URBAN STREAMS PERSPECTIVES*

Ecological resistance in urban streams: the role of natural and legacy attributes

Ryan M. Utz1,9, Kristina G. Hopkins2,10, Leah Beesley3,11, Derek B. Booth4,12, Robert J. Hawley5,13, Matthew E. Baker6,14, Mary C. Freeman7,15, and Krista L. Jones8,16

1Falk School of Sustainability, Chatham University, 6035 Ridge Road, Gibsonia, Pennsylvania 15044 USA 2National Socio-Environmental Synthesis Center, University of Maryland, 1 Park Place Suite 300, Annapolis, Maryland 21401 USA 3Centre of Excellence in Natural Resource Management, University of Western Australia, Albany, Western Australia 6332 Australia 3Cooperative Research Centre for Water Sensitive Cities, Clayton 3800 Australia 4Bren School of Environmental Science and Management, University of California Santa Barbara, Santa Barbara, California 93106 USA 5Sustainable Streams, LLC, 1948 Deer Park Avenue, Louisville, Kentucky 40205 USA 6Department of Geography and Environmental Systems, University of Maryland-Baltimore County, Baltimore, Maryland 21250 USA 7US Geological Survey Patuxent Wildlife Research Center, Athens, Georgia 30602 USA 8US Geological Survey Oregon Water Science Center, 2130 SW 5th Avenue, Portland, Oregon 97201 USA

Abstract: Urbanization substantially changes the physicochemical and biological characteristics of streams. The trajectory of negative effect is broadly similar around the world, but the nature and magnitude of ecological responses to urban growth differ among locations. Some heterogeneity in response arises from differences in the level of urban development and attributes of urban water management. However, the heterogeneity also may arise from variation in hydrologic, biological, and physico- chemical templates that shaped stream ecosystems before urban development. We present a framework to develop hypotheses that predict how natural watershed and channel attributes in the pre-urban-development state may confer ecological resistance to urbanization. We present 6 testable hypotheses that explore the expression of such attributes under our framework: 1) greater water storage capacity mitigates hydrologic regime shifts, 2) coarse substrates and a balance between erosive forces and sediment supply buffer morphological changes, 3) naturally high ionic concentrations and pH pre-adapt biota to water- quality stress, 4) metapopulation connectivity results in retention of species richness, 5) high functional redundancy buffers trophic function from species loss, and 6) landuse history mutes or reverses the expected trajectory of eutrophication. Data from past comparative analyses support these hypotheses, but rigorous testing will require targeted investigations that account for confounding or interacting factors, such as diversity in urban infrastructure attributes. Improved understanding of the susceptibility or resistance of stream ecosystems could substantially strengthen conservation, management, and monitoring efforts in urban streams. We hope that these preliminary, conceptual hypotheses will encourage others to explore these ideas further and generate additional explanations for the heterogeneity observed in urban streams. Key words: urbanization, ecological resistance, water storage capacity, sediment supply, connectivity, conductance, biodiversity and ecosystem function, land use history

A consistent suite of physical, chemical, and ecological nels. The altered hydrologic regime, geomorphic adjust- changes constitute the now well-defined urban stream ment, and chemical degradation result in biodiversity loss syndrome (Meyer et al. 2005, Walsh et al. 2005). In urban (Brown et al. 2005). Consistent patterns of environmental watersheds, impervious surfaces disrupt the natural flow degradation in urban streams reflect the importance of regime and route water and pollutants directly to chan- impervious surface as a fundamental landscape stressor.

E-mail addresses: [email protected]; [email protected]; [email protected]; [email protected]; 13bob.hawley@sustainablestreams .com; [email protected]; [email protected]; [email protected] *This section of the journal is for the expression of new ideas, points of view, and comments on topics of interest to aquatic scientists. The editorial board invites new and original papers as well as comments on items already published in Freshwater Science. Format and style may be less formal than conventional research papers; massive data sets are not appropriate. Speculation is welcome if it is likely to stimulate worthwhile discussion. Alternative points of view should be instructive rather than merely contradictory or argumentative. All submissions will receive the usual reviews and editorial assessments.

DOI: 10.1086/684839. Received 11 February 2015; Accepted 9 September 2015; Published online 8 December 2015. Freshwater Science. 2016. 35(1):380–397. © 2016 by The Society for Freshwater Science. Volume 35 March 2016 | 381

However, natural streams exhibit substantial heterogene- ity in their physical, chemical, and biological properties, both within and among landscapes (Cushing et al. 2006, Schueler et al. 2009). Therefore, one might reasonably expect that some streams will be less susceptible than others to environmental changes resulting from urbaniza- tion, particularly in watersheds with low-to-moderate levels of impervious surface where the landscape stressor may not be sufficient to obscure the variability of natural watershed and channel attributes. A growing body of evidence highlights heterogeneity in the ecological responses of streams to urbanization. The se- verity of chemical (Beaulieu et al. 2012), biological (Bryant and Carlisle 2012), and hydrogeomorphic (Fitzpatrick and Peppler 2010, Hopkins et al. 2015) changes in urban stream ecosystems vary among metropolitan areas, perhaps re- flecting heterogeneity of the physiographic templates where these cities are located. Differences also are observed among urban streams within a single physiographic area. For in- stance, Snyder et al. (2003) observed distinctly greater losses in fish diversity in relatively high-gradient urban streams of the North American Ridge and Valley physiographic re- gion relative to their lower-gradient counterparts. Burns et al. (2005) highlighted how large areas of natural ripar- ian wetlands within a region could substantially buffer the hydrologic effect of urban development. Utz and Hilder- brand (2011) demonstrated disparate geomorphic patterns among urban streams in 2 adjacent physiographic regions. Figure 1. A conceptual depiction of how variation in the nat- ural environmental attributes can protect streams from urbani- Changes to channel structure and sediment regime of ur- zation via resistance created by the ability of natural landscape banized streams were reduced in lowland Coastal Plain wa- factors to buffer changes in attribute A associated with urbaniza- tersheds compared to in similarly urbanized streams in the tion (hypotheses 1, 2, 4, 5) (A) or by pre-adaptation associated upland Piedmont province. King et al. (2016) found that with previous exposure to natural or anthropogenic influences organisms were more sensitive to impervious surface cover resembling conditions in urban streams (hypotheses 3, 6) (B). in semipermanent than in perennial streams. The observed heterogeneity of responses to urbaniza- tion collectively suggest that certain watershed and chan- ization to degrade stream ecosystems if they already exist nel attributes (hydrologic, biological, and geological) in the in a state that resembles an urban stream because of nat- pre-urban state might consistently influence the degree of ural or anthropogenic influences (hereafter, a legacy at- physicochemical and ecological degradation in urbanizing tribute), which may pre-adapt biota to urban conditions streams. The concept that the pre-urbanized physicochem- (Fig. 1B). The concept of legacy attributes might not fit ical template of an ecosystem might influence the extent consistently within an ecological resistance framework, but to which streams are altered by urbanization fits within a we consider any attribute that buffers streams from envi- framework of ecological resistance and resilience (Grimm ronmental change to be conferring a desired effect and, at and Wissel 1997). Ecological resistance is defined as the least as we have defined the term above, qualified as con- degree to which an ecosystem withstands changes to func- ferring ecological resistance. tion and structure despite effects induced by a persistent We present 6 hypotheses constructed based on a frame- (press) disturbance, whereas ecological resilience refers to work that incorporates natural (hypotheses 1–5) or legacy the ability of an ecosystem to return to a reference state (hypothesis 6) attributes when considering how an ecosys- after a severe but temporally limited disturbance (Grimm tem service might degrade in response to low-to-moderate and Wissel 1997). Urbanization is a permanent and perva- degrees of urbanization. The framework (Fig. 2, Table 1) sive driver of stress in streams under most circumstances, expands upon traditional frameworks by integrating nat- so we consider natural landscape attributes that buffer ural or legacy heterogeneity that may confer ecological re- streams from urban stressors to be agents that provide re- sistance to urbanization with the extent and configuration sistance, rather than resilience, to degradation (Fig. 1A). of urban infrastructure. We include abiotic stream eco- System properties also may reduce the potential of urban- system services in our resistance examples for 2 reasons: 382 | Ecological resistance in urban streams R. M. Utz et al.

Figure 2. Structural framework to assess how pre-urban-development watershed and channel attributes may confer ecological resistance to degradation in urbanizing streams. Pathways are shown for the ecological-resistance and the traditional frameworks. The traditional framework primarily considers the intensity and configuration of urbanization when assessing or predicting ecological impact. The uppermost 4 boxes are bolded because these reflect the novel attributes of our updated framework.

1) abiotic attributes, such as flood mitigation capacity and could affect how urban infrastructure grows within a wa- channel stability, are highly valuable ecosystem services tershed, thereby creating an interactive effect between ur- provided by streams (Daily 1997, Brauman et al. 2007) and ban and pre-urban environmental attributes. Addressing 2) physicochemical degradation in urban stream ecosys- each hypothesis will require careful consideration of urban tems directly affects biotic components. Thus, natural or infrastructure attributes that vary among metropolitan re- legacy attributes that confer resistance to abiotic stressors gions and those that covary with the natural environmen- (such as retention of a natural flow regime despite urbani- tal factors of interest. We think that overcoming such zation) also could preserve existing biological resources. challenges may lead to insights applicable to urban stream Urban land cover presents many simultaneous stressors to monitoring and management. We hope our examples serve aquatic organisms (chemical, physical, and thermal) at mul- as inspiration for further consideration of how the pre- tiple spatiotemporal scales. Therefore, identifying ecosystem urban environmental setting may affect stream ecosystems. attributes that introduce variability in how physicochemical properties respond to urbanization is necessary when ex- ploring disparate biological patterns among urban streams Hypothesis 1: Watershed storage capacity influences (i.e., Chadwick et al. 2012, Kollaus et al. 2014). resistance to hydrologic regime changes caused Our hypotheses are based on evidence observed at mul- by urbanization tiple spatial scales, offer some degree of mechanistic ratio- Urban impervious area seals soil surfaces, reduces infil- nale, and posit conceptual lines of inquiry that might prove tration and recharge, and rapidly delivers runoff directly worthy of further investigation. We limit in-depth treat- to stream channels (Arnold and Gibbons 1996). As a re- ment of the following hypothesis-based discussion to po- sult of increased hydraulic efficiency and connectivity, ur- tential influences of natural or legacy attributes for the ban streams typically exhibit flow regime changes including sake of brevity and focus. The attributes we discuss also elevated magnitude and frequency high flows, increased Table 1. Six hypothetical ways in which pre-urban environmental heterogeneity may affect ecological resistance to urbanization in streams. Potential effect of environmental Hypothesis Ecosystem attribute Environmental heterogeneity Typical effect of urbanization heterogeneity on resistance to urbanization

1 Flood mitigation Flood mitigation varies with watershed Flood mitigation declines as urbanization Greater watershed water storage capacity water storage capacity increases because of reduced infiltration; may help attenuate runoff from impervi- drainage infrastructure quickly routes ous surface by promoting infiltration and runoff to stream channels suppressing overland flow 2 Geomorphic stability Channel stability naturally varies with bed- Channel stability decreases as urbanization Relatively coarser channel beds and a material resistance and balance between increases the erosive potential of the flow balance between flow regime and sedi- resistive and erosive hydrogeomorphic regime ment supply may enhance channel forces stability by resisting channel evolution despite increased flooding 3 Conductivity regimes Background conductivity and pH vary Urbanization elevates pH and conductivity, Chemical regimes may exhibit dilution or that support biota substantially among watersheds; organ- corresponding to biodiversity loss no discernable response to urbanization; isms may be adapted to high ionic biota adapted to high conductivity may concentrations where they naturally exist show muted response 4 Biodiversity Organisms have differing among- and Biodiversity declines in urban settings Urban streams that retain migration within-ecosystem life-history attributes corridors and species with migratory life related to migration and recolonization histories may retain a greater degree of biodiversity 5 Ecosystem functions Functional redundancy varies among Functional redundancy decreases with the Trophic function of streams with naturally streams with natural variation in species urban loss of sensitive species high functional redundancy will be richness buffered from species loss because tolerant species will compensate 6 Nutrient concentrations Past landuse practices may elevate nutrient N, C, and P concentrations rise with input If urban infrastructure successfully limits concentrations, even if land use reverts to from treatment facilities, sewage leaks, nutrient export, urbanization in areas seminatural cover and lawn fertilization; eutrophication with human-centered landuse legacies ensues may have negligible eutrophication or nutrient concentrations may decline 384 | Ecological resistance in urban streams R. M. Utz et al. rise and fall of storm hydrographs, and reduced lag times to peak flows (Shuster et al. 2005, Walsh et al. 2005). The volume of surface runoff and the speed at which it moves through a watershed also is influenced by physical factors that pertain to intrinsic watershed storage capacity, in- cluding topographic relief, geologic materials, and surface- water impoundments (Sayama et al. 2011). In undevel- oped watersheds, such physical attributes can promote infiltration and suppress overland flow. In watersheds with greater water storage capacity, whether surface (in lakes or wetlands) or subsurface (via infiltration), we hypothe- size that these attributes may impart resistance to the hydrologic changes common in urban watersheds. Such storage potential could limit the extent of flow magni- tude during events by: 1) infiltration and temporary stor- age of water in porous riparian sediments during spates, and 2) limiting the extent to which nonpiped hydrologic linkages between impervious surfaces and channels, such as 0-order swales, deliver water. Resistance to flow alteration in urbanizing streams would be highest in watersheds with greater ability to promote fi in ltration and to provide greater surface-water storage. Figure 3. Mean high-flow event frequency along a gradient Broad differences in hydrologic responses to urbanization of road density in metropolitan areas with differing sediment among metropolitan areas correspond to systematic differ- regimes with respect to glacial history (data from Hopkins et al. ences in natural water storage capacity or stormwater man- 2015). agement. Urban streams in metropolitan areas in regions with extensive deposits of thick, permeable ice-sheet-derived sediments tend to express fewer high-flow events than ur- stormwater management, showed the interplay of water- ban streams in nonglaciated landscapes (Hopkins et al. 2015; shed imperviousness and natural wetland- and groundwater- Fig. 3). Hydrologic responses to moderate urbanization in dominated runoff patterns in western Washington (USA) low-gradient streams of the Coastal Plain physiographic prov- (Fig. 4B). The Jenkins Creek watershed drains highly infil- ince, with unconsolidated sediments, are muted relative to trative, gravel-rich lands that provided a primary path for responses of higher-gradient, shallow bedrock streams of the glacial meltwater during the retreat of the last ice sheet. Piedmont in the Mid-Atlantic Ecoregion (USA) (Utz et al. The glacial legacy resulted in a very low-gradient channel 2011). Walsh et al. (2012) and Walsh and Webb (2014) docu- (slope = 0.007), highly permeable soils, and discharge- mented streams that retain elements of their natural flow attenuating lakes and wetlands covering 13% of the water- regime even with substantial watershed urbanization as long shed (Mullineaux 1970). Despite a moderate (13% total im- as their watersheds lack conventional stormwater drainage pervious area [TIA]) degree of urbanization, the hydrograph networks and, instead, have informal drainage of impervi- of Jenkins Creek is quite subdued in comparison to that of ous surfaces to pervious areas that allow more retention the highly developed Miller Creek (50% TIA) and to that and infiltration of stormwater. of Bear Creek, whose watershed has only 7% TIA but has Comparisons within a single physiographic setting also neither an equally extensive a set of wetlands nor highly in- highlight the potential for water storage capacity to influ- filtrative soils. Infiltrative soils and extensive wetlands prob- ence urban degradation. Janke et al. (2014) suggested that ably provide greater water storage capacity in Jenkins Creek, extensive surface-water connections to lakes and ponds slow thereby imparting resistance to hydrologic changes via sup- the movement of stormflow through their urban study wa- pression of overland flow and discharge attenuation. tersheds in St Paul, Minnesota (USA), where watersheds Building upon these examples, we hypothesized that with greater surface-water connections to lakes had longer resistance to flow alteration in urbanizing streams would runoff peaks with flatter hydrographs compared to sim- be highest in watersheds with greater water storage capac- ilarly developed watersheds with fewer connections to lakes, ity. We recognized that differences in climate and drainage indicating that water storage and release in lakes can de- infrastructure may confound the expression of water stor- lay flow response to storm events in urban watersheds age capacity in urban hydrographs (Brown et al. 2009a, (Fig. 4A). Data published by Booth et al. (2004), who com- Booth et al. 2016). Despite these complications, the grow- pared the hydrologic attributes of 3 urbanized streams ing interest and investment in urban stormwater manage- with similar (open roadside ditches, piped drainage) urban ment and stream restoration projects that increase infil- Volume 35 March 2016 | 385

et al. 2014). Preserving existing storage features and attri- butes within a developing watershed and preferentially locat- ing developments to make use of existing storage capacity could be a logical management implication related to this hypothesis.

Hypothesis 2: Coarse substrates and settings with greater balance between erosive forces and sediment supply convey morphological resistance to urban stream channels The disturbance regime of streambed material and the relative stability of benthic habitats are ecosystem attributes that influence geomorphic dynamics of stream channels and the biological communities that reside in them (Poff 1992, Holomuzki and Biggs 2000, Wohl et al. 2015). Such geo- morphic traits vary naturally among hydrogeomorphic set- tings. For example, cobble-dominated snowmelt streams tend to have seasonally predictable periods of bed-material mobility (Poff et al. 1997), whereas flashy sand-dominated streams tend to have more erratic disturbance regimes with relatively prolonged periods of channel evolution and in- stability (Schumm et al. 1984). Increased geomorphic instability has been documented in urban streams across multiple hydroclimatic settings (Booth 1990, Bledsoe and Watson 2001, Leopold et al. 2005), but channel responses to urbanization are not universal in magnitude or trajectory (Chin 2006, Fitzpatrick and Pep- pler 2010). Some urban channels incise and have initially smaller width-to-depth ratios (W/D; Park 1977, Simon and Rinaldi 2000, Vietz et al. 2014), whereas others aggrade and develop larger W/D ratios (Chin and Gregory 2001, Hawley et al. 2012). Magnitude of channel enlargement also varies, with enlargement ratios (post-urban-development cross-sectional area relative to pre-urban-development cross- sectional area) of 2 to 8 in humid temperate urban streams (Hammer 1972, Chin 2006), up to 14 in southern Califor- Figure 4. Intraregional heterogeneity in responses to precipi- nia (USA) (Hawley and Bledsoe 2013), and reduced ratios tation events among urban watersheds. A.—A watershed with in Nigeria (Odemerho 1992). We hypothesize that the dis- greater pond density (PD; n/km2) in the St Paul, Minnesota parate urban channel responses to changes in flow may be (USA) metropolitan region exhibits a muted response to rainfall attributable to: 1) relative channel resistance expressed as relative to nearby watersheds with comparable road densities a function of bed-material grain size, and 2) differences in 2 — (RD; km/km ); data are from Janke et al. (2014). B. Percent the relative balance between sediment supply and erosive wetland cover (%WL) is correlated with attenuation of hydrologic forces. We anchor the concepts on Lane’s (1955) balance responses across watersheds in the greater Seattle, Washington that assumes that streams trend toward an equilibrium: (USA) metropolitan area, despite differing levels of watershed- scale total impervious area (%TIA). The watershed sizes for Bear, ∝ ; Miller, and Jenkins Creeks are 36.3, 18.9, and 42.6 km2, respectively. SQ d50Qs (Eq. 1) Dates are formatted dd/mm and data are from Booth et al. (2004). where the dominant elements of erosive force, channel slope (S)andflow (Q), are proportional to the dominant tration and water storage make addressing this hypothesis resistive properties of an alluvial channel, the median bed particularly worthwhile. Increasing water storage capacity material grain size (d50) and sediment supply (Qs). Urban- through artificial structures is a common approach to ur- ization can move all 4 components of the Lane framework ban stream mitigation and can be partly successful at re- away from equilibrium, which helps explain why channel ducing runoff contributions at the small-watershed scale instability is a nearly universal symptom of urban streams. (e.g., Booth and Jackson 1997, Davis et al. 2009, Fletcher For example, urbanization increases the erosive potential 386 | Ecological resistance in urban streams R. M. Utz et al. of the hydrologic network (Q+) by increasing the magni- tude, frequency, and duration of erosive events (e.g., Mac- Rae and Rowney 1992, Konrad and Booth 2002, Hawley and Bledsoe 2011). The enhanced water transport effi- ciency of urban drainage networks, including reduced hy- draulic roughness and increased slope (S+), may also elevate erosion potential (ASCE and WEF 1992). Urbanization can −/+ also change sediment supply (Qs ) and can lead to fining −/+ or coarsening of bed material sediment (d50 ), with the effect varying with the age of the urban infrastructure, ex- tent of effective erosion control practices during construc- tion phases (Wolman 1967, Pizzuto et al. 2000, Hawley et al. 2013), and geologic setting. Our hypothesis posits that the natural attributes related to the resistive components within the Lane framework (d50 and the balance between Q and Qs)influences the resis- tance to channel evolution, or lack thereof, in urban set- tings. First, we suggest that geomorphic resistance can be primarily expressed as bed-material resistance to mobili- Figure 5. Channel enlargement as a function of impervious fl zation as expressed by d50. The altered ow regimes in surface cover in the finer, less resistant systems of southern urban catchments (see hypothesis 1) cannot induce com- California, USA (n = 66, median particle diameter = 3.8 mm; data mon incision-driven channel-evolution trajectories (e.g., from Hawley and Bledsoe 2013) compared to the coarser, more Schumm et al. 1984, Simon 1989) without first exceeding resistant systems of northern Kentucky, USA, (n = 88, median flow thresholds necessary for particle mobility. For exam- particle diameter = 55 mm; data from Hawley et al. 2013). ple, in a boulder canyon where the flow threshold for bed- material mobility is on the order of the 100-y flow (200– 1000% of the 2-y flow; Booth et al. 2016), urbanization transport supplied sediment may be more susceptible to may not induce the same scale of geomorphic effect as aggradation and channel reductions, particularly during the fi in a sand- or gravel-dominated stream where sediment- initial construction phase of urbanization when ne sedi- moving flow thresholds are several orders of magnitude ment loads can increase by factors of 10 to 100 (Wolman fi more frequent (e.g., ∼0.01–1% of the 2-y flow; Hawley 1967), which could explain the urban channel in lling ob- and Vietz 2016). Thus, the magnitude of urban channel served in Nigeria (Ebisemiju 1989). These 3 examples show instability (expressed by the channel enlargement ratio) that the relative balance between Qs and Q is nuanced and should be greater in settings with relatively finer bed mate- warrants detailed mechanistic explorations. rial (Fig. 5). Second, we propose that channel structure will change Hypothesis 3: Naturally high ionic concentrations and less in urbanizing streams with a balance between trans- pH convey water-quality and potential biotic resistance port capacity (= flow; Q)andQs, but change in different ways in urban streams where Q or Qs is more dominant. For example, streams in Urbanization severely alters stream chemical regimes semi-arid southern California carry naturally high Qs but in well-documented ways (Herlihy et al. 1998, Paul and exhibit a large increase in channel size when their water- Meyer 2001). In classic conceptual models, total dissolved sheds are urbanized, especially in unconfined valleys (e.g., solids (i.e., salinity), specific conductivity, and pH increase Jordan et al. 2010, Hawley et al. 2012). Here, naturally high relative to pre-urban conditions (Prowse 1987, Paul and sediment loads are exceeded by greater changes in the mag- Meyer 2001, Walsh et al. 2005). However, ambient total nitude and duration of erosive flows (Hawley and Bledsoe dissolved solids, conductivity, and alkalinity vary naturally 2011) coupled with sediment discontinuities created by ur- among watersheds as a consequence of regional variation ban infrastructure (Chin and Gregory 2001) and reduced sup- in bedrock, differential weathering, evaporative demand, plies of coarse sediment (e.g., Gurnell et al. 2007, Bledsoe atmospheric inputs, and varying water contact time with et al. 2012). At smaller scales, stream reaches with bank- bedrock (Williams 1987, Olson and Hawkins 2012, Casey derived sources of bed sediment are more resistant than et al. 2013, Griffith 2014). For example, reference conduc- reaches from similar settings that lack autochthonous coarse- tance levels are generally lower and dominated by differ- grained sediment sources (Vietz et al. 2014), results sug- ent ions in eastern and northwestern ecoregions than in gesting that local sources of sediment are helping to main- central and southwestern ecoregions in the USA (Griffith tain the balance between Qs and Q. Alternatively, streams 2014). Stream chemistry can vary notably within regions, in extremely low-energy settings with very little ability to even among sites with minimal landuse disturbance (Olsen Volume 35 March 2016 | 387 and Hawkins 2012). A key insight is that natural variation tebrate drift) to maintain homeostasis or endure increased does not simply distinguish waters with high and low con- osmoregulatory energy demands and risk death (Bradley centrations—variation may include differences in major 2009). Like Hobbs et al. (2006) and King and Baker (2010), ions that contribute to conductance as well as temporal we expect that sharp community changes will be more variation in concentrations associated with distinct water likely when urbanizing conditions depart radically from sourcing and transport pathways. pre-urban regimes, even among resistant communities. The As urbanization proceeds, dissolved ion concentra- importance of chemical vs biotic resistance is likely to differ tions may increase periodically from diffuse sources (e.g., depending on the magnitude and stability of natural con- land disturbance, fertilizer, building materials, or deicers; ductance. For instance, in regions with naturally high but Prowse 1987, Wright et al. 2011, Corsi et al. 2015) or chron- stable conductivity, chemical resistance is likely to dominate ically from point sources (e.g., sanitary sewers, treated waste because urbanization at low levels probably will dilute back- water, industrial effluent; Paul and Meyer 2001, Piscart ground sources. Conversely, in regions with spatially or tem- et al. 2005). Symptoms of the former tend to occur under porally variable conductivity, biotic resistance should be more low-to-moderate forms of urbanization and mirror the shift common because the organisms have been exposed to ele- toward circumneutral to alkaline pH, whereas those of the vated ionic concentrations at various points in their life latter emerge with greater population density (Tu et al. cycle or evolutionary history. 2007, Kaushal and Belt 2012). Here, we focus on resistance From a chemical standpoint, several observations lend to chemical changes that occur under low-to-moderate support to our hypotheses. For example, Dow and Zam- forms of urbanization. We distinguish chemical resistance pella (2000) and Conway (2007) showed that naturally di- (the ability to maintain chemical integrity despite low-to- lute and acidic streams exhibited clear chemical responses moderate levels of urbanization) from biotic resistance to to low levels of impervious cover (e.g., <4%). Long-term altered chemistry (the ability to maintain pre-urban com- records of conductivity in Maryland’s Piedmont reveal sub- munity composition and structure in the face of chemical stantial differences in response to low and moderate lev- changes). els of urbanization depending on local geologic variation We hypothesize that streams with naturally high or var- (Fig. 6). Similarly, Brown et al. (2009a) showed that urban iable ionic concentrations will be more resistant to chemical areas in naturally low-conductivity regions (away from changes caused by urbanization than those with naturally heavy use of road salt) showed strong positive correlations low and relatively stable ionic concentrations. Similarly, between urbanization and specific conductivity (e.g., At- streams with pre-urban ionic compositions that approxi- lanta, Georgia, and Raleigh, North Carolina), whereas ur- mate or mirror the ions in moderate urban landscapes ban areas in high-conductivity regions (e.g., Denver, Col- should be more resistant than those with markedly differ- orado, and Dallas, Texas) did not. Global comparisons ent compositions because water delivered by urban chan- indicate that regional differences in background salinity nels may already chemically resemble typical conditions. influence species sensitivity distributions (Kefford et al. Moreover, natural waters whose pH or conductivity dif- 2012), but limited evidence of biotic resistance to urban fers strongly from urban sources will be more susceptible chemistry has been found, possibly because few compari- to transformation. For chemical resistance, high natural sons have been made at low levels of urbanization and be- ionic concentrations typically indicate prolonged exposure cause data often are confounded by biophysical covari- of stream water to bedrock or evaporation (Olson and Haw- ates across longer urbanization gradients (e.g., Walsh et al. kins 2012, Griffith 2014). These systems may resist urbaniza- 2001, King et al. 2005). Nonetheless, indirect evidence con- tion because reduced infiltration and accelerated runoff sistent with our hypothesis comes from comparisons of from urbanizing lands (Walsh and Kunapo 2009, Kaushal urbanization gradients across regional and national scales. and Belt 2012, Jones et al. 2014) limits detention storage King et al. (2011) showed differences in distributions of (see hypothesis 1), exposure to bedrock, or reduces head taxa sensitivity along gradients of impervious surface among pressure on ion-rich aquifers. Such dilution may initially regions ranked according to background conductivity (Grif- offset ionic increases associated with urban pollution (e.g., fith 2014). Across the conterminous USA, negative relation- Walsh et al. 2001). ships between urbanization and invertebrate species rich- We also hypothesize that biota in streams with natu- ness were stronger for streams in regions with low than rally high or variable ionic concentrations or high pH will with high background conductivity (Brown et al. 2009a, be more resistant to urbanization than those in streams Cuffney et al. 2010). With the growing salinization of global with naturally stable ionic concentrations or low pH. Bi- freshwater (Cañedo-Argüelles et al. 2013), such patterns otic resistance is imparted by a system’s chemical resistance suggest that chemical heterogeneity can be an important or by physiological mechanisms for coping with steep ionic consideration when interpreting biological responses along gradients (Evans 2008). In systems where natural chemical urbanization gradients, particularly as stormwater mitiga- regimes have not pre-adapted biota to urban-induced changes tion techniques increasingly decouple hydrologic and chem- in water chemistry, organisms must emigrate (e.g., inver- ical responses to urban development (Loperfido et al. 2014). 388 | Ecological resistance in urban streams R. M. Utz et al.

brate communities may likewise be variably influenced by dispersal depending on location within stream networks (Brown and Swan 2010). Dispersal also is expected to be more influential in streams that experience more frequent and intense disturbances (Palmer et al. 1996), including anthropogenic disturbances. For example, macroinverte- brate drift from undisturbed headwaters may maintain sensitive taxa in streams affected by coal mining and val- ley fill (Pond et al. 2014). Therefore, we hypothesize that dispersal is a primary mechanism for maintaining some degree of biodiversity in urban systems where it persists. Native species richness commonly declines as streams be- come more urban, but to varying degrees (Brown et al. 2009a, Bryant and Carlisle 2012). Such variability may be partly attributable to differences in organism dispersal among populations. In particular, native species may be more likely to persist in urban streams when traits that promote mobility are common and when streams retain Figure 6. Intensity duration (exceedance probability) curves connections to potential source habitats. representing a time-series (1986–2014) of monthly specific Several studies suggest or provide direct evidence that fl conductivity from 20 watersheds in the Maryland Piedmont, metapopulation connectivity in uences biological characteris- USA (chemical data from Maryland Department of Natural tics of urban streams. Adult dispersal from adjacent, rural Resources core/trend monitoring program; annual impervious headwaters may support stream insect populations in ur- cover maps from Sexton et al. 2013). Solid lines represent ban headwater streams (Urban et al. 2006, Smith et al. 2009, watersheds in Precambrian schist, quartzite, and gneiss (Geo1; 2016) though upstream connection to forested tributaries 13 watersheds), whereas dashed lines represent watersheds that may not be sufficient to mitigate effects of urban hydro- include substantial fractions of marble, shale, and dolomite logic changes (Violin et al. 2011). Most fishes lack capabil- (Geo2; 7 watersheds). In Geo1, increases from low to medium ities for overland movement. Thus, populations are partic- (med) % impervious cover are associated with increased conduc- ularly vulnerable to culverts, dams, and other barriers that tivity across all exceedance frequencies so that previously rare high conductivities have become common, whereas high % imper- impede dispersal. Brown et al. (2009b) cite stream frag- viousness is more closely associated with increases in low- mentation by numerous reservoirs and resulting low po- fi fi exceedance-frequency pulses (predominantly winter storms). tential for stream sh dispersal as a reason that sh com- Geo2 watersheds have elevated conductivity, even at low % imper- munity metrics were related only weakly to local urban vious cover, and small increases in low-to-medium-frequency intensity in one metropolitan area. In contrast, high faunal exceedance events associated with moderate % imperviousness are mobility can have the opposite effect of allowing species associated with interannual droughts. Reference conductivity to persist despite intense urban development. For exam- for the region ranges from 80 to 120 μS/cm (Griffith 2014). ple, platypus (Ornithorhynchus anatinus) persistence in urbanized streams has been associated with movements from neighboring stream sites less affected by stormwater runoff (Martin et al. 2014). Native, diadromous stream Hypothesis 4: Greater metapopulation connectivity fishes also show high persistence in Puerto Rican urban confers biological resistance to urbanization in streams streams as long as streams remain connected by barrier- The potentially strong influence of organism dispersal free channels to estuaries (Ramírez et al. 2009). However, on composition of local communities is well established sensitive stream insect taxa not dependent on estuarine (MacArthur and Wilson 1967, Brown and Kodric-Brown environments to disperse are similar between urban streams 1977, Hanski and Gilpin 1991), including in streams (Palmer in Puerto Rico and continental analogs (de Jesus-Crespo et al. 1996, Grant et al. 2010, Brown et al. 2011). How- and Ramírez 2011). ever, dispersal potential in stream communities varies with Persistence of native stream biota in urban streams that faunal mobility and proximity to source populations. Col- maintain connectivity to source habitats carries manage- onists from large adjacent channels can enhance local spe- ment restoration implications. Faunal resistance to urbani- cies richness of fishes in tributary streams. As a result, di- zation could be enhanced by retaining undisturbed landscape versity in similarly sized tributaries may differ depending and riparian corridors that promote aerial dispersal by on fluvial distance from larger streams (Osborne and Wiley stream insects (Urban et al. 2006), by limiting stream sys- 1992, Hitt and Angermeier 2011). Stream macroinverte- tem fragmentation (e.g., by replacing roadway culverts with Volume 35 March 2016 | 389 bridges), and by maintaining barrier-free connections to this magnitude of species loss can have a marked impact nearby undeveloped habitats. Local restoration actions in- on trophic function (Hooper et al. 2012). For this reason, tended to improve ecological condition in urban streams streams with naturally high functional redundancy in a pre- may fail if barriers to organism dispersal prevent coloniza- urban state should demonstrate greater resistance than tion of restored habitats (Bond and Lake 2003). The costs streams with low functional redundancy to changes in and benefits of enhanced connectivity may require case- trophic function associated with species loss. Under this specific analysis, particularly where isolation by barriers pro- hypothesis, resistance arises because assemblages that nat- tects populations and habitats from invading non-native urally contain many functionally equivalent species will re- species. In addition, if many urban stream taxa are main- tain function posturbanization because the remaining ∼60% tained by dispersal from less disturbed areas, biological as- of macroinvertebrate taxa will be able to fill all functional sessments based on species occurrences in those streams roles. Populations of tolerant species that remain could may fail to reflect actual urbanization effects on organism increase and compensate for the loss of sensitive species, survival and reproduction (Hitt and Angermeier 2011, Mar- per the insurance hypothesis (Elmqvist et al. 2003, Balva- tin et al. 2014). Nonetheless, urban streams could be popu- nera et al. 2006). Under such circumstances, trophic func- lation sinks for species maintained by periodic immigra- tion will change relatively little (Fig. 7). tion, but those species could still influence and contribute Few studies conducted in an urban setting suitably to local ecosystem function and services. address this hypothesis. One exception is the experimental We suggest that measures of connectivity relative to work by Wang et al. (2014), who found that a diverse non- faunal mobility should provide additional power for explain- urban microbial community maintained denitrification po- ing and predicting variation in biological status of urban tential in the face of multiple urban stressors (heavy metals, streams (Smith et al. 2015). Do urban areas situated in temperature, and elevated salt concentrations), whereas a downstream portions of dendritic systems, where connec- less diverse urban community did not. Extrapolated to tivity is higher than in headwater locations (Fagan 2002), pre-urban streams, which naturally vary drastically in mi- support a greater proportion of native species? Does closer crobial community composition (Hahn 2006), the poten- proximity to streams supporting intact native communities tial for variable denitrifying ability in posturban settings increase biotic integrity of urban streams? Are highly mo- after a proportion loss of biodiversity is substantial. Most bile fauna, such as diadromous species and long-distance other pertinent studies related to leaf-litter breakdown dispersers, consistently more persistent in urban streams typically were correlative and conducted within a single than less vagile species? Tests of these ideas using interre- system (urban to nonurban gradient) and did not assess gional data sets should improve our understanding of how the functional redundancy of the biotic assemblage. As a connectivity enhances ecological resistance to urbanization. consequence, the inference that can be gained from such studies is limited. Nevertheless, positive relationships be- tween leaf-litter breakdown and macroinvertebrate spe- Hypothesis 5: High functional redundancy buffers the cies richness or shredder abundance are reported often in trophic function of urban streams from species loss urban streams (see Chadwick et al. 2006, Cook and Hoel- In species-rich systems, multiple species often have sim- lein 2016, but see Imberger et al. 2008) suggesting that ilar trophic roles (Peterson et al. 1998, Hooper et al. 2005). species–trophic-function relationships warrant the atten- The overlap of species function is called functional redun- tion of urban stream researchers. dancy, an attribute considered important because it enables Addressing this hypothesis will depend on the outcomes streams to maintain trophic function (i.e., nutrient process- of related lines of research. The hypothesis is contingent on ing) in the face of species loss (Hooper et al. 2005). The the assumption that urbanization reduces the functional re- number of species varies broadly across streams among dundancy of freshwater assemblages, so clear quantification and within regions, so we expect functional redundancy to of this relationship is an obvious starting point. Such in- vary likewise. For example, Mediterranean streams sup- vestigations could follow the approach taken by Laliberté port greater macroinvertebrate diversity than temperate et al. (2010), who assessed the functional redundancy of streams in Europe (Bonada et al. 2007), temperate streams plants along a gradient of increasing landuse intensification. support more shredders than tropical streams (Boyero et al. The disparate levels of biodiversity loss observed within 2011, but see Cheshire et al. 2005), and arid streams sup- (Snyder et al. 2003, King et al. 2005, Smith and Lamp port a depauperate macroinvertebrate assemblage (Gray 2008) and among regions (Utz et al. 2009, 2010, Cuffney 1981). et al. 2010) suggest that ecological function among simi- The urbanization of streams can affect functional redun- larly urbanized ecosystems may vary to a comparable degree. dancy through species loss. Many urban streams lose >40% We encourage field studies assessing functional re-dundancy of their macroinvertebrate species (Brown et al. 2005, and trophic function (primary productivity, leaf-litter break- Walsh et al. 2005) and considerable evidence exists that down, N uptake) across urban gradients and controlled 390 | Ecological resistance in urban streams R. M. Utz et al.

Figure 7. Theoretical illustration of the implications of the loss of functional redundancy associated with species loss for trophic function in urbanizing streams. Each black bar represents niche space occupied by a species. The side graphs represent trophic function in each system. A.—Little change in trophic function with 50% species loss because all ecological functions (represented by the dimensions of the black bars) are still being performed. B.—Decline in trophic function with 50% species loss because some functions are now lost.

laboratory studies of real-world urban macroinvertebrate as agriculture or secondary forest (del Mar López et al. or microbial assemblages. 2001, Drummond and Loveland 2010). The nutrient con- Several ecological and management implications exist if centrations of such streams are likely to be elevated al- this hypothesis is supported. Urban streams will benefit ready (Lenat and Crawford 1994, Coulter et al. 2004, Mor- from management strategies that retain or enhance biodi- gan and Kline 2011), which may mute or distort nutrient versity (Dolédec and Bonada 2013). Certain urban stream enrichment associated with urbanization. functions may be improved by the presence of tolerant Resistance to nutrient enrichment in urbanizing streams nonnative species if they fill currently unoccupied ecolog- should occur when: 1) legacy effects of land use, such as ical niches. Last, management to protect species will be par- excessive fertilization that results in decadal-scale persis- ticularly important in metropolitan areas that have naturally tence of groundwater contamination (Tomer and Burkart depauperate assemblages. Work pertaining to how urban- 2003), already have elevated nutrient stores substantially, induced species loss alters the structure of stream food webs or 2) nutrient export from urban land is minimized by best and the ability of urban streams to process nutrients and practices or preventative urban infrastructure, such that nu- energy (Cardinale et al. 2006, Hooper et al. 2012) would in- trient export is equal to or lower than that resulting from form this hypothetical assertion. other anthropogenic land uses. These situations are likely to arise when urbanization encroaches on agricultural land or replaces secondary-growth forest, if the legacy effect of Hypothesis 6: Historic land use may mute or distort prior land use (i.e., agriculture) is strong (Harding et al. 1998, effects of urban expansion on nutrient enrichment Foster et al. 2003). Thus, our final hypothesis posits that Urban expansion typically elevates nutrient concentra- landuse legacies have the potential to distort the expected tions in ground water and storm runoff because treated elevation in nutrient concentrations predicted by the ur- wastewater, sewage leaks, and fertilized lawns deliver ex- ban stream syndrome model. cess N and P to stream ecosystems. This increase can lead Several observational studies lend support to the hy- to eutrophication of receiving aquatic ecosystems (Hatt pothesis outlined above. For example, in a study of ef- et al. 2004, Brett et al. 2005, Kaushal et al. 2008). How- fects of current and past land use on Maryland (USA) ever, urbanization rarely takes place in undisturbed land streams, Maloney and Weller (2011) used structural equa- that supports pristine streams. Rather, an expanding ur- tion modeling to show that agricultural land use was a − ban center typically replaces human-centered land use, such better predictor of elevated NO3 concentrations than ur- Volume 35 March 2016 | 391 ban development. Furthermore, contemporary urban de- regime and supply, high ionic concentrations, metapop- velopment was not strongly related to elevated stream ulation connectivity, and high functional redundancy of − NO3 concentrations on parcels of land that were primar- species could potentially attenuate urban effects on stream ily forested in 1952, perhaps because contemporary low- hydrology, channel dynamics, water chemistry, species per- impact development was effective at preventing nutrient sistence, and trophic function to some degree. In addition, export.IntheTorontometropolitanregion(Canada),where historical land use can be a modifier of responses that can salt is used to deice roads, a long-term decline in stream distort expected trajectories in urbanizing streams. An argu- − P concentrations was correlated with rising Cl , suggest- ment could be made that certain concepts presented in Ta- ing that the transition from agricultural to urban land use ble 1 do not convey true ecological resistance because they reduced P loading (Raney and Eimers 2014). In the Phoe- simply represent conditions that resemble urban streams nix metropolitan region (USA), soil C, N, and organic prior to urban development, but we think that, no matter matter were >2× greater in residential lawns built on for- what their label, all such factors potentially contribute to mer agricultural land compared to lawns built on native how an ecosystem will respond to urbanization and, there- desert scrubland (Lewis et al. 2014). Such patterns pre- fore, are worth considering when assessing or predicting sumably result in greater nutrient loads in urban streams impact. We hope that our hypotheses will be used by oth- with agricultural legacies relative to those located in for- ers to explore natural variability in catchment attributes and merly natural habitat. their links to the heterogeneity in ecological, physical, and Given the limited number of studies addressing the role hydrological responses to urbanization. of legacy effects and the long period over which they oper- Our list of hypotheses is not exhaustive. Rather, we ate (decades to centuries), research in this field has the offer these hypotheses as a starting point with the hope potential to make a substantial contribution to urban stream that our conceptual framework will serve as a springboard ecology. Future research could extend the hypothesis pre- for discussion of additional important natural attributes. sented here to other ecosystem attributes, such as biotic Two potentially fruitful areas to explore are the role of cli- assemblages (Brown et al. 2009a, Maloney and Weller 2011), mate and vegetation in explaining heterogeneity in urban and should ensure that the hypothesis encompasses legacy responses. For example, urban streams in areas where the effects beyond agriculture. For example, clear cutting has rainfall is naturally variable, such as arid or tropical re- legacy effects on stream dissolved organic C that exceed gions, may be pre-adapted to flashy flows (Osterkamp and 30 y (Yamashita et al. 2011). Challenges facing investi- Friedman 2000, Ramírez et al. 2009), whereas streams in gators interested in this line of research will be acquisition areas with prolonged dry periods may be more susceptible of the historical data essential to accurate quantification to urban water-quality stress (McNeil and Cox 2007, see of past land use and the difficulty of uncoupling legacy Hale et al. 2016 for a detailed discussion of climate and effects from the cumulative effects of multiple land-cover the urban stream syndrome). Vegetation in riparian areas conversions. To overcome such challenges, Bain et al. can regulate temperatures and serve as important inputs (2012) suggest separating legacy effects arising from lags of organic matter to streams (Benfield 1997, Webster and in transport associated with slow flow-paths, such as ground- Meyer 1997, Lyons et al. 2000). Therefore, streams with water, from those arising from rearrangement of physical naturally wooded riparian zones may be more susceptible systems within the landscape that alter material interac- to loss of riparian vegetation associated with urbanization tions, i.e., limitation of the colonization of aquatic or semi- than streams draining grasslands or shrublands (Cuffney aquatic species by fragmentation of remnant vegetation or et al. 2011). refuge pools. The same regional variation in pre-urban environmen- tal attributes that directly confers resistance to urbaniza- tion also may affect urban stream degradation indirectly DISCUSSION by influencing the type of urban infrastructure or its func- Growing awareness of the heterogeneous nature of tion on the landscape. For example, urban planners in flat, urbanized systems is an emerging theme in urban stream semi-arid landscapes with high soil permeability may di- ecology (Booth et al. 2016, Parr et al. 2016). Our concep- rect local stormwater runoff to detention basins or under- tual framework (Fig. 2) provides a starting point to begin ground sumps rather than piping it to the nearest drainage incorporating the role of pre-urban environmental attri- lines as a means of reducing stormwater runoff and main- butes into assessments of the cascading effects of factors taining water tables. Such urban designs may help mitigate that contribute to the observed heterogeneity. Concepts concerns associated with rapid and extreme surface flows, listed in Table 1 identify some potential natural or legacy but may further alter flow regimes with steeper subsurface factors influencing the degree of physicochemical and hydraulic gradients and enhanced base flows (Meyer et al. ecological resistance to degradation in urbanizing streams. 2005, Loperfido et al. 2014, see Bhaskar et al. 2016). In a Five of our hypotheses posit that the natural attributes 2nd example, regions subjected to Pleistocene glaciation related to high water-storage capacity, balanced sediment may have naturally elevated conductivity, and sufficient 392 | Ecological resistance in urban streams R. M. Utz et al. modern snow and ice accumulation to warrant application agement, and monitoring efforts. We close by stressing of road salt deicers and may develop an urban chemical that concepts listed in Table 1 are hypothetical and that signal (e.g., Daley et al. 2009, Corsi et al. 2010, Cooper et al. only investigations targeted to specifically address their va- 2014) distinctly different from the signal in areas where lidity will confirm or refute such ideas. We think that the Pleistocene glaciation did not occur, such as Atlanta and high degree of environmental heterogeneity of streams Melbourne (e.g., Roy et al. 2003, Hatt et al. 2004). In a 3rd ex- warrants consistent integration into urban stream ecologi- ample, new developments may preferentially replace agri- cal studies. cultural lands rather than forested areas in regions where pre-existing land-cover patterns are driven by predictable ACKNOWLEDGEMENTS patterns of soil depth (e.g., ridges and valleys), water supply The concepts presented here expanded upon a discussion ses- rd (e.g., mountain blocks), or poor drainage (e.g., sand plains). sion held at the 3 Symposium on Urbanization and Stream Here, the same attribute that drives landuse history, nutri- Ecology (SUSE) in Portland, Oregon, USA, in May 2014, which ent concentrations, and subsequent ecological resistance was funded in part by US National Science Foundation (NSF) award DEB 1427007. The authors also acknowledge support from to rising nutrient loads also affects patterns of urban infra- the Commonwealth of Australia through the Cooperative Re- structure. Although beyond the scope of our paper, the in- ff fl search Centre Program. Je Burkley at the King County Water uence of natural factors on infrastructure systems and man- and Land Resources Division and Benjamin Janke of the Univer- fi agement practices ts well into our framework. sity of Minnesota generously provided land cover and hydrologic Thevariableresistanceofstreamstourbanizationmeans data for the examples shown in Fig. 4. We thank the SUSE orga- that extrapolation of management guidelines across all sys- nizing committee for coordinating the event. Two anonymous tems may be unwise. Impervious-area thresholds have been referees, Amy Rosemond, Christopher Walsh, and Kit Wheeler applied to the management of urban watersheds on the greatly improved the quality of the manuscript through editorial basis of a presumption of little/no impact <10% total im- and conceptual suggestions. pervious area (e.g., Klein 1979) to a more nuanced recogni- tion of potential effects, particularly biological, at levels of LITERATURE CITED effective imperviousness <2 or 3% (e.g., Booth and Jackson Arnold, C. L., and C. J. Gibbons. 1996. Impervious surface cover- 1997,KingandBaker2010).However,naturalvariationin age: the emergence of a key environmental indicator. Journal of the American Planning Association 62:243–258. the attributes listed in Table 1 together with characteristics ASCE and WEF (American Society of Civil Engineers and the of the urban infrastructure itself (Bhaskar et al. 2016, Parr Water Environment Federation). 1992. Design and construc- et al. 2016) are likely to lead to variation in the relationship tion of urban stormwater management systems. ASCE Man- between impervious cover and ecological integrity, thereby ual and Reports on Engineering Practice, Book 77. American reducing the generality of management thresholds. Thus, Society of Civil Engineers, New York. relationships should be assessed for individual metropolitan Bain, D. J., M. B. Green, J. L. Campbell, J. F. Chamblee, S. Chaoka, areas. J. M. Fraterrigo, S. S. Kaushal, S. L. Martin, T. E. Jordan, A. J. The first 5 hypotheses and the stage-setting influence Parolari, W. V. Sobczak, D. E. Weller, W. M. Wollheim, of historical land use also carry implications for monitor- E. R. Boose, J. M. Duncan, G. M. Gettel, B. R. Hall, P. Kumar, ing and restoration of degraded stream systems. The wa- J. R. Thompson, J. M. Vose, E. M. Elliott, and D. S. Leigh. ff fl ter storage capacity hypothesis highlights natural features 2012. Legacy e ects in material ux: structural catchment changes pre-date long-term studies. BioScience 62:575–584. that could be preserved or mimicked to maintain or im- fi ff Balvanera, P., A. B. P sterer, N. Buchmann, J.-S. He, T. Nakashizuka, prove the natural hydrologic bu ering capacity of a water- D. Raffaelli, and B. Schmid. 2006. Quantifying the evidence shed. Such a concept could be realized by preserving wet- for biodiversity effects on ecosystem functioning and services. lands or instituting management practices that create new Ecology Letters 9:1146–1156. features (i.e., rain gardens) to infiltrate runoff at the source. Beaulieu, J. J., A. H. Bell, and J. F. Coles. 2012. Variability in Such approaches may reduce hydrologic disturbances and stream chemistry in relation to urban development and bio- provide additional opportunities for nutrient processing in logical condition in seven metropolitan areas of the United urban watersheds. The conductance hypothesis highlights States, 1999-2004. Scientific Investigations Report 2012-5170. the importance of knowing the background level of con- US Geological Survey, Reston, Virginia. fi ductivity and natural patterns of variation, especially for Ben eld, E. F. 1997. Comparison of litterfall input to streams. – streams with naturally higher or more variable concentra- Journal of the North American Benthological Society 16:104 108. tions. Similar levels of conductivity across urban regions ff Bhaskar, A. S., L. Beesley, M. J. Burns, T. Fletcher, P. Hamel, can carry di erent implications for biota depending on C. Oldham, and A. H. Roy. 2016. Will it rise or will it fall? whether biota are pre-adapted or not. As a whole, improv- Managing the complex effects of urbanization on base flow. ing our understanding of the natural resistance of stream Freshwater Science 35:293–310. ecosystems to urbanization could strengthen our knowl- Bledsoe, B. P., E. D. Stein, R. J. Hawley, and D. Booth. 2012. edge of the drivers of degradation and restoration, man- Framework and tool for rapid assessment of stream suscep- Volume 35 March 2016 | 393

tibility to hydromodification. Journal of the American Water Brown, L. R., T. F. Cuffney, J. F. Coles, F. Fitzpatrick, G. Resources Association 48:788–808. McMahon, J. Steuer, A. H. Bell, and J. T. May. 2009a. Urban Bledsoe, B. P., and C. C. Watson. 2001. Effects of urbanization streams across the USA: lessons learned from studies in 9 on channel instability. Journal of the American Water Resources metropolitan areas. Journal of the North American Bentholog- Association 37:255–270. ical Society 28:1051–1069. Bonada, N., S. Dolédec, and B. Statzner. 2007. Taxonomic and Brown, L. R., R. H. Gray, R. M. Hughes, and M. R. Meador. biological trait differences of stream macroinvertebrate com- 2005. Effects of urbanization on stream ecosystems. Sympo- munities between Mediterranean and temperate regions: im- sium 47. American Fisheries Society, Washington, DC. plications for future climatic scenarios. Global Change Biology Brown, L. R., M. B. Gregory, and J. T. May. 2009b. Relation of 13:1658–1671. urbanization to stream fish assemblages and species traits in Bond, N. R., and P. S. Lake. 2003. Local habitat restoration in nine metropolitan areas of the United States. Urban Eco- streams: constraints on the effectiveness of restoration for systems 12:391–416. stream biota. Ecological Management and Restoration 4:193– Bryant, W. L., and D. M. Carlisle. 2012. The relative importance 198. of physicochemical factors to stream biological condition in Booth, D. B. 1990. Stream-channel incision following drainage- urbanizing basins: evidence from multimodel inference. Fresh- basin urbanization. Water Resources Bulletin 26:407–417. water Science 31:154–166. Booth, D. B., and C. R. Jackson. 1997. Urbanization of aquatic Burns, D., T. Vitvar, J. McDonnell, J. Hassett, J. Duncan, and systems: degradation thresholds, stormwater detection, and C. Kendall. 2005. Effects of suburban development on runoff the limits of mitigation. Water Resources Bulletin 33:1077– generation in the Croton River basin, New York, USA. Jour- 1090. nal of Hydrology 311:266–281. Booth, D. B., J. R. Karr, S. Schauman, C. P. Konrad, S. A. Cañedo-Argüelles, M., B. J. Kefford, C. Piscart, N. Prat, R. B. Morley, M. G. Larson, and S. J. Burges. 2004. Reviving ur- Schäfer, and C.-J. Schulz. 2013. Salinisation of rivers: an ur- ban streams: land use, hydrology, biology, and human be- gent ecological issue. Environmental Pollution 173:157–167. havior. Journal of the American Water Resources Association Cardinale, B. J., D. S. Srivastava, J. Emmett Duffy, J. P. Wright, 40:1351–1364. A. L. Downing, M. Sankaran, and C. Jouseau. 2006. Effects Booth, D. B., A. R. Roy, R. Smith, and C. Capps. 2016. Global of biodiversity on the functioning of trophic groups and perspectives on the urban stream syndrome. Freshwater Sci- ecosystems. Nature 443:989–992. ence 35:412–420. Casey, R. E., S. M. Lev, and J. W. Snodgrass. 2013. Stormwater Boyero, L., R. G. Pearson, D. Dudgeon, M. A. S. Graça, M. O. ponds as a source of long-term surface and ground water Gessner, R. J. Albariño, V. Ferreira, C. M. Yule, A. J. Boulton, salinisation. Urban Water Journal 10:145–153. M. Arunachalam, M. Callisto, E. Chauvet, A. Ramírez, J. Chará, Chadwick, M. A., D. R. Dobberfuhl, A. C. Benke, A. D. Huryn, M. S. Moretti, J. F. Gonçalves, J. E. Helson, A. M. Chará-Serna, K. Suberkropp, and J. E. Thiele. 2006. Urbanization affects A. C. Encalada, J. N. Davies, S. Lamothe, A. Cornejo, A. O. Y. stream ecosystem function by altering hydrology, chemistry, Li, L. M. Buria, V. D. Villanueva, M. C. Zúñiga, and C. M. and biotic richness. Ecological Applications 16:1796–1807. Pringle. 2011. Global distribution of a key trophic guild con- Chadwick, M. A., J. E. Thiele, A. D. Huryn, A. C. Benke, and D. R. trasts with common latitudinal diversity patterns. Ecology 92: Dobberfuhl. 2012. Effects of urbanization on macroinverte- 1839–1848. brates in tributaries of the St. Johns River, Florida, USA. Ur- Bradley, T. J. 2009. Animal osmoregulation. Oxford University ban Ecosystems 15:347–365. Press, Oxford, UK. Cheshire, K., L. Boyero, and R. G. Pearson. 2005. Food webs in Brauman, K. A., G. C. Daily, T. K. Duarte, and H. A. Mooney. tropical Australian streams: shredders are not scarce. Fresh- 2007. The nature and value of ecosystem services: an over- water Biology 50:748–769. view highlighting hydrologic services. Annual Review of Envi- Chin, A. 2006. Urban transformation of river landscapes in a ronment and Resources 32:67–98. global context. Geomorphology 79:460–487. Brett, M. T., G. B. Arhonditsis, S. E. Mueller, D. M. Hartley, Chin, A., and K. J. Gregory. 2001. Urbanization and adjustment J. D. Frodge, and D. E. Funke. 2005. Non-point-source impacts of ephemeral stream channels. Annals of the Association of on stream nutrient concentrations along a forest to urban gra- American Geographers 91:595–608. dient. Environmental Management 35:330–342. Conway, T. M. 2007. Impervious surface as an indicator of pH Brown, B. L., and C. M. Swan. 2010. Dendritic network struc- and specific conductance in the urbanizing coastal zone of ture constrains metacommunity properties in riverine ecosys- New Jersey, USA. Journal of Environmental Management 85: tems. Journal of Animal Ecology 79:571–580. 308–316. Brown, B. L., C. M. Swan, D. A. Auerbach, E. H. Campbell Grant, Cook, A. R., and T. J. Hoellein. 2016. Environmental drivers of N. P. Hitt, K. O. Maloney, and C. Patrick. 2011. Metacom- leaf breakdown in an urban watershed. Freshwater Science munity theory as a multispecies, multiscale framework for 35:311–323. studying the influence of river network structure on riverine Cooper, C. A., P. M. Mayer, and B. R. Faulkner. 2014. Effects of communities and ecosystems. Journal of the North American road salts on groundwater and surface water dynamics of Benthological Society 30:310–327. sodium and chloride in an urban restored stream. Biogeo- Brown, J. H., and A. Kodric-Brown. 1977. Turnover rates in chemistry 121:149–166. insular biogeography: effect of immigration on extinction. Corsi, S. R., L. A. De Cicco, M. A. Lutz, and R. M. Hirsch. Ecology 58:445–449. 2015. River chloride trends in snow-affected urban water- 394 | Ecological resistance in urban streams R. M. Utz et al.

sheds: increasing concentrations outpace urban growth rate Evans, D. H. 2008. Osmotic and ionic regulation: cells and and are common among all seasons. Science of the Total animals. CRC Press, Boca Raton, Florida. Environment 508:488–497. Fagan, W. F. 2002. Connectivity, fragmentation, and extinction Corsi, S. R., D. J. Graczyk, S. W. Geis, N. L. Booth, and K. D. risk in dendritic metapopulations. Ecology 83:3243–3249. Richards. 2010. A fresh look at road salt: aquatic toxicity and Fitzpatrick, F. A., and M. C. Peppler. 2010. Relation of urbani- water-quality impacts on local, regional, and national scales. zation to stream habitat and geomorphic characteristics in Environmental Science and Technology 44:7376–7382. nine metropolitan areas of the United States. Scientific Investi- Coulter, C. B., R. K. Kolka, and J. A. Thompson. 2004. Water gations Report 2010-5056. US Geological Survey, Reston, quality in agricultural, urban, and mixed land use watersheds. Virginia. Journal of the American Water Resources Association 40: Fletcher, T. D., G. Vietz, and C. J. Walsh. 2014. Protection of 1593–1601. stream ecosystems from urban stormwater runoff: the mul- Cuffney, T. F., R. A. Brightbill, J. T. May, and I. R. Waite. 2010. tiple benefits of an ecohydrological approach. Progress in Responses of benthic macroinvertebrates to environmental Physical Geography 38:543–555. changes associated with urbanization in nine metropolitan Foster, D., F. Swanson, J. Aber, I. Burke, N. Brokaw, D. Tilman, areas. Ecological Applications 20:1384–1401. and A. Knapp. 2003. The importance of land-use legacies to Cuffney, T. F., R. Kashuba, S. S. Qian, I. Alameddine, Y. K. Cha, ecology and conservation. BioScience 53:77–88. B. Lee, J. F. Coles, and G. McMahon. 2011. Multilevel re- Grant, E. H. C., J. D. Nichols, W. H. Lowe, and W. F. Fagan. gression models describing regional patterns of invertebrate 2010. Use of multiple dispersal pathways facilitates amphib- and algal responses to urbanization across the USA. Journal ian persistence in stream networks. Proceedings of the Na- of the North American Benthological Society 30:797–819. tional Academy of Sciences of the United States of America Cushing, C. E., K. W. Cummins, and G. W. Minshall. 2006. River 107:6936–6940. and stream ecosystems of the world. University of California Gray, G. 1981. Species composition and life histories of aquatic Press, Oakland, California. insects in a lowland Sonoran Desert stream. American Mid- Daily, G. 1997. Nature’s services: societal dependence on natu- land Naturalist 106:229–242. ral ecosystems. Island Press, Washington, DC. Griffith, M. B. 2014. Natural variation and current reference for Daley, M. L., J. D. Potter, and W. H. McDowell. 2009. Saliniza- specific conductivity and major ions in wadeable streams of tion of urbanizing New Hampshire streams and groundwa- the conterminous USA. Freshwater Science 33:1–17. ter: effects of road salt and hydrologic variability. Journal of Grimm, V., and C. Wissel. 1997. Babel, or the ecological stabil- the North American Benthological Society 28:929–940. ity discussions: an inventory and analysis of terminology and Davis, A. P., W. F. Hunt, R. G. Traver, and M. Clar. 2009. Bio- a guide for avoiding confusion. Oecologia (Berlin) 109:323– retention technology: overview of current practice and future 334. needs. Journal of Environmental Engineering 135:109–117. Gurnell, A., M. Lee, and C. Souch. 2007. Urban rivers: hydrology, de Jesús-Crespo, R., and A. Ramírez. 2011. Effects of urban- geomorphology, ecology and opportunities for change. Geog- ization on stream physicochemistry and macroinvertebrate raphy Compass 1:1118–1137. assemblages in a tropical urban watershed in Puerto Rico. Hahn, M. W. 2006. The microbial diversity of inland waters. Journal of the North American Benthological Society 30:739– Current Opinion in Biotechnology 17:256–261. 750. Hale, R. L., M. Scoggins, N. J. Smucker, and A. Suchy. 2016. Ef- del Mar López, T., T. M. Aide, and J. R. Thomlinson. 2001. fects of climate on the expression of the urban stream syn- Urban expansion and the loss of prime agricultural lands in drome. Freshwater Science 35:421–428. Puerto Rico. Ambio 30:49–54. Hammer, T. R. 1972. Stream channel enlargement due to ur- Dolédec, S., and N. Bonada. 2013. So what? Implications of loss banization. Water Resources Research 8:1530–1540. of biodiversity for ecosystem functioning. Pages 169–193 in Hanski, I., and M. Gilpin. 1991. Metapopulation dynamics: brief S. Sabater and A. Elosegi (editors). River conservation: chal- history and conceptual domain. Biological Journal of the Lin- lenges and opportunities. Fundación BBVA, Bilbao, Spain. nean Society 42:3–16. Dow, C. L., and R. A. Zampella. 2000. Specific conductance and Harding, J. S., E. F. Benfield, P. V. Bolstad, G. S. Helfman, and pH as indicators of watershed disturbance in streams of the E. B. D. Jones. 1998. Stream biodiversity: the ghost of land New Jersey Pinelands, USA. Environmental Management 26: use past. Proceedings of the National Academy of Sciences 437–445. of the United States of America 95:14843–14847. Drummond, M. A., and T. R. Loveland. 2010. Land-use pressure Hatt, B. E., T. D. Fletcher, C. J. Walsh, and S. L. Taylor. 2004. and a transition to forest-cover loss in the eastern United The influence of urban density and drainage infrastructure on States. BioScience 60:286–298. the concentrations and loads of pollutants in small streams. Ebisemiju, F. S. 1989. Patterns of stream channel response to Environmental Management 34:112–124. urbanization in the humid tropics and their implications for Hawley, R. J., and B. P. Bledsoe. 2011. How do flow peaks and urban land use planning: a case study from southwestern durations change in suburbanizing semi-arid watersheds? A Nigeria. Applied Geography 9:273–286. southern California case study. Journal of Hydrology 405: Elmqvist, T., C. Folke, M. Nyström, G. Peterson, J. Bengtsson, 69–82. B. Walker, and J. Norberg. 2003. Response diversity, ecosys- Hawley, R. J., and B. P. Bledsoe. 2013. Channel enlargement in tem change, and resilience. Frontiers in Ecology and the semiarid suburbanizing watersheds: a southern California case Environment 1:488–494. study. Journal of Hydrology 496:17–30. Volume 35 March 2016 | 395

Hawley, R. J., B. P. Bledsoe, E. D. Stein, and B. E. Haines. 2012. Jordan, B. A., W. K. Annable, C. C. Watson, and D. Sen. 2010. Channel evolution model of semiarid stream response to Contrasting stream stability characteristics in adjacent ur- urban-induced hydromodification. Journal of the American ban watersheds: Santa Clara Valley, California. River Re- Water Resources Association 48:722–744. search and Applications 26:1281–1297. Hawley, R. J., K. R. MacMannis, and M. S. Wooten. 2013. Bed Kaushal, S. S., and K. T. Belt. 2012. The urban watershed con- coarsening, riffle shortening, and channel enlargement in tinuum: evolving spatial and temporal dimensions. Urban urbanizing watersheds, northern Kentucky, USA. Geomor- Ecosystems 15:409–435. phology 201:111–126. Kaushal,S.S.,P.M.Groffman,L.E.Band,C.A.Shields,R.P. Hawley, R. J., and G. Vietz. 2016. Addressing the urban stream Morgan,M.A.Palmer,K.T.Belt,C.M.Swan,S.E.G.Findlay, disturbance regime. Freshwater Science 35:278–292. and G. T. Fisher. 2008. Interaction between urbanization and Herlihy, A. T., J. L. Stoddard, and C. B. Johnson. 1998. The climate variability amplifies watershed nitrate export in Mary- relationship between stream chemistry and watershed land land. Environmental Science and Technology 42:5872–5878. cover data in the Mid-Atlantic region, U.S. Water, Air, and Kefford, B. J., G. L. Hickey, A. Gasith, E. Ben-David, J. E. Dun- Soil Pollution 105:377–386. lop, C. G. Palmer, K. Allan, S. C. Choy, and C. Piscart. 2012. Hitt, N. P., and P. L. Angermeier. 2011. Fish community and Global scale variation in the salinity sensitivity of riverine bioassessment responses to stream network position. Jour- macroinvertebrates: Eastern Australia, France, Israel and nal of the North American Benthological Society 30:296– South Africa. PLoS ONE 7:e35224. 309. King, R. S., and M. E. Baker. 2010. Considerations for analyzing Hobbs, R. J., S. Arico, J. Aronson, J. S. Baron, P. Bridgewater, ecological community thresholds in response to anthropo- V. A. Cramer, P. R. Epstein, J. J. Ewel, C. A. Klink, A. E. Lugo, genic environmental gradients. Journal of the North Ameri- D. Norton, D. Ojima, D. M. Richardson, E. W. Sanderson, can Benthological Society 29:998–1008. F. Valladares, M. Vilà, R. Zamora, and M. Zobel. 2006. Novel King, R. S., M. E. Baker, P. F. Kazyak, and D. E. Weller. 2011. ecosystems: theoretical and management aspects of the new How novel is too novel? Stream community thresholds at ecological world order. Global Ecology and Biogeography 15: exceptionally low levels of catchment urbanization. Ecologi- 1–7. cal Applications 21:1659–1678. Holomuzki, J. R., and B. J. F. Biggs. 2000. Taxon-specific re- King, R. S., M. E. Baker, D. F. Whigham, D. E. Weller, T. E. sponses to high-flow disturbance in streams: implications Jordan, P. F. Kazyak, and M. K. Hurd. 2005. Spatial con- for population persistence. Journal of the North American siderations for linking watershed land cover to ecological in- Benthological Society 19:670–679. dicators in streams. Ecological Applications 15:137–153. Hooper, D. U., E. C. Adair, B. J. Cardinale, J. E. K. Byrnes, B. A. King, R. S., M. S. Scoggins, and A. Porras. 2016. Stream biodiver- Hungate, K. L. Matulich, A. Gonzalez, J. E. Duffy, L. Gamfeldt, sity is disproportionately lost to urbanization when flow per- and M. I. O’Connor. 2012. A global synthesis reveals biodiver- manence declines: evidence from southwestern North Amer- sity loss as a major driver of ecosystem change. Nature 486: ica. Freshwater Science 35:340–352. 105–108. Klein, R. 1979. Urbanization and stream water quality impair- Hooper, D. U., F. S. Chapin, J. J. Ewel, A. Hector, P. Inchausti, S. ment. Water Resources Bulletin 15:948–963. Lavorel, J. H. Lawton, D. M. Lodge, M. Loreau, S. Naeem, Kollaus, K. A., K. P. K. Behen, T. C. Heard, T. B. Hardy, and B. Schmid, H. Setälä, A. J. Symstad, J. Vandermeer, and D. A. T. H. Bonner. 2014. Influence of urbanization on a karst terrain Wardle. 2005. Effects of biodiversity on ecosystem function- stream and fish community. Urban Ecosystems 18:293–320. ing: a consensus of current knowledge. Ecological Mono- Konrad, C. P., and D. B. Booth. 2002. Hydrologic trends associ- graphs 75:3–35. ated with urban development for selected streams in the Hopkins, K. G., N. B. Morse, D. J. Bain, N. D. Bettez, N. B. Puget Sound Basin, Western Washington. Water-Resources Grimm, J. L. Morse, M. M. Palta, W. D. Shuster, A. R. Bratt, Investigations Report 02-4040. US Geological Survey, Tacoma, and A. K. Suchy. 2015. Assessment of regional variation in Washington. stream flow responses to urbanization and the persistence Laliberté, E., J. A. Wells, F. DeClerck, D. J. Metcalfe, C. P. of physiography. Environmental Science and Technology 49: Catterall, C. Queiroz, I. Aubin, S. P. Bonser, Y. Ding, J. M. 2724–2732. Fraterrigo, S. McNamara, J. W. Morgan, D. Sanchez Merlos, Imberger, S., C. Walsh, and M. Grace. 2008. More microbial activ- P. A. Vesk, and M. M. Mayfield. 2010. Land-use intensifica- ity, not shredder abundance or abrasive flow increases the tion reduces functional redundancy and response diversity breakdown rate of labile leaf litter in urban streams. Journal of in plant communities. Ecology Letters 13:76–86. the North American Benthological Society 27:549–561. Lane, E. W. 1955. The importance of fluvial morphology in Janke, B. D., J. C. Finlay, S. E. Hobbie, L. A. Baker, R. W. hydraulic engineering. Proceedings of the American Society Sterner, D. Nidzgorski, and B. N. Wilson. 2014. Contrasting of Civil Engineering 81:1–17. influences of stormflow and baseflow pathways on nitrogen Lenat, D. R., and J. K. Crawford. 1994. Effects of land use on and phosphorus export from an urban watershed. Biogeo- water quality and aquatic biota of three North Carolina Pied- chemistry 121:209–228. mont streams. Hydrobiologia 294:185–199. Jones, D. K., M. E. Baker, A. J. Miller, S. T. Jarnagin, and D. M. Leopold, L. B., R. Huppman, and A. Miller. 2005. Geomorphic Hogan. 2014. Tracking geomorphic signatures of watershed effects of urbanization in forty-one years of observation. suburbanization with multitemporal LiDAR. Geomorphol- Proceedings of the American Philosophical Society 149:349– ogy 219:42–52. 371. 396 | Ecological resistance in urban streams R. M. Utz et al.

Lewis, D. B., J. P. Kaye, and A. P. Kinzig. 2014. Legacies of agri- Park, C. C. 1977. Man-induced changes in stream channel mor- culture and urbanization in labile and stable organic carbon phology. Pages 121–144 in K. J. Gregory (editor). River chan- and nitrogen in Sonoran Desert soils. Ecosphere 5:59. nel changes. John Wiley and Sons, Chichester, UK. Loperfido, J. V., G. B. Noe, S. T. Jarnagin, and D. M. Hogan. Parr, T. B., N. J. Smucker, M. W. Neale, and C. N. Bentsen. 2014. Effects of distributed and centralized stormwater best 2016. Potential roles of past, present, and future urbaniza- management practices and land cover on urban stream hy- tion characteristics in producing varied stream responses. drology at the catchment scale. Journal of Hydrology 519, Freshwater Science 35:436–443. Part C:2584–2595. Paul, M. J., and J. L. Meyer. 2001. Streams in the urban land- Lyons, J., S. W. Trimble, and L. K. Paine. 2000. Grass versus scape. Annual Review of Ecology and Systematics 32:333– trees: managing riparian areas to benefit streams of central 365. North America. Journal of the American Water Resources Peterson, G., C. R. Allen, and C. S. Holling. 1998. Ecological Association 36:919–930. resilience, biodiversity, and scale. Ecosystems 1:6–18. MacArthur, R. H., and E. O. Wilson. 1967. The theory of island Piscart, C., J.-C. Moreteau, and J.-N. Beisel. 2005. Biodiversity biogeography. Princeton University Press, Princeton, New and structure of macroinvertebrate communities along a small Jersey. permanent salinity gradient (Meurthe River, France). Hydro- MacRae, C. R., and A. C. Rowney. 1992. The role of moderate biologia 551:227–236. flow events and bank structure in the determination of chan- Pizzuto, J. E., W. C. Hession, and M. McBride. 2000. Compar- nel response to urbanization. 45th Annual Conference. Resolv- ing gravel-bed rivers in paired urban and rural catchments ing Conflicts and Uncertainty in Water Management. Proceed- of southeastern Pennsylvania. Geology 28:79–82. ing of the Canadian Water Resources Association, Kingston, Poff, N. L. 1992. Why disturbances can be predictable: a per- Ontario. spective on the definition of disturbance in streams. Journal Maloney, K. O., and D. E. Weller. 2011. Anthropogenic disturbance of the North American Benthological Society 11:86–92. and streams: land use and land-use change affect stream eco- Poff, N. L., J. D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegaard, systems via multiple pathways. Freshwater Biology 56:611–626. B. D. Richter, R. E. Sparks, and J. C. Stromberg. 1997. The Martin, E. H., C. J. Walsh, M. Serena and J. A. Webb. 2014. natural flow regime. BioScience 47:769–784. Urban stormwater runoff limits distribution of platypus. Aus- Pond, G. J., M. E. Passmore, N. D. Pointon, J. K. Felbinger, tral Ecology 39:337–345. C. A. Walker, K. J. G. Krock, J. B. Fulton, and W. L. Nash. McNeil, V. H., and M. E. Cox. 2007. Defining the climate signal 2014. Long-term impacts on macroinvertebrates downstream in stream salinity trends using the interdecadal Pacific oscil- of reclaimed mountaintop mining valley fills in central Ap- lation and its rate of change. Hydrology and Earth System palachia. Environmental Management 54:919–933. Sciences 11:1295–1307. Prowse, C. W. 1987. The impact of urbanization on major ion Meyer, J. L., M. J. Paul, and W. K. Taulbee. 2005. Stream eco- flux through catchments: a case study in Southern England. system function in urbanizing landscapes. Journal of the Water, Air, and Soil Pollution 32:277–292. North American Benthological Society 24:602–612. Ramírez, A., R. de Jesús-Crespo, D. M. Martinó-Cardona, N. Morgan, R. P., and K. M. Kline. 2011. Nutrient concentrations Martínez-Rivera, and S. Burgos-Caraballo. 2009. Urban streams in Maryland non-tidal streams. Environmental Monitoring in Puerto Rico: what can we learn from the tropics? Journal of and Assessment 178:221–235. the North American Benthological Society 28:1070–1079. Mullineaux, D. R. 1970. Geology of the Auburn, Renton, and Raney, S. M., and M. C. Eimers. 2014. Unexpected declines in Black Diamond quadrangles, King County, Washington. US stream phosphorus concentrations across southern Ontario. Geological Survey Professional Paper 672. US Geological Canadian Journal of Fisheries and Aquatic Sciences 71:337– Survey, Reston, Virginia. 342. Odemerho, F. O. 1992. Limited downstream response of stream Roy, A. H., A. D. Rosemond, M. J. Paul, D. S. Leigh, and J. B. channel size to urbanization in a humid tropical basin. Pro- Wallace. 2003. Stream macroinvertebrate response to catch- fessional Geographer 44:332–339. ment urbanization (Georgia, U.S.A.). Freshwater Biology 48: Olson, J. R., and C. P. Hawkins. 2012. Predicting natural base- 329–346. flow stream water chemistry in the western United States. Sayama, T., J. J. McDonnell, A. Dhakal, and K. Sullivan. 2011. Water Resources Research 48:W02504. How much water can a watershed store? Hydrological Pro- Osborne, L. L., and M. J. Wiley. 1992. Influence of tributary spa- cesses 25:3899–3908. tial position on the structure of warmwater fish communities. Schueler, T., L. Fraley-McNeal, and K. Cappiella. 2009. Is im- Canadian Journal of Fisheries and Aquatic Sciences 49:671– pervious cover still important? Review of recent research. 681. Journal of Hydrologic Engineering 14:309–315. Osterkamp, W. R., and J. M. Friedman. 2000. The disparity be- Schumm, S. A., M. D. Harvey, and C. C. Watson. 1984. Incised tween extreme rainfall events and rare floods—with empha- channels: morphology, dynamics, and control. Water Resources sis on the semi-arid American West. Hydrological Processes Publications, Littleton, Colorado. 14:2817–2829. Sexton, J. O., X.-P. Song, C. Huang, S. Channan, M. E. Baker, and Palmer, M. A., J. D. Allan, and C. A. Butman. 1996. Dispersal as J. R. Townshend. 2013. Urban growth of the Washington, a regional process affecting the local dynamics of marine and DC–Baltimore, MD metropolitan region from 1984 to 2010 stream benthic invertebrates. Trends in Ecology and Evolu- by annual, Landsat-based estimates of impervious cover. Re- tion 11:322–326. mote Sensing of Environment 129:42–53. Volume 35 March 2016 | 397

Shuster, W. D., J. Bonta, H. Thurston, E. Warnemuende, and Vietz, G. J., M. J. Sammonds, C. J. Walsh, T. D. Fletcher, I. D. D. R. Smith. 2005. Impacts of impervious surface on water- Rutherfurd, and M. J. Stewardson. 2014. Ecologically rele- shed hydrology: a review. Urban Water Journal 2:263–275. vant geomorphic attributes of streams are impaired by even Simon, A. 1989. A model of channel response in disturbed alluvial low levels of watershed effective imperviousness. Geomor- channels. Earth Surface Processes and Landforms 14:11–26. phology 206:67–78. Simon, A., and M. Rinaldi. 2000. Channel instability in the Violin, C. R., P. Cada, E. B. Sudduth, B. A. Hassett, D. L. loess area of the Midwestern United States. Journal of the Penrose, and E. S. Bernhardt. 2011. Effects of urbanization American Water Resources Association 36:133–150. and urban stream restoration on the physical and biological Smith, R. F., L. C. Alexander, and W. O. Lamp. 2009. Dispersal structure of stream ecosystems. Ecological Applications 21: by terrestrial stages of stream insects in urban watersheds: a 1932–1949. synthesis of current knowledge. Journal of the North Ameri- Walsh, C. J., T. D. Fletcher, and M. J. Burns. 2012. Urban can Benthological Society 28:1022–1037. stormwater runoff: a new class of environmental flow prob- Smith, R. F., R. J. Hawley, M. W. Neale, G. J. Vietz, E. Diaz- lem. PLoS ONE 7:e45814. Pascacio, J. Herrmann, A. C. Lovell, C. Prescott, B. Rios-Touma, Walsh, C. J., and J. Kunapo. 2009. The importance of upland B. Smith, and R. M. Utz. 2016. Urban stream renovation: in- flow paths in determining urban effects on stream ecosys- corporating societal objectives to achieve ecological improve- tems. Journal of the North American Benthological Society ments. Freshwater Science 35:364–379. 28:977–990. Smith, R. F., and W. O. Lamp. 2008. Comparison of insect com- Walsh, C. J., A. H. Roy, J. W. Feminella, P. D. Cottingham, P. M. munities between adjacent headwater and main-stem streams Groffman, and R. P. Morgan. 2005. The urban stream syn- in urban and rural watersheds. Journal of the North Ameri- drome: current knowledge and the search for a cure. Journal can Benthological Society 27:161–175. of the North American Benthological Society 24: 706–723. Smith, R. F., P. D. Venugopal, M. E. Baker, and W. O. Lamp. Walsh, C. J., A. K. Sharpe, P. F. Breen, and J. A. Sonneman. 2015. Habitat filtering and adult dispersal determine the taxo- 2001. Effects of urbanization on streams of the Melbourne nomic composition of stream insects in an urbanizing land- region, Victoria, Australia. I. Benthic macroinvertebrate com- scape. Freshwater Biology 60:1740–1754. munities. Freshwater Biology 46:535–551. Snyder, C. D., J. A. Young, R. Villella, and D. P. Lemarié. 2003. Walsh, C. J., and J. A. Webb. 2014. Spatial weighting of land use Influences of upland and riparian land use patterns on and temporal weighting of antecedent discharge improves pre- stream biotic integrity. Landscape Ecology 18:647–664. diction of stream condition. Landscape Ecology 29:1171–1185. Tomer, M. D., and M. R. Burkart. 2003. Long-term effects of Wang, S.-Y., E. S. Bernhardt, and J. P. Wright. 2014. Urban nitrogen fertilizer use on groundwater nitrate in two small stream denitrifier communities are linked to lower functional watersheds. Journal of Environmental Quality 32:2158–2171. resistance to multiple stressors associated with urbanization. Tu, J., Z.-G. Xia, K. C. Clarke, and A. Frei. 2007. Impact of urban Hydrobiologia 726:13–23. sprawl on water quality in eastern Massachusetts, USA. Envi- Webster, J. R., and J. L. Meyer. 1997. Organic matter budgets for ronmental Management 40:183–200. streams: a synthesis. Journal of the North American Ben- Urban, M. C., D. K. Skelly, D. Burchsted, W. Price, and S. Lowry. thological Society 16:141–161. 2006. Stream communities across a rural–urban landscape Williams, W. D. 1987. Salinization of rivers and streams: an gradient. Diversity and Distributions 12:337–350. important environmental hazard. Ambio 16:180–185. Utz, R. M., K. N. Eshleman, and R. H. Hilderbrand. 2011. Varia- Wohl, E., B. P. Bledsoe, R. B. Jacobson, N. L. Poff,S.L. tion in physicochemical responses to urbanization in streams Rathburn, D. M. Waters, and A. C. Wilcox. 2015. The natu- between two Mid-Atlantic physiographic regions. Ecological ral sediment regime in rivers: broadening the foundation for Applications 21:402–415. ecosystem management. BioScience 65:358–371. Utz, R. M., and R. H. Hilderbrand. 2011. Interregional variation Wolman, M. G. 1967. A cycle of sedimentation and erosion in in urbanization-induced geomorphic change and macroin- urban river channels. Geografiska Annaler, Series A, Physi- vertebrate habitat colonization in headwater streams. Jour- cal Geography 49:385–395. nal of the North American Benthological Society 30:25–37. Wright, I. A., P. J. Davies, S. J. Findlay, and O. J. Jonasson. Utz, R. M., R. H. Hilderbrand, and D. M. Boward. 2009. Identi- 2011. A new type of water pollution: concrete drainage in- fying regional differences in threshold responses of aquatic frastructure and geochemical contamination of urban waters. invertebrates to land cover gradients. Ecological Indicators Marine and Freshwater Research 62:1355–1361. 9:556–567. Yamashita, Y., B. D. Kloeppel, J. Knoepp, G. L. Zausen, and R. Utz, R. M., R. H. Hilderbrand, and R. L. Raesly. 2010. Regional Jaffé. 2011. Effects of watershed history on dissolved organic differences in patterns of fish species loss with changing matter characteristics in headwater streams. Ecosystems 14: land use. Biological Conservation 143:688–699. 1110–1122.