International Association for Ecology

Effect of Periphyton Biomass on Hydraulic Characteristics and Nutrient Cycling in Streams Author(s): P. J. Mulholland, A. D. Steinman, E. R. Marzolf, D. R. Hart and D. L. DeAngelis Reviewed work(s): Source: Oecologia, Vol. 98, No. 1 (1994), pp. 40-47 Published by: Springer in cooperation with International Association for Ecology Stable URL: http://www.jstor.org/stable/4220664 . Accessed: 06/03/2013 19:47
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Springer and International Association for Ecology are collaborating with JSTOR to digitize, preserve and extend access to Oecologia.

http://www.jstor.org

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions

Oecologia (1994) 98:40^7 ORIGINAL PAPER ? E. R. Marzolf

? Springer Verlag 1994

? A. D. Steinman P. J. Mulholland ? D. R. Hart D. L. DeAngelis Effect and of nutrient periphyton cycling

biomass in streams

on

hydraulic

characteristics

Received: 13 January 1994 / Accepted 3 February 1994

The effect of periphyton biomass on hydraulic Abstract characteristics and nutrient cycling was studied in laborsnail herbivores. atory streams with and without Hysuch as average water velocity, draulic characteristics, and relative volume of transient coefficients, dispersion storage zones (zones of stationary water), were quantiof a conservafied by performing short-term injections model to tive tracer and fitting an advection-dispersion the conservative tracer concentration profile downstream from the injection site. Nutrient cycling was quantified by measuring two indices: (1) uptake rate of phosphorus to gross primary producfrom stream water normalized tion (GPP), a surrogate measure of total ? demand, and in the periphyton matrix. (2) turnover rate of phosphorus These measures indicate the importance of internal cycling (within the periphyton matrix) in meeting the ? deDense growths of filamentous diamands of periphyton. in the streams toms and blue-green algae accumulated with no snails (high-biomass streams), whereas the periin communities streams with snails consisted alphyton most entirely of a thin layer of basal cells of Stigeoclonium sp. (low-biomass coefficients streams). Dispersion were significantly storage zones greater and transient were significantly streams larger in the high-biomass the streams. Rates of GPPto low-biomass compared ? uptake from water and rates of ? turnover normalized in periphyton were significantly lower in high biomass than in low biomass periphyton communities, suggesting that a greater fraction of the ? demand was met by recyP. J. Mulholland (IS) ?E. R. Marzolf ?D. L. DeAngelis Environmental Sciences Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN, 37831-6036, USA A. D. Steinman Department of Research, South Florida Water Management District, West Palm Beach, FL 33416-4680, USA D. R. Hart Department of Mathematics and Computer Science, Knox College, Galesburg, IL 61401, USA

in communities. Increases cling in the high biomass ? concentration streamwater increased significantly ? uptake in high biomass communities, GPP-normalized of nutrient transfer from diffusion limitation suggesting stream water to algal cells in these communities. Our results demonstrate that accumulations of periphyton bioof streams, mass can alter the hydraulic characteristics transient by increasing particularly storage zones, and can increase internal nutrient cycling. a They suggest of hydraulic characteristics close coupling and nutrient cycling processes in stream ecosystems. Stream periphyton ? Nutrient cycling Key words Stream hydraulics ? Transient storage zones

Introduction communities can be important in periphyton in and stream uptake cycling ecosystems Paul and Duthie Triska et al. 1987; 1989; (Grimm 1989a), relatively little attention has been devoted to the factors that influence nutrient cycling within the periphyton matrix. Several studies have shown that nutrient limitation and nutrient cycling within periphyton communities are intensified as periphyton biomass increases Paul and Duthie Mulholland et al. 1989; 1989; (Bothwell 1991; Peterson and Grimm 1992). This finding is generally attributed to reduced advective transport of nutrients from flowing water to the lower layers of cells within the increased dematrix, and as a consequence, periphyton on diffusive slow propendence transport, a relatively cess. Thus, Bothwell has noted that (1989) high-biomass stream periphyton can be nutrient-limited communities even when nutrient concentrations in stream water are well above those that saturate the growth rates of individual cells. Nutrient cycling characteristics of stream periphyton communities therefore be may closely linked to stream of the effects of characteristics. Past studies hydraulic on nutrient dynamics in streams hydraulic characteristics Although nutrient

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions

41 have focused primarily on the influence of water velocity and Schumacher on nutrient uptake (Whitford 1964; Horner and Welch 1981), and on phosphorus uptake et al. 1991). However, few studies halength (D'Angelo other ve addressed the role of hydraulic characteristics An exception than water velocity on stream periphyton. has been a recent study of a stream in Arizona where difvs. ferences in streambed hydraulic gradient (upwelling areas) were strongly related to periphyton downwelling biomass (Valett et al. 1994). characteristics Stream hydraulic may also be influenced by periphyton biomass. Using an approach involvof an tracer injections and application ing conservative model to the tracer data, Kim et al. advection-dispersion on artificial accumulations (1990) found that periphyton substrata in flumes placed within a natural stream produced stagnant zones of temporary water storage, termed transient storage zones. These transient storage zones, in turn, influenced the overall transport velocity of nutrients downstream. we have developed a simulation model of Recently, stream ecosystems that includes a zone of stationary water adjacent to the stream bottom in which the periphyton biota reside (DeAngelis and other attached et al. in press). The stationary water zone included in our stream model can be viewed conceptually as a transient storage zone in the advection-dispersion model of stream solute by Bencala and Walters (1983) and dynamics developed used by Kim et al. (1990) in the study cited above (see also Stream Solute Workshop 1990 and references therebetween this nearin). Soluble nutrients are exchanged bottom, stationary water zone and the overlying, free-flowing stream water via diffusion. Uptake and release of nutrients by periphyton occur only within the stationary water zone. Steady state simulations of this model have indicated that nutrient limitation is more intense and nutrient cycling with larger stationary greater in systems water zones (DeAngelis et al. in press). The objective of this study was to experimentally determine the effect of periphyton biomass on hydraulic and nutrient cycling in streams. The folcharacteristics were tested: lowing hypotheses 1. Transient storage zones are created by accumulations of periphyton and the size of these zones increases with increasing periphyton biomass. 2. Nutrient cycling within periphyton communities increases with increasing biomass. The study was performed in periphyton-dominated laboratory streams. Transient storage zones were identified tracer injections and nutrient cycling using conservative was determined from metabolism-normalized uptake and turnover of radiolabeled in phosphorus periphyton growing in the streams. that had been passed through a sand filter to remove most paniculate material. Water depth was approximately 4 cm and average water velocity was approximately 7 cm/s in the channels. Light was supplied by a bank of 400-W metal halide lamps suspended above the streams, which provided about 100 ?p??? quanta nr2 s-1 to the water surface. The photoperiod was set at 10L:14D. About 18,000 solid unglazed ceramic cylinders ( 1.6 cm diameter ? 1.6 cm length) were placed on the bottom of each stream to serve as substrata for periphyton. The streams were inoculated with periphyton by placing rocks collected from Walker Branch, a nearby stream, in each channel. Periphyton also was supplied in the incoming water to each stream because the sand filters were not completely effective in excluding small cells. To achieve contrasting levels of periphyton biomass, snails (Elimia clavaeformis, a prosobranch snail) were added to four of the eight streams at densities of 1000 individuals/m2. Elimia is the dominant herbivore in most of the natural streams in the area, and densities similar to those used in the laboratory streams maintain very low periphyton biomass in the natural streams (Hill et al. 1992; Rosemond in press). Periphyton had been growing in the experimental channels for more than 2 years prior to the study, although the channels with no snails were occasionally scoured to remove large accumulations of periphyton. The streams were scoured approximately 5 months prior to the beginning of the experiments reported here. To determine hydraulic characteristics of the channels, shortterm injections of a 0.3 m NaCl solution were conducted in each channel (Stream Solute Workshop 1990). Each injection consisted of pumping the NaCl solution into the upstream end of each channel at a constant rate for a 30- to 60-min period and measuring electrical conductance (EC, determined with a YSI Model 32 conductivity meter) at 15-s intervals at the downstream end of each channel during and after the injection. The length of injections was varied in order to reach steady state EC values at the downstream station. Pre-injection EC measurements were made at the downstream station and subtracted from EC readings during and after the injection to obtain the increase in EC (EC?) due to the injection. Average streamwater velocity was calculated by dividing the distance between injection and the downstream station by the time to reach one-half of the steady state EC? value at the downstream station. Other hydraulic characteristics, including the dispersion coefficient, exchange coefficient between the freeflowing stream water and the transient storage zone, and the relative volume of the transient storage zone, were estimated by fitting an advection-dispersion model (see below) to the EQ data. The effect of periphyton biomass on different hydraulic characteristics was determined by one-way analysis of variance (SAS 1988). Data were log-transformed where necessary to achieve homogenous variances. The advection-dispersion model used is similar to that presented by Bencala and Walters (1983) and the Stream Solute Workshop (1990) and consists of two partial differential equations describing the rate of change in tracer concentration over time (t) and distance (x): ^+?^=i?(AD^ dc5 dt ?k2(C-Cs) (1)

(2)

Methods Eight indoor fiberglass channels, each 22 m long and 0.3 m wide, were used in the study. Each channel received continuous, oncethrough flow of water (0.37 1/s), supplied from a spring-fed pond,

C=tracer concentration in the free-flowing water Cs=tracer concentration in the stationary water zone (transient storage zone) Q=tracer concentration in lateral inputs Q=stream flow (1/s) A=cross sectional area of the flowing water zone (m2) D=dispersion coefficient (m2/s) /c,=exchange coefficient from free-flowing water to the stationary water zone (s_1) &2=exchange coefficient from stationary water zone to the freeflowing water (s_i) fc,=lateral inflow rate coefficient (s_l)

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions

42 The model is solved analytically using an efficient, converging infinite series approach and the parameters D, /:,, and k2 are estimated by fitting the model to empirical data from conservative tracer injection experiments. The ratio k,: k2 is equivalent to the volume of the transient storage zone relative to the volume of the free-flowing water zone, and is equivalent to the ratio As: A presented by Bencala and Walters (1983). Phosphorus cycling in periphyton communities was estimated by measuring two indices: (1) the ratio of uptake rate of ? from stream water to total metabolic ? demand, the latter estimated by gross primary production (GPP), and (2) turnover rate of ? in the periphyton matrix. These measures indicate the importance of internal cycling (within the periphyton matrix) in meeting the ? demand of periphyton. Cylinders with attached periphyton were collected from one stream with snails (low periphyton biomass) and one stream without snails (high periphyton biomass) and placed into clear, plexiglass chambers to which pumps were attached to continuously circulate water (0.85 1 total volume). From each stream type (low or high periphyton biomass) 45 cylinders were collected, and 15 cylinders of each level of periphyton biomass placed into three chambers, giving three replicates of each biomass level. Filtered water from the appropriate stream (biomass level) was added to each chamber and the chambers were placed into a controlled temperature water bath maintained at 15? C (within 2? C of the ambient stream water) and illuminated by an overhead metal halide lamp; the lamp provided about 100 ?p??? quanta m 2s '. Rates of ? uptake from stream water and GPP were determined simultaneously in each chamber. Approximately 25.7x 103 Bq of carrier-free H333P04 were added initially, and 1-ml subsamples of water were removed from each chamber at 10, 20, 30, 45 and 60 min. The subsamples were assayed for 33P by liquid scintillation spectroscopy and the rate constant for 33P uptake was calculated. Rates of ? uptake from stream water were computed from the 33P uptake rate constants and the concentrations of soluble reactive phosphorus (SRP) in stream water. SRP concentrations were determined by the standard molybdenum blue procedure (American Public Health Association 1989). Rates of GPP were determined by measuring changes in dissolved oxygen concentration (Orbisphere Model 2607 dissolved oxygen meter) during a 3-h period in light and a subsequent 2-h period in which the chambers were darkened. After 24 h of exposure to 33P in the chambers, > 90% of the isotope had been removed from the water in the chambers and taken up by periphyton (past studies have shown that sorption to chamber walls is negligible). At this time, the cylinders were removed and returned to the downstream ends of the channels to measure the turnover rate of ? from the periphyton community as a whole. Three streams without snails were used in this portion of the study, with replicate cylinders from one high biomass and one low biomass chamber being placed into each stream. Only streams that lacked snails were used in the ? turnover study because we wished to preclude the confounding effect of snail grazing on ? turnover. Three cylinders from each biomass level were removed randomly from each stream after 1 h, 1 day, 3 days, 6 days, and 10 days. The cylinders were dried at 80?C for 48 h, weighed, combusted at 500?C for 20 h, and reweighed. The ash free dry mass (AFDM) on each cylinder was computed as the difference between dry mass and mass following combustion. Each combusted cylinder was then leached in 5 ml of 2 m HC1 for 24 h to extract 33P The leachate was diluted 1:1 with distilled water, and 1-ml subsamples were removed and assayed for 33Pby liquid scintillation spectroscopy. The quantities of 33P were corrected for background radioactivity and isotopie decay and normalized to AFDM. The ? turnover rate for each chamber group was computed as the first-order rate coefficient (k) of the decline in 33P activity per unit AFDM over time, according to the following relationship: 33/>=33/^? (3) A second phase of this experiment examined the effect of increased concentration of P04 in water on phosphorus cycling in high biomass periphyton communities. We hypothesized that nutrient cycling is greater in high biomass periphyton communities because of mass transport constraints within transient storage zones formed by periphyton accumulations. Consequently, increases in nutrient concentrations in the free-flowing water should reduce nutrient cycling as a result of increases in mass transport of nutrients into storage zones. To test this prediction, two plastic rain gutters (each approximately 0.5 m long and 10 cm wide) were placed side by side on the bottom of three streams with no snails. Twenty ceramic cylinders with attached periphyton from that stream were placed in each gutter. A concentrated solution of K2HP04 was dripped continuously into the upstream end of one of the two gutters in each stream to increase the P04 concentration in the stream water of this gutter by 10-20 times. The ambient concentrations of P04 (measured as SRP) in the manipulated portions of the stream were 5?2 ?gP/l (mean?range). Four days after the P04 additions began, all cylinders from each gutter were transferred to the plexiglass chambers with filtered water from the appropriate gutter. Rates of ? uptake from stream water and GPP were measured using the methods described previously. Following ? uptake and GPP measurements, the cylinders were removed from the chamber and returned to their respective gutters in the streams. Turnover of 33Pfrom the periphyton communities was determined by randomly removing 4 cylinders 1 h, 3 days, 6 days, and 9 days later. AFDM and 33P content of periphyton on each cylinder were determined as described above. The enrichment of the stream water by P04 in one gutter in each stream was maintained throughout this period. T-tests were used to determine whether the ratio of ? uptake from water to GPP and ? turnover rate were different between high and low biomass periphyton communities in the first phase of the experiment (SAS 1988). ? uptake:GPP ratios were log-transformed prior to statistical analysis to achieve homogeneous variances. A randomized block analysis of variance, with each stream treated as a block, was used to determine effects of ? enrichment on ? uptake:GPP ratios and on ? turnover rate in the second phase of the experiment (SAS 1988). Again ? uptake:GPP ratios were log-transformed to achieve homogeneous variances.

Results Transient storage zones

where 33P, is the 33P activity at any time t and 33/>0 is the initial 33P activity.

Dense growths of filamentous in the algae accumulated streams with no snails, forming an almost continuous mat across the top of the ceramic substrata on the stream bottom (Fig. la). The periphyton in these communities streams were composed of the filamentous primarily Phormidium retzii and the diatoms Achcyanobacterium nanthes minutissima, and Eunotia Amphora perpusilla, minor. In var. the contrast, pectinalis periphyton in streams with snails consisted almost entirely of a thin layer of basal cells of the green alga Stigeoclonium sp. (Fig. of lb). This alga dominates the periphyton communities natural streams in the area because of its resistance to herbivory (Steinman 1992, Rosemond in press). biomass influenced the hydraulic characPeriphyton teristics of the streams as determined by the conservative tracer injection data (EC?) and application of the advecmodel to the EC? data (Table 1). Values tion-dispersion of ECj in the low-biomass streams reached approximate state about 10 min after the injections steady began, whereas EC? in the high-biomass streams did not reach

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions

43 Fig. 1 a, b Photographs showing the difference in periphyton biomass in a stream with no snails, and b stream with snails

Fig. 2 Plot comparing the change in electrical conductance (pre-injection values subtracted from all data points, ?C?) over time in a low biomass stream (with snails) and a high biomass stream (no snails) during an NaCl injection experiment. Note the greater time needed to reach approximate steady state in EC? for the high biomass stream, indicating a larger transient storage zone. Steady state EC? values are not identical in the streams because the injection rates were slightly different

140 Low Biomass Stream; 30 minute injection.

120-

100-

High Biomass Stream; 62 minute injection.

e GO a. CJ

80-

60-

40-

20

-20

I ' ^*-1-'-!-'-G -10 0 10 20

30

1 40 50 60 Time (minutes)

70

80

?" ? I ' ??? 90 100 110

120

Table 1 Hydraulic characteristics estimated from the conservative tracer injections using the advection-dispersion model in low biomass and high biomass channels (mean? SD, n=4). Results of one-way ANOVA are also given Hydraulic characteristic ? (cm/s) D(m2/min) ?.(min-1) ^(min-1) Low biomass 6.6(0.3) 0.77(0.12) 0.013(0.007) 0.438(0.204) 0.030 (0.004) High biomass 6.2(0.6) 1.14(0.20) 0.044(0.010) 0.242(0.052) 0.179 (0.007) F16 1.57 10.3 26.4 3.44 1542 ?

0.257 0.018 0.002 0.113 <0.001

45 min (Fig. 2). The steady state until approximately taken reach time to longer steady state in the high-biomass streams was not the result of different water velocities; the average water velocity, as determined by the time required to reach one-half the steady state EC? value, was similar among all streams (Table 1). Although discoefficients were in the (D) persion significantly greater streams to the low-biomass strehigh-biomass compared ams, the longer time required to reach steady state in EC? in the high-biomass streams was primarily the result of 6 times larger transient storage zone volumes (k{. k2 ratios) in the high-biomass streams compared with the low-bioIn the high-biomass mass streams. transient streams,

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions

44 Storage zone volumes were about 0.18 times as large as zone volumes, whereas in the low-biomass free-flowing were only 0.03 streams transient storage zone volumes zone volumes. The average times as large as free-flowing water residence time in the transient storage zones (l/k2) streams was also somewhat longer in the high-biomass in than the low-biomass streams (2.3 min). (4.1 min) The average distance traveled by a water molecule before entering the transient storage zone can be calculated by dividing the average water velocity by the coefficient for uptake from free-flowing water to the storage zone On water molecules would travel 305 m average, (k{). before entering the transient storage zone in the low-biomass streams, but only 85 m before entering the transient streams, further highstorage zone in the high-biomass biomass on stream hylighting the effect of periphyton draulic characteristics. 6 E 4 2 O (b)PUptake/GPP 10 e ?! 5 h O CJ) E CD =3. 0 .03 .02 T ?b .01 Phosphorus cycling Low Biomass High Biomass 4 0 Ambient High ? Cone. ? Cone. Rate (e)? Turnover (a)Periphyton Biomass (AFDM)

100 80 ^> 60 40 20 0

Stream SRP Concentration (a)

ftj PUptake/GPP

The mean AFDM on ceramic substrata collected from a was apstream with no snails (high-biomass community) 10 times greater than that on substrata colproximately lected from a stream with snails (low-biomass communi? The ratio of from water to GPP 3a). ty) (Fig. uptake was significantly lower for the high biomass community than it was for the low biomass (r=9.81, community P=0.0023), indicating greater internal ? cycling in the high biomass community (Fig. 3b). The SRP concentrations in the stream water of each community type were similar (6 ?g/l). The mean turnover rate of 33P in the high biomass perwas lower than that in the low bioiphyton community between commass community, although the difference munities was reduced when the low biomass community rates were corrected downward to account for growth dilution of the isotope within the biomass (Fig. 3c). Mean increased during AFDM of the low biomass community the turnover portion of the study because snails were excluded from the gutters in which the cylinders were placed (periphyton was released from grazing pressure). Be33P was normalized cause periphyton to AFDM (to reduce variability in periphyton 33P among cylinders due to in AFDM), variability higher 33P turnover rates for the were due, in part, to the reduclow-biomass community tion in 33P/AFDM caused by algal growth. To correct for this growth dilution of 33P, the mean rate of AFDM increase on the cylinders in each low-biomass community replicate (gutter) was computed and subtracted from the 33P turnover rate calculated for that replicate. The high biomass periphyton did not increase in AFDM during the turnover study period, so no growth correction was necessary. Even after correcting the low biomass community turnover rates for growth dilution of 33P, the mean 33P was lower turnover rate for the high biomass community 33P turnover rate for the than the mean growth-corrected lower biomass community (?=2.73, P=0.053). In the second phase of the experiment which involved in SRP concentrations only the high biomass community,

Fig. 3 a-c Comparison of a periphyton biomass expressed as AFDM, b ratio of ? uptake from stream water to GPP, and c phosphorus turnover rates in low biomass (exposed to snail grazers and dominated by prostrate cells) and high biomass (no exposure to snails and dominated by mat-forming species) periphyton communities grown in laboratory streams. Bars are means and vertical lines denote 1 SE Fig. 4 a-c Comparison of a stream water SRP concentration, b ratio of ? uptake from stream water to GPP by periphyton, and c phosphorus turnover rates in periphyton under ambient and phosphorus-enriched conditions. Periphyton communities used in the study were from laboratory streams that lacked snails (high biomass). Bars are means and vertical lines denote 1 SE

treatment were about 18 times greater the P-enriched than those in the ambient treatment (Fig. 4a). The mean ratio of ? uptake to GPP was 3.5 times greater in the Pconcentration than in the ambient enriched treatment treatment (Fig. 4b). The effect of streamwater ? on phosratio was significant phorus uptake:GPP (F12=112.6, of the blocked variable (streand the effect F=0.0088), In contrast, am) was not significant (F22=5.00, P=0.167). were mean turnover rates of ? for the two treatments commusimilar (Fig. 4c). The high biomass periphyton nities showed no evidence of net biomass accrual during this phase of the experiment in either treatment. Thus, no correction for dilution of 33P by algal growth was applied rates of 33P turnover. to the data prior to calculating

Discussion tracer injection The results of our conservative experiaccumulations can alter ments indicate that periphyton of streams, in particular disthe hydraulic characteristics and transient storage zone volumes. persion coefficients Although the average size of transient storage zones was

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions

45 that of flowing water zones in the only about one-fifth streams, average turnover time of water in high-biomass was probably zones transient (4 min) long storage of scarce nutrients to for allow enough partial recycling such as phosphorus. tracer injecconservative Previous studies involving tions have shown that streambed sediments are an important zone for transient storage of water and uptake of reactive materials (Kennedy and biologically chemically et al. 1984; Bencala et al. 1984; Bencala 1984; Triska et et al. 1989b, 1993; Harvey and Bencala 1993; D'Angelo acal. 1993). Results of our study show that periphyton on substratum surfaces can also function as cumulations important zones of transient storage of water and nutrient uscycling. In a study of nutrient uptake by periphyton al. found Kim transient instream et flumes, (1990) ing storage zones ranging from 7% to 29% of the flume cross-sectional area and attributed them to periphyton mats growing on slides in the flumes. The transient storage zones identified by Kim et al. (1990) are similar in in our high-biomass size to those identified streams, alin al. the Kim biomass et though periphyton study (2.3 was intermediate between our low biomg AFDM/cm2) streams (Fig. 3a). While our permass and high-biomass iphyton biomass levels were somewhat higher than those found in most shaded natural streams, similar and higher levels have been reported for well-lit biomass streams flow periods (Busch and Fisher 1981, Biggs low during and Close 1989). Our results showing ratios and lower ? uptake:GPP lower ? turnover rates in the high biomass communities with the low biomass communities suggest compared that internal ? cycling was greater in the high biomass communities. is dependent on However, this conclusion in use inherent of these the measures as inassumptions dices of internal ? cycling. The first index (P uptake:GPP ratio) assumes that the nutrient stoichiometry for primary production in stream is relatively communities constant, and that periphyton smaller ? uptake:GPP ratios indicate that a greater frac? demand is met by cycling tion of the total autotrophic of ? rather than by uptake from the free-flowing stream water. In a previous study using the same laboratory streams (Mulholland et al. 1991), ? content of periphyton in streams with snails (mostly was about Stigeoclonium) 1.5 times higher than that in streams with no snails and filamentous diatoms higher levels of biomass (mostly and blue-green algae). In a natural stream, periphyton ? content was observed to vary by as much as a factor of 2 in response to seasonal changes in streamwater ? concentration with very little change in species composition and Rosemond of the (Mulholland 1992). Measurements ? content of intact periphyton communities, however, inof the periphyvariably include non-living components ton (e.g., detritus) and therefore may not accurately reflect the stoichiometric demands of primary production. But more importantly, the nearly ten fold difference in ? the low and high biomass ratios between uptake:GPP communities in the present study is considerably greater measured in the differences than biomass stoichiometry ratio of previous studies. Also, the mean ? uptake:GPP conwas the high biomass community (1.3 ?gP/mg02) from 7.6 lower than the expected siderably ?gP/mg02 the Redfield ratio (using a photosynthesis quotient of 1.2), a difference unlikely to be entirely due to stoichiometry differences. The second index of ? cycling (turnover rate of ? in the periphyton matrix) assumes that ? turnover from individual algal cells is similar between community types, in 33P loss rate from the entire bioand thus differences in the efficiency mass matrix are the result of differences of 33P reassimilation by algae. Although we have no dion rates of ? turnover from individual rect information algal cells, rates of GPP per unit AFDM for the high biowere about twice those for the low bimass communities that the lower 33P turnoomass communities, suggesting ver rates in the high biomass communities were not the result of lower mass-specific metabolism. We must also emphasize that our measure of ? turnorate reflecting turnver rate in periphyton is a composite over from different cellular ? pools that likely have difturnover rates. Further, the 33P labeling ferent specific was h) (24 period certainly too short to achieve isotopie ? and all cellular ? between streamwater equilibrium those with slower turnover rates, so pools, particularly our measured turnover rates reflect disproportionately the more rapidly cycling pools. Nonetheless, by using a standard period for 33P labeling of both periphyton comin measured 33P turnover rates munity types, differences in the turnover of ? within should reflect real differences the communities. The greater transient storage zones associated with the high biomass communities should have had little influence on the isotopie labeling of these communities because the mean water turnover time for these storage zones (about 4 min) was short compared with the 24-h 33P labeling period. Our results are consistent with the results of several and other studies indicating greater nutrient limitation within stream periphyton as they communities cycling accumulated biomass. Bothwell (1989) showed that diffusion limitations to mass flux into periphyton mats can result in ? limitation over a greater range of streamwater ? concentrations for entire periphyton communities than for individual cells. Paul and Duthie (1989) reported more rapid reassimilation of ? in older, more well-develothan in younger, poorly decommunities ped periphyton in the Matamek communities River. Peterson veloped and Grimm (1992), nutrient dynamics during studying in a desert stream, noted a shift in periphyton succession the primary supply of nutrients to the periphyton from streamwater nutrients in early successional, low biomass to internal cycling in late successional, communities high biomass communities. et al. (in press) have developed a stream DeAngelis simulation model with a transient storage zone containing algae and detritus and shown that increases in the size of the transient storage zone and increases in algal biomass should both result in increases in nutrient cycling.

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions

46 Increases in transient storage zone size increase nutrient residence time in the storage zone and hence the opporof regenerated nutrients. Intunity for reassimilation creases in algal biomass increase the demand for nutrients and hence lower the available nutrient concentrations in the storage zone, thus increasing the probability that a regenerated nutrient would be reassimilated. Alresults reported here, we canthough in the experimental not separate the effects of biomass from those of transient storage zone size on nutrient cycling, we have shown that biomass and storage zone size are closely in a strong positive relationship becoupled, resulting tween biomass and nutrient cycling. The conclusion that nutrient cycling is directly related to transient storage zone size is based on the assumption that mass transfer of soluble nutrients to algal cells in stream periphyton communities is controlled by diffusion through a zone of stationary water surrounding the cells. In addition to the evidence from the study of Bothwell (1989) mentioned above, studies showing the strong influence of water velocity on uptake of solutes from stream water by attached streambed communities suggest that diffusion through a stationary water zone limits mass transfer to these communities (Horner and Welch 1981, Gantzer et al. 1988). Relatively high denitrification rates reported for periphyton in strecommunities ams with dissolved concentrations near saturaoxygen tion levels also suggest strong limitations on mass transfer from stream water to periphyton (Triska and Oremland 1981). Results from our ? enrichment experiment ratios for periphyton exshowing higher ? uptake:GPP to elevated with concentrations, posed P04 compared to ambient stream water (Fig. 4b), periphyton exposed to mass transfer were suggests that diffusion limitations in controlling nutrient cycling. important Higher P04 in free-flowing concentrations stream water should increase the diffusive flux of nutrients into transient storzones thus reage surrounding periphyton communities, ducing the relative importance of cycling in meeting periphyton nutrient demands. The lack of an observed effect of streamwater on ? turnover rates P04 concentrations with the hypothesis (Fig. 4c), however, is not consistent that diffusion to mass flux is the mechanism limitation for the relationship between biomass and nuresponsible trient cycling in stream periphyton. Our results suggest important links between trophic and nutrient cycling at the scale of structure, hydraulics, stream reaches. Mulholland et al. (1991) a presented model two for nutrient conceptual identifying pathways in periphyton-dominated streams: (1) spiraling, cycling which involves advective of nutrients displacement and (2) internal (in-place) which downstream, cycling, involves diffusion-controlled within the periprocesses have been phyton matrix. Snail herbivores previously shown to have an important effect on nutrient spiraling in periphyton-dominated streams et al. (Mulholland 1983). Here we show that snails also can have an important effect on internal cycling of nutrients in streams by of periphyton biomass preventing large accumulations and limiting the development of transient storage zones. McCormick and Stevenson also demonstrated (1991) that snails greatly affected nutrient dynamics within stream periphyton communities by preventing the development of a thick periphyton matrix consisting of understory and overstory cells. Thus, factors such as herbivory, that control periphyton biomass and physiognomy at the scale of the periphyton matrix, also have important effects on reach-scale and nutrient cycling in hydraulics streams. In general, our results were consistent with a conceptual model of stream ecosystems in which benthic organisms are contained within zones of relatively stationary water (transient storage zones) that limit the mass transfer of soluble nutrients from the flowing water to organisms and enhance the importance of internal nutrient cycling in meeting nutrient demand. Internal nutrient cycling may be particularly important in supporting primary production in streams or stream reaches with limited lateral input of new nutrients via groundwater or surface water inflows because nutrients are retained within the reach. We thank Ramie Wilkerson for assistance Acknowledgements with sample processing and laboratory analysis. The manuscript benefitted from reviews by Art Stewart, Walter Hill, and two anonymous reviewers. This work -was sponsored by the National Science Foundation's Ecosystem Studies Program through Interagency Agreement No. DEB-9013883 with the U.S. Department of Energy. E.R. Marzolf was supported by an appointment to the Oak Ridge National Laboratory Postdoctoral Research Associates Program administered jointly by the Oak Ridge National Laboratory and the Oak Ridge Institute for Science and Education. The work was performed at Oak Ridge National Laboratory which is managed by Martin Marietta Energy Systems, Inc. under contract DEAC05-84OR21400 with the U. S. Department of Energy. Publication Number 4219, Environmental Sciences Division, Oak Ridge National Laboratory.

References American Public Health Association (1989) Standard methods for the examination ofwater and wastewater, 17th edn. American Public Health Association, Washington DC Bencala KE, Kennedy VC, Zellweger GW, Jackman AP, Avanzino RJ (1984) Interactions of solutes and streambed sediment 1. an experimental analysis of cation and anion transport in a mountain stream. Water Resour Res 20: 1797-1803 Bencala KE ( 1984) Interactions of solutes and streambed sediment 2. A dynamic analysis of coupled hydrologie and chemical processes that determine solute transport. Water Resour Res 20: 1804-1814 Bencala KE, Walters RA (1983) Simulation of solute transport in a mountain pool-and-riffle stream: a transient storage model. Water Resour Res 19: 718-724 Biggs BJF, Close ME (1989) Periphyton biomass dynamics in gravel bed rivers: the relative effects of flows and nutrients. Freshwater Biol 22: 209-231 Bothwell ML (1989) Phosphorus-limited growth dynamics of lotie periphytic diatom communities: areal biomass and cellular growth rate responses. Can J Fish Aquat Sci 46: 1293-1301 Busch DE, Fisher SG (1981) Metabolism of a desert stream. Freshwater Biol 11: 301-307 D'Angelo DJ, Webster JR, Benfield EF (1991) Mechanisms of stream phosphorus retention: an experimental study. J ? Am BentholSoc 10:225-237

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions

47 D'Angelo DJ, Webster JR, Gregory SV, Meyer JL (1993) Transient storage in Appalachian and Cascade mountain streams as related to hydraulic characteristics. J ? Am Benthol Soc 12: 223-235 DeAngelis DL, Loreau M, Neergaard D, Mulholland PJ, Marzolf ER (in press) Modeling nutrient-periphyton dynamics in streams: the importance of transient storage zones. Ecol Modeling Gantzer CJ, Rittmann BE, Herricks EE (1988) Mass transport to streambed biofilms. Water Res 22: 709-722 Grimm NB (1987) Nitrogen dynamics during succession in a desert stream. Ecology 68: 1157-1170 Harvey JW, Bencala KE (1993) The effect of streambed topography on surface- subsurface water exchange in mountain catchments. Water Resour Res 29: 89-98 Hill WR, Boston HL, Steinman AD (1992) Grazers and nutrients simultaneously limit lotie primary productivity. Can J Fish Aquat Sci 49: 504-512 Horner RR, Welch EB (1981) Stream periphyton development in relation to current velocity and nutrients. Can J Fish Aquat Sci 38: 449-457 Kennedy VC, Jackman AP, Zand SM, Zellweger GW, Avanzino RJ (1984) Transport and concentration controls for chloride, strontium, potassium and lead in Uvas Creek, a small cobblebed stream in Santa Clara County, California, USA. JHydrol75:67-110 Kim BK, Jackman AP, Triska FJ (1990) Modeling transient storage and nitrate uptake kinetics in a flume containing a natural periphyton community. Wat Resour. Res 26: 505-515 McCormick PV, Stevenson RJ (1991) Grazer control of nutrient availability in the periphyton. Oecologia 86: 287-291 Mulholland PJ, Rosemond AD (1992) Periphyton response to longitudinal nutrient depletion in a woodland stream: evidence of upstream-downstream linkage. J ? Am Benthol Soc 11: 405^19 Mulholland PJ, Newbold JD, Elwood JW, Horn CL (1983) The effect of grazing intensity on phosphorus spiralling in autotrophic streams. Oecologia 58: 358-366 Mulholland PJ, Steinman AD, Palumbo AV, Elwood JW, Kirschtel DB (1991) Role of nutrient cycling and herbivory in regulating periphyton communities in laboratory streams. Ecology 72: 966-982 Paul BJ, Duthie HC (1989) Nutrient cycling in the epilithon of running waters. Can J Bot 67: 2302-2309 Peterson CG, Grimm NB (1992) Temporal variation in enrichment effects during periphyton succession in a nitrogen-limited desert stream ecosystem. J ? Am Benthol Soc 11: 20-36 Rosemond AD (in press) Multiple factors limit seasonal variation in periphyton in a forest stream. J ? Am Benthol Soc SAS (1988) SAS user's guide, Version 6.03. SAS Institute, Cary, NC, USA Steinman AD (1992) Does an increase in irradiance influence periphyton in a heavily-grazed woodland stream? Oecologia 91: 163-170 Stream Solute Workshop (1990) Concepts and methods for assessing solute dynamics in stream ecosystems. J ? Am Benthol Soc 9: 95-119 Triska FJ, Oremland RS (1981) Denitrification associated with periphyton communities. Appi Environ Microbiol 42: 745-748 Triska FJ, Kennedy VC, Avanzina RJ, Zellweger GW, Bencala KE (1989a) Retention and transport of nutrients in a third-order stream in northwestern California: channel processes. Ecology 70:1877-1892 Triska FJ, Kennedy VC, Avanzina RJ, Zellweger GW, Bencala KE (1989b) Retention and transport of nutrients in a third-order stream in Northwestern California: hyporheic processes. Ecology 70: 1893-1905 Triska FJ, Duff JH, Avanzino RJ (1993) Patterns of hydrological exchange and nutrient transformation in the hyporheic zone of a gravel-bottom stream: examining terrestrial-aquatic linkages. Freshwater Biol 29: 259-274 Valett HM, Fisher SG, Grimm NB, Camill ? (1994) Vertical hydrologie exchange and ecological stability of a desert stream ecosystem. Ecology 75: 548-560 Whitford LA, Schumacher GJ (1964) Effect of a current on respiration and mineral uptake in Spirogyra and Oedogonium. Ecology 45: 168-170

This content downloaded on Wed, 6 Mar 2013 19:47:11 PM All use subject to JSTOR Terms and Conditions