of consumer-driven nutrient recycling subsidizes ambient nutrient levels, and can affect the community structure of phytoplankton in lakes , , control the availability of nutrients in phytotelmata , and control the spatial distribution of nutrients in river systems . The rate at which nutrients are recycled within a system depends heavily on the physical characteristics of that system and characteristics of the organisms that inhabit it.
Interactions among producers, herbivores, and predators are important determinants of nutrient dynamics in aquatic ecosystems. Changes in nutrient availability can have direct effects on producer biomass , as well as potential for future growth through nutrient storage . As the base of the food web, producers in streams can drive ecosystem productivity and can be directly and indirectly affected by a range of animals. Herbivores can directly limit producer growth and biomass through grazing pressure . Higher trophic level animals indirectly affect autotroph abundance by stimulating higher producer productivity through nutrient recycling and preying on herbivores, which in turn reduces grazing pressure , –.
To date, investigations of the effects of consumer identity and abundance on nutrient recycling in streams have focused mainly on fish assemblages (e.g. , , –, but see –). While fish species represent a large fraction of animal biomass in many stream ecosystems, there are myriad other taxa that make up a substantial component of secondary production, including aquatic macroinvertebrates and other vertebrates (amphibians, reptiles, mammals, birds). Predatory amphibian populations in particular, often rival the abundance and biomass of fishes in many systems, and may have similarly large effects on nutrient recycling. While some previous work has estimated the potential for nutrient recycling by a limited number of amphibians at broad scales , , excretion data for salamanders is non-existent for most species, and they have seldom been considered as potentially important nutrient recyclers . In temperate tributary streams such as those in the Pacific Northwest, amphibian populations can reach extremely high densities –, even surpassing those of fishes , . In this region, two main vertebrate groups dominate predator assemblages; larval salamanders of the Genus Dicamptodon and juvenile salmonids of the Genus Oncorhynchus. Larval salamanders can often occur at densities exceeding all other predators, accounting for up to 99% of the vertebrate biomass in some areas . Dicamptodon tenebrosus (coastal giant salamander; ) is one of the largest salamanders in North America, and often co-occurs with Oncorhynchus mykiss (steelhead trout). Both vertebrates are similarly opportunistic predators , . As large bodied vertebrates, these predators have may fill similar ecological niches and may contribute similarly to ecosystem processes such as nutrient recycling.
Here we examined the effects of species identity in nutrient recycling between two taxonomically diverse top-predators, D. tenebrosus and O. mykiss, and explored the relative importance of ecosystem-scale N and P recycling by these predators across a range of potential real-world densities. We conducted a field study to estimate diet composition, nutrient excretion rates (N, P), and densities for each species, as well as the elemental body composition (C, N, P) of their prey in a coastal stream in Northern California. We hypothesize that, due to their large body size and high abundance , D. tenebrosus would dominate predator-mediated nutrient recycling in tributary ecosystems where both species co-occur. We also hypothesize that changes in the relative abundance of each species could affect ecosystem-level nutrient availability.
To quantify the magnitude and importance of nitrogen and phosphorus recycling by vertebrates in tributary biogeochemistry, we conducted a survey of the size distribution, abundance, and diet composition of D. tenebrosus and O. mykiss. We estimated mass-specific excretion rates by two methods; in situ incubations for D. tenebrosus to establish novel excretion rate estimates, and bioenergetics modeling for O. mykiss. Bioenergetics models are often substituted for direct excretion measurements when appropriate models and parameter estimates exist , –. Using a bioenergetics model to estimate excretion rates has been shown to be a very equitable surrogate for direct measurement . However, due to a lack of available data on D. tenebrosus, direct measurements were necessary to establish excretion estimates. Despite the uncertainties of using two methodologies for determining excretion rates, both methods can be considered comparable if adequate care is taken with handling and incubation time to minimize the effects of stress and fasting , , . By using a bioenergetics model for fish, we were able to avoid the error in estimating normal excretion rates caused by handling stress, and reduce the need to experimentally manipulate additional animals.
Although aquatic vertebrates can produce similar daily quantities of N and P through egestion and excretion , nutrients in fecal matter must undergo further microbial processing before they are available for uptake by autotrophs. We therefore do not consider egestion as an instantaneous contribution to nutrient recycling in this study. As such, we acknowledge that our estimates of excretion represent underestimates of total nutrients recycled by vertebrates. We parameterized an O. mykiss bioenergetics model using surveyed diet composition, prey whole-body nutrient compositions (carbon, nitrogen, phosphorus), and summer water temperatures collected from several tributary streams. We combined estimates of O. mykiss and D. tenebrosus population densities with mass-specific excretion rates to estimate the relative magnitude of nutrient recycling by each predator. Using these excretion estimates and literature values for local nutrient uptake rates, we determined total vertebrate contribution to nutrient demand, and lastly explore the consequences of different relative abundances of D. tenebrosus and O. mykiss for ecosystem level nutrient availability.
This work was conducted with the approval California Department of Fish and Game (#11077), NOAA (#14904), and Simon Fraser University Animal Care (920B-09) permits. All animals were anaesthetized using Tricaine methanesulfonate (MS-222) before handling, and all efforts were made to minimize stress and suffering during this study.
Like many coastal tributaries in Northern California, Oregon, and Washington, our study site (Fox Creek, South Fork Eel river watershed, UTM: 10S 445880E, 4399070N) has two classes of vertebrate predators, stream salamanders (Genus Dicamptodon) and juvenile salmonid fishes (Genus Oncorhynchus). Dicamptodon tenebrosus exhibits life-history plasticity; after two to three years as aquatic juveniles, D. tenebrosus can either remain aquatic as paedomorphic adults or metamorphose into terrestrial adults . The geographic range of D. tenebrosus includes coastal watersheds from the southern extremes of British Columbia to central California with the exception of the Olympic Peninsula . Oncorhynchus mykiss populations in Pacific coastal watersheds are usually comprised of a mix of anadromous (i.e. steelhead) and resident (i.e. rainbow trout) individuals, with juveniles rearing in freshwaters for one to two years typically followed by divergence into one of the two dominant life-histories , . To determine the mass-frequency distribution and total biomass represented by each species, we conducted a survey of D. tenebrosus and O. mykiss in a tributary typical of coastal watersheds in the Pacific Northwest. We conducted our survey on Fox Cr. (2.6 km2 drainage area), a perennial tributary of the South Fork Eel River (SF Eel), Mendocino County, California, whose watershed lies entirely within the University of California’s Angelo Coast Range Reserve. The Fox Cr. watershed is dominated by Douglas fir (Pseudotsuga menziesii), Coast redwood (Sequoia sempervirens), and mixed conifer-deciduous forests. The channel is moderately incised with steep banks and is heavily shaded, and frequent woody debris jams are indicative of high winter discharge (November-March) despite low dry-season base flow (April-October, 5–7 L⋅s−1; ). During summer base flow, the stream is reduced to a series of short riffles and shallow pools. The streambed alluvium is a mixture of cobbles and boulders, embedded with sand and pebbles.
To estimate the body size distribution and total biomass of D. tenebrosus and O. mykiss, we surveyed all individuals of both species in a sub-set of pools within the first 1.3 km of Fox Cr. Both O. mykiss and D. tenebrosus co-occur in Fox Cr. from the confluence with the SF Eel to approximately 1.3 km upstream, at which point fish passage is restricted, and only D. tenebrosus is present. Based on previous surveys, these two species constitute the vast majority of vertebrate biomass in tributary streams in the region and are present year round –, –. Animals were surveyed by serial depletion in 32 tributary reaches (total stream distance of 170 m) blocked with nets from the top of the upstream riffle to the downstream end of each pool, using a combination of methods including snorkeling, electro-shocking, and hand-capture using the Parker-stick method . After capture, animals were anesthetized with MS-222, weighed (g), and measured for total length (TL), snout to vent length (SVL, for D. tenebrosus) and standard length (SL, for O. mykiss). We used SVL and SL measurements as the primary measures of body length due to the frequency of tail injuries that can bias TL measurements. To compare dietary intake of N and P, a subset of captured individuals were sampled for diet using non-lethal gastric lavage (O. mykiss n = 86, D. tenebrosus n = 55). We adopted a minimum size threshold of 100 mm SL for O. mykiss and 50 mm SVL for D. tenebrosus based on lavage apparatus to avoid injuring small individuals during diet collection. Individuals with extensive injuries were not dieted. All individuals in the diet survey were captured in the daytime in pools, controlling for diel and habitat variability in gut contents. Individuals were dieted opportunistically across several concurrent studies, which is reflected by the inconsistent sample sizes. Diets were preserved in 70% EtOH until enumeration, measurement, and identification to the highest level of taxonomic resolution possible (e.g. Genus or Family). Dry biomass of individual diet items was estimated using measured length (mm) and taxon-specific length-weight regressions developed from our survey data and the literature (Table S1). Diet data were supplemented with a longer-term data set from our study stream and nearby tributaries. Water content of prey items was determined from the literature , and used to convert invertebrate dry weight composition of C, N, and P to wet weight composition for use in the O. mykiss bioenergetics model. After processing and recovery from anesthesia, all animals were released live at the point of capture.
Quantification of Nutrient Excretion
D. tenebrosus field study.
To estimate nutrient-specific excretion rates and ratios for larval D. tenebrosus, we incubated 18 individuals (SVL range: 56–133 mm) with one of three common diet items , ; terrestrial invertebrates (Orthoptera adults), aquatic invertebrates (Odonata larvae), or aquatic vertebrates (young-of-the-year O. mykiss). Individuals for excretion trials were collected by hand from Fox Cr. and placed in window screen-covered flow-through buckets to prevent the introduction of additional prey. Minimum dietary throughput for D. tenebrosus in this system is estimated to be 60 hours (Munshaw unpublished data), therefore salamanders were incubated with food treatments for 60 hours (±2 hours) to allow excretion rates and ratios to reflect dietary treatments. Food treatments were assigned to control for unknown diet composition that may introduce additional variation into our estimates of excretion rates and ratios. To allow study salamanders to feed ad libitum, we kept each enclosure stocked with an excess of treatment diet items. This reflects the high availability of prey items in the ecosystem, and low occurrence of individuals with empty stomachs . After 60 hours, salamanders were removed from flow-through enclosures, gently rinsed with filtered water to remove adhered particles, and immediately placed in individual 2 L acid-washed containers with 1 L of 0.7 µm-filtered (Whatman GF-F) stream water. Containers were covered with loose plastic lids to prevent addition of airborne particles, and incubated in the stream margins (approx. 5 cm depth) to maintain ambient stream temperature for the duration of the incubation. In addition to incubating salamanders to measure nutrient excretion, we also incubated 3 control containers without salamanders to evaluate the change in nutrient concentrations due to microbial processes, despite our care with filtration. After 120 minutes, approximately 100 mL of water was filtered (0.7 µm, Nalgene 190 syringe filter) to measure ammonium (NH4) and soluble reactive phosphorus (SRP), and animals were removed, weighed, measured as above and released. Water samples were kept dark in coolers with ice packs and were processed within 4 hours. SRP concentrations were determined using spectrophotometry, and ammonium concentrations were determined using fluorometry .
To evaluate the effects of mass and diet group on D. tenebrosus excretion rates, we compared the likelihood of several competing linear regression models using Akaike’s Information Criterion for small sample sizes (AICc). We fit linear models by maximum likelihood including all combinations of mass (log10(g)) and diet treatment (terrestrial invertebrate, aquatic invertebrate, aquatic vertebrate), including a mass by diet treatment interaction and an intercept only model, assuming normally distributed errors. We used the best-supported model by AICc to estimate daily excretion rates (N, P) for each individual from our survey reach and summed across all individuals (n = 348 D. tenebrosus) to estimate the total amount and ratio of N and P recycled by D. tenebrosus in the 1 km reach.
O. mykiss bioenergetics model.
We estimated O. mykiss mass-specific excretion rates using an established bioenergetics model for salmonids . This model was built using a range of body masses that fully encompass those found in our study. Nutrient specific (total dissolved N, total dissolved P) bioenergetics models estimate mass-specific excretion rates (Ei) for each individual (i) in grams per day based on animal mass (Mi), water temperature (T), and N and P content of diet items (N).
Schindler and Eby  determined coefficients (a, b, c, d) for a generalized salmonid bioenergetics model through meta-analysis and literature review including several species of salmonids, and we applied those estimates to our model (N: a = −3.256, 0.084 SE, b = 0.893, 0.021 SE, c = −15.049, 5.526 SE, d = 0.014, 0.003 SE, P: a = −4.776, 0.068 SE, b = 0.902, 0.024 SE, c = 96.801, 19.288 SE, d = 0.008, 0.003 SE). Each time we applied the model to an individual, we selected coefficients from the error distribution associated with each estimate. To parameterize the remaining terms of our O. mykiss bioenergetics model, we surveyed the diet composition of O. mykiss (>50 mm TL) in Fox Cr. as described above. Diets were converted to biomass pooled by order (as described above), and the average C, N, and P composition for each order was estimated based on percent C, N, and P measured from invertebrates collected during a 2008 survey of tributary sites within the SF Eel watershed (J. Hood, unpublished data). Invertebrate C and N composition was determined using an elemental analyzer on dried and homogenized individuals or composited samples of multiple individuals. Samples for P were ashed at 550°C and hydrolyzed with HCl followed by colorimetric determination of PO4 . Multiple values for the same genus were averaged, and values for genera not represented in the survey were summarized from the literature (Table 1 footnote). Diet biomass and nutrient compositions (C, N, and P) were converted to wet-mass using length to wet mass regressions for terrestrial and aquatic diet items separately . We used the combination of the diet composition from surveyed fishes and invertebrate body composition to determine the nitrogen and phosphorus composition of an average steelhead diet, and used these values in the O. mykiss bioenergetics model. Hourly stream temperatures were recorded in 11 pools within the 1.3 km study reach using iButton temperature loggers (Maxim Integrated Products Inc., Sunnyvale, CA) in Fox Cr. for a 7-week period (July-August) in 2010 during the peak season of biological productivity. Average daily temperatures at 11 stream locations were averaged to calculate stream-wide mean temperature (T). We used the average dietary nitrogen and phosphorus values (N, P), along with individual weight (Mi), and average summer stream temperature (T) to estimate the excretion rates of each individual sampled in our survey of Fox Cr. for both N and P. Excretion rates (N, P) for each individual were converted to daily rates (per 24 hr.) and summed across all individuals (n = 527 O. mykiss) to estimate the total amount and ratio of N and P recycled by O. mykiss in the 1 km study reach. In freshwater teleosts, urea and ammonium generally constitute the majority of the total nitrogenous end-products of metabolism , and in O. mykiss specifically, the greatest portion is excreted as ammonium , . To directly compare O. mykiss excretion to D. tenebrosus excretion (NH4), we considered in our results only the portion of total excreted nitrogen by O. mykiss that can be attributed to ammonium (59.3%, 3.99 SE in O. mykiss; ). Ammonium also constitutes the bulk of nitrogenous excretion in aquatic amphibians –. We therefore consider ammonium as a conservative proxy for total N excretion in this study. It is generally established that SRP constitutes the bulk of excreted P, and can therefore be considered an acceptable estimate of total excreted P (e.g. , , ).
Ecosystem Nutrient Recycling
By combining measured excretion rates of D. tenebrosus, with bioenergetics estimates for O. mykiss, we predicted the total amount and ratios of N and P recycled by these two predators across the 1 km study reach. We performed a bootstrap numerical simulation  on the data using R (version 2.12.1) to provide a mean and a 95% confidence interval for the estimate of total excreted nitrogen and phosphorus for each species. In each bootstrap simulation (n = 10,000), individuals of each species from our survey dataset were randomly selected with replacement until the summed biomass of selected individual (Mt-sim) equaled that observed for each species in our field survey of Fox Cr. for the study reach (O. mykiss Mt-obs = 1733 g wet mass, D. tenebrosus Mt-obs = 4542 g wet mass). Mass specific nutrient excretion rates were estimated for each selected individual as above, and summed to estimate the species-specific total N and total P recycled, as well as the N to P ratio. Each individual’s mass-specific excretion rates were estimated including parameter variability, as reported in the literature bioenergetics model in the case of O. mykiss, and from our empirically derived regressions in the case of D. tenebrosus. To assess the importance of the relative abundance of the two predators for nutrient recycling, we also ran simulations that varied the proportion of the predator biomass made up by salamanders and fish, assuming that the combined predator biomass we observed in Fox Cr. is a conservative estimate of the maximum sustainable regardless of predator identity. As such, we varied the predator composition from 100% O. mykiss to 100% D. tenebrosus biomass in 10% increments (n = 1,000 simulations for each increment).
To evaluate the magnitude of nutrient recycling by top predators relative to ecosystem demand, we calculated rates of community nutrient uptake in the study reach. Areal uptake rates were calculated based on Schade et al.  to provide estimates of nutrient demand during the summer growth period. Uptake rate (U: µg⋅m−2 s−1) can be calculated by multiplying discharge rate (Q: Ls−1) by nutrient concentration (C: µg⋅L−1), and dividing by reach width (w: m) and nutrient-specific spiral length (Sw: m).
The model was parameterized using measurements from 32 replicated pools (w, Q) measurements combined with data presented by Schade et al. (SwN: 200,SwP: 540, CN: 10.1, CP: 15) . Aerial uptake rates (U) can be used to estimate the proportion of nutrient demand supplied by O. mykiss and D. tenebrosus excretion in the study reach by difference. These estimates consider the largest single component of dissolved inorganic nitrogen (i.e. NH4) and total dissolved phosphorus (i.e. SRP), and so can thus be considered a conservative estimate of uptake demand. Proportion of demand supplied by O. mykiss and D. tenebrosus recycling was determined by dividing areal recycling rate (total recycling rate divided by total study area) by uptake rate (U).
During the field survey of the study area in Fox Cr., we captured 356 D. tenebrosus and 595 O. mykiss, at an average density of 1.05 and 1.76 individuals/m2 respectively (Table 2). Despite lower numerical abundance, the total biomass of D. tenebrosus was 2.5 times higher than O. mykiss (13.7 and 5.4 g wet mass/m2 respectively; Table 2). While the length range of captured individuals of each species was quite similar (Fig. 1, left panels), mean D. tenebrosus body mass was nearly 6 times higher (Fig. 1, right panels; D. tenebrosus = 13.05 g wet mass, 0.79 SE, O. mykiss = 3.28 g wet mass, 0.37 SE).
Size- and mass-frequency distributions for O. mykiss and D. tenebrosus from a 1 km study reach of Fox Creek. Standard length (mm) was used for fish (n = 528) and snout-vent length (mm) for salamanders (n = 348) to exclude the size variability generated by tail injuries. Note differences in y-axis scales.
Our summary of invertebrate C, N, and P composition and predator diet composition showed that both species of predator had diets with a high degree of taxonomic overlap and similar elemental composition. Both O. mykiss and D. tenebrosus diets were dominated by Order Ephemeroptera (>25% by mass; Table 1). Carbon (min = 34.6%, max = 53.6% dry mass), N (min = 5.8%, max = 12.7% dry mass), and P (min = 0.3%, max = 1.5% dry mass) content varied substantially among orders (Table 1), and in total our analysis accounted for 90% of D. tenebrosus and 67% of O. mykiss diet items by mass (Table 1). The majority of unaccounted-for diet was composed of unidentifiable remains. We found that on average, D. tenebrosus diets were 8.5% C, 1.9% N, and 0.2% P by wet mass, and O. mykiss diets were composed of 7.9% C, 1.6% N, and 0.2% P by wet mass (Table 1). Only a small percentage of the salamanders surveyed had empty stomachs (5%), which is similar to the 1% occurrence of empty stomachs reported by Parker , which served as the basis for our decision to feed excretion trial salamanders ad libitum.
Quantification of Excreted Nutrients
D. tenebrosus field study.
Control containers showed no significant change in nutrient concentration over time (N: t = 0.56, df = 2, p-value = 0.63, P: t = 0.10, df = 2, p-value = 0.93) indicating that any changes we observed in trials containing salamanders were attributable to animal excretion, and not microbial processes. Model selection by AICc of N and P excretion rates for D. tenebrosus from our experiment identified clear top models for both nutrients, with ΔAICc scores of all other models >3 (Table S2). The best-supported model for D. tenebrosus P (SRP) excretion rate included both intercept and mass (log10[µgP·min−1] = −3.12+1.60·log10[mass], SEintercept = 0.73, SEmass coefficient = 0.56; Fig. 2), as did the best-supported model for N (NH4) excretion rate (log10[µgN·min−1] = −2.04+1.41·log10[mass], SEintercept = 0.24, SEmass coefficient = 0.18; Fig. 2). We found little support for models that included diet treatment (terrestrial invertebrate, aquatic invertebrate, and aquatic vertebrate), where all models including diet had ΔAIC ≥3. Consequently, dietary treatment was disregarded for D. tenebrosus in subsequent analyses.
Nitrogen (NH4) and phosphorus (SRP) nutrient excretion rates (ug·min−1) of D. tenebrosus. Lines represent the fit of the top model selected by AICc for P (log10[µgP·min−1] = −3.12+1.60(log10[mass]), r2 = 0.31, P = 0.01), and N excretion rates (log10[µgN·min−1] = −2.04+1.41(log10[mass]), r2 = 0.79, P<<0.001).
O. mykiss bioenergetics model.
Results from the diet survey and the invertebrate body C, N, and P survey were used to establish the N and P composition of an average steelhead diet (Table S3; N = 1.8%, P = 0.2% wet weight). Diet information was combined with average stream temperature (14.2°C, 0.004 SE, min = 12.0°C, max = 16.5°C) in our bioenergetics model. Using these inputs, rate of excretion of N and P was calculated for each individual and used for the ecosystem analysis.
Ecosystem Nutrient Recycling
We estimated the daily N and P recycled by individual D. tenebrosus and O. mykiss, by species, and by both species combined within the study reach. We estimated that O. mykiss recycled nearly twice as much N per day (0.42 g·day−1±0.028 95% CI) as that recycled by D. tenebrosus (0.25 g·day−1±0.017 95% CI, Fig. 3) within the study reach. The phosphorus recycled by O. mykiss (0.049 g·day−1±0.0021 95% CI) was approximately 19% greater than that recycled by D. tenebrosus (0.041 g·day−1±0.0039 95% CI). We found that when estimates for both species were summed, 0.67 g·day−1 of N was recycled (±0.032 95% CI, Fig. 3A), and 0.089 g day−1 of P (±0.0044 95% CI). The N:P ratio of nutrients recycled by O. mykiss (8.7±0.6 95% CI, Fig. 4) was higher than that of D. tenebrosus (6.0±0.2 95% CI) and the N:P ratio of both species combined (7.5±0.3 95% CI) was intermediate between the two.
Estimated total daily excreted N (NH4) and P (SRP) by O. mykiss (filled), D. tenebrosus (grey), and both predators combined (open) within the Fox Cr. study reach. Bars represent mean ±95%CI. Note the log scaled y-axis.
Estimates of the ratio of excreted N:P for O. mykiss (filled), D. tenebrosus (grey), and both predators combined (open). Bars represent mean ±95%CI.
We found that varying the relative abundance of the two vertebrate predators (assuming constant total biomass) led to large differences in total excreted nutrients (N and P). When we simulated the predator biomass as being entirely comprised of D. tenebrosus, N and P excretion rates were estimated to be 0.34 and 0.06 g·day−1 respectively (Fig. 5). These rates increased along the relative abundance continuum to a maximum N and P excretion rate of 1.53 and 0.18 g·day−1 respectively when O. mykiss comprised all the predator biomass. The overall N:P ratio of predator-recycled nutrients ranged from 6.04 when all biomass was D. tenebrosus, to 8.72 when all biomass was O. mykiss (Fig. 5).
Estimated total recycled N (NH4) and P (SRP) excretion (g·day−1) in Fox Cr. due to simulated changes in the relative abundance (by biomass) of O. mykiss and D. tenebrosus. Simulations assumed a fixed total biomass of predators (6275 g) within the study reach, and estimated total excretion rates (left y-axis) and ratios (right y-axis) by bootstrapped re-sampling of surveyed individuals. Predator relative abundance (x-axis) varies by 10% increments from 100% O. mykiss composition to 100% D. tenebrosus composition, expressed as the proportion of predator biomass (salamander:fish). Vertical line indicates the observed ratio of predators in Fox Cr.
We calculated that uptake rates (U) for N and P were 0.1 µg·m−2 s−1 and 0.04 µg·m−2 s−1 respectively. At the observed densities of D. tenebrosus and O. mykiss, we estimated that nutrient recycling by predators in the study reach represented 20.3% of N (0.0023 µg·m−2 s−1±1.1*10−4 95% CI) and 7.3% of P (3.1*10−4 µg·m−2 s−1±1.5*10−5 95% CI) demand by producers in the study reach.
This study suggests that differences in predator in abundance, body size distribution, and species identity directly influence nutrient recycling rates and ratios in headwater streams. These findings emphasize the importance of species identity in predator-driven nutrient recycling. As such, we found that biomass alone is not an adequate proxy for estimating species contributions to the total rate of nutrient recycling, which depends on physical, physiological, and ecological traits of different members of the predator guild. Despite more than two and a half times higher biomass, the total amount of nutrients recycled by D. tenebrosus in Fox Cr. was only a fraction (58% N, 84% P) of that recycled by O. mykiss. Despite a lower relative contribution, the D. tenebrosus population in Fox Cr. did provide 37% of the total N and 46% of the total P recycled by vertebrate predators, providing support for the importance of vertebrates other than fish in stream nutrient recycling. We also found that the relative abundance of salamander and fish biomass altered the total amount of nutrients recycled by top predators per day. Our simulations suggest that when predator biomass is comprised completely of D. tenebrosus, the amount of recycled N and P are only approximately half (51% N, 63% P relative to background) that recycled by the natural assemblage of salamanders and fish in Fox Cr. When the total predator biomass in our simulation was comprised entirely of O. mykiss, both N and P recycling rates were estimated to be approximately twice as high as background (2.3×N, 2.0×P).
The large differences we found in estimated stream-wide excretion rates between O. mykiss and D. tenebrosus are in contrast to the many ecological similarities we found between the species. As ambush predators, D. tenebrosus feed mostly on aquatic benthic prey , whereas O. mykiss feed throughout the water column, and consume a larger proportion of allochthonous (e.g. terrestrial) prey. While each predator feeds in different stream microhabitats, they have a high degree of overlap in diet composition (Table 1, also see ), with similar elemental composition. Our data suggest that dietary or stoichiometric differences in prey are not likely to explain the observed differences in species-specific N:P ratios (Table S3). Similarly, data from a complementary study found that growth rates of D. tenebrosus and O. mykiss in the same study reach did not differ significantly over the summer growth period (Atlas et al., unpublished data), indicating that differences we identified in nutrient excretion rates are likely driven by differences in metabolic requirements. At zero activity level, O. mykiss is estimated to have a 50% higher metabolic demand as measured by oxygen consumption  than an average salamander of the same weight . D. tenebrosus is a sit-and-wait predator and is only marginally active at night and even less so during the day, whereas O. mykiss is an active water-column predator and expends large amounts of energy for locomotion and foraging when not resting or seeking cover in interstitial spaces . Larger expenditures of energy by O. mykiss likely necessitate higher metabolic rates and faster production of metabolic wastes, which is excreted at a higher mass-specific rate than by D. tenebrosus. An additional factor that likely contributes to the difference in excretion ratios is species specific elemental requirements for growth and maintenance. Stoichiometric theory suggests that as the required N:P ratio of an organism increases at any given ingested N:P ratio, the ratio of excreted nutrients will decrease . Though we do not have elemental composition data for D. tenebrosus from our study system, values from the literature show that eastern salamanders that fill a similar ecological niche have organismal N:P ratios of 2.4–4.8 , whereas juvenile steelhead in our study system had an average N:P ratio of 9.8. However, our findings that O. mykiss recycles N:P at a higher ratio than D. tenebrosus suggests that the body N:P ratio of D. tenebrosus may be much higher than eastern species.
Metabolic scaling theory predicts that as mass increases, mass specific excretion rate should decrease , which may further explain the disparity in total excreted nutrients between predator species. The higher numbers but lower biomass of O. mykiss as compared to D. tenebrosus (Table 2), resulted from a smaller average size of O. mykiss individuals in our study system (mean O. mykiss = 3.3 g wet mass, mean D. tenebrosus = 13.1 g wet mass). The size-frequency distribution of O. mykiss (Fig. 1), illustrates two distinct nodes representing young of the year (age 0) and one-year or older age-classes that make up the vast majority of O. mykiss individuals. This distribution is typical of populations consisting of primarily anadromous individuals that migrate out of freshwaters after approximately 1.5 years (summarized by ), and is also indicative of the consistent growth pattern exhibited by most salmonids –. By contrast, there is a near continuous distribution of D. tenebrosus body size, with no distinct nodes and a higher frequency of larger individuals than O. mykiss (Fig. 1). Coastal giant salamanders can achieve life spans in excess of 20 years, and are characterized by slow growth, leading to difficulty in assigning age based on body size. Ultimately, the larger average mass achieved by D. tenebrosus compared to O. mykiss is likely to contribute to the observed lower mass-specific excretion rate and lower total species-specific rate of nutrient recycling.
Our test of the influence of diet on excretion ratios in D. tenebrosus showed that despite the differing N:P ratios of the items fed to captive individuals (aquatic invertebrate = 12.7, terrestrial invertebrate = 13.4, aquatic invertebrate = 9.9), we found no support for differences in N or P excretion rates of D. tenebrosus fed these prey. Our results were likely inconsistent with our stoichiometric prediction for several reasons. Despite holding salamanders with food items for a full iteration of their estimated digestive throughput time (60 hours), individual excretion rates may not have had time to equilibrate to reflect the homogenous experimental diets. Additional factors may have included a wide range of body sizes (6.3–72.8 mm SVL), individuals at different stages of reproductive maturity , and small sample size (n = 18).Our findings demonstrate that vertebrate excretion can affect nutrient availability in light-limited forested streams. However, the repercussions of altering nutrient availability depend on the nutrient requirements of producers in the system. By increasing the availability of limiting nutrients, algal growth and elemental composition can be altered. Using nutrient addition experiments, Schade et al.  demonstrate that Fox Cr. is N-limited, and has an external dissolved inorganic NH4-N:SRP ratio of 0.67. The ratio of recycled nutrients from predators examined in this study is modestly higher than ambient levels (7.5; Fig. 3), suggesting the potential for partial alleviation of the N-limitation experienced by autotrophs in Fox Cr. Due to the higher ambient concentrations of inorganic nutrients in Fox Cr. compared to other similar tributary streams , it is likely that the contribution by predators to nutrient availability in other streams that support similar predator densities is greater than in Fox Creek. This suggests that our estimates of the percent of nutrient demand supplied by vertebrates in Fox Cr. may be lower than other streams in the region.
River networks are characterized by strong downstream gradients in light, temperature, disturbance frequency, and physical channel attributes, each of which can differentially affect the movement, growth, and persistence of riverine species , . Despite having highly overlapping regional distributions, O. mykiss and D. tenebrosus vary in their relative abundance across the heterogeneity present in river networks at local-scales . Our simulation results suggest that such differences, combined with spatial and temporal variation in ecosystem nutrient demand, are likely to affect broad-scale patterns of nutrient limitation. While the study system we worked in is within a protected area, and has been well buffered from local anthropogenic impacts, the relative abundance of these two species across broader watersheds and larger coastal regions is likely affected by species-specific responses to anthropogenic stressors in the past as well as the future. In the past, steelhead abundance was very high along the entire west coast of North America, but overexploitation, especially along the California coastline, has drastically diminished abundance . Future thermally limiting events may also affect abundances. Salmonids are known to have much lower critical thermal maxima, both as juveniles (26.2–27.9°C; ) and older classes (28.25–29.85; ), than many salamanders (30.1–37.3°C; ). As summer maximum temperatures increase with future climate change and continued land-cover change (e.g. forest clearing), amphibian populations may be able to persist in thermal environments that cause acute mortality or exclude salmonid populations. Our study suggests that in such cases, large salamander populations may act to partially buffer the impacts of declining fish populations via nutrient recycling. By not considering amphibians and other animals as nutrient recyclers, we may be missing critical components of nutrient dynamics in stream ecosystems.
This study provides an estimate of excretion rates for a dominant stream predator in the Pacific Northwest that has not been previously studied. The bulk of literature in the field has examined highly productive tropical streams or lentic ecosystems (e.g. , , , , ). We show here that nutrient recycling by top predators is an important component of ecosystem nutrient dynamics even in small temperate streams. Though amphibians have been previously considered, their importance as nutrient recyclers has been discounted due to low densities compared to other vertebrate taxa (but see , ). Salamanders in many stream ecosystems have strong top-down effects on prey abundance –, and act as sources of energy for higher level predators , , . This study demonstrates that amphibians can also contribute to ecosystem-level dynamics through recycling of nutrients, in this case accounting for 37% of N and 46% of P recycled by vertebrates. Our findings suggest that studies of consumer-driven nutrient recycling will continue to benefit by identifying additional key recyclers in various systems. Continued work in this field will help expand our knowledge of the ecological roles of many organisms, and foster a better understanding of nutrient dynamics in aquatic systems.
Sources for invertebrate length-weight regression used to determine the mass of individual insects for the diet survey.
AICc scores for models predicting N and P excretion rates.
Raw excretion data from all Dicamptodon tenebrosus individuals used in the excretion study. Excretion rates were measured at t = 0 min and t = 120 min. Diet treatment codes are as follows: AM: Aquatic macroinvertebrate, YOY: young-of-the-year steelhead, TM: Terrestrial macroinvertebrate.
We thank Peter Steel and Mary Power for their tireless help and support at the University of California’s Angelo Coast Range Reserve, Grace Wilkinson, Jill Welter, and Sandra Brovold for their assistance with analytical chemistry, Zachary Monteith and Will Atlas for help wrangling salamanders, Camille McNeely for assistance with invertebrate taxonomy, John Schade for help designing the experiment, and Jim Hood for his contribution of unpublished elemental invertebrate data.
Conceived and designed the experiments: RGM WJP. Performed the experiments: RGM WJP DMC JCF. Analyzed the data: RGM WJP. Wrote the paper: RGM WJP.
- 1. Schindler DW (1978) Factors regulating phytoplankton production and standing crop in the world’s freshwaters. Limnology and Oceanography: 478–486.
- 2. Morris DP, Lewis WM (1988) Phytoplankton nutrient limitation in Colorado mountain lakes. Freshwater Biology 20: 315–327 .
- 3. Hecky RE, Kilham P (1988) Nutrient Limitation of Phytoplankton in Freshwater and Marine Environments: A Review of Recent Evidence on the Effects of Enrichment. Limnology and Oceanography 33: 796–822.
- 4. Elser JJ, Bracken MES, Cleland EE, Gruner DS, Harpole WS, et al. (2007) Global analysis of nitrogen and phosphorus limitation of primary producers in freshwater, marine and terrestrial ecosystems. Ecology Letters 10: 1135–1142 .
- 5. Finlay JC, Hood JM, Limm MP, Power ME, Schade JD, et al. (2011) Light-mediated thresholds in stream-water nutrient composition in a river network. Ecology 92: 140–150 .
- 6. Sanches LF, Guariento RD, Caliman A, Bozelli RL, Esteves FA (2011) Effects of nutrients and light on epiphytic biomass and nutrient stoichiometry in a tropical black-water aquatic ecosystem. Hydrobiologia 669: 35–44 .
- 7. Veraart AJ, Romaní AM, Tornes E, Sabater S (2008) Algal response to nutrient enrichment in forested oligotrophic stream. Journal of Phycology 44: 564–572 .
- 8. Bowman MF, Chambers PA, Schindler DW (2005) Changes in stoichiometric constraints on epilithon and benthic macroinvertebrates in response to slight nutrient enrichment of mountain rivers. Freshw Biol 50: 1836–1852 .
- 9. Bormann FH, Likens GE (1967) Nutrient cycling. Science 155: 424 .
- 10. Filippelli GM (2002) The global phosphorus cycle. Reviews in mineralogy and geochemistry 48: 391–425 .
- 11. Vanni MJ (2002) Nutrient cycling by animals in freshwater ecosystems. Annual Review of Ecology and Systematics: 341–370. https://doi.org/10.1146/annurev.ecolysis.33.010802.150519.
- 12. Vanni MJ, Layne CD (1997) Nutrient recycling and herbivory as mechanisms in the “top-down” effect of fish on algae in lakes. Ecology 78: 21–40 .
- 13. Schindler DE, Knapp RA, Leavitt PR (2001) Alteration of Nutrient Cycles and Algal Production Resulting from Fish Introductions into Mountain Lakes. Ecosystems 4: 308–321 .
- 14. Ngai JT, Srivastava DS (2006) Predators accelerate nutrient cycling in a bromeliad ecosystem. Science 314: 963–963 .
- 15. McIntyre PB, Flecker AS, Vanni MJ, Hood JM, Taylor BW, et al. (2008) Fish distributions and nutrient cycling in streams: can fish create biogeochemical hotspots. Ecology 89: 2335–2346 .
- 16. Schindler DW (1974) Eutrophication and recovery in experimental lakes: implications for lake management. Science 184: 897 .
- 17. Azad HS, Borchardt JA (1970) Variations in phosphorus uptake by algae. Environmental Science & Technology 4: 737–743 .
- 18. Feminella JW, Hawkins CP (1995) Interactions between stream herbivores and periphyton: a quantitative analysis of past experiments. Journal of the North American Benthological Society: 465–509. https://doi.org/10.2307/1467536.
- 19. Power ME, Matthews WJ, Stewart AJ (1985) Grazing minnows, piscivorous bass, and stream algae: dynamics of a strong interaction. Ecology 66: 1448–1456 .
- 20. Attayde JL, Hansson LA (1999) Effects of nutrient recycling by zooplankton and fish on phytoplankton communities. Oecologia 121: 47–54 .
- 21. Hillebrand H (2005) Light regime and consumer control of autotrophic biomass. Journal of Ecology 93: 758–769 .
- 22. Leroux SJ, Loreau M (2010) Consumer-mediated recycling and cascading trophic interactions. Ecology 91: 2162–2171 .
- 23. Mather ME, Vanni MJ, Wissing TE, Davis SA, Schaus MH (1995) Regeneration of nitrogen and phosphorus by bluegill and gizzard shad: effect of feeding history. Canadian Journal of Fisheries and Aquatic Sciences 52: 2327–2338 .
- 24. Schaus MH, Vanni MJ, Wissing TE, Bremigan MT, Garvey JE, et al. (1997) Nitrogen and Phosphorus Excretion by Detritivorous Gizzard Shad in a Reservoir Ecosystem. Limnology and Oceanography 42: 1386–1397.
- 25. McIntyre PB, Jones LE, Flecker AS, Vanni MJ (2007) Fish extinctions alter nutrient recycling in tropical freshwaters. Proceedings of the National Academy of Sciences 104: 4461 .
- 26. Knoll LB, McIntyre PB, Vanni MJ, Flecker AS (2009) Feedbacks of consumer nutrient recycling on producer biomass and stoichiometry: separating direct and indirect effects. Oikos 118: 1732–1742 .
- 27. Villéger S, Grenouillet G, Suc V, Brosse S (2012) Intra- and interspecific differences in nutrient recycling by European freshwater fish. Freshwater Biology 57: 2330–2341 .
- 28. Taylor JM, Jason M (2011) Nutrient enrichment effects on stream periphyton stoichiometry, algal and fish assemblage structure, and grazing fish-periphyton interactions. Available: https://beardocs.baylor.edu:8443/xmlui/handle/2104/8284. Accessed 9 January 2013.
- 29. Vanni MJ, Flecker AS, Hood JM, Headworth JL (2002) Stoichiometry of nutrient recycling by vertebrates in a tropical stream: linking species identity and ecosystem processes. Ecology Letters 5: 285–293 .
- 30. Wilson HF, Xenopoulos MA (2011) Nutrient recycling by fish in streams along a gradient of agricultural land use. Global Change Biology 17: 130–139 .
- 31. McManamay RA, Webster JR, Valett HM, Dolloff CA (2010) Does diet influence consumer nutrient cycling? Macroinvertebrate and fish excretion in streams. The Society for Freshwater Science. Available: http://www.jnabs.org/doi/abs/10.1899/09-152.1. Accessed 9 January 2013.
- 32. Atkinson CL, Opsahl SP, Covich AP, Golladay SW, Conner LM (2010) Stable isotopic signatures, tissue stoichiometry, and nutrient cycling (C and N) of native and invasive freshwater bivalves. The Society for Freshwater Science. Available: http://www.jnabs.org/doi/abs/10.1899/09-083.1. Accessed 9 January 2013.
- 33. Moslemi JM, Snider SB, MacNeill K, Gilliam JF, Flecker AS (2012) Impacts of an Invasive Snail (Tarebia granifera) on Nutrient Cycling in Tropical Streams: The Role of Riparian Deforestation in Trinidad, West Indies. PLoS ONE 7: e38806 .
- 34. Liess A, Kahlert M (2009) Gastropod grazers affect periphyton nutrient stoichiometry by changing benthic algal taxonomy and through differential nutrient uptake. The Society for Freshwater Science. Available: http://www.jnabs.org/doi/abs/10.1899/08-115.1. Accessed 9 January 2013.
- 35. Whiles MR, Hall RO, Dodds WK, Verburg P, Huryn AD, et al.. (2012) Disease-Driven Amphibian Declines Alter Ecosystem Processes in a Tropical Stream. Ecosystems. Available: http://link.springer.com/content/pdf/10.1007%2Fs10021-012-9602-7. Accessed 9 January 2013.
- 36. Burton TM, Likens GE (1975) Energy flow and nutrient cycling in salamander populations in the Hubbard Brook Experimental Forest, New Hampshire. Ecology: 1068–1080.
- 37. Davic RD, Orr LP (1987) The relationship between rock density and salamander density in a mountain stream. Herpetologica: 357–361.
- 38. Corn PS, Bury BR (1989) Logging in western Oregon: responses of headwater habitats and stream amphibians. Forest Ecology and Management 29: 39–57 .
- 39. Parker MS (1991) Relationship between cover availability and larval Pacific giant salamander density. Journal of Herpetology 25: 355–357 .
- 40. Hawkins CP, Murphy ML, Anderson NH, Wilzbach MA (1983) Density of fish and salamanders in relation to riparian canopy and physical habitat in streams of the northwestern United States. Canadian Journal of Fisheries and Aquatic Sciences 40: 1173–1185.
- 41. Lau MR (1984) Habitat utilization, density, and growth of steelhead trout, coho salmon, and pacific giant salamander in relation to habitat types in a small coastal redwood stream [MSc Thesis]. Sacramento: California State University.
- 42. Davic RD, Welsh HH Jr (2004) On the ecological roles of salamanders. Annual Review of Ecology, Evolution, and Systematics: 405–434. https://doi.org/10.1146/annurev.ecolsys.35.112202.130116.
- 43. Good DA (1989) Hybridization and cryptic species in Dicamptodon (Caudata: Dicamptodontidae). Evolution: 728–744. https://doi.org/10.2307/2409302.
- 44. Antonelli AL, Nassbaum RA, Smith SD (1972) Comparative food habits of four species of stream-dwelling vertebrates (Dicamptodon ensatus, D. copei, Cottus tenuis, Salmo gairdner). Northwest Sci 46: 277–289.
- 45. Parker MS (1994) Feeding ecology of stream-dwelling Pacific giant salamander larvae (Dicamptodon tenebrosus). Copeia: 705–718.
- 46. Kraft CE (1992) Estimates of Phosphorus and Nitrogen Cycling by Fish Using a Bioenergetics Approach. Canadian Journal of Fisheries and Aquatic Sciences 49: 2596–2604 .
- 47. He X, Kitchell JF, Carpenter SR, Hodgson JR, Schindler DE, et al. (1993) Food Web Structure and Long-Term Phosphorus Recycling: A Simulation Model Evaluation. Transactions of the American Fisheries Society 122: 773–783 .
- 48. Schindler DE, Kitchell JF, He X, Carpenter SR, Hodgson JR, et al. (1993) Food Web Structure and Phosphorus Cycling in Lakes. Transactions of the American Fisheries Society 122: 756–772 .
- 49. Hood J, Vanni M, Flecker A (2005) Nutrient recycling by two phosphorus-rich grazing catfish: the potential for phosphorus-limitation of fish growth. Oecologia 146: 247–257 .
- 50. Whitledge GW, Hayward RS (1997) Laboratory Evaluation of a Bioenergetics Model for Largemouth Bass at Two Temperatures and Feeding Levels. Transactions of the American Fisheries Society 126: 1030–1035 .
- 51. Whiles MR, Huryn AD, Taylor BW, Reeve JD (2009) Influence of handling stress and fasting on estimates of ammonium excretion by tadpoles and fish: recommendations for designing excretion experiments. Limnology and Oceanography: Methods 7: 1–7 .
- 52. Nussbaum RA, Clothier GW (1973) Population structure, growth, and size of larval Dicamptodon ensatus (Eschscholtz). Northwest Science 47: 218–227.
- 53. Nussbaum RA, Brodie ED Jr, Storm RM (1983) Amphibians and reptiles of the Pacific Northwest. Idaho: University Press Idaho.
- 54. Shapovalov L, Taft AC (1954) The Life Histories of the Steelhead Rainbow Trout (Salmo Gairdneri Gairdneri) and Silver Salmon (Oncorhynchus Kisutch): With Special Reference to Waddell Creek, California, and Recommendations Regarding Their Management. California Department of Fish and Game.
- 55. Nehlsen W (1997) Coastal salmon status and trends – A coastwide perspective. Pacific Salmon and Their Ecosystem: Status and Future Options. New York: Chapman and Hall.
- 56. Hartwell H, Ollivier LM, Welsh Jr HH (1998) Stream amphibians as indicators of ecosystem stress: a case study from California’s redwoods. Ecological Applications 8: 1118–1132 .
- 57. Roni P (2002) Habitat use by fishes and Pacific giant salamanders in small western Oregon and Washington streams. Transactions of the American Fisheries Society 131: 743–761 .
- 58. Welsh HH Jr, Lind AJ (2002) Multiscale habitat relationships of stream amphibians in the Klamath-Siskiyou region of California and Oregon. The Journal of wildlife management: 581–602. https://doi.org/10.2307/3803126.
- 59. Benke AC (1993) Concepts and patterns of invertebrate production in running water. Proceedings of the International Association of Theoretical and Applied Limnology 25: 15.
- 60. Parker MS (1993) Predation by Pacific giant salamander larvae on juvenile steelhead trout. Northwestern Naturalist 74: 77–81.
- 61. Schade JD, MacNeill K, Thomas SA, McNeely CF, Welter JR, et al. (2011) The stoichiometry of nitrogen and phosphorus spiralling in heterotrophic and autotrophic streams. Freshwater Biology 56: 424–436 .
- 62. Schindler DE, Eby LA (1997) Stoichiometry of fishes and their prey: implications for nutrient recycling. Ecology 78: 1816–1831 .
- 63. DeMott WR, Gulati RD, Siewertsen K (1998) Effects of phosphorus-deficient diets on the carbon and phosphorus balance of Daphnia magna. Limnology and Oceanography: 1147–1161.
- 64. Sabo JL, Bastow JL, Power ME (2002) Length-mass relationships for adult aquatic and terrestrial invertebrates in a California watershed. Journal of the North American Benthological Society 21: 336–343 .
- 65. Smith HW (1929) The excretion of ammonia and urea by the gills of fish. Journal of Biological Chemistry 81: 727.
- 66. Evans DH (1999) The Physiology of Fishes. Florida: CRC Press.
- 67. Kajimura M, Croke SJ, Glover CN, Wood CM (2004) Dogmas and controversies in the handling of nitrogenous wastes: the effect of feeding and fasting on the excretion of ammonia, urea and other nitrogenous waste products in rainbow trout. Journal of experimental biology 207: 1993–2002 .
- 68. Munro AF (1953) The ammonia and urea excretion of different species of Amphibia during their development and metamorphosis. Biochem J 54: 29–36.
- 69. Underhay EE, Baldwin E (1955) Nitrogen excretion in the tadpoles of Xenopus laevis Daudin. Biochem J 61: 544–547.
- 70. Wright PM, Wright PA (1996) Nitrogen Metabolism and Excretion in Bullfrog (Rana catesbelana) Tadpoles and Adults Exposed to Elevated Environmental Ammonia Levels. Physiological Zoology 69: 1057–1078.
- 71. Small GE, Pringle CM, Pyron M, Duff JH (2011) Role of the fish Astyanax aeneus (Characidae) as a keystone nutrient recycler in low-nutrient Neotropical streams. Ecology 92: 386–397 .
- 72. Efron B, Tibshirani RJ (1993) An Introduction to the Bootstrap. New York: Chapman and Hall.
- 73. Huey RB, Pianka ER (1981) Ecological consequences of foraging mode. Ecology: 991–999. https://doi.org/10.2307/1936998.
- 74. Whitford WG, Hutchison VH (1967) Body size and metabolic rate in salamanders. Physiological zoology 40: 127–133.
- 75. Rao GMM (1968) Oxygen consumption of rainbow trout (Salmo gairdneri) in relation to activity and salinity. Canadian Journal of Zoology 46: 781–786 .
- 76. Elser JJ, Urabe J (1999) The stoichiometry of consumer-driven nutrient recycling: theory, observations, and consequences. Ecology 80: 735–751 .
- 77. Brown JH, Gillooly JF, Allen AP, Savage VM, West GB (2004) Toward a metabolic theory of ecology. Ecology 85: 1771–1789 .
- 78. Griffiths D (1994) The size structure of lacustrine Arctic charr (Pisces: Salmonidae) populations. Biological Journal of the Linnean Society 51: 337–357 .
- 79. Thorpe JE, Talbot C, Villarreal C (1982) Bimodality of growth and smolting in Atlantic salmon, Salmo salar L. Aquaculture. 28: 123–132 .
- 80. Ewing RD, Ewing GS (2002) Bimodal length distributions of cultured chinook salmon and the relationship of length modes to adult survival. Aquaculture 209: 139–155 .
- 81. Fillatre EK, Etherton P, Heath DD (2003) Bimodal run distribution in a northern population of sockeye salmon (Oncorhynchus nerka): life history and genetic analysis on a temporal scale. Mol Ecol 12: 1793–1805.
- 82. Fitzpatrick LC (1973) Energy Allocation in the Allegheny Mountain Salamander, Desmognathus Ochrophaeus. Ecological Monographs 43: 43–58 .
- 83. Vannote RL, Minshall GW, Cummins KW, Sedell JR, Cushing CE (1980) The river continuum concept. Canadian journal of fisheries and aquatic sciences 37: 130–137.
- 84. Power ME, Dietrich WE (2002) Food webs in river networks. Ecological Research 17: 451–471 .
- 85. Busby PJ, Wainwright TC, Bryant GJ, Lierheimer LJ, Waples RS, et al.. (1996) Status review of west coast steelhead from Washington, Idaho, Oregon, and California. US Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Northwest Fisheries Science Center, Coastal Zone and Estuarine Studies Division. Available: http://www.nwfsc.noaa.gov/publications/techmemos/tm27/tm27.htm. Accessed 12 October 2012.
- 86. Grande M, Andersen S (1991) Critical thermal maxima for young salmonids. Journal of Freshwater Ecology 6: 275–279 .
- 87. Lee RM, Rinne JN (1980) Critical thermal maxima of five trout species in the southwestern United States. Transactions of the American Fisheries Society 109: 632–635 .
- 88. Zweifel RG (1957) Studies on the critical thermal maxima of salamanders. Ecology 38: 64–69 .
- 89. Spooner DE, Vaughn CC (2008) A trait-based approach to species’ roles in stream ecosystems: climate change, community structure, and material cycling. Oecologia 158: 307–317 .
- 90. Beard KH, Vogt KA, Kulmatiski A (2002) Top-down effects of a terrestrial frog on forest nutrient dynamics. Oecologia 133: 583–593 .
- 91. Davic RD (1983) An investigation of salamander guild predation in a North Carolina stream: an experimental approach [PhD Thesis]. Kent State University.
- 92. Parker MS (1992) Feeding ecology of the larvae of the Pacific giant salamander (Dicamptodon tenebrosus) and their role as top predator in a headwater stream benthic community [PhD Thesis]. Davis: University of California.
- 93. Broadman R, Ogger J, Kolaczyk M, Pulver RA, Long AJ, et al. (2003) Mosquito control by pond-breeding salamander larvae. Herpetological Review 34: 116–118.
- 94. Feder ME (1983) Integrating the ecology and physiology of plethodontid salamanders. Herpetologica: 291–310.
- 95. Pough FH (1983) Amphibians and reptiles as low-energy systems. Behavioral energetics: the cost of survival in vertebrates. 141–188.