Document Yj2Zwkw7bVMDDvGk9aJ3BanqD

Limnol. Oceanogr., 51(1. part 2), 2006. 671680 q 2006. by the American Society of Limnology and Oceanography. Inc. Eutrophication and trophic state in rivers and streams Walter K. Dodds1 Division of Biology, Kansas State University, Manhattan, Kansas 66506 Abstract Many natural streams are net heterotrophic, so I propose that trophic state be divided into autotrophic and heterotrophic state. This division allows consideration of the in uence of external carbon sources as well as nutrients such as nitrogen and phosphorus. Empirical results suggest that phosphorus and nitrogen are the most important nutrients regulating autotrophic state in owing waters and that benthic algal biomass is positively correlated to gross primary production in streams. Reference (minimally in uenced by human activities) nutrient concentrations and correlations of nutrients with algal biomass are used to characterize reference distributions of stream autotrophic state. Only when reference nutrient concentrations are in the upper one third of those expected in the United States, is maximum benthic chlorophyll projected to exceed 100 mg m22 (a concentration commonly used to indicate nuisance levels) . 30% of the time. Average reference nutrient concentrations lead to sestonic chlorophyll concen trations above those considered typical of eutrophic lakes ( . 8 mg m23) less than half the time. Preliminary analysis suggests that autotrophic state is variable in small pristine streams because it is in uenced by canopy cover (light), but heterotrophic state is less variable because it can be based on allochthonous or autochthonous production. Nitrogen and phosphorus enrichment can in uence both heterotrophic and autotrophic state, and these effects could cascade to animal communities. Stoichiometry should be considered because carbon, nitrogen, and phosphorus are all involved in trophic state. The proposed doSnition of trophic state offers a starting conceptual framework for such considerations. The evolution of concepts regarding enrichment in streams In its course from the source to the sea, the progressive eu trophication of a river water by drainage from cultivated and inhabited districts is an almost inevitable natural process. D Butcher 1947 Although current concerns about stream eutrophication mainly focus on nitrogen (N) and phosphorus (P) enrichment (e.g., Smith 2003), early water quality and nutrient enrich ment studies in lotic systems focused on carbon (C) enrich ment from untreated sewage. Excessive loading of biochem ical oxygen demand (BOD) made rivers completely anoxic downstream of sewage treatment plants. Hynes (1960) con sidered the physical, biological, and chemical effects of sew age loading to create a general conceptual model on the basis of the research of Butcher (1946) and others. The conceptual model of Hynes in part considered the in uence of increased organic C on dissolved oxygen (O2) and subsequently on hypoxia- and anoxia-sensitive animals. He noted that most animals immediately downstream from a sewage outfall dis appear under anoxic conditions and that, as O2 enters the stream via aeration, high densities of pollution-tolerant fauna 1 Corresponding author (wkdodds@ksu.edu). Acknowledgments I thank Dolly Gudder for technical assistance. Conversations with Vai Smith, Jack Jones, Kirk Lohman, and Gene Welch helped for mulate many of the ideas presented here. Two anonymous reviewers and Bob Hecky provided many excellent suggestions for improving the text. I am thankful for support from the U.S. National Science Foun dation Konza Long Term Ecological Research grant and award DEB 0111410. This is contribution 04-301-J of the Kansas Agricultural Exper iment Station, Manhattan, Kansas. could be found. Eventually, as the in uence of the sewage diminished downstream, Hynes predicted a return to the clean water animal communities found upstream of the sew age outfall. Enrichment by N and P were also considered in the Hynes model. He noted a substantial increase in ammonium, phos phate, and nitrate immediately downstream from the sewage outfall that diminished farther downstream. In the anoxic zone, the prevalence of cyanobacteria {Oscillatoria and Phormidium) and Euglena, and further downstream substan tial biomass of Cladophora, were predicted. A decade later, Hynes (1970) speciDcally noted that nutrient enrichment could occur in rivers and streams as a direct result of human alteration of land use (in addition to sewage input). He de scribed the amounts of increase in fertilizer use and made preliminary calculations of how much fertilizer might reach rivers and streams. At that time, however, Hynes document ed a paucity of studies on the effects of nutrient enrichment alone but predicted that planktonic algae in large rivers would be stimulated. There has been conceptual progress related to denning trophic state and characterizing lotic eutrophication on sev eral fronts in the last three decades. Omemik (1977) pro posed that various regions are expected to have distinct base line amounts of nutrients related to geology, topography, and land uses. He described areas of the United States that would be expected to have relatively greater concentrations of nu trients in streams, and he degned the concept of nutrient ecoregions. The idea that a reference baseline trophic level occurs naturally in a region forms the basis of many current efforts to regulate stream nutrients. Over the last three de cades, numerous research programs were designed to link nutrient enrichment to increases in autotrophic biomass in rivers and streams by methods that included the "clay pot" nutrientdiffusing substrata experiments, experimental 671 17cv1906 Sierra Club v. EPA ED_001523_00003569-00001 672 Dodds stream channel enrichment experiments, whole-stream en richments, and a deiition of nuisance amounts of algae (Welch et al. 1988). At a more fundamental level, there have been few at tempts to deie the trophic state of lotic ecosystems and provide a comprehensive definition of eutrophication appli cable to rivers and streams. Thus, I initially attempt to pro vide such a definition, and under this framework, I describe how prior research can be viewed given my deShition. 100 's oCl tlj 10 DeBhing trophic state and eutrophication in streams The deShition of tropic state I develop here is designed to include both autotrophic and heterotrophic components; thus, there is a "heterotrophic state" and an "autotrophic state'' of a stream or river. Heterotrophic state can be deSned as the metabolic activity of the stream (typically measured as average O2 demand [respiration, R] during dark periods and scaled to 24 h). Autotrophic state is the gross primary production (GPP) during lighted periods (typically measured as production and scaled to 24 h). The delineation of het erotrophic and autotrophic state in owing waters was pio neered by Odum (1956). I propose that eutrophication in lotic habitats be deSned as an increase in a nutritive factor or factors that leads to greater whole-system heterotrophic or autotrophic metabolism. Heterotrophic state and autotrophic state are not mutually exclusive; a system with substantial autotrophic activity will likely have high heterotrophic activity and certainly have high respiration. This link between autotrophy and respira tion can lead to a positive correlation between respiration and GPP (Fig. 1). But, a system with more heterotrophic activity does not necessarily have more autotrophic activity (e.g., the condition obtained with high BOD loading). Thus, GPP:R can indicate the balance between heterotrophic and autotrophic state. Considering both autotrophic and hetero trophic components accounts for enrichment by organic C in addition to N and P, and accounts for the observation that lotic food webs can be based on consumption of autotrophic or heterotrophic organisms. My proposed definition of lotic trophic state is based on total heterotrophic and autotrophic production and in uenced by emerging research on lakes. Although production of lakes has often been linked to planktonic biomass (usually expressed as chlorophyll concentrations), lakes can be net heterotrophic and highly in uenced by terrestrial C inputs (e.g., Cole et al. 1994). Thus, solely emphasizing autotrophic biomass might not accurately describe trophic structure in lentic ecosystems. Rivers and streams are likely to be more dominated by heterotrophic processes than lakes given their stronger linkage to terrestrial systems as a source of organic C and the greater likelihood that light is intercepted. In small streams, the riparian canopy often shades the stream bottom, turbidity greatly attenuates light in many large, well-mixed rivers, and in some streams (blackwater streams), dissolved organic C colors the water and retards primary production. In many rivers and streams, much allochthonous organic matter enters seasonally and through storm water runoff. The net production of most streams is negative (i.e., GPP:R , & s 10 100 1000 Benthic chlorophyll (mg m"^) 10000 100 o'Cst 10 100 Gross primary production (g O2 m'^ d"^) Fig. 1. (A) Relationships between benthic chlorophyll and gross primary production and (B) between gross primary production and community respiration. Data are taken from literature compiled by Bott et al. (1985); ranges were reported for values at one site, and the mean of the minimum and maximum is plotted. The relationship in (A) is significant by linear regression (p , 0.05, r- 5 0.24) and in (B) (p , 0.05, r2 5 0.80). 1), even in open-canopy, shallow, clear-water streams (Mul holland et al. 2001). Thus, any delnition of eutrophication in streams should consider heterotrophic activity. Autotrophic activity can also be important in rivers and streams. Some streams with open canopies are net autotro phic (Mulholland et al. 2001). Phytoplankton production can supply a significant portion of the productivity in medium to large rivers that are not highly turbid and do not com 17cv1906 Sierra Club v. EPA ED_001523_00003569-00002 Lotic eutrophication 673 pletely mix because they have zones with limited water re placement (e.g., Thorp et al. 1998; Wehr and Descy 1998). Thus, allochthonous and autochthonous sources of C both should be considered, as well as inorganic and organic forms of nutrients such as N and P, when deShing trophic status of lotic ecosystems. Historically, trophic state in lakes was deShed on the basis of clear delineation between anoxic hypolimnia and oxygen ated waters (i.e., the difference between a mesotrophic and a eutrophic lake) and subsequent increases in the prevalence of cyanobacterial blooms, eutrophication- resistant animals, decreased water clarity, and taste and odor problems. Fore most, biogeochemical processes favor increased internal loading of P, leading to a positive feedback that stabilizes the eutrophic state with an anoxic hypolimnion (Dodds 2002). Such clear delineation of eutrophic conditions does not occur in shallow lakes, wetlands, and lotic systems for a variety of reasons. Rivers and streams are relatively shallow and have con siderably greater rates of atmospheric exchange compared with lentic systems, except under very low ow conditions when they become similar to small, shallow lentic systems. Thus, it is dilcult for biota to consume all the O2 in the water column without substantial inputs of BOD and ade quate nutrients to support very rapid rates of heterotrophic activity. Anoxia is rare in the water column of natural rivers and streams, even in forested streams under deciduous can opies immediately after leaf fall. In most lotic systems, in ternal loading of P and N tends to be dominated by remin eralization, groundwater inputs, and erosion. Subsequently, alternative methods are required for describing trophic dis tributions in lotic ecosystems. An approach that uses statistical distribution of benthic chlorophyll and water column nutrients was proposed to classify trophic state in streams given a lack of breakpoints (Dodds et al. 1998). Trophic categories by statistical distri butions signify the probabilities of each trophic state. How ever, Dodds et al. (1998) used distributions from data sets that included affected sites; thus, the proposed categories do not represent natural trophic distributions. Many regions of developed countries completely lack such reference sites. However, a broad deShition of stream trophic state requires consideration of the historical condition of streams before substantial modication that might in uence heterotrophic or autotrophic state. Which nutrients might be expected to control trophic state in lotic systems? Before I propose trophic categories, it is important to jus tify which nutrients need to be considered to classify trophic state (i.e., if respiration and production are the response var iables, what are the driver variables?). The most in-uential limnologist in modem times, G. E. Hutchinson (1957), stated "Phosphorus is in many ways the element most important to the ecologist, since it is more likely to be decient, and therefore to limit the biological productivity of any region of the earth's surface, than are the other major biological elements." This has led to the view that "Excessive con centrations of P is [sic] the most common cause of eutro phication in freshwater lakes, reservoirs, streams, and in the headwaters of estuarine systems" (Correll 1999). These as sertions are not based on specic experimental and empirical observations of lotic ecosystems. How well do the data sup port the statement that P is the primary agent of autotrophic eutrophication in rivers and streams? One of the common methods for assessing nutrient limi tation of benthic algae in owing waters is measuring chlo rophyll accumulation on nutrient-diffusing substrata (e.g., Pringle et al. 1986; Winterboum 1990). Francoeur (2001) performed a meta-analysis of 237 nutrient enrichment stud ies in temperate streams and found that 16.5% indicated an N response, 18.1% indicated a P response, 23.2% required N and P be added together for a response, 5% had N or P inhibition, and 43% had no response to N or P. Tank and Dodds (2003) tested for autotrophic nutrient lim itation across 10 North American streams with the use of silica (glass ber lters) or wood (thin layers of wood ve neer) substrata in nutrient-diffusing agar devices. Algae re sponded differently to nutrients, depending on the substrata on which they were growing, and there was no primary pos itive response of algae to P enrichment alone at any site. No response to nutrient enrichment was a common result of these experiments, but N or N and P stimulated algal bio mass at unshaded sites. The lack of nutrient response was closely linked to sites with limited light and a large amount of canopy cover. Tank and Dodds (2003) also presented a literature review that closely mirrored that of Francoeur (2001) in the percent responses to N and N1 P treatments. Few nutrient-releasing substrata experiments have tested for nutrient responses other than N and P. Chessman et al. (1992) tested for trace nutrient concentrations in eight streams during two seasons in Australia. They found no ev idence for trace nutrients (Ca, Mg, K, S, Fe, Mn, Co Cu, Zn, thiamine, biotin, or B12) stimulating algal growth in any stream during any season. The most common response was to N addition alone, with secondary P limitation occurring frequently. Whole-stream fertilization experiments are rarely con ducted under natural conditions, but Stockner and Shortreed (1978) used streamside enrichment troughs in British Co lumbia and demonstrated a strong P enrichment effect on algal biomass, with a secondary N enrichment effect. En richment of the Nechako River in British Columbia indicated N limitation (Perrin and Richardson 1997). Enrichment of a tundra river with P for four consecutive summers rst stim ulated algal biomass and productivity and eventually stim ulated sh production (Peterson et al. 1993). An alternative, empirical approach for determining the re sponse of algal biomass to nutrients is to establish the sta tistical relationship between in-stream nutrients and algal biomass. This approach was applied across Missouri streams of varied nutrient enrichment, and positive relationships be tween water column total N and benthic chlorophyll were found, with a similar relationship between total P and mean benthic chlorophyll (Lohman et al. 1992). Nutrient-enriched sites in this study had more rapid chlorophyll accrual after a scouring ood than nutrient-poor sites. Lohman et al. (1992) speculated that N was more important in these 17cv1906 Sierra Club v. EPA ED_001523_00003569-00003 674 Dodds Table 1. Some studies of owing waters reporting nitrogen and phosphorus enhancement of heterotrophic activities. Location Response variables Response Reference Shaded New Zealand 14C glucose incorporation, endo- N and P stimulated endocellulase activi- Tank and Winterboum 1996 stream cellulase activity ty, but not glucose uptake Small Appalachian streams Microbial respiration, fungal N and P colimitation Tank and Webster 1998 biomass, extracellular en zyme activity Laboratory with stream as Leaf degradation by bacteria N and P stimulated degradation Gulis and Suberkropp 2002 semblages and fungi 10 North American streams Fungal biomass (ergosterol) 2 streams, no effect; 2, P effect; 4, N Tank and Dodds 2003 stimulation; 2, P stimulation Laboratory experiment Leaf mass loss Both N and P effects found B&locher and Corkum 2003 Coastal wetland soil Bacterial activity P stimulated bacteria, whereas N stimu Sundareshwar et al. 2003 lated macrophytes Laboratory experiments on Bacterial counts Both N and P stimulated bacteria Mallin et al. 2004 plankton from blackwa ter streams of North Car olina Carolina streams Fungal biomass (ergosterol) High N- and P-sites had greater ergoster Padgett et al. 2000 ol accumulation streams than P. Similar relationships were siibseqnenily es tablished for 13 rivers in southern Ontario (ChAclal et al. 1999). A cross-system analysis of temperate streams established that total N and total P in the water column were signicantly related to benthic algal biomass (Dodds et al. 1997). This relationship is relevant to trophic state because produc tion is positively correlated with algal biomass (Fig. 1). Sub sequent analysis of an expanded data set suggested that mean and maximum algal biomass were signicantly correlated with total N and, to a lesser extent, with total P in the water column and that the best predictive model for algal biomass included both N and P. This analysis also suggested that in excess of a threshold value of total N and total P, there are no increases in mean benthic chlorophyll, thus indicating that nutrient limitation is overcome when water column nu trient concentrations are great enough (Dodds et al. 2002). Positive correlations also exist between planktonic chlo rophyll and water column nutrients in lotic systems. An anal ysis of suspended chlorophyll in temperate rivers and streams showed a positive relationship between water col umn total P and suspended chlorophyll (Van Nieuwenhuyse and Jones 1996), with an apparent decrease in planktonic chlorophyll yield per unit P when total P was in excess of approximately 300 mg m23. These authors did not consider total N, so the relative importance of N and P could not be assessed from their data. However, Basu and Pick (1996) studied 31 Canadian Shield rivers and found positive cor relations of sestonic chlorophyll with total N in addition to total P, but did not demonstrate any decrease in chlorophyll yield at high nutrient concentrations. Nutrient enrichment experiments on heterotrophic activity are less numerous. However, the existing laboratory and eld experiments suggest that nutrients can limit heterotrophic ac tivity (N, P, or both can be important, Table 1). A survey of stream metabolism across eight streams from various North American biomes indicated that soluble reactive P concen trations were positively correlated with both GPP and res piration (Mulholland et al. 2001). The forms that N and P are in might not be extremely important determinants of heterotrophic or autotrophic state. Algae can use organic N as a primary N source (Antia et al. 1991). In addition, heterotrophic bacteria can be strong com petitors for dissolved inorganic nutrients as well as nutrients in dissolved organic compounds (Dodds 2002). Both N and P (in organic and inorganic forms) could be important determinants of autotrophic and heterotrophic ac tivity in rivers and streams. This is borne out by manipula tive experimental approaches and empirical analyses. There is little experimental support for minor nutrients stimulating heterotrophic or autotrophic microbial activity. As discussed in the introductory section, it is clear that C additions will have a strong in uence on system heterotrophic activity (O2 consumption). It is probably unwise to assume a priori that P is the limiting nutrient of the autotrophic state in any par ticular stream. Determining boundaries of trophic state Data presented in the previous section suggest that water column N and P should be considered when characterizing the autotrophic state of rivers and streams and perhaps when determining heterotrophic state. The relative trophic state should be based on the frequency distribution of relatively pristine lotic waters because anthropogenic inputs change over time, as will trophic boundaries. Whereas Dodds et al. (1998) considered total N, total P, and benthic chlorophyll across a wide variety of streams, they did not account for streams that are naturally heterotrophic and did not attempt to use only reference streams to create an expected distri bution in the absence of anthropogenic effects. Reference nutrient data can be used to establish rough limits on the autotrophic state of streams with regard to nu trients; I present one possible approach. Reference nutrient 17cv1906 Sierra Club v. EPA ED_001523_00003569-00004 Lotic eutrophication 675 Table 2. Lower one-third and upper one-third of the distribution of stream total N and total P pooled across 14 ecoregions according to reference values determined for each individual ecoregion by Smith et al. (2003), 13 ecoregions for total P,and 12 ecoregions for total N from Dodds and Oakes (2004) and the relationship of the boundary numbers from Smith et al. (2003) data to cumulative fre quency distribution of benthic chlorophyll (Chi) as a function of total N or total P (Fig. 1) expressed as the percentage of benthic chlorophyll mean or maximum values exceeding 100 mg m22 when nutrient values were less than the boundary value. For example, when seasonal mean of total N was ,714 mg m23, then 10% of the streams had mean benthic chlorophyll values exceeding 100 mg m22 and 29% had maximum values exceeding that amount. Nutrient Total N Total P Autotrophic state boundary Lower one-third Upper one-third Lower one-third Upper one-third Concentration (mg m23) Smith et al. 2003 Dodds and Oakes 2004 285 370 714 659 29 23 71 48 Cases exceeding 100 mg m22 (%) Maxi Mean mum Chi Chi 7 27 10 29 5 17 13 25 concentrations from modeling, including a correction for at mospheric N deposition, have been proposed for 14 nutrient ecoregions across the United States (Smith et al. 2003). I ranked the median values (one for each ecoregion), and the distribution was divided into the lower, middle, and upper one third (oligotrophic, mesotrophic, and eutrophic, respec tively, following limnological convention) of the reference nutrient values (Table 2). The distribution of reference nu trient values roughly agreed with those provided by Dodds and Oakes (2004), who corrected for anthropogenic in uences (as represented by human population density and land use characteristics) on stream nutrient concentrations with analysis of covariance across the same ecoregions (Table 2). There is a positive correlation between autotrophic activ ity and benthic chlorophyll concentrations in rivers and streams (Fig. 1). Therefore, I initially base autotrophic boundaries on standing stocks of algal biomass, as is the convention in lakes. To accomplish this, the reference nutri ent values from Smith et al. 2003 are applied to observed frequency distributions of seasonal mean and maximum ben thic chlorophyll, plotted against water column nutrients (Fig. 2). These frequency distributions are used to calculate the probability that a stream will have a given amount of chlo rophyll at a specic level of nutrients (Table 2). Relation ships derived from those developed by Dodds et al. (2002, corrected for errors Dodds made when entering data from Lohman) also can be used to calculate expected mean and maximum values for benthic chlorophyll on the basis of the nutrient boundaries presented in Table 2 (Table 3). Benthic chlorophyll values . 100 mg m22 previously have been considered a nuisance (Welch et al. 1988). This analysis suggests that a mean value of 100 mg m22 of chlorophyll is attained in , 7% of oligotrophic streams and in 1013% of eutrophic systems. The regression analyses also suggest that oligotrophic systems should exhibit maximum benthic chlo rophyll values . 100 mg m22 only 27% of the time. Other approaches are possible (e.g., Dodds et al. 1998), but the method presented in this paper considers the dynamic nature of chlorophyll in streams and is reference based. A similar approach to determining reference trophic state can be taken with regard to planktonic chlorophyll in rivers and streams. A large data set (n 5 292) of lotic planktonic chlorophyll and water column total P was assembled for temperate rivers and streams, and associated regression equations can be used to link nutrients and phytoplankton biomass (Van Nieuwenhuyse and Jones 1996). A smaller data set from 31 rivers in southern Ontario and western Que bec related total N (mg m23) and total P to planktonic chlo rophyll (mg m23; Basu and Pick 1996). This paper presented a regression equation for total P, but regression of their raw data yielded the following relationship. log 10(planktonic chlorophyll) 5 21.247 1 0.676 log,,,(total N) r2 5 0.65 The distribution of reference values from Smith et al. (2003) can then be used to calculate autotrophic categories from these equations (Table 4). These data agree roughly with both the Van Nieuwenhuyse and Jones (1996) and the Basu and Pick (1996) equations for total P, but the total N boundaries derived from the Basu and Pick chlorophylltotal N relationship were substantially lower than those derived for total P from the same data set. The data suggest that planktonic chlorophyll only exceeds values considered typ ical of eutrophic lakes (8 mg m23; Dodds 2002) when nu trients are abundant relative to the reference condition. The data also are consistent with the idea that the amount of planktonic chlorophyll per unit total N or total P is less in lotic waters than in lentic waters (Suballe and Kimmel 1987). More limited data are available for whole-stream esti mates of autotrophic and heterotrophic state, but some idea of the ranges expected for the trophic states can be gleaned from analysis of the results of a cross-system study (Mul holland et al. 2001). Although this study and an additional data point (P. Mulholland pers. comm.) only covers nine streams, it has three important characteristics. First, all the measurements were done the same way at each site with methods likely to give the best results (two-station diel O2 method, corrected for groundwater in uences). Second, all the sites studied but one were relatively pristine small streams, so the data can be used to determine trophic bound aries mostly in the absence of human effects. Third, the streams were located in a variety of biomes, including one desert, one prairie, one tropical, one arid montane, one mesic montane, and four temperate deciduous biomes (Mulholland et al. 2001). Whole-stream autotrophic state varied over 150 fold in this data set (very high rates of GPP were associated with the lighted desert stream), with the central one third of the distribution falling between 0.4 and 1.8 g O2 m22 d21 (Table 5). Heterotrophic state was considerably less variable, ranging about 10-fold with the central one third of the dis tribution falling between 6.7 and 8.3 g O2 m22 d21 (Table 5). Bott et al. (1985) reviewed studies of ; 70 streams with 17cv1906 Sierra Club v. EPA ED_001523_00003569-00005 676 Dodds 10 -3 Total N (mg m ) -3 Total P (mg m ) Fig. 2. Relationships between seasonal mean water column nutrients (total N and total P) and proportion of instances in which seasonal mean and maximum chlorophyll exceed 50, 100, or 150 mg m23. Data are from literature sources compiled in Dodds et al. (2002), mostly for shallow rivers and streams. This compilation previously had incorrect values for data reported by Lohman et al. (1992). Those values now match the original source, n 5 250 for total P and n 5 199 for total N. maximum rates of 48 and 50 g O2 m22 d21 for GPP and respiration, respectively. These rates were from streams with human effects and were several-fold higher than the maxi mum from more pristine streams. This indicates that both autotrophic state and heterotrophic state can be in uenced by eutrophication. Maximum rates of GPP are probably lim ited by light under nutrient-replete conditions, whereas res piration is probably limited by O2 aeration rate in streams with high loading of biochemical oxygen demand. I speculate that light limits autotrophic state of streams (interception by the canopy), but not heterotrophic state, be cause although light is intercepted by riparian vegetation, it Table 3. Corrected regression equations for data presented in Dodds et al. (2002) and expected autotrophic state mean and maximum benthic chlorophyll (Chi) values calculated from nutrient concentrations in Table 1 with these equations. Equations are of the form log10(mg chlorophyll m22) 5 Intercept 1 Bl log10(mg m23 total N or total P) 1 B2 [log10(mg m23 total N or total P)]2. Relationship Mean Chi versus total N Maximum Chi versus total N Mean Chi versus total P Maximum Chi versus total P Intercept 22.638 0.438 20.608 0.216 2.460 0.613 1.486 1.680 b2 20.320 20.255 20.297 R2 0.401 0.295 0.402 0.371 Expected chlorophyll (mg m2) Lower Upper 30 60 88 154 36 65 109 204 17cv1906 Sierra Club v. EPA ED_001523_00003569-00006 Lotic eutrophication 677 Table 4. Autotrophic state boundaries for suspended chlorophyll in temperate rivers and streams as calculated from the reference nutrient concentrations from Smith et al. (2003) and regression equations based on VanNieuwenhuyse and Jones (1996) and Basu and Pick (1996). Nutrient Total N Total P Planktonic chlorophyll (mg m23) Autotrophic state boundary VanNieu Nutrient wenhuyse cone, and Jones Basu and (mg m23) (1996) Pick (1996) Lower one-third 285 2.4 Upper one-third 714 4.5 Lower one-third 29 4.6 6.4 Upper one-third 71 11.9 12.3 does not substantially in uence rates of C input. I predict that the amount of C xed by the riparian canopy that enters the streams to fuel heterotrophic activity is approximately equal to what would enter by autochthonous production in a lighted stream without canopy cover. Small streams in forested biomes are shaded, have sub stantial amounts of organic C input from nearby riparian areas fueling heterotrophic activity, and have minimal au totrophic production (except in deciduous seasonal forests in which light can penetrate the canopy when leaves are not present). Prairie, tundra, or desert streams have limited ri parian canopy and substantial autotrophic production fueling heterotrophic activity. An independent measure of total met abolic activity, N uptake rates, also varied little across the range of biomes studied by Mulholland et al. (2001), sup porting the concept of relatively constant heterotrophic ac tivity in small pristine streams (Webster et al. 2003). Het erotrophic state might be more variable in rivers; canopy has less of an in uence, and turbidity could substantially inter fere with riverine C production. Although the approach taken here might provide useful in setting boundaries for autotrophic and heterotrophic state, more comprehensive measurements of stream metabolism are required. Until such comprehensive measurements are made, the values for boundaries presented here should be used with caution. In addition, whole-river metabolism rates are dicult to measure, and data are diflcult to come by for such rivers. Very few large rivers remain in temperate regions that are relatively weakly in uenced by humans, so it might not be possible to set deSiitive autotrophic and heterotrophic state boundaries for larger lotic systems in some regions. Although determining trophic boundaries could be useful in describing fundamental ecosystem processes, changes in trophic state must be linked to other aspects of stream eco systems for such boundaries to be relevant. Furthermore, it is important to explore how stream eutrophication is prop agated through the food web to in uence biotic integrity and community structure. Effects of eutrophication Producers D Stevenson and Pan (1999) reviewed the uses of diatoms for assessing environmental conditions in rivers Table 5. Distribution of whole-stream metabolism rates from nine small, relatively pristine streams (data from Mullholland et al. [2001] plus one point from Ball Creek, North Carolina [Mulholland pers. comm.]). Respiration rates are corrected for groundwater in put. Distribution Upper one-third Lower one-third Minimum Maximum Metabolism (g O, m22 d23) Gross primary production Respiration Net primary production 1.8 8.3 24.2 0.4 6.7 26.7 0.06 2.4 229 15 29 6.7 and streams. They traced the use of species compositions of algae to infer amount of pollution to work by Kolkwitz and Marsson in the early 1900s, with substantial contributions by Ruth Patrick in the 1940s and 1950s (as cited by Steven son and Pan 1999). Studies that use algal assemblages as indicators of the extent of pollution rely on the concept that predictable species shifts occur with set amounts of enrich ment (e.g., Kelly 2002). Detailed work has been carried out relating nutrients to diatom and other algal assemblages in several places, mostly in temperate, developed countries. The green alga Cladophora has often been associated with eutrophication events (Hynes 1960) and is ubiquitous in nu trient-rich owing waters (Dodds and Gudder 1992). Large streamers of Cladophora develop under nutrient-rich con ditions. These streamers potentially lead to low O2 events at night, alter the community structure, snag sh lures, slow water ow in canals, and clog industrial and domestic water intakes (Dodds and Gudder 1992). One of the problems with predicting eutrophication effects in streams is that variability caused by ooding can in uence autotrophic state. At one extreme, algal biomass might not accrue with ample light and nutrients if oods always scour biomass. On the other end of the spectrum, attached algae| might be able to attain impressive biomass in nutrient-poor| |water because periphyton can use the small amounts of nu-| (trients that continuously ow by. Biggs (2000) developed a comprehensive model linking hydrologic regime and nutri ents to accrual of algal biomass. This model was developed with a database from New Zealand rivers and streams across a wide range of land use practices and hydrologic patterns. Regressions considering only dissolved inorganic nutrients could predict algal biomass with r2 values of approximately 30%. Consideration of the time of accrual (time since the last scouring ood) increased r2 values to about 70%. The work of Biggs (2000) supports the proposition that eutro phication effects will be stronger under stable ow regimes. The effects of eutrophication on macrophytes in owing waters have been poorly studied, and the effects of nutrient reductions on macrophyte biomass are difficult to predict (Chambers et al. 1999). Biomass of macrophytes declined in the Bow River (Alberta) in response to nutrient control (par ticularly N) from municipal wastewater sources (Sosiak 2002). Sewage ef uent led to substantially greater macro phyte biomass in the Saskatchewan River (Saskatchewan), 17cv1906 Sierra Club v. EPA ED_001523_00003569-00007 678 Dodds and this was correlated with somewhat decreased dissolved O2 concentration (Chambers and Prepas 1994). In some rivers and streams with reduced water replace ment times, phytoplankton blooms can become problematic, with cyanobacterial blooms more likely in excess-nutrient conditions (Smith 2003). Shorter water turnover time (hy draulic residence time) leads to a decreased amount of sus pended chlorophyll per unit concentration of P (Suballe and Kimmel 1987). Problems occur with phytoplankton blooms in European and other rivers around the world (Wehr and Descy 1998). In the Murray Darling river system in South Australia, water withdrawals reduce ow to a near standstill in the river, and excess amounts of nutrients, stratication, and wann temperature stimulate algal blooms (Maier et al. 2001). These blooms are commonly dominated by the hepatotoxic Microcystis. Other slow- owing rivers in the world suffer a similar fate, particularly those with limited quantities of light-intercepting ne sediments. Microbial heterotrophs D Although enrichment experi ments have documented that rates of microbial heterotrophic processing of organic materials can be stimulated by_nutrients (as previously discussed), less is known about in uences on the heterotrophic microbial community. If the primary source of organic C to a stream or river is leaf material, N and P need to be obtained from the water column, and nu trient enrichment will increase C utilization rates. One study documented that nutrient enrichment causes shifts in fungal taxa associated with decomposing leaf litter (Gulis and Suberkropp 2002). Presumably, some bacteria that decompose organic matter are better competitors for organic nutrients than others, leading to shifts in community structure in re sponse to nutrient enrichment. Future studies are likely to document this effect, given the recent expansion of molec ular techniques. Clear increases in the rates of heterotrophic microbial biogeochemical cycling (denitrication) related to nutrient enrichment by agricultural practices have been dem onstrated (Kemp and Dodds 2002). Food web effects D Effects of C and, particularly, N and P loading on animals in streams are less clear. The effects of C on the animal community are obvious, with greater rates of organic C loading leading to dominance by pollu tion-tolerant invertebrates (such as Tubifex, Limnodrilus, Chironomus), decreases in diversity, and increases in raw abundance (Hynes 1960). With the advent of BOD treatment in sewage and industrial ef uents in developed nations, less attention has been paid to the effects of BOD loading. Enrichment effects related to N and P are less well estab lished. Macroinvertebrate assemblage structure has been cor related statistically with P concentration (Miltner and Rankin 1998). Nutrient enrichment can cause increases in inverte brate abundance and alters assemblage structure (Bourassa and Cattaneo 1998). The clearest study to date on the im portance of sustained nutrient loading to the food web oc curred on the Lawrence River downstream of Montreal, Quebec. This study used the distinctive isotopic signal of 15N to establish that nutrients from the sewage outfall sig nificantly enriched macroinvertebrates and production of both macroinvertebrates and shes (deBruyn et al. 2003). The sewage was treated for BOD, but stimulated secondary production over vefold in spite of the small amount of N and P that entered the food web in the sewage plume 10 km down from the sewage outfall. Control of cultural eutrophication D Given the de<fhition of the trophic state proposed, and the potential effects of autotrophic and heterotrophic eutrophication, what consid erations are important in controlling eutrophication? Mech anistic methods are only beginning to be established for link ing in-stream nutrient concentrations to watershed activities. Empirical methods have prevailed (e.g., Dodds et al. 1997) until recently. Modeling efforts are beginning to relne nu trient concentration and loading estimates for rivers, but there still is some difficulty in linking models created for small streams with larger river systems (e.g., Alexander et al. 2002). Ultimately, linking land use practices, including both point and nonpoint sources of nutrients, to instream nutrient concentrations will be necessary to control cultural eutrophication that in uences autotrophic state, and poten tially kTuences heterotrophic state. Nutrient control is, on one level, simple. Agricultural practices, atmospheric loading, and human sewage outfall increase inorganic and organic nutrients in rivers and streams. Technology is available to decrease that input (but nonpoint sources of nutrients such as atmospheric deposition and runoff from cropland remain difficult to control). Best management practices of cropland include riparian buffer strips, cropland terracing, and the use of only the necessary amounts of fertilizer. Ef uent from human sewage and live stock-handling facilities can be treated with existing tertiary treatment methods (e.g., denitrication facilities, P precipi tation) to reduce N and P loads to lotic waters. The effective reduction of BOD into the waters of most developed coun tries exemplffies the technical ability of water treatment en gineers and managers to remove potentially harmful pollut ants at acceptable costs. The challenge now is to determine what lengths are necessary to control point and nonpoint source pollution, and to what degree the benelts of nutrient control justify the costs. Determining the reference trophic state provides a starting point for costbenet and feasibility analyses of eutrophication control schemes. Nutrient cycles do not occur in isolation, and colimitation of algal and heterotrophic activity is commonly seen in bio assays (Tank and Dodds 2003). We are only beginning to understand the implications of the effects of humans on the stoichiometry of nutrient loading (Turner 2002). Stoichio metric changes could alter algal assemblages and relative rates of material ux (e.g., Woodruff et al. 1999). Changes in stoichiometry could then cascade to higher trophic levels (Frost et al. 2002). Given the broad deShition of eutrophication presented herein, organic C enrichment should be considered, as well as anthropogenic processes causing shifts in the relative het erotrophic and autotrophic states. For example, increased BOD from sewage has deShite in uences on stream hetero trophic state. In addition, shifts in riparian vegetation, such as loss of riparian forests, might increase the autotrophic state and decrease the heterotrophic state. In systems such as tailgrass prairies, historically dominated by little riparian 17cv1906 Sierra Club v. EPA ED_001523_00003569-00008 Lotic eutrophication 679 vegetation, increases in riparian vegetation could alter the fundamental ecosystem and community structure (Dodds et al. 2004). Finally, organic C enrichment might interact with N and P enrichment. The highest rates of C consumption and the greatest biomass of heterotrophic organisms are ex pected when loading of N, P, and C are simultaneously high. Water retention times might alter nutrient stoichiometry and heterotrophic and autotrophic states by in uencing de position and nutrient processing rates. Small and large im poundments that were not historically present are now a ubiquitous feature on many river networks. Such impound ments could also alter the balance between heterotrophic and autotrophic states because many recalcitrant C-rich particu late organic materials can settle in the reservoir, and plank ton with relatively low values of C:N and C:P could dom inate reservoir tail waters (Whiles and Dodds 2002). Humans will affect ever more river miles with hydrologic modication, alter the inputs of organic C and its form to lotic waters through alteration of riparian vegetation and in put of BOD in sewage from humans and livestock. Increased fertilizer to grow the crops necessary to feed an expanding human population and increases in industrial livestock op erations resulting in vast production of animal waste will cause further eutrophication of already affected rivers and streams. These effects will continue to spread into the few relatively pristine watersheds that remain on earth, altering water quality and in uencing the biotic integrity of these waters. Understanding the fall implications of these effects will require further knowledge of the native trophic state of streams as a baseline. More complete comprehension of how nutrient interactions in uence trophic state, and determina tion of trophic states of medium to large rivers will improve the scientic basis for managing eutrophication of lotic wa ters. References Alexander, R. B., P. J. Johnes, E. W. Boyer, and R. A. Smith. 2002. A comparison of models for estimating the riverine ex port of nitrogen from large watersheds. Biogeochemistry 57/ 58: 295339. Antia, N. J., P. J. Harrison, and L. Oliveira. 1991. The role of dissolved organic nitrogen in phytoplankton nutrition, cell bi ology and ecology. Phycologia 30: 189. BSrlocher, F., and M. Corkum. 2003. Nutrient enrichment over whelms effects in leaf decomposition by stream fungi. Oikos 101: 247252. Basu, B. K., and F.R. Pick. 1996. Factors regulating phytoplank ton and zooplankton biomass in temperate rivers. Limnol. Oceanogr. 41: 15721577. Biggs, B. J. F.2000. Eutrophication of streams and rivers: dissolved nutrient-chlorophyll relationships for benthic algae. J. N. Am. Benthol. Soc. 19: 1731. Bott, T. L., J. T. Brock, C. S. Dunn, R. J. Naiman, R. W. Ovink, and R. C. Petersen. 1985. Benthic community metabolism in four temperate stream systems: an inter-biome comparison and evaluation of the river continuum concept. Hydrobiolgia 123: 345. Bourassa, N., and A. Cattaneo. 1998. Control of periphyton biomass in Laurentian streams (Quebec). J. N. Am. Benthol. Soc. 17: 420429. Butcher, R. W. 1946. The biological detection of pollution. Insti tute of Sewage. --------- . 1947. Studies in the ecology of rivers: VII. The algae of organically enriched waters. J. Ecol. 35: 186191. Chambers, P. A., R. E. DeWreede, E. A. Irlandi, and H. Vandemeulen. 1999. Management issues in aquatic macrophyte ecology: a Canadian perspective. Can. J. Bot. 77: 471487. --------- , and E. E. Prepas. 1994. Nutrient dynamics in riverbeds: the impact of sewage ef"uent and aquatic macrophytes. Water Res. 28: 453464. Chessman, B. C., P.E. Hutton, and J. M. Burch. 1992. Limiting nutrients for periphyton growth in sub-alpine, forest, agricul tural and urban streams. Freshw. Biol. 28: 349361. ChAtelat, J., F.R. Pick, A. Morin, and P. B. Hamilton. 1999. Periphyton biomass and community composition in rivers of different nutrient status. Can. J. Fish. Aquat. Sci. 56: 560569. Cole, J. J., N. F. Caraco, G. W. Kling, and T. K. Kratz. 1994. Carbon dioxide supersaturation in the surface waters of lakes. Science 265: 15681570. Correll, D. L. 1999. Phosphorus: a rate limiting nutrient in surface waters. Poultry Sci. 78: 674682. deBruyn, A. M. H., D. J. Marcogliese, and J. B. Rasmussen. 2003. The role of sewage in a large river food web. Can. J. Fish. Aquat. Sci. 60: 13321344. Dodds, W. K. 2002. Freshwater ecology: concepts and environ mental applications. Academic Press. --------- , K. Gido, M. Whiles, K. Fritz, and W. Mathews. 2004. Life on the edge: ecology of Great Plains prairie streams. Bio science 53: 207218. --------- , and D. A. Gudder. 1992. The ecology of Cladophora. J. Phycol. 28: 415427. --------- , J. R. Jones, and E. B. Welch. 1998. Suggested classifi cation of stream trophic state: distributions of temperate stream types by chlorophyll, total nitrogen, and phosphorus. Water Res. 32: 14551462. --------- , and R. M. Oakes. 2004. A technique for establishing ref erence nutrient concentrations across watersheds impacted by humans. Limnol. Oceanogr. Methods 2: 333341. --------- , V. H. Smith, and K. Lohman. 2002. Nitrogen and phos phorus relationships to benthic algal biomass in temperate streams. Can. J. Fish. Aquat. Sci. 59: 865874. --------- , V. H. Smith, and B. Zander. 1997. Developing nutrient targets to control benthic chlorophyll levels in streams: a case study of the Clark Fork River. Water Res. 31: 17381750. Francoeur, S. N. 2001. Meta-analysis of lotic nutrient amendment experiments: detecting and quantifying subtle responses. J. N. Am. Benthol. Soc. 20: 358368. Frost, P. C., R. S. Stelzer, G. A. Lamberti, and J. J. Elser. 2002. Ecological stoichiometry of trophic interactions in the benthos: understanding the role of C:N:P ratios in lentic and lotic habitats. J. N. Am. Benthol. Soc. 21: 515528. Gulis, V, and K. Suberkropp. 2002 Effect of inorganic nutrients on relative contributions of fungi and bacteria to carbon ow from submerged decomposing leaf litter. Microb. Ecol. 45: 11 19. Hutchinson, G. E. 1957. A treatise on limnology. Geography, physics and chemistry. V. 1. Wiley. Hynes, H. B. N. 1960. The biology of polluted waters. Liverpool Univ. Press. --------- . 1970. The ecology of running waters. Univ, of Toronto Press. Kelly, M. G. 2002. Role of benthic diatoms in the implementation of the urban wastewater treatment directive in the River Wear, north-east England. J. Appl. Phycol. 14: 918. Kemp, M. J., and W. K. Dodds. 2002. Comparisons of nitrication 17cv1906 Sierra Club v. EPA ED_001523_00003569-00009 680 Dodds and denitrication in pristine and agriculturally in uenced streams. Ecol. Appl. 12: 9981009. Lohman, K., J. R. Jones, and B. D. Perkins. 1992. Effects of nutrient enrichment and ood frequency on periphyton biomass in northern Ozark streams. Can. J. Fish. Aquat. Sci. 49: 1198 1205. Maier, H. R., M. D. Burch, and M. Bormans. 2001. Flow man agement strategies to control blooms of the cyanobacterium, Anabaena circinalis, in the River Murray at Morgan, South Australia. Reg. Riv. Res. Manage. 17: 637650. Mallin, M. A., M. R. McIver, S. H. Ensign, and L. B. Cahoon. 2004. Photosynthetic and heterotrophic impacts of nutrient loading to blackwater streams. Ecol. Appl. 14: 823838. Millner, R. J., and E. T. Rankin. 1998. Primary nutrients and the biotic integrity of rivers and streams. Freshw. Biol. 40: 145 158. Mulholland, P. J., and others. 2001. Inter-biome comparison of factors controlling stream metabolism. Freshw. Biol. 46: 1503 1517. _ Odum, H. T. 1956. Primary production in owing waters. Limnol. Oceanogr. 1: 102117. Omernik, J. M. 1977. Nonpoint source-stream nutrient level rela tionships: a nationwide study. EPA-600/3-77-105. U.S. Envi ronmental Protection Agency. Padgett, D. E., M. A. Mallin, and L. B. Cahoon. 2000. Eval uating the use of ergosterol as a bioindicator for assessing wa ter quality. Environ. Monit. Assess. 65: 547558. Perrin, C. J. and J. S. Richardson. 1997. N and P limitation of benthos abundance in the Nechako River, British Columbia. Can. J. Fish. Aquat. Sci. 54: 25742583. Peterson, B. J., and others. 1993. Biological responses of a tun dra river to fertilization. Ecology 74: 653672. Pringle, C. M., P. Paab, P. D. Vaux, and C. R. Goldman. 1986. In situ nutrient assays of periphyton growth in a lowland Costa Rican stream. Hydrobiologia 134: 207213. Smith, R. A., R. B. Alexander, and G. E. Schwarz. 2003. Nat ural background concentrations of nutrients in streams and riv ers of the conterminous United States. Environ. Sci. Technol. 37: 20393047. Smith, V. H. 2003. Eutrophication of freshwater and coastal marine ecosystems. A global problem. Environ. Sci. Pollut. Res. 10: 126139. Suballe, D. M., and B. L. Kimmel. 1987. A large-scale compar ison of factors in uencing phytoplankton abundance in rivers, lakes, and impoundments. Ecology 68: 19431954. Sosiak, A. 2002. Long-term response of periphyton and macro phytes to reduced municipal nutrient loading to the Bow River (Alberta, Canada). Can. J. Fish. Aquat. Sci. 59: 9871001. Stevenson, J., and Y. Pan. 1999. Assessing environmental con ditions in rivers and streams with diatoms, p. ll40. In E. F. Stoermer and J. P. Smol [eds.], The diatoms: applications for the environmental and earth sciences. Cambridge Univ. Press. Stockner, J. G., and K. R. S. Shortreed. 1978. Enhancement of autotrophic production by nutrient addition in a coastal rain forest stream on Vancouver Island. J. Fish. Res. Board Can. 35: 2834. SUNDARESHWAR, P.V., J. T.MORRIS, E. K. KOEPFLER, AND B. FOR- walt. 2003. Phosphorus limitation of coastal ecosystem pro cesses. Science 299: 563565. Tank, J . L .and W.K. Dodds. 2003. Nutrient limitation of epilithic and epixylic biolms in 10 North American streams. Freshw. Biol. 48: 10311049. --------- , and J. R. Webster. 1998. Interaction of substrate and nutrient availability on wood biolm processes in streams. Ecology 79: 21682179. --------- , and M. J. Winterbourn. 1996. Microbial activity and invertebrate colonisation of wood in a New Zealand forest stream. N. Z. J. Mar. Freshw. Res. 30: 271280. Thorp, J. FL, M. D. Delong, K . S . Geenwood, and A. F.Casper. 1998. Isotopic analysis of three food web theories in constrict ed and oodplain regions of large river. Oecologia 117: 551 563. Turner, R. E. 2002. Element ratios and aquatic food webs. Estu aries 25: 694703. Van Nieuwenhuyse, E. E., and J. R. Jones. 1996. Phosphorus chlorophyll relationship in temperate streams and its variation with stream catchment area. Can. J. Fish. Aquat. Sci. 53: 99 105. Webster, J. R., and others. 2003. Factors affecting ammonium uptake in streams Dan inter-biome perspective. Freshw. Biol. 48: 13291352. Wehr, J. D., and J.-P.Descy. 1998. Use of phytoplankton in large river management. J. Phycol. 34: 741749. Welch, E. B., J. M. Jacoby, R. R. Horner, and M. R. Seeley. 1988. Nuisance biomass levels of periphytic algae in streams. Hydrobiologia 157: 161168. Whiles, M. R., and W. K. Dodds. 2002. Relationships between stream size, suspended particles, and lter-feeding macroin vertebrates in a Great Plains drainage network. J. Environ. Qual. 31: 15891600. Winterbourn, M. J. 1990. Interactions among nutrients, algae, and invertebrates in a New Zealand mountain stream. Freshw. Biol. 23: 463474. Woodruff, S. L., W. A. House, M. E. Callow, and B. S. C. Leadbeater. 1999. The effects of a developing biolm on chemical changes across the sedimenttwater interface in a freshwater environment. Int. Rev. Hydrobiol. 84: 509532. Received: 26 July 2004 Accepted: 4 November 2004 Amended: 19 November 2004 17cv1906 Sierra Club v. EPA ED_001523_00003569-00010