Freshwater Biology (2008) 53, 509–520 doi:10.1111/j.1365-2427.2007.01916.x Does the Redfield ratio infer nutrient limitation in the macroalga Spirogyra fluviatilis? S . A . T O W N S E N D * , J . H . S C H U L T †, M . M . D O U G L A S † A N D S . S K I N N E R ‡ *Department of Natural Resources, Environment and the Arts, Darwin, Northern Territory, Australia † Charles Darwin University, Darwin, Northern Territory, Australia ‡ Royal Botanic Gardens of Sydney, Sydney, Australia SUMMARY 1. The cellular nutrient contents of microalgae, when growing at or approaching maximum rates, approximate the Redfield C : N : P (molar) ratio of 106 : 16 : 1. Deviations from this optimal ratio can be used to infer nutrient limitation of microalgal growth. However, this ratio may not be applicable to macroalgae, which are distinguished from microalgae by forming a thallus that is a discrete structure visible to the naked eye. The utility of the Redfield ratio to infer nutrient limitation of the growth of macroalgae was tested for Spirogyra fluviatilis in a field experiment conducted in tropical Australia. 2. The optimal cellular C : N : P ratio for S. fluvialitis was estimated by means of in situ nutrient addition. This was compared with S. fluvialitis cellular ratios determined from eight sites with a wide range of soluble N concentrations (<1–90 lg L)1), a smaller range of soluble P concentrations (5–12 lg L)1), and soluble molar N : P ratios of 0.11– 27. 3. Spirogyra fluviatilis had an optimal molar C : N : P ratio of 1800 : 87 : 1 which differs substantially from the Redfield ratio, and suggests that the latter ratio is not applicable to this macroalga. Concentrations of N and P in the river deviated from the optimal N : P ratio of 87 : 1, inferring nutrient limitation of growth. 4. C : P and C : N ratios of S. fluviatilis varied in accordance with general stoichiometric relationships for autotrophs under nutrient limitation of growth. Ratios of C : P and C : N increased, respectively, with increased severity of P- and N-limitation. Additionally, C : P ratios increased with increased N : P ratios, whilst the C : N ratio increased with decreased N : P ratios. The C : N molar ratio however was an insensitive indicator of nutrient depletion compared with the C : P ratio. Under N-limitation of growth, luxury amounts of P were stored by S. fluviatilis. 5. In aquatic environments where macroalgae are sufficiently abundant to be sampled, their cellular carbon, nitrogen and phosphorus stoichiometry can be used to infer nutrient limitation of growth when their optimal C : N : P ratio is known. Keywords: ecological stoichiometry, eutrophication, macroalgae, nitrogen, phosphorus Introduction Ecological stoichiometry provides a theoretical framework for understanding the mass balance of biologCorrespondence: Simon A. Townsend, Department of Natural Resources, Environment and the Arts, PO Box 30, Palmerston, Northern Territory 0831, Australia. E-mail: simon.townsend@nt.gov.au ically significant elements within ecosystems. This approach has increased mechanistic understanding of trophic interactions, nutrient cycling, population dynamics and the role of species in ecosystems (Cross et al., 2005). Photoautotrophs are ultimately the basis of nearly all food webs, converting inorganic nutrients to living biomass with energy supplied by photosynthesis. An imbalance typically exists between the relative proportions and amounts of nutrients Ó 2007 Dept. of Natural Resources, Environment and the Arts, Journal compilation Ó 2007 Blackwell Publishing Ltd 509 510 S. A. Townsend et al. supplied by the environment and those required for photoautotroph growth. This constrains the growth of photoautotrophs and can prompt an uncoupling between carbon fixation and nutrient acquisition to produce a wide variation in the nutrient contents of photoautotrophs, notably their C : N : P stoichiometry (Sterner & Elser, 2002). Under conditions of maximum growth, when nutrient limitation is not discernible, the elemental content of marine phytoplankton attains an optimal C : N : P molar ratio that approximates the Redfield ratio of 106 : 16 : 1 (Goldman, McCarthy & Peavey, 1979). Theoretically, only nutrient replete phytoplankton growing at a maximum rate can attain the Redfield ratio (Goldman, 1986), whilst the absolute growth rate is both a function of nutrient supply as well as abiotic factors such as temperature and light (Sterner & Elser, 2002). The Redfield ratio is also applicable to freshwater phytoplankton (Kilham, 1990; Hecky, Campbell & Hendzel, 1993), and resembles the optimal C : N : P cellular ratio for marine benthic microalgae of 119 : 17 : 1 (Hillebrand & Sommer, 1999). Deviations from the optimal cellular C : N : P ratio for microalgae have been proposed to infer nutrient limitation of growth (Healey, 1978). Growth rates increase with the internal cellular concentration of a limiting nutrient until a maximum internal content is reached, as described by the Droop (1974) model of cell quota and growth kinetics. When this maximum is reached, a second nutrient becomes limiting. The transition point between the growth limitations of two nutrients is marked by an optimal cellular ratio of the two nutrients (Elrifi & Turpin, 1985). Marine phytoplankton N : P ratios have been shown by Goldman et al. (1979) to be more responsive to growth rate under P-limitation than under N-limitation, resulting in a wider range of N : P ratios when P limits growth. Macroalgae, in common with microalgae, are nonvascular photoautrophs. Stream macroalgae are distinguished from microalgae by forming a mature thallus that is a discrete structure recognizable with the naked eye (Sheath & Cole, 1992). They belong to the following algal divisions: Bacillariophyta (e.g. Didymosphenia geminata (Lyngbye) Schmidt), Chlorophyta (e.g. Cladophora, Oedogonium, Spirogyra, Chara and Nitella), Chrysophyta (e.g. Chrysocapsa), Cyanophyta (e.g. Nostoc, Stigonema, Tolypothtix), Rhodophyta (e.g. Batrachospermum), Phaeophyta (predominately marine kelps), Tribophyta (e.g. Tribonema minus (Wille) Hazen) and Xanthophyta (e.g. Vaucheria). The algal divisions Cryptophyta, Dinophyta, Glaucophyta, Prymnesiophyta and Raphidophyta do not have macroscopic species. Most freshwater macroalgae are attached to stream and lake substrata, although some are free floating (e.g. Hydrodictyon). Macroalgae can be prominent primary producers in eutrophic (e.g. Cladophora, Whitton, 1970) and oligotrophic freshwaters (e.g. Spirogyra, Townsend & Padovan, 2005). The concept of an optimal cellular nutrient ratio should be applicable not only to microalgae but also to macroalgae, although the optimal ratios of macroalgae may differ from the Redfield ratio. Environmental samples of freshwater and marine macroalgae are carbon rich relative to their nitrogen and phosphorus content, and compared to microalgal ratios (Atkinson & Smith, 1983; Duarte, 1992). Whether this reflects a greater degree of nutrient deficiency than microalgae or is because of a different elemental composition of macroalgae is not clear. In this paper, we address this issue by examining the nutrient stoichiometry of the filamentous green macroalga Spirogyra fluviatilis Hilse in the Douglas–Daly region of tropical Australia. More specifically, we examine the quantitative relationships between carbon, nitrogen and phosphorus. We also estimate the optimal C : N : P ratio of S. fluviatilis, compare this ratio to the Redfield ratio, and test the utility of the S. fluviatilis optimal nutrient ratio to infer nitrogen or phosphorus limitation. Methods Site description The Douglas–Daly region (14°S, 131°E) is located in the wet/dry tropics of northern Australia (Fig. 1), and is 180 km south from the coastal city of Darwin in the Northern Territory. It comprises the lower reaches of the Douglas River catchment and the adjacent Daly River catchment, and approximates 6100 km2 (Hill, 2004). The principal land-use in the region is cattle grazing on native grass pastures and Eucalyptus savanna woodlands, with about 15% of the region developed for agriculture. The Daly River and its tributaries are considered to be in a near-undisturbed state compared with temperate Australian rivers, Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 The Redfield ratio and Spirogyra fluviatilis 131°20'0"E Study region 511 131°30'0"E 13°40'0"S ee k Tropic of capricorn H ay es cr AUSTRALIA Do 13°50'0"S DG2 r ive sr la ug HC DG4 MC Mid dle 13°50'0"S cre ek Daly river Fig. 1 Macroalgae and water quality sample sites in the Douglas–Daly Region, northern Australia. DR, Daly River sites; DG, Douglas River sites; HC, Hayes Creek site; MC, Middle Creek site. DR6 DR5 DR4 14°0'0"S although regionally high nitrate concentrations in the lower Douglas River maybe of anthropogenic origin. The flow regime of rivers and streams in the region is highly seasonal. Wet season flow between January and April in the Daly River averages 90% of total annual discharge, and is dominated by surface runoff. In contrast, dry season flows (June–October) originate from groundwater sources, and constitute an extended period of baseflow that is rarely supplemented by surface run-off (Townsend & Padovan, 2005). The Daly River is typically 5–20 m deep during the wet season, but in the dry season rapids and runs are <1 m deep, whilst pools are 2– 4 m deep. The highly seasonal flow of the Daly River and its tributaries underpins a seasonal pattern of Spirogyra biomass (Townsend & Padovan, 2005). During the wet season, high current speeds and turbidity prevent the growth of macroalgae, but dry season hydraulic and optical conditions favour macroalgal growth. In 2001, Spirogyra first became visible in the middle reaches of the Daly River in mid-May, then grew to reach a maximum biomass of 28 mg m)2 of chlorophyll a (Townsend & Padovan, 2005) when mats of the alga covered areas of up to 200 m2 with at least 30% areal cover. Over the following 2 months, the biomass of Spirogyra halved, before being removed from the river by the first major runoff event of the wet season in November. Spirogyra can constitute a significant proportion of the river’s total primary productivity. In 2001, the alga contributed up to 40% of the river’s 14°0'0"S N 0 2.5 5 10 km DR2 131°10'0"E 131°20'0"E 131°30'0"E total primary production along a 10-km reach of the Daly River (Webster et al., 2005). Eight study sites were selected in the Douglas– Daly region, where nutrient concentrations during the dry season indicate oligotrophic conditions (Townsend & Padovan, 2005) based on the classification scheme of Dodds, Jones & Welch (1998). Dry season nitrate concentrations in the region vary by two orders of magnitude, whilst soluble phosphorus concentrations vary by a factor of two. Sites for the collection of macroalgae and water samples were selected where hydraulic conditions, incident light and river substrata were favourable for Spirogyra growth, and provided a wide range of nitrate concentrations. Four sites were selected (DR2, DR4, DR5, DR6; Fig. 1) along an 18-km reach of the Daly River, previously studied by Townsend & Padovan (2005) and Webster et al. (2005), where the river is 50 m wide with banks 20 m high. Four additional sites were located on the Douglas River (DG2, DG4; Fig. 1), and two tributaries, Hayes and Middle Creeks (HC, MC; Fig. 1). Sample collection Macroalgae and water samples were collected monthly between May and November 2005, with additional water samples collected from Daly River sites to provide a fortnightly sampling frequency. Samples collected for taxonomic identification were preserved in 4% formalin. Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 512 S. A. Townsend et al. Water samples were collected for the analysis of nitrite, nitrate and filterable reactive phosphorus, hereafter referred to as soluble phosphorus. Ammonia in these well-oxygenated waters during the dry season is not detectable, being <5 lg L)1 (S.A. Townsend unpublished data). A representative water sample was collected at each site, located along either a run or riffle with well-mixed, turbulent flow. Water samples were filtered through a 0.45-lm membrane in the field, stored on ice and then frozen until their analysis by standard methods. These were [with APHA (1998) method numbers in parentheses], the automated cadmium method for nitrate and nitrite (4500-P F), and automated ascorbic acid reduction method for filterable reactive phosphorus (4500-P F). These analyses had a detection limit of 1 lg L)1, with all nitrite concentrations less than this limit. Water samples were also collected for alkalinity, which was determined by standard titration (2320B) with a detection limit of 1 mg L)1 as CaCO3. Field measurements were made of temperature, dissolved oxygen and conductivity using a DataSonde multi-parameter probe (Hydrolab Corporation, Austin, TX, U.S.A.) and turbidity with a Hach 2100P turbidity meter (Hach Company, Loveland, CO, U.S.A.). The attenuation of photosynthetically available radiation (PAR) was measured in the Daly River with a Licor (Lincoln, NE, U.S.A.) meter and scalar bulb, but not at the other sites because of their restricted depth. The attenuation of PAR through the water column was determined from a regression of the natural log-transformed PAR (dependent variable) and depth (independent variable). colour and its slimy feel, and could not be confused with other macroalgae present in the Daly River, primarily Nitella and Chara. The algal samples were processed in a laboratory within 3 days of collection. The samples retained their integrity during storage as no statistically significant relationship was found between the duration of sample storage and the chlorophyll a to phaeophytin ratio (S.A. Townsend unpublished data) which would have indicated macroalgal decomposition. The samples were rinsed with distilled water to remove macroinvertebrates, debris and sediment, then gently squeezed by hand to remove water, and finally placed on blotting paper to absorb the remaining water. We acknowledge the samples included epiphytic algae, although these were a negligible proportion of the total cell volume (<0.1%). The sample was then weighed and oven dried (3 days at 60 °C), then re-weighed to determine its dry weight. The sample was analysed by APHA (1998) methods for total organic carbon by the high temperature combustion method (5310 A & B), for total nitrogen following sulphuric acid digestion (4500-Norg B) and the cadmium reduction method (4500-NO 3 F) and for total phosphorus by Inductively Coupled Plasma Mass Spectrometry. The percentage dry weight contents of carbon (C%), nitrogen (N%) and phosphorus (P%) were then calculated, which is the mass of C, N or P expressed as a percentage of the sample dry weight. All ratios expressed in the paper are molar ratios of the dry weight sample rounded to two significant figures, and used the atomic weights of C, N and P to convert masses to molar (atomic) ratios. Nutrient addition experiment Spirogyra nutrient analysis Spirogyra samples collected for nutrient analyses were stored in plastic bottles and kept on ice whilst in the field. Samples were collected in waters 0.05–0.5 m deep on either gravel or bedrock substrata with observable flow. The collection of samples from deeper waters was restricted due to safety concerns about the presence of estuarine crocodiles (Crocodylus porosus Schneider) in the river. Up to four replicate samples were collected randomly at each site, depending on the biomass available and ease of sampling. Spirogyra was readily identified in the field by its filamentous growth form, luxuriant green A simple experiment was conducted to evaluate the C : N : P ratio of Spirogyra in the Daly River exposed to higher N, P and trace element concentrations than ambient river values. Approximately 2 kg of the granulated slow-release garden fertilizer Multicote8Ò (Haifa Chemicals Ltd., Haifa Bay, Israel) was placed in each of three porous bags and deployed on gravel, where Spirogyra was absent, at site DR6 in mid-June. Three bags of washed sand were used as controls. The depths of the test and control sites were 0.1–0.6 m and current velocities were within the range favourable for Spirogyra growth (0.1–0.5 m s)1; Townsend & Padovan, 2005). Incident radiation (290–2800 nm) Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 The Redfield ratio and Spirogyra fluviatilis during the experiment was measured continuously by the Australian Bureau of Meteorology at Darwin Airport, which is representative of the region’s radiation during the dry season, and reported as half hour total radiation values. The granules are encapsulated in a polymer coating that controls the release of nutrients, allowing a continuous supply to the river. The granules comprise by weight 16.0% N (urea, ammonium nitrate, potassium nitrate, ammonium phosphate) and 3.5% P (calcium phosphate, ammonium phosphate), and yield a N : P supply molar ratio of 10 : 1. The quantities of other nutrients present were 10.0% K (potassium nitrate), 1.9% S, 1.2% Mg and small amounts (<0.2%) of Fe, Mn, Mo, Zn and B. The growth of S. fluviatilis was observed weekly, and samples collected for nutrient analysis were collected 3 m downstream of the bags, after 4 weeks of growth. 513 Table 1 Average soluble nitrogen and phosphorus concentrations Site n Soluble N (lg L)1) Soluble P (lg L)1) Soluble molar N:P DG2 DG4 DR2 DR4 DR5 DR6 HC MC 7 8 10 14 5 11 6 5 1* 60 3 8 8 9 7 7 7 7 8 9 8 6 7 9 0.29* 22 0.94 1.9 2.3 2.5 2.4 1.8 (0.2) (6) (0.1) (0.8) (0.4) (0.9) (0.4) (0.1) (0.3) (0.3) (0.8) (0.5) (0.7) (0.6) (0.5) (0.8) (0.059) (0.27) (0.30) (0.21) (0.24) (0.24) (1.4) (0.22) The standard error of the mean is presented in parentheses. n, number of samples. *Most soluble nitrogen concentrations were less than the detection limit of 1 lg L)1. To calculate an average, concentrations <1 lg L)1 were assigned a value equal to the detection limit. The true mean, therefore, will be <1 lg L)1. nutrient exhibited a pronounced temporal trend, although nitrate concentrations tended to be higher at the beginning of the study at some sites. Results Water quality Stoichiometric relationships The Daly River and its tributaries were warm (24– 34 °C), clear (1.1–4.1 NTU) and well-oxygenated with concentrations between 5.5 mg L)1 (74% saturation) early in the morning to 8.5 mg L)1 (114% saturation) late in the afternoon. Light attenuation in the Daly River averaged 0.53 m)1, equivalent to a euphotic depth of 8.7 m (range 6.0–15.9 m). Ionic chemistry was dominated by dolomite-derived ground water at all sites, except site DG2. The alkalinity and conductivity of the dolomitic riverine waters averaged, respectively, 305 mg L)1 as CaCO3 and 566 lS cm)1, with a median pH of 7.5. At site DG2 alkalinity (average 25 mg L)1 as CaCO3), conductivity (average 52 lS cm)1) and pH (median 7.3) were lower because of their sandstone aquifer origins. Concentrations of soluble phosphorus were similar at each site, with averages between 6 and 9 lg L)1 (Table 1) and an overall range of 5–12 lg L)1. In contrast, average soluble nitrogen concentrations for each site, measured as nitrate, spanned at least two orders of magnitude, ranging from <1 lg L)1 to 60 lg L)1 (Table 1). The maximum nitrate concentration measured was 90 lg L)1. Because of the variation in soluble nitrogen concentrations, soluble molar N : P ratios also spanned two orders of magnitude (Table 1), with an overall range from 0.1 to 26. Neither The overwhelmingly dominant taxon within the macroalgal samples at all sites was the S. fluviatilis complex, with three morphotypes defined by vegetative cell width. These morphotypes range from a narrow form, 24–29 lm in vegetative cell diameter, as described for S. fluviatilis in Czurda (1932), through to a medium form, 29–37 lm diameter, fitting S. rhizopus Jao (1936) and S. corrugata Transeau et al. (1934), to a wide form, 39–48 lm or more in diameter, similar to S. fluviatilis var. africana Fritsch & Stephens (1921). The medium form was the most common morphotype, comprising 72% of taxonomic samples, with the longest form the next most common and constituting 22% of samples. The species is distributed throughout most of Australia (S. Skinner unpublished data) and globally with records in China (Jao, 1936) and north America (Borchardt, Hoffmann & Cook, 1994). Other algae present amongst the S. fluviatilis mats, although in negligible cell numbers, were Mougetia, Cylindrospermum, Oedogonium, Zygnema, Zygnemopsis and epiphytic diatoms. The mean nutrient content of S. fluviatilis was 30% C, 1.8% N and 0.051% P, which is within the nutrient range of 46 macroalgal species reported by Duarte (1992). The nutrient contents of S. fluviatilis covered a wide range, varying twofold for C%, threefold for Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 S. A. Townsend et al. (a) (a) 4 50 n = 83 Avg = 30 CV = 18 Range = 18–43 40 30 2 20 1 10 0 0.00 0 10 20 30 40 C (% dry weight) 30 0.10 P% 0.15 0.20 0.05 0.10 P% 0.15 0.20 DG2 DG4 DR2 DR4 DR5 DR6 HC MC 40 C% n = 83 Avg = 1.8 CV = 32 Range = 0.72–3.3 40 0.05 (b) 50 30 20 20 0.00 10 (c) 0 1 (c) 50 2 3 N (% dry weight) 4 80 60 40 n = 85 Avg = 0.051 CV = 53 Range = 0.025–0.19 30 20 0.5 1.0 1.5 2.0 2.5 3.0 3.5 N% 20 0 0.00 40 C% Relative frequency (%) (b) Site legend: 3 N% 514 0.05 0.10 0.15 P (% dry weight) 0.20 Fig. 2 Frequency distribution of (a) carbon, (b) nitrogen and (c) phosphorus (% dry weight) of Spirogyra fluviatilis in the Douglas–Daly region (northern Australia). n, number of samples; Avg, average; CV, coefficient of variation (%). N% and 10-fold for P% (Fig. 2). The average C : N : P ratio of S. fluviatilis samples was 1700 : 88 : 1, with ratios ranging from 410 : 18 : 1 to 3800 : 170 : 1. The quantitative relationships between elements of biological significance underpin ecological stoichiometry. These relationships are presented in Fig. 3 and summarized in Table 2. The C%, N% and P% (dry weight) ranges for each site tended to overlap with other sites. A notable exception, however, was the P content at DG2 and, to a lesser extent, sites MC and HC, which had high P contents (0.083–0.19%; Fig. 3) compared with S. fluviatilis at other sites (0.019– 0.074%), thus reducing the explanatory power of the linear regressions between P% and other variables (Table 2). The regressions with most explanatory power were between N% and C% (r2 ¼ 0.24), and Fig. 3 Relationships between (a) P% and N%, (b) P% and C%, and (c) N% and C% for Spirogyra fluviatilis samples collected from eight sites in the Douglas–Daly region (northern Australia). The regressions for (a) and (b) exclude site DG2. Further information about the regressions is presented in Table 2. P% and N% when DG2 data were excluded (r2 ¼ 0.30). The carbon : nutrient ratio of aquatic plants can be scaled according to their nutrient content to explore further quantitative relationships (Duarte, 1992). The nutrient content of S. fluviatilis explained c. 80% of the variation in carbon : nutrient ratios (Fig. 4). Increases in P% and N% were associated with exponential decreases in the C : P and C : N ratios respectively (Fig. 4), which closely resemble the relationships for macroalgal species reported by Duarte (1992). The asymptotic nature of the relationship suggests a maximum S. fluviatilis carbon content relative to the nutrient (i.e. low C : nutrient ratio). This approximates ratios of 670 : 1 for C : P as P% approaches 0.12%, and 10 : 1 for C : N as N% approaches 4%. The N : P ratio of S. fluviatilis was closely linked to the C : P ratio, explaining 36% of variation (Fig. 5a; Table 2). A weaker relationship existed between N : P Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 The Redfield ratio and Spirogyra fluviatilis 515 Table 2 Linear regressions between N, P and C contents (percentage dry weight), and nutrient ratios for Spirogyra fluviatilis X y n Slope Intercept r2 Probability Power P% P%* P% P%* N% N:P N:P N% N% C% C% C% C:P C:N 82 78 82 79 81 81 81 4.44 20.8 29.1 74.8 4.53 12.4 )0.118 1.63 0.900 10.8 26.3 21.1 641 31.1 0.05 0.30 <0.01 0.05 0.24 0.37 0.19 0.059 <0.001 0.65 0.05 <0.001 <0.001 <0.001 0.47 0.99 0.07 0.50 0.99 0.99 0.98 Plots of the dry weight data and significant regressions are shown in Fig. 3. Plots of the ratio data are presented in Fig. 5. (All tests satisfied the assumptions of normality and equal variance) *Data from site DG2 excluded, and plotted in Fig. 3. (a) 4000 (c) 100 1000 50 0 C:N (b) 0.05 60 0.10 P% 0.15 0.20 0.00 0.05 (d) Y = 29.7 X–0.768 0.10 P% 0.15 N:P 2000 0.00 Site legend: 150 0 0.20 200 Y = 38.1 + 27.2 N% 150 40 100 20 DG2 DG4 DR2 DR4 DR5 DR6 HC MC N:P C:P 3000 200 Y = 120 – 655 P% Y = 92.0 X–0.935 50 0 0 0 1 2 N% 3 4 0 1 2 N% 3 4 Fig. 4 Relationships between nutrient content (percentage dry weight) and nutrient molar ratios, (a) P% and C : P (n ¼ 82, r2 ¼ 0.82), (b) N% and C : N (n ¼ 81, r2 ¼ 0.78), (c) P% and N : P (n ¼ 83, r2 ¼ 0.31) and (d) N% and N : P (n ¼ 83, r2 ¼ 0.27) for Spirogyra fluviatilis sampled from eight sites in the Douglas–Daly region (northern Australia). All regressions satisfied the assumptions of normality and equal variance, and were significant at <0.001% level and had power of 0.99. and C : N however (Fig. 5b; Table 2), which was influenced by the relatively high C : N values at site DR2. The high P% content of S. fluviatilis at site DG2 coincided with a low C : P ratio. Spirogyra fluviatilis optimal nutrient content During the nutrient addition experiment, incident radiation averaged 680 Wm)2 between 9:00 and 17:00 hours when the riverbed was exposed to direct sunlight, and the attenuation of PAR in the Daly River averaged 0.55 m)1. Assuming that PAR comprised 45% of incident radiation and 1 Wm)2 ¼ 4.57 lmol ph m)2 s)1 (Townsend & Padovan, 2005), the PAR at depths of <0.5 m would have exceeded 100 lmol ph m)2 s)1. Such high light intensities exceed the lower range of photosaturation intensities of benthic algae (Hill, 1996), and indicate that photosynthesis would have been light saturated, although selfshading would have exposed some S. fluviatilis filaments to lower PAR intensities. Current speeds, which can affect S. fluviatilis growth (Borchardt, 1994) though not C : N : P ratios under nutrient replete conditions, approximated 0.5 m s)1. A luxuriant mat of S. fluviatilis grew immediately downstream of the fertilizer bags after 4 weeks of growth, but not at the control sites. The d15N of S. fluviatilis downstream of the fertilizer treatments (mean 3.6&) was significantly lower than S. fluviatilis in the upper reaches of the river (mean 6.9&; Schult Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 S. A. Townsend et al. (a) 4000 C:P 3000 2000 1000 0 0 C:N (b) 50 100 N:P 150 60 Site legend: DG2 DG4 DR2 DR4 DR5 DR6 HC 200 MC Optimum N : P ratio 40 20 0 0 50 100 N:P 150 200 Fig. 5 Relationships between (a) N : P and C : P ratios (r2 ¼ 0.37; P < 0.001), and (b) N : P and C : N ratios (r2 ¼ 0.19; P < 0.001) for Spirogyra fluviatilis at eight sites in the Douglas– Daly region (northern Australia). The optimum S. fluviatilis molar N : P ratio is 87 : 1. Information about the regressions is presented in Table 2. et al., 2007), indicating the nitrogen originated from fertilizer which had a low d15N value close to zero (Udy & Dennison, 1997). The average nutrient content of the fertilizer-enriched S. fluviatilis, as a percentage of dry weight, was 21% C, 1.3% N and 0.031% P, with an average C : N : P ratio of 1800 : 87 : 1. Assuming S. fluviatilis was growing at or near maximum rates, the C : N : P ratio of 1800 : 87 : 1 indicates the optimal C : N : P ratio for S. fluviatilis. Approximately half of the S. fluviatilis samples had cellular N : P ratios <87 : 1, indicating nitrogen limitation, notably at sites DG2, HC and MC (Fig. 6). Nitrogen limitation in S. fluviatilis was associated with low soluble N : P ratios of <2 (Fig. 6) and soluble N concentrations of <5 lg L)1. Conversely, phosphorus limitation in S. fluviatilis (N : P > 87) was often associated with high soluble N : P ratios, for example at site DG4 (mean soluble N : P ratio 18). Discussion Spirogyra fluviatilis nutrient content and optimal C : N : P ratio Macroalgae, in common with other aquatic plants, have a wide range of nutrient contents (see Atkinson S. fluviatilis molar N : P 516 160 P limitation 120 80 40 0 0.01 N limitation 0.1 1 10 Soluble molar N : P Legend: DG2 DG4 DR2 DR4 DR5 DR6 HC MC N : P optimum mean ratio. Standard error of the mean optimum N : P ratio. Fig. 6 Soluble molar N : P ratio and Spirogyra fluviatilis cellular N : P ratio for eight sites in the Douglas–Daly region (northern Australia). Soluble nitrogen comprised of nitrate, as nitrite and ammonia concentrations were below detection (see Methods). Nitrate concentrations less than the detection limit were assigned a value equal to the detection limit for calculation purposes. & Smith, 1983; Duarte, 1992; Ferández-Aláez, Ferández-Aláez & Bécares, 1999) which is attributable to both species-specific and environmental factors. The range of S. fluviatilis N and C nutrient contents approximated 70% of the C and N ranges of the Duarte (1992) data set for 46 macroalgal species, whereas the P% range of S. fluviatilis approximated a smaller proportion (c. 20%). Moreover, the relationships between nutrient contents and C : P and C : N ratios (Fig. 4) were similar between S. fluviatilis and the Duarte (1992) data set. The high variation of nutrient contents for S. fluviatilis, compared with the Duarte (1992) data set, indicates that environmental variation can account for a large proportion of the variation in nutrient content in a single macroalgal species. Variation within the S. fluviatilis complex because of the three morphotypes, however, may have also contributed to the overall nutrient content variability. The optimal C : N : P ratio for S. fluviatilis, which approximated 1800 : 87 : 1, differed substantially from the Redfield ratio (106 : 16 : 1), being deficient in N, and even more deficient in P relative to C. This ratio however was estimated from a field experiment and assumed near maximum growth rates. If growth rates when S. fluviatilis was sampled were not near the maximum, then the optimal ratio for the alga will be richer in N and P relative to C. Carbon limitation of growth was considered unlikely during the experiment because the river has been shown to have carbon Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 The Redfield ratio and Spirogyra fluviatilis dioxide concentrations between 15 and 30 times saturated concentrations (Webster et al., 2005). Laboratory experiments to determine the optimal macroalgal C : N : P ratio, when growth is at a maximum have not been published for freshwater macroalgae. Borchardt (1994) calculated N : P ratios for S. fluviatilis from minimum cell quotas, when growth was zero rather than at a maximum. Several studies of the nutrient limitation of the growth of Cladophora and its N and P cellular content has been undertaken (e.g. Gerloff & Fitzgerald, 1976; Neil & Jackson, 1982; Rosemarin, 1982), although these did not relate maximum growth rates and nutrient stoichiometric ratios. Macroalgae probably have an optimal cellular C : N : P ratio which is rich in carbon relative to N and P, and deficient in P relative to N when compared with the optimal ratio for microalgae based on this field experiment. However, only experiments that measure growth rate and determine the C : N : P cellular ratios of macroalgal species at maximum growth rates will confirm this proposition. The richer carbon content of macroalgae may reflect the greater cellular structural requirements for carbon of macroalgae. We conclude that the application of the Redfield ratio cannot be assumed to be applicable to macroalgae, and is probably inappropriate to assess macroalgal nutrient deficiency. Nutrient limitation inferred from stoichiometric ratios The N and P contents of plants are closely related because these two elements are found in common biological compounds (Sterner & Elser, 2002), such as nucleic acids (e.g. DNA) and energetic nucleotides (e.g. ATP). Nitrogen is also present in proteins and chlorophyll, whilst phosphorus occurs in polyphosphates. Carbon is common to these compounds, and is also present in carbohydrates and lipids. The relative and absolute amounts of the C, N and P in cellular material indicate the plant’s nutrient status, as well as its response to nutrient limitation. One response is luxury nutrient consumption (or uptake) which occurs when growth that is limited by one factor coincides with the uptake of a non-limiting nutrient (Sterner & Elser, 2002). Planktonic luxury consumption differs between nutrients and is growth dependent, and has been shown to be greater for P than for N (Elrifi & Turpin, 1985). 517 Evidence for P luxury consumption exists for S. fluviatilis at site DG2, and to a lesser extent sites MC and HC, where the cellular N : P ratio was considerably lower than the optimal ratio (Fig. 6), indicating N limitation, and the P% was higher than at the other sites (Fig. 3). The P% contents of S. fluviatilis were high and C : P ratios low (e.g. site DG2; Fig. 4) when growth was N-limited. The low soluble N concentrations at site DG2, and low soluble N : P ratios provide further evidence that S. fluviatilis growth was N-limited, although these ratios need to be interpreted with caution as discussed below. Spirogyra fluviatilis was most P rich, and therefore of greatest nutritional value to herbivores, where N limits growth. Whilst P luxury uptake has been demonstrated for phytoplankton (Sterner & Elser, 2002), it has not been demonstrated for freshwater macroalgae. There was no evidence of N luxury uptake by S. fluviatilis in the Douglas–Daly region. Conditions most conducive for N luxury consumption existed at site DG4 where the soluble N : P ratio was at least 10 times that at other sites due to locally high nitrate concentrations. Spirogyra fluviatilis N : P ratios at this site were marginally higher than the optimal ratio (Fig. 6), suggesting P limitation, although S. fluviatilis N% and C : N ratios did not differ substantially from other sites but were nevertheless amongst the highest N% and lowest C : N ratios. Spirogyra fluviatilis may not be capable of luxury consumption of N, or may have taken up only small quantities of N not identified by this study. Sterner & Elser (2002, p. 105) proposed the generalization that ‘for any given nutrient element X, the C : X of autotroph biomass increases with the severity of growth limitation of nutrient X’. This has been demonstrated for phytoplankton (Goldman et al., 1979) and benthic microalgae (Hillebrand & Sommer, 1999) grown in laboratory cultures under a range of P limiting conditions. In each case, the severity of growth limitation was measured by the growth rate relative to a maximum rate when nutrients were not limiting. The P and N contents of S. fluviatilis, relative to their optimal values, can also be used as indicators of growth limitation, with P or N limitation decreasing with increasing nutrient content. As shown in Fig. 4, C : P and C : N ratios increase, respectively, with increased P and N limitation in accordance with Sterner & Elser’s (2002) generalization. Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 518 S. A. Townsend et al. A second generalization proposed by Sterner & Elser (2002, p. 205) is that the C : X ratio for nutrient element X ‘increases with decreased availability of X compared with alternative limiting nutrients (i.e. Y : X)’. The N : P ratios and C : nutrients ratios of S. fluviatilis concurred with this generalization (Fig. 5). The C : P ratio of S. fluviatilis increased with decreased availability of P relative to N, as indicated by higher N : P ratios (i.e. C : P increased with increasing N : P). This also concurs with the findings of Hillebrand & Sommer (1999) that N : P and C : P ratios increase with P limitation. Similarly, with decreased availability of N relative to P and hence decreased N : P ratios, the C : N ratio increased (Fig. 5). This relationship was weak for S. fluviatilis however, compared with the N : P versus C : P regression. Spirogyra fluviatilis C : N ratios were a less-sensitive indicator of nutrient limitation than C : P ratios, which could be due to negligible N luxury consumption. Hillebrand & Sommer (1999), however, reported increased C : N ratios for four species of benthic microalgae, regardless of the limiting nutrient, and concluded that this ratio was indicative of general nutrient limitation. Knowledge about the nutrients limiting primary production in the aquatic environment underpins the management of eutrophication, and is fundamental to understanding the constraints that limit ecosystem productivity (Douglas, Bunn & Davies, 2005). This paper has tested the use of the cellular N : P ratio to evaluate nutrient limitation of S. fluviatilis. The river’s soluble N : P ratios provided a means of testing the validity of cellular nutrient contents and ratios. This must be undertaken with some caution, however, because concentrations of soluble N and P provide limited information about nutrient availability. Importantly, nutrient concentrations do not infer nutrient supply or turnover rate (Dodds, 2003), and low concentrations may be interpreted either as evidence of rapid nutrient turnover and sufficient nutrient supply, or as resulting from the demand for a nutrient that exceeds its supply. In the Douglas–Daly Region, low cellular N : P ratios (which indicate N limitation of growth) coincided with low soluble N : P ratios and low soluble N. Conversely, high cellular N : P ratios (notably at site DG4, Fig. 4), which indicate P limitation, coincided with high soluble N : P ratios and high soluble nitrogen concentrations. Further- more, P limitation of S. fluviatilis inferred from high cellular N : P ratios at sites DR4, DR5 and DR6 in October 2005 was supported by the results of a nutrient addition experiment that concluded S. fluviatilis growth was P limited at these sites (Schult et al., 2007). Macroalgae can be a prominent feature of rivers, streams and the littoral zone of lakes (Wetzel, 1983). This paper examined the utility of using cellular stoichiometric ratios and the Redfield ratio to evaluate nutrient limitation of S. fluviatilis in a region of tropical Australia. An estimate of the optimal C : N : P ratio of S. fluviatilis was determined from a nutrient addition experiment conducted in the Daly River, which assumed near maximum growth rates had been attained when the samples for cellular composition were collected. The optimal cellular C : N : P ratio for S. fluviatilis (1800 : 87 : 1) was substantially higher than the Redfield ratio (106 : 16 : 1) which is applicable to aquatic microalgae. This paper has demonstrated the significant potential for macroalgal stoichiometric ratios to be used to evaluate macroalgal N- and P-limitation. The Redfield ratio, however, is unlikely to be applicable to freshwater macroalgae, although this needs to be tested for other macroalgae. Acknowledgments Financial support for this project was provided by the National Action Plan for Salinity and Water Quality. The paper was improved by the thoughtful comments of two anonymous reviewers. References APHA (1998) Standard Methods for the Examination of Water and Wastewater, 20th edn. APHA/AWWA/WEF, U.S.A. Atkinson M.J. & Smith S.V. (1983) C:N:P ratios of benthic marine plants. Limnology and Oceanography, 28, 568– 574. Borchardt M.A. (1994) Effects of flowing water on nitrogen- and phosphorus-limited photosynthesis and optimum N:P ratios by Spirogyra fluviatilis (Charophyceae). Journal of Phycology, 30, 418–430. Borchardt M.A., Hoffmann J.P. & Cook P.W. (1994) Phosphorus uptake kinetics of Spirogyra fluviatilis (Charophyceae) in flowing water. Journal of Phycology, 30, 403–417. Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 The Redfield ratio and Spirogyra fluviatilis Cross W.T., Benstead J.P., Frost P.C. & Thomas S.A. (2005) Ecological stoichiometry in freshwater benthic systems: recent progress and perspectives. Freshwater Biology, 50, 1895–1912. Czurda V. (1932) Zygnemales. Volume 9 in Pascher, A. Die Süsswasser-Flora Mitteleuropas. Verlag von Gustav Fischer, Jena. Dodds W.K. (2003) Misuse of inorganic N and soluble reactive P concentrations to indicate nutrient status of surface waters. Journal of the North American Benthological Society, 22, 171–181. Dodds W.K., Jones J.R. & Welch E.B. (1998) Suggested classification of stream trophic state: distributions of temperate stream types by chlorophyll, total nitrogen and phosphorus. Water Research, 32, 1455– 1462. Douglas M.M., Bunn S.E. & Davies P.M. (2005) River and wetland food webs in Australia’s wet-dry tropics: general principles and implications for management. Marine and Freshwater Research, 56, 329–342. Droop M.R. (1974) The nutrient status of algal cells in continuous culture. Journal of the Marine Biology Association U.K., 55, 541–555. Duarte C.M. (1992) Nutrient concentration of aquatic plants: patterns across species. Limnology and Oceanography, 37, 882–889. Elrifi I.R. & Turpin D.H. (1985) Steady-state luxury consumption and the concept of optimum nutrient ratios: a study with phosphate and nitrate limited Selenastrum minutum (Chlorophyta). Journal of Phycology, 21, 592–602. Ferández-Aláez M., Ferández-Aláez C. & Bécares E. (1999) Nutrient content in macrophytes in Spanish shallow lakes. Hydrobiologia, 408/409, 317–326. Fritsch F.E. & Stephens E. (1921) Contributions to our knowledge of the freshwater algae of Africa. 3. Freshwater Algae (exclusive of Diatoms), mainly from the Transkei Territories, Cape Colony. Transactions of the Royal Society of South Africa, 9, 1–72. Gerloff G.C. & Fitzgerald G.P. (1976) The Nutrition of Great Lakes Cladophora. E.P.A., Michigan. Goldman J.C. (1986) On phytoplankton growth rates and particulate C:N:P ratios at low light. Limnology and Oceanography, 31, 1358–1363. Goldman J.C., McCarthy J.J. & Peavey D.G. (1979) Growth rate influence on the chemical composition of phytoplankton in oceanic waters. Nature, 279, 210– 215. Healey F.P. (1978) Physiological indicators of nutrient deficiency in algae. Mitteilungen der Internationale Vereinigung für Theoretische und Angewandte Limnologie, 21, 34–41. 519 Hecky R.E., Campbell P. & Hendzel L.L. (1993) The stoichiometry of carbon, nitrogen and phosphorus in particulate matter of lakes and oceans. Limnology and Oceanography, 38, 709–724. Hill W. (1996) Effects of light. In: Algal Ecology: Freshwater Benthic Ecosystems. (Eds R.J. Stevenson, M.L. Bothwell & R.L. Lowe), pp. 128–148. Academic Press, San Diego, California. Hill J. (2004) The Extent of Developed and Undeveloped Agricultural Land in the Douglas Daly Region and the Application of Stream Buffers. Report 13/2004D. Department of Infrastructure, Planning and Environment, Darwin, Australia. Hillebrand H. & Sommer U. (1999) The nutrient stoichiometry of benthic microalgal growth: Redfield proportions are optimal. Limnology and Oceanography, 44, 440–446. Jao C. (1936) New Zygnemataceae collected in China. American Journal of Botany, 23, 53—60. Kilham S.S. (1990) Relationship of phytoplankton and nutrients to stoichiometric measures. In: Large Lakes: Ecological Structure and Function (Eds M.M. Tilzer & C. Serruya), pp. 403–413. Springer-Verlag, New York. Neil J.H. & Jackson M.B. (1982) Monitoring Cladophora growth conditions and the effect of phosphorus additions at a shoreline site in northeastern Lake Erie. Journal of Great Lakes Research, 8, 30–34. Rosemarin A.S. (1982) Phosphorus nutrition of two potentially competing filamentous algae, Cladophora glomerata (L.) Kütz. and Stigeoclonium tenur (Agardh) Kütz. from Lake Ontario. Journal of Great Lakes Research, 8, 66–72. Schult J., Townsend S.A., Douglas M.M., Webster I.T., Skinner S. & Casanova M. (2007) Recommendations for Nutrient Resource Condition Targets for the Daly River. Charles Darwin University, Darwin, Australia. Sheath R.G. & Cole K.M. (1992) Biogeography of stream macroalgae in North America. Journal of Phycology, 28, 228–260. Sterner R.W. & Elser J.J. (2002) Ecological Stoichiometry: The Biology of Elements from Molecules to the Biosphere. Princeton University Press, Princeton, U.S.A., 439 pp. Townsend S.A. & Padovan A.V. (2005) The seasonal accrual and loss of benthic algae (Spirogyra) in the Daly River, an oligotrophic river in tropical Australia. Marine and Freshwater Research, 56, 317–327. Transeau E.N., Tiffany L.H., Taft C.E. & Li L.C. (1934) New Species of Zygnemataceae. Transactions of the American Microscopical Society, 53, 208–230. Udy J.W. & Dennison W.C. (1997) Physiological responses of seagrasses used to identify anthropogenic Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520 520 S. A. Townsend et al. nutrient inputs. Marine and Freshwater Research, 48, 605–614. Webster I.T., Rea N., Padovan A.V., Dostine P., Townsend S.A. & Cook S. (2005) An analysis of primary production in the Daly River, a relatively unimpacted tropical stream in Northern Australia. Marine and Freshwater Research, 56, 303–316. Wetzel R.G. (1983) Periphyton of Freshwater Ecosystems. Dr W. Junk Publishers, The Hague. Whitton B.A. (1970) Biology of Cladophora in freshwaters. Water Research, 4, 457–476. (Manuscript accepted 18 October 2007) Ó 2007 Dept. of Natural Resources, Environment and the Arts Journal compilation Ó 2007 Blackwell Publishing Ltd, Freshwater Biology, 53, 509–520