MARINE ECOLOGY PROGRESS SERIES Vol. 149: 267-277. 1997 1 Published April 10 Mar Ecol Prog Ser

Leaf versus root nitrogen uptake by the surfgrass Ph yllospadix torreyi

J. Terrados*, S. L. Williams

Department of Biology. San Diego State University. San Diego. California 92182-4614. USA

ABSTRACT- Phyllospadix torreyiS. Watson ('surfgrass')is one of very few species that grow on rocks. Thus, unlike most other , nutrient uptake across leaves rather than across roots might be very important for nutrient acquisition. Ammonlum and nitrate+nitrite uptake by surfgrass leaves was measured under flowing water and modelled (Michaelis-Menten)as a functlon of external concentrations. Leaf NH,' uptake showed a Kmof 17.4PM and a V,,,,, of 125.1 pm01 N g dry wt-' h-' Kmand V,,, values for leaf NO3 +NO2-uptake were 10.1 pM and 54.5 pm01 N g dry wt.' h-', respec- tively. Ammonium available to the roots had little discernible effect on NH,' uptake by leaves Low NH,' uptake rates by roots (~0.2pm01 N g dry wt.' h-'] suggest that surfgrass acquires most of its nitro- gen via its leaves.

KEY WORDS: Ammonium . Nitrate Leaf uptake Root uptake Phyllospadix . Seagrasses

INTRODUCTION (McRoy & Barsdate 1970. McRoy & Goering 1974, McRoy & McMillan 1977, Penhale & Thayer 1980, The seagrass species of the genus Pl~yllospadix Short & McRoy 1984, Thursby & Harlin 1984), the Hook. grow on rocky substrata in high-energy envi- importance of leaf uptake has been stressed lately ronments (Dudley 1894, den Hartog 1970). This habitat (Patriquin 1972, Borum et al. 1989, Hemminga et al. is distinctly different from that of most other seagrasses 1991, 1994, Pedersen & Borum 1992, 1993). Some of that grow on unconsolidated substrata (mud, sand), the features of Phyllospadix habitat (surf exposure, and where the amount of physical energy imposes lim- rocky substratum) suggest that Phyllospadix would its on the development of meadows (Fonseca & Ken- take up the nutrients mainly from the water column via worthy 1987).These physical differences of the habitat leaf uptake. Seagrasses apparently have complex have implications for community processes such as source-sink interactions between leaves and below- nutrient cycling, community stability and succession ground organs and the nutrient uptake rate and tissue (McRoy & Lloyd 1981), and suggest that Phyllospadix concentration of one can influence those of the other beds could be more similar to macroalgae communities (Thursby & Harlin 1982, 1984) The relative contribu- than to other seagrass beds. In fact, Phyllospadix is tion of leaves and roots to whole acquisition of able to compete successfully with macroalgae for nutrients must depend on kinetic attributes of the space and form communities with high persistence different tissues, nutrient concentrations available to (Turner 1983a, b, 1985, Turner & Lucas 1985). each type of tissue, their biomasses, and nutrient flux Seagrasses can take up inorganic nutrients present rates. both in the water column and the sediment via leaf Phyllospadixspecies show some anatomical features and/or root uptake. Although the sediments have been that have been interpreted as adaptations to surf expo- considered the main nutrient source for seagrasses sure (Cooper & McRoy 1988, Barnabas 1994). Exten- sive root-hair development (Cooper & McRoy 1988), a 'Present address: Centro de Estudios Avanzados de Blanes, well-developed mechanical layer (i.e. thick-walled C.S.I.C., Cami de Santa Barbara, s/n, E-17300 Blanes compactly arranged cells without any air spaces) in the (Glrona).Spain. E-mall: [email protected] outer cortex of the root, a thickened outer wall of root

0 Inter-Research 1997 Resale of ft111arl~cle not permitted 268 Mar Ecol Prog Ser 149: 267-277. 1997

epidermal cells, and the production of a mucilaginous- like material by the roots as an adhesive to the substra- tum (Barnabas 1994) are some of the potentially adap- tive features for life on rocks with high surf exposure. The essential function of the roots of Phyllospadix is considered to be attachment, not nutrition (Gibbs 1902, C. McMillan, p 236, in Stewart 1989). However, the extensive development of root hairs suggests that surf- grass could obtain a significant portion of inorganic nutrients through its roots (Cooper & McRoy 1988). Further, the pectin-rich layer present in the outer wall of root epidermal cells of Hook. could be involved in nutrient absorption (Barnabas 1994). Presently there is no experimental evidence to support or reject a role of Phyllospadix roots in nutrient acquisition. The roots of Thalassodendron cjljatum, one of the other few seagrass species that can grow on a rocky substrate, are surmised to have limited nutrient absorption capacity (Barnabas 1991). This study aims to describe the nitrogen (ammonium, nitrate+nitnte) uptake by the leaves and roots of Phyllo- spadix torreyi S.Watson, and their interaction, through answering the following questions: Is P. torreyi able to acquire nutriehts via the roots? How do ambient nitro- gen concentrations compare to uptake kinetic para- meters? Does leafhoot ammonium uptake change Fig. 1. Station where surfgrass for uptake experiments when the roots/leaves are supplied with ammonium? were collected ('surfgrass'),and stations where plant nltrate concentrations ('Bird Rock', 'shore', 'False Po~nt'),tempera- ture, NH,' and NOJ-+NO2 concentrations In the water col- umn ('surfgrass', 'surface', 'shore') and the interstlt~alwater of MATERIALS AND METHODS the sediments ('surfgrass') were measured at False Polnt, La Jolla (CA, USA) Study site. Surfgrass S. Watson is abundant on the upper subtidal rocky bottoms of the proximately 200 m offshore from the first location; and San Diego County, Southern California (USA), coast (3) inside the surfgrass canopy ('surfgrass') in the same (Stewart & Myers 1980, Stewart 1989); all plants used vicinity as the 'surface' location (Fig. 1). The 'shore' here were collected in the False Point area, La Jolla location was the only one sampled for NO,-+NO2 (32" 48' N, 117" 16' W). This stretch of the coast is analysis (n = 23) between July 1993 and May 1994 (see openly exposed to the ocean swell. Different types of Fig 2);all 3 locations were sampled for both NH,' (n = substrata can be found in the area: rocky outcrops, 21 or l?) and NO3-+NO2. (n = 18) analysls from June coarse sands, and cobble beaches. Surfgrass is distrib- 1994 to December 1994. Five replicate samples were uted from the mean-lower-low-water (MLLW) level taken in individual acid-washed polycarbonate bottles into the upper subtidal, = -5 m (Ricketts et al. 1985), at each sampling date at each location. The samples forming patches always located on rocks or big boul- (50 ml) were immediately filtered through 2.5 cm ders, sometimes buried under a thin (less than 2 cm) Whatman GF/C filters and NH4+reagents were added layer of coarse sand. The plants used in the expen- to a 10 m1 subsample, which was kept cool and dark ments came from surfgrass patches present in an area while transported to the laboratory. Nitrate+nitrite of roughly 100 X 200 m along the shore at a depth samples were frozen until analysis (less than 1 mo after (MLLW) of -3 m (Fig. 1).The study was done between collection). Ammonium samples were kept in the February 1993 and December 1994. refrigerator and measured within 24 h. Water temper- Water samples for analysis of dissolved inorganic ature was measured each sampling date at each loca- nitrogen (DIN) concentrations (NH4+,NO3-+ NO2-) tion (see Fig. 2). were collected at 3 locations within the study site: (1) at Interstitial water was collected from the water imme- the water's surface near the shallowest surfgrass diately above unvegetated sediments, from sediments patches ('shore'); (2) at the water's surface where without surfgrass, and In sedlments with surfgrass by plants were collected for experiments ('surface'), ap- insertion of a 2-end hypodermic needle wrapped In a Terrados & Wllllams: Sur.fgrass nitrogen uptake 269

30 pm mesh into the sediment; when the serum stopper tal chambers, plus 7 control chambers with no plant in- on an evacuated 4 m1 glass tube (Vacutainer) was side). The mean (+SE) leaching rate was 0.13 + 0.02 punched through one of the needles, the interstitial wa- pm01 N g dry wt-' h-'. For NO, +NO, uptake experi- ter was drawn from the sediment. Samples for NH,' ments the plants were kept in seawater at 10 ].]M NO3- analysis (sediment with surfgrass, n = 9; sediment with- to induce the nitrate reductase activity (Guerrero et al. out surfgrass, n = ?; above sediment, n = 3; 2 to 5 repli- 1981, Thursby & Harlin 1984, Roth & Pregnall 1988). cates per sample) were obtained in July and September The seawater used in the experiments was collected 1994; samples for N0,+N02- analysis were collected at the shore and filtered once through Whatman GF/C on 2 different days in January and February 1994. filters and then through Millipore GS filters (pore size, Nitrate+nitrite was determined with a flow-injection 0.22 1.m). One batch of each of the initial NO3-+NOz- autoanalyzer (Lachat Instruments, Milwaukee, WI, or NH,' concentrations needed according to the USA) using a Cd-Cu reduction method (Parsons et al. planned experiment was prepared by adding small 1984). Ammonium was determined using a phenol- volumes (ml) of concentrated KNOj or (NH,)2S0, solu- hypochlorite method (Crasshoff et al. 1983).The redox tions to large volumes (1) of water kept in plastic con- potential (Eh)of the interstitial water of sediments with tainers. Based on the Eh values obtained (see and without surfgrass (n = 8 and 3, respectively; 2 to 5 'Results'), the oxygen content of the water used in the replicates per sample) and of the water immediately root compartment was not reduced. above unvegetated sediment (n = l) was also mea- Uptake experiments. Nitrogen uptake experiments sured on 4 different days during August and Septem- were performed in 2-compartment chambers made ber 1994 using a platinum electrode [calibrated with from a clear acrylic pipe of 25 mm (internal diameter, ZoBell's (1946) solution] and an Ag/AgCl reference i.d.); the leaf compartment was 70 cm in length, the electrode with saturated KC1 electrolyte (American root compartment was 7.5 cm. A silicone stopper with Public Health Association 1992). The Eh value of the a small hole drilled at the centre and a slit to facilitate water samples, corrected for the potential of the refer- plant placement held the 2 compartments together. ence electrode, is used here to qualitatively describe The leaves without attached rhizome and roots were the degree of anaerobiosis of Phyllospadjx torreyi fixed to a 20 cm plastic-coated wire inserted in the hole sediments (Whitfield 1969, 1974). of the silicone stopper; the leaves with attached i-hi- Surfgrass collections for uptake experiments. Both zome and roots were inserted through the hole in the the plants and the water to be used in an experiment stopper until the shoot insertion point on the rhizome. were collected 24 h beforehand. One surfgrass clump The seal between compartments was formed by plac- (>50shoots) from 15 to 20 distinct surfgrass patches at ing a small amount of 'UHU HOLD-IT Removable Plas- the 'surfgrass' location was haphazardly collected by tic Adhesive' (Eberhard Faber, Inc., Lewisburg, TN, hand and transported to the laboratory in seawater. USA) around the leaves inside the stopper hole to fill in One whole shoot was selected from each clump to min- gaps and covering the root end of the hole with non- imize using clonal material. The leaves were cut uni- toxic anhydrous lanolin. A water pump (Maxi-jet formly to the basal 45 cm. The bundle of 2 to 4 leaves in MJ250, Aquarium Systems, Mentor, OH, USA) placed the shoot was either pinched off the rhizome for leaf up- at one end of the leaf compartment pumped water take experiments or left attached to the rhizome with across the leaves and recirculated it within the cham- up to 2 rhizome nodes and with roots on either side of ber via a 1 cm i.d. PVC tubing This design forced a the shoot for leaf versus root uptake experiments. After unidirectional flow of the water inside the leaf com- leaves were cleaned of macroepiphytes, plant frag- partment from the base of the leaves to the apex at an ments were kept in a greenhouse inside a plastic average velocity of 14 cm S-'. A small hole drilled In the bucket filled with Whatman GF/C filtered seawater chamber wall next to the pump and equipped with a which was mixed and aerated by a small aquarium air red rubber septum stopper was used for sampling pump and kept at field temperatures. To insure that using 5 m1 syringes. The root compartment was plants recovered from collection before being used in opaque, unstirred, and had a sampling port. the uptake experiments, the respiration rate of whole The experiments were performed during mldday (leaves with attached rhizome and roots) shoots (n = 6) under direct sunlight. The chambers were placed hori- was measured 6, 21 and 27 h after collection. The respi- zontally inside a shallow water bath. Water tempera- ration rates became constant after 21 h (t-test for paired ture was controlled (+2OC) by 2 cooling units (Aqua- comparisons between 21 and 27 h: t = 0.0934, df = 5, p = netics Systems, San Diego, CA) that recirculated the 0.93); thus all nutrient uptake experiments began after bath water and by a layer of shadecloth placed over the 21 h. Ammon~umleaching from the cut surfaces of the chambers. Even under the shade, light inside the rhizome and roots was measured in the experimental chambers was always between 140 and 300 pm01 m-2 chambers (see 'Uptake experiments'; n = 7 experimen- S-'. The surfgrass habitat at the study site was fre- 270 Mar Ecol Prog Ser 149: 267-277, 1997

quently quite turbid due to resuspension of detritus without the rhizome (n = 9 replicates). There was no and sediments in the surf zone, and light was 69 pm01 sign~ficantdifference between rates (p > 0.05, Mann- photons m-' SS' over the leaf canopy and 29 pm01 pho- Whitney U-test). tons m -' S' within the canopy on the one ocassion light Nitrate+nitnte uptake of leaves without attached was measured in situ (4~quantum sensor, Biospheri- rhizome and roots was measured at the initial nitrate cal, Inc., San Diego, CA). The total CO, content of the concentrations of 2, 4, 8, 24 and 32 pM. Two plants water inside the leaf chambers decreased by 0.2 to (2 chambers, each with a shoot inside) and 2 controls 0.4 mm01 COz 1-' during a leaf N03-+N02- uptake (2 chambers without plants) were used at each concen- experiment, and final pH values were always ~8.6; tration. Experiments were performed on 13 and 4, 11, thus, carbon should not have been limiting. 18 and 25 May 1994. Ammonium and nitrate+nitrite uptake rates were Nitrogen uptake rates are controlled by the different estimated following the 'multiple flask' method (De internal N pools (D'Elia & De Boer 1978, Harrison & Boer 1981) from the decline in concentration in the Druehl 1982. Thomas & Harrison 1985, Hwang et al. water inside the chambers during periods of 2 to 3 h. 1987, Pedersen 1994, McGlathery et al. 1996). We Water samples (5 ml; n = 3) were collected from the measured the internal No3- pool in surfgrass leaves chambers at constant time intervals (45 to 60 min) and (n = 6 each) from 2 points (about 100 m apart) at each kept cool (4°C)and dark until analysis (less than 24 h). of 3 places within the study area (Fig 1) following Ammonium reagents were added either immediately methods of Williams & Herbert (1989).Dissolved NO3- or within 2 h after finishing the experiment. After the concentrations ranged from 0.57 + 0.23 and 1.22 + experiment, the leaf surface a.rea of each plant frag- 0.03 pm01 N g dry wt-' and there were not significant ment was measured using a Li-3100 Area Meter (Li- differences between places (Kruskal-Wallis ANOVA, Cor, Inc., Lincoln, NE, USA) and then the plant frag- p = 0.17).We assumed, therefore, that our results are ment was dried at 65'C for 24 h and weighed. not affected by differences in the internal NO3- pools A llnear regression between the nutrient concentra- between shoots. However, this is only a partial analysis tion inside the chamber and time was obtained for of the problem (Pedersen 1994, McGlathery et al. each chamber. If there were non-linearities in the time 1996), and we cannot exclude the possibility that our course, only the initial linear portion was used. The results might be affected by differences in other inter- mean of the slopes obtained in the 2 control chambers nal N pools (NH,", amino compounds) between shoots. (without a plant inside) at each concentration was sub- To assess the effect of NH,+ availability to the roots tracted from the slope obtained in each of the 2 cham- on leaf NH,' uptake rates, 3 experiments were per- bers with a plant inside at the same concentration to formed (during November and December 1994) where calculate the net apparent uptake rates, corrected for the leaf and root uptake rates of NH4+were measured volume changes along the experiment. The uptake in shoots with or without NH4' available to the roots. rates were standardized both by biomass (g dry wt) Each experiment had 5 replicates for each of the and shoot leaf surface (cm2,2 sides). following conditions: (1) shoots with attached rhizome The uptake of NH4+and NO3'+ NO2- was described and roots that had 16 pM in the leaf compartment and using a Michaelis-Menten type model V = (V,,,,, [q)/ 20 pM in the root compartment, (2) shoots with (K,,, + [q),where is the initial concentration of the attached rhizome and roots that had 16 pM in the leaf nutrient inside the chamber (PM), Vis the apparent net compartment and

RESULTS -- MJJASONDJFMAMJJASONDJ Surfgrass habitat characterization 1993 1994

Water temperature at the study site varied from 14.5-16°C in January-March 1994 to 22-23 5°C in July- September 1994 (Fig. 2a). The average concentration of

NO,+ NO2- in the water column was 0.40 + 0.10 (mean rc_ SE) PM, and 0.42 + 0.11 pM for NH,' (see Table 4). Nitratei-nitrite concentrations reached maximum values (=l or 2 pM) in July and August for 1993 and in August to 'r S I ';1 L 8 October for 1994 (Fig. 2b). Ammonium concentration in h, the water column showed no seasonal pattern from May to December 1994 (Fig. 2c). The availability of DIN In the JASONDJFMAMJJASONDJ sediment was 1 order of magnitude higher than in the water column. Eh values were always positive, ranglng from +258 to +326 mV (Table 1).

Uptake experiments

Typical time-courses of NH,' and No3-+ NO2- con- centrations inside the chambers are presented for the experiments performed on 17 August 1994 (Fig. 3a) and 13 April 1994 (Fig 3b), respectively. Leaf NH,+ uptake fitted a Michaelis-Menten type model both for MJJASOND 1994 Table 1. Ambient nutrient levels at the study site (False Point, Fig. 2. (a) Temperature; (b) NO,-+NO2-; and (c) NH; concen- La Jolla, CA, USA). n: number of samples tl-ations of the water at False Point during the period of study. Locations- (0) 'shore', [m]'surface', and [A)'surfgrass' Bars Sampling station Mean * SE (PM)n - indicate * 1 SE (nvalues in Table 1) Water column Nitrate+nitrite Shore 0 22 i 0.05 41 Surface 0.42 + 0.12 18 each individual experiment and for the pooled ]results Surfgrass 0.40+0.10 18 of the 4 experiments (Table 2, Fig. 4). The coefficient of Ammonium Shore 0.85 + 0.33 21 determination of the Woolf plot of the data standard- Surface 038 + 0.10 17 ized by dry mass was R2= 0.65, resulting in a signifi- Surfgrass 0.42 r 0.1 1 17 cant regression coefficient (p < 0.0001) and constant Interstitial water (p < 0.0001). The results were similar when standard- Nitrate+nitt-lte Phyllospadixpatch 4.78 2 0.91 5 ized by surface area. The mean (*SE) K,,, for the 4 Ammonium Ph}/llospadix patch 9.58 * 1.43 9 experiments was 17.4 * 5.60 pM, and the mean V,,,,, Bare sand 12.29 t 2.54 7 was 125.1 + 26.43 pm01 N g dry wt-' h-' Over sed~mentsurface 3 05 + 0.59 3 Leaf NO3-+NO2- uptake also fitted a Michaelis- E h PhyLlospadlxpatch +285 + 19 mV 8 Menten type model for each individual experiment Bare sand +326*201nV 3 Over sediment surface +258 mV 1 and for the pooled results of the 5 experiments (Table 2, Fig. 5). The coefficient of determination of the 272 Mar Ecol Prog Ser 149: 267-277, 1997

l l I IIII

- a -

-

0 V,, = 110.1 prnol-h'g DW' h.' 0.0 0.5 1.0 1.5 2.0 Time (h)

Time (h)

Fig. 3. Typical time-courses of (a) NH4+(17 Aug 1994) and (b) NO3-+NO2- (13 Apr 1994) concentration inside the cham- bers during the experiments. (A,v) Control chambers; (a,m) plant chambers. Bars represent *l SE (n = 3) Fig. 4.Leaf NH,' uptake rates of Phyllospadix torrey~.(a) Vvs plot of all the experiments performed. The line represents Woolf plot of the data standardized by dry mass was the Michaelis-Menten equation obtained for the pooled RZ = 0.59, resulting in a significant regression coeffi- results (see Table 2). (b) Woolf plots. All points with the same symbol belong to the same experiment: (0) 27 Jul 1994, (0) cient (p < 0.0001) and constant (p < 0.01). The results 3 Aug 1994, (U) 10 Aug 1994. and (m)17 Aug 1994. Bars rep- were similar when standardized by surface. The mean resent i l SE (n = 2) (*SE) K,,, for the 5 experiments was 10.1 & 2.17 PM, and the mean V,,, was 54.5 & 9.16 pm01 N g dry wt-' leaves changed maximally by 58% and differed little, h-'. K, values increased from the first experiment (13 was lower, or was higher than when the roots were not April 1994) to the last (25 May 1994); V,,, values supplied with NH,'. Only in 1 of the 3 experiments was showed no clear trend (Table 2). the difference significant. Hence no strong effect of Ammonium depletion in the root chambers was NH,' availability to the roots was evident, as the small (Fig. 6) and on the order of the estimated lower pooled results of the 3 experiments show (Table 3). The limit of sensitivity of the experimental setup used lack of a trend in these results was not surprising given (0.5 yM h-', or 0.1 pm01 g dry wt-' h-'). The results that roots had limited capacity for taking up NH4+. from 2 other experiments supported the same pattern The density of shoots in October (probably varies of minimal or undetectable (within limits of experi- seasonally) within surfgrass patches was 6232 * 416 mental protocol) root uptake when roots were supplied (n = 11) shoots m-'. The total biomass of surfgrass was with 20 pM NH4+,and suggest that roots and rhizomes 2576 * 255 g (ash-free dry wt) m-' within the patches contributed little to N acquisition by surfgrass. sampled. Surfgrass allocates the majority of its biomass Leaf uptake of NH,' was not affected by the avail- to leaves; the percentages of biomass allocated to ability of NH,' in the root chambers (Table 3). When leaves, rhizomes and roots were 76.6 r 1.96%, 19.9 + NH4+was supplied to the roots, the uptake rate by the 1.82%, and 3.5 0.34%, respectively. Terrados & Williams: Surfgrass nitrogen uptake 273

Table 2. Linear regressions of Iq/V vs (Woolf transformation) and denved parameter estimates of the Michaelis-Menten model for NH,' and No3-+NO2 uptake by leaves of Phyllospadix torrey~S. Watson Results of each experiment performed and the pooled results of the experiments al-e presented. n: number of points used in the regression model

Ammonium Nitrate + nitrate

Date K", vm,, Date K,, L'mrLk Regression parameters [PM) (pm01 g dry wt"' h-')- Regression parameters (PM) (pm01 g DW-' h-') 27 Jul1994 13 Apr 1994 Constant (SE)0.125 (0.051) 12.64 100 76 Constant (SE) 0 176 (0.099) 4.39 24.97 Coeff. (SE) 0.010 (0.005) Coeff (SE) 0.040 (0.004) RL 0.643 R2 0.973 n 4 n 5 3 Aug 1994 4 May 1994 Constant (SE) 0 166 (0.027) 33.97 204.32 Constant (SE) 0 123 (0.098) 7.42 60.56 Coeff. (SE) 0.005 (0.003) Coeff. (SE) 0.016 (0.004) R2 0.619 R2 0.868 n 4 n 5 10 Aug 1994 11 May 1994 Constant (SE) 0.093 (0.027) 9.30 99.61 Constant (SE) 0.214 (0.233) 9.22 42.97 Coeff. (SE) 0.010 (0.001) Coeff. (SE) 0.023 (0.009) R2 0.943 R" 0.679 n 5 n 5 17 Aug 1994 18 May 1994 Constant (SE) 0.143 (0.084) 13.65 95.69 Constant (SE) 0.168 (0.115) 12.70 75.47 Coeff. (SE) 0.010 (0.007) Coeff. (SE) 0.013 (0.004) 0.539 R2 0 749 n 4 n 5 25 May 1994 Constant (SE) 0.248 (0.043) 16.98 68.50 Coeff. (SE) 0.015 (0.002) R' 0 959 n 5 Pooled experiments Pooled experiments Constant (SE) 0.129 (0.002) 14.24 110.16 Constant (SE) 0.186 (0.065) 8.69 46.67 Coeff (SE) 0 009 (0.002) Coeff (SE) 0.021 (0.004) R' 0.652 R2 0 591 n 17 n 25

DISCUSSION other seagrass meadows (reviewedin Moriarty & Boon 1989). The low values of DIN observed in the water Surfgrass lives in a nutrient-poor environment at the column were comparable to those previously reported study site. Concentrations of DIN in both the water col- for Southern California coastal waters (Jackson 1977, umn and interstitial waters of the sediment (Fig. 2, Zimmerman & Kremer 1984). The Eh values measured Table 1) are in the low range of the values found in in the interstitial water (Table 1)are characteristic of

Table 3 Leaf NH,' uptake rates of surfgrass shoots w~thor ~v~thoutNH.,' available to the roots. In~t~alNH,' concentration in the leaf chambers: 16 PM. Initial NH,' concentration in the root chambers:

Date Ammonium in the root chamber Leaf ammonlum uptake rate: (PM) mean 2 SE (pmol g -DW-' - h-') 3 Nov 94 20 1 72 + 0.41 < 1 1.47 t 0.24 17 Nov 94 20 < 1 8 Dec 94 20 < 1 Pooled experiments Mar Ecol Prog Ser 149: 267-277, 1997

g - K,,. = 8.7FM a - - VmaX= 46.7pmol-N g DW' K' --,- 70 - - n3 60- Cg 50 -

0 5 10 15 20 25 30 35 40 0.0 0.5 l 0 1.5 2.0 2.5 3 0 [NO3'tNO2'], pM Time (h)

Fig. 6 Mean change In NH,4'concentrations ~nsideroot cham- [S] I V = 0.186 +0.021 [S] bers wth (a,m) or without (A)NH,+ ava~lablein the leaf com- r2=0.59.P< OOOOI partment, and In control chambers (with no plant ~nside,U) 1.6 durlng the 3 Nov 1994 experiment Bars [represent *SE of 5 different root chambers

Our results suggest that surfgrass leaves are the pri- mary source of acquired nitrogen and that the roots function primarily in attachment. Although surfgrass roots might take up small amounts of nitrogen, we could not accurately quantify the uptake rate because the changes in concentration in the root chambers were too small (0.1 to 0.2 pm01 N g dry wt-' h-". The contention that surfgrass must obtain most of its N via the leaves is supported by the uptake kinetic results and by the allocation of most of the plant biomass to leaves, in contrast to other seagrass species that live on Fig. 5. Leaf N03-+N02-uptake of Phyllospadix torreyi. (a) V soft substrata (Stevenson 1988, Hillman et al. 1989). vs [S] plot of all the experiments performed. The line repre- sents the Michaelis-Menten equation obtained for the pooled The independence of leaf uptake of NH,' from its results (see Table 2). (b) Woolf plots. All points with the availability to the roots (Table 3) also supports the con- same symbol belong to the same experiment: (0) 13 Apr clusion that most of N acquired by the plant must be 1994, (0) 4 May 1994, (0) 11 May 1994, (m)18 May 1994, and channelled through the leaves. (A) 25 May 1994. Bars represent *l SE (n = 2) Seagrasses tend to allocate more biomass to rhi- zomes and roots in sediments where the availability of coarse, well-oxygenated sediments or sediments poor nutrients is low (Short 1983, 1987, Perez et al. 1994). in organic matter (ZoBell 1946), and are similar to Although the surfgrass we studied grows in a nutrient- those found in the upper 2 to 3 cm of wave-disturbed poor environment, the shallow sediment layer and the Thalassia testudinum sediments (Patriquin 1972). presence of rocky substratum beneath might impose Ammonium and nitrate+nitrite uptake rates by overall restrictions to surfgrass rhizome and root bio- leaves of surfgrass are within the range of those found mass allocation. The strength of the root/rhizome in other seagrasses (Iizumi & Hattori 1982, Thursby & attachment in surfgrass is essential for survival in its Harlin 1982, 1984, Short & McRoy 1984, Paling & high-energy habitat (Williams 1995) and indicates that McComb 1994, Stapel et al. 1996, Pedersen et al. 1997) both biomechanical and physiological functions need and macroalgae (Haines & Wheeler 1978, O'Bnen & to be included in considerations of the potentially Wheeler 1987, Pedersen 1994, McGlathery et al. 1996). adaptive significance of biomass allocation stra.tegies Surfgrass leaf uptake rates of NH,' were higher than in seagrasses, particularly in high-energy environ- those of NO3-+NOz-(Table 2) as has been also found ments (Williams 1995). in Zostera marina (Short & McRoy 1984) and Amphibo- The rapid water motion of surfgrass habitat probably lis antarctica (Paling & McComb 1994). assures a rapid flux of nutrients to the plants. If the Terrados & Williams: Surl'grass nltrogen uptake 275

concentrations of NO:,-+ NO,- and NH,' in the water LITERATURE CITED column at the collection site (Table 1, Fig. 2) are Alpert P (1991) Nitrogen sharlng anlong raniets Increases compared with the values of K,,, and V,,, obtained clonal growth in Fragana chjloensis. Ecology 72 69-80 (Table 2, Figs. 4 & 5), leaf uptake of DIN by surfgrass Alpel-t P, Mooney HA (1986) Resource sharing among ramets would not be saturated under natural conditions. In the clonal herb, Fragaria ch~loens~s.Oecologla 70: Plants could exploit, however, sporadic nutrient pulses 227-233 advected to the surfgrass beds as has been suggested American Public Health Association (l992) Standard methods for the examination of water and wastewater American for Amphibolis antarctica (Pedersen et al. 1997). A Publlc Hedlth Association, LVashington, DC complete evaluation of the nutrient supply to surfgrass Atklnson MJ, Bilger RM' (1992) Effects of water velocity on leaves requires a study of how uptake rates vary with phosphate uptake in coral reef-flat coniniunities. Llmnol water flow speed. Oceanogr 37:273-279 Seagrasses are clonal plants and the existence of Barnabas AD (1991) Thalassodendron cilia tun^ (Forssk.)den Idartog: root structure and hlstochelnistry in ~relatlonto interactions between ramets is undoubtedly important, dpoplastlc transport. Aquat Bot 40:129-143 as has been extensively shown in their land counter- Barndbas AD (1994) Anatomical, histochemical and ultra- parts (Hartnett & Bazzaz 1983, Pitelka & Ashmun 1985, structural features of the seagrass Phyllospadlx scoulen Alpert & Mooney 1986, Slade & Hutchings 1987a, b, c, Hook. Aquat Bot 49:167-182 Borum J, Murray L,Michael Kemp W (1989) Aspects of nitro- Alpert 1991). Interactions occur both within (leaf-root gen acquisition and conservation in eelgrass plants. Aquat interactions; Thursby & Harlin 1982, 1984) and among Bot 35 289-300 (interactions between connected shoots; Tomasko & Cooper LW, McKoy CP (1988) Anatomical adaptations to Dawes 1989, Tomasko et al. 1991) seagrass ramets. rocky substrates and surf exposure by the seagrass genus The presence of interactions between leaves and roots Phyllospadix. Aquat Bot 32:365-381 DeBoer JA (1981) Nutrients. In: Lobban CS, Wynne MJ (eds) in nutrient acquisition has been demonstrated for The biology of seaweeds. Blackwell Scientific Publica- Zostera marina (Thursby & Harlin 1982) and tions, London. p 356-392 maritima (Thursby & Harlin 1984),but not for Thalassia D'Elia CF. DeBoer JA (1978) Nutritional stuciies of two red hemprichii (Stapel et al. 1996).Our results suggest that algae. 11. Klnetics of ammonlum and nltrate uptake. J Phy- c01 14.266-272 these interactions do not occur in Phyllospadix torreyi. den Hdrtog C (1970) The seagrasses of the world North- Although nutrient uptake kinetics prov~desuseful Holland Publishing, Amsterdam insight into the physiological ecology of seagrasses, a Dowd JE. Rlggs DS (1965) A colnparlson of estimates of dynamic understanding of nutrient supply and de- hl~chaelis-Mentenkinetic constants from various linear mand is critically lacking. For example, the nutrient transformations. J Biol Chem 240:863-869 Dudley WR (1894) Phyllospadix, its systematic characters and supply rate to the leaves is a function of both concen- dlstrlbutlon. Zoe 4:381-385 tration and water flow speeds (Lapointe & Ryther 1979, Fonseca MS. Kenworthy CVJ (1987) Effects of cul-rent on Parker 1981, Harrison & Druehl 1982, Fujita 1985, Fon- photosynthesis and distribution of seagrasses. Aquat Bot seca & Kenworthy 1987. Atkinson & Bilger 1992) just as 27.59-78 Fujita RM (1985) The role of nitrogen status in reyulat~ny supply to the roots is in part a function of remineraliza- transient ammonium uptake and nitrogen storage by tion rates. Also, the nutrient concentration gradient macroalgae. J Exp Mar Biol Ecol92:283-301 and perhaps flux differential experienced between the Gibbs RE (1902) Phyllospadlx as a beach-builder. Am Nat 36: aboveground and belowground biomass hypotheti- 101-109 cally mlght influence which is the primary site for Grasshoff K, Ehrhardt M, Kremling K (1983) Methods of sea- water analysis. Verlag Chemlt., CVeinheim nutrient acquisition. In turn, this would influence Guerrero MG, Vega JM, Losada M (1981) The asslmllatory soul-ce-sink activity within a clone and the relationship nitrate-reducing system and ~tsregulation. Ann Rev Plant between ramets. Physiol 32:169-204 Haines KC, Wheeler PA (1978) Ammonium and nitrate uptake by the marine macrophytes Hypnea rnusciforrnis Acknowledgements. This study was funded by the hlinisterio (Rhodophyta) and Macrocystis pyrifera (Phaeophyta). de Educacion y Ciencla, Programa de Becas de Formacion de J Phycol 14:319-324 Personal Investiyador en el Extranjero through a fellowship Harrlson PJ. Druehl LD (1982) Nutrlent uptake and growth in awarded to J.T We thank Dr Joy Zedler for providing access the Larnlnariales and other Macrophytes: a conslderatlon to the flow-injection autoanalyxer and the spectrophotometer. of methods. In: Srivastava LM (ed)Synthetic and degrada- Jim Zlrnmer provided valuablp assistance in the construction tive processes in marine rnacrophytes. W de Gruvter & CO, of the incubation chambers and the water bath, as dld Angela New York, p 99-120 Andrews for the autoanalyzer. Pat Ewanchuk, Jeanne Burch, Hartnett DC. Bazzaz FA (1983) Physiological integration Amy Sewell. Phillip Larkin and Dr Thorsten Reusch helped among intraclonal ramets in Solidago canadensis. Ecology in the field and laboratory. We are also thankful for the 64:?79-788 comments and thorough editing of a previous version of the Hemminga MA, Harrison PG, van Lent F (1991) The balance manuscript by Dr Morten F Pedersen, as well as those of 2 of nutrlent losses and gains in seagrass meadows. Mar anonymous reviewers. This 1s contribution 280 from the Ecol Prog Ser 71:85-96 Coastal and h4arine Instltute, San Diego State Un~versity. Hemminga MA, Koutstaal BP, van Soelen J, Merks AGA Mar Ecol Prog Ser 149: 267-277, 1997

(1994) The nitrogen supply to intertidal eelgrass (Zostera 101:169-177 marina). Mar Biol 118:223-227 Pedersen MF, Paling EI, Walker D I (1997) Nitrogen uptake Hillman K, Walker DI, Larkum AWD, McComb AJ (1989)Pro- and allocation in the seagrass A~nphibolis antarctica. ductivity and nutrient limitation. In: Larkum AWD, Aquat Bot (in press) McComb AJ, Shepherd SA (eds) Biology of seagrasses. Penhale PA, Thayer GW (1980) Uptake and transfer of carbon Elsevier, Amsterdam, p 635-685 and phosphorus by eelgrass (Zostera marina L.) and its Hwang SP, Williams SL. Brinkhuis BH (1987) Changes in epiphytes. J Exp Mar Biol Ecol42:113-123 dissolved nitrogen pools as related to nitrate uptake and Perez M, Duarte CM, Romero J, Sand-Jensen K, Alcoverro T assimilation in Gracilaria tikvahiae McLachlan (Rhodo- (1994) Growth plasticity in Cymodocea nodosa stands: the phyta). Bot Mar 30:ll-19 importance of nutrient supply. Aquat Bot 47:249-264 ~izumiH, Hattori A (1982) Growth and organic production of Pitelka LF, Ashmun JW (1985) Physiology and integration of eelgrass (Zostera marina L.) in temperate waters of the ramets in clonal plants. In: Jackson JBC. Buss LW, Cook Pacific coast of Japan. 111. The kinetics of nitrogen uptake. RE (eds) Population biology and evolution of clonal organ- Aquat Bot 12:245-256 isms. Yale University Press, New Haven, p 399-435 Jackson GA (1977) Nutrients and production of giant kelp, Ricketts EG, Calvin J, Hedgpeth J. Phillips DW (1985) Macrocystis pyrifera, off southern California. Limnol Between Pacific tides. Stanford University Press, Palo Alto Oceanogr 22:979-995 Roth NC, Pregnall AM (1988) Nitrate reductase activity in Lapointe BE, Ryther JH (1.979) The effects of nitrogen and Zostera marina. Mar Biol 99:45?-463 seawater flow rate on the growth and b~ochemicalcompo- Segel 1H (1968) Biochemical calculations. John Wiley & Sons, sition of Gracilaria foljifera var angustissima in mass out- Inc, New York door cultures. Bot Mar 22:529-537 Short FT (1983)The seagrass Zostera marina L.: plant mor- McGlathery KJ, Pedersen MF, Borum J (1996) Changes in phology and bed structure in relation to sediment ammo- intracellular nitrogen pools and feedback controls on nium in Izembek Lagoon, Alaska. Aquat Bot 16:149-161 nitrogen uptake in Chaetomorpha linum (Chlorophyta). Short FT (1987) Effects of sediment nutrients on seagrasses: J Phyco132:393-401 literature review and mesocosm experiment. Aquat Bot McRoy CP, Barsdate RJ (1970) Phosphate absorption in eel- 27:41-57 grass. Limnol Oceanogr 15:6-13 Short FT, McRoy CP (1984) Nitrogen uptake by leaves and McRoy CP, Goenng JJ (1974) Nutnent transfer between the roots of the seagrass Zostera marina L. Bot Mar 27(12). seagrass Zostera marina and its epiphytes. Nature 248 547-555 173-174 Slade AJ, Hutch~ngsMJ (1987a) The effects of light intens~ty McRoy CP. Lloyd DB (1981) Comparative function and stabil- on foraging in the clonal herb Glechoma hederacea. J Ecol ity of macrophyte-based ecosystems. In: Longhurst AR 75:639-650 (ed)Analysis of marine ecosystems. Academic Press, Lon- Slade AJ, Hutchings MJ (1987b) Clonal integration and plas- don, p 473-489 ticity in foraging behaviour in Glechoma hederacea. J Ecol McRoy CP, h4cMillan C (1977) Production ecology and physi- 75:1023-1036 ology of seagrasses. In: McRoy CP, Helfferlch C (eds) Sea- Slade AJ, Hutchings MJ (1987~)The effects of nutrient avail- grass ecosystems: a scientific perspective. Marcel Dekker, ability on foraging in the clonal herb Glechoma heder- Inc, New York, p 53-87 area. J Ecol75.95-112 Moriarty DJW, Boon PI (1989) lnteract~onsof seagrasses with Sokal RR, Rohlf FJ (1981) Biometry. The principles and prac- sediment and water In: Larkum AWD, McCornb AJ, tice of statistics in biological research. WH Freeman and Shepherd SA (eds) Biology of seagrasses. Elsevier, Ams- CO,New York terdam, p 500-535 Stapel J, Aarts TL, van Duynhoven BHM, de Groot JD. van O'Brien MC. Wheeler PA (1987) Short term uptake of nutri- den Hoogen PHW, Hemminga MA (1996) Nutrient uptake ents by Enteromorpha prolifera (Chlorophyceae).J Phycol by leaves and roots of the seagrass Thalassia hemprichii 23:547-556 (Ehrenb.) Aschers. in the Spermonde Archipielago, Paling El, McComb AJ (1994) N~trogen and phosphorus Indonesia Mar Ecol Prog Ser 134:195-206 uptake in seedlings of the seagrass An~ph~bol~santarctlca Stevenson JC (1988) Comparative ecology of submersed in Western Australia. Hydrobiologia 294:l-4 grass beds In freshwater, estuarine and marine environ- Parker HS (1981) Influence of relative water motion on the ments. Limnol Oceanogr 33:867-893 growth, ammonlum uptake, and carbon and nltrogen com- Stewart JG (1989) Maintenance of a balanced, shifting position of Ulva latuca (Chlorophyta). Mar Biol63:309-318 boundary between the seagrass Phyllospad~xand algal Parsons TR. Maita Y, Lalli CM (1984) A manual of chemical turf. Aquat Bot 33:223-241 and, biological methods for seawater analysis. Pergamon Stewart JG, Myers B (1980) Assemblages of algae and inver- Press, Oxford tebrates in Southern Californ~aPhyllospadjx-dominated Patriquin DG (1972) The origin of nitrogen and phosphorus intertidal habitats. Aquat Bot 9.73-94 for growth of the marine angiosperm Thalassia tes- Thomas TE, Harnson PJ (1985) Effect of nitrogen supply on tudinum. Mar Biol 15:35-46 nitrogen uptake, accumulation and assimilation in Por- Pedersen MF (1994) Transient ammonium uptake in the phyra perforata (Rhodophyta). Mar B101 85:269-278 macroalga Ulva lactuca (Chlorophyta): nature, regulation Thursby GB, Harlin MM (1982)Leaf-root interaction in the up- and the consequences for choice of measuring technique. take of ammonia by Zostera marina. Mar Biol 72:109-112 J Phycol30:980-986 Thursby GB, Harlin MM (1984)Interaction of leaves and roots Pedersen MF, Borum J (1992) N~trogendynamics of eelgrass of Ruppia maritima in the uptake of phosphate, ammonia Zostera marina during a late summer period of high and nitrate. Mar Biol 83:61-67 growth and low nutrient ava~labllity.Mar Ecol Prog Ser 80: Tomasko DA, Dawes CJ (19891 Evtdence for phys~ologl- 65-73 cal integration between shaded and unshaded short shoots Pedersen MF, Borum J (1993) An annual nitrogen budget for of Thalassia testudinum. Mar Ecol Prog Ser 54:299-305 a seagrass Zostera marina population. Mar Ecol Prog Ser Tomasko DA, Dawes CJ, Hall MO (1991) Effects of the num- Terrados & Williams: Surfgrass nitrogen uptake 277

ber of short shoots and presence of the rhizome apical Whitfield M (1974) Thermodynamic limitations on the use of meristem on the survival and growth of transplanted sea- the platinum electrode in Eh measurements. Limnol grass Thalassia tesludinum. Contrib Mar Sci 32:41-48 Oceanogr 19:857-865 Turner T (1983a)Complexity of early and middle successional Williams SL (1995) Surfgrass (Phyllospadix torreyi) reproduc- stages in a rocky intertidal surfgrass community. Oecolo- tion: reproductive phenology, resource allocation and gia 60.56-65 male rarity. Ecology 76:1953-1970 Turner T (1983b) Facilitation as a success~onalmechanism in Wllliarns SL, Herbert SK (1989) Transient photosynthetic a rocky intertidal comn~unity.Am Nat 121 729-738 responses of nitrogen-deprived Petalonia fascia and Lam- Turner T (1985) Stability of rocky intertidal surfgrass beds: ~nariasaccharina (Phaeophyta) to ammonium resupply. Persistence, preemption and recovery. Ecology 66:83-92 J Phycol 25515-522 Turner T. Lucas J (1985) Differences and similarities in the Zlrnrnerrnan RC, Kremer JN (1984) Episodic nutrient supply community roles of three rocky intertidal surfgrasses. to a kelp forest ecosystem in Southern California. J Mar J Exp Mar Biol Ecol89:175-189 Res 42:591-604 Whitfield M (1969) Eh as an operational parameter in estuar- ZoBell CE (1946) Studies on redox potential of marine sedi- ine studies. Limnol Oceanogr 14:547-558 ments. Bull Am Assoc Petrol Geol 30:477-513

This article was submitted to the editor Manuscript first received: November 7, 1996 Revised verslon accepted: February 14, 1997