the freshwater biome gradient framework: predicting ......ways to delineate freshwater biomes....

SYNTHESIS & INTEGRATION

The freshwater biome gradient framework: predicting macroscaleproperties based on latitude, altitude, and precipitation

WALTER K. DODDS ,1,� LINDSEY BRUCKERHOFF,1 DAROLD BATZER,2 ANNE SCHECHNER,1 CASEY PENNOCK,1

ELIZABETH RENNER,1 FLAVIA TROMBONI ,3 KARI BIGHAM,4 AND SAMANTHA GRIEGER5

1Division of Biology, Kansas State University, Manhattan, Kansas 66506 USA2Department of Entomology, University of Georgia, Athens, Georgia 30602 USA

3Global Water Center and Biology Department, University of Nevada, Reno, Nevada 89557 USA4Department of Biological and Agricultural Engineering, Kansas State University, Manhattan, Kansas 66506 USA

5School of the Environment, Washington State University, Vancouver, Washington 98686 USA

Citation: Dodds, W. K., L. Bruckerhoff, D. Batzer, A. Schechner, C. Pennock, E. Renner, F. Tromboni, K. Bigham, andS. Grieger. 2019. The freshwater biome gradient framework: predicting macroscale properties based on latitude, altitude,and precipitation. Ecosphere 10(7):e02786. 10.1002/ecs2.2786

Abstract. Understanding global ecological patterns and processes, from biogeochemical to biogeo-graphical, requires broad-scale macrosystems context for comparing and contrasting ecosystems. Climategradients (precipitation and temperature) and other continental-scale patterns shape freshwater environ-ments due to their influences on terrestrial environments and their direct and indirect effects on the abi-otic and biotic characteristics of lakes, streams, and wetlands. We combined literature review, analyses ofopen access data, and logical argument to assess abiotic and biotic characters of freshwater systems acrossgradients of latitude and elevation that drive precipitation, temperature, and other variability. Weexplored the predictive value of analyzing patterns in freshwater ecosystems at the global macrosystemsscale. We found many patterns based on climate, particularly those dependent upon hydrologic character-istics and linked to characteristics of terrestrial biomes. For example, continental waters of dry areas willgenerally be widely dispersed and have higher probability of drying and network disconnection, greatertemperatures, greater inorganic turbidity, greater salinity, and lower riparian canopy cover relative toareas with high precipitation. These factors will influence local community composition and ecosystemrates. Enough studies are now available at the continental or global scale to start to characterize patternsunder a coherent conceptual framework, though considerable gaps exist in the tropics and less developedregions. We present illustrative global-scale trends of abiotic, biotic, and anthropogenic impacts in fresh-water ecosystems across gradients of precipitation and temperature to further understanding of broad-scale trends and to aid prediction in the face of global change. We view freshwater systems as occurringacross arrays of multiple gradients (including latitude, altitude, and precipitation) rather than areas withspecific boundaries. While terrestrial biomes capture some variability along these gradients that influencefreshwaters, other features such as, slope, geology, and historical glaciation also influence freshwaters.Our conceptual framework is not so much a single hypothesis as a way to logically characterize patternsin freshwaters at scales relevant to (1) evolutionary processes that give rise to freshwater biodiversity, (2)regulatory units that influence freshwater ecosystems, and (3) the current scope of anthropogenic impactson freshwaters and the vital ecosystem services they provide.

Key words: altitude; climate change; freshwater biomes; lakes; latitude; macroscale properties; macrosystems;precipitation; rivers; streams; temperature; wetlands.

Received 29 April 2019; accepted 6 May 2019; final version received 28 May 2019. Corresponding Editor: Debra P. C.Peters.

❖ www.esajournals.org 1 July 2019 ❖ Volume 10(7) ❖ Article e02786

https://orcid.org/0000-0002-6666-8930

https://orcid.org/0000-0002-6666-8930

https://orcid.org/0000-0002-6666-8930

https://orcid.org/0000-0001-7287-8048

https://orcid.org/0000-0001-7287-8048

https://orcid.org/0000-0001-7287-8048

info:doi/10.1002/ecs2.2786

Copyright: © 2019 The Authors. This is an open access article under the terms of the Creative Commons AttributionLicense, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.� E-mail: [email protected]

INTRODUCTION

Characteristics of terrestrial ecosystems con-verge across global precipitation and tempera-ture gradients and ecologists use the concept ofbiomes to classify terrestrial systems across thesegradients (Holdridge 1947, Whittaker 1970).Here, we ask whether such an approach wouldbe useful to freshwater ecologists concerned withpattern and process at the broadest (macrosys-tem) scales. For example, an ecologist consider-ing a mesic temperate grassland expectsdominance by grasses, a role of fire in maintain-ing grass dominance, an evolutionary historywith large grazing mammals, adaptations to sea-sonality, ability to withstand hot and dry condi-tions as well as times of ample moisture, andmoderately high productivity. If we consider anArctic freshwater habitat, we expect strong sea-sonal effects; it may freeze completely and willhave long days with low light intensity andmonths of darkness. All species in these Arcticwaters must be adapted to withstand cold tem-peratures, low-stature plants dominate wetlands,and groundwater is shallow. Limnologistsalready make assumptions that specific charac-teristics link to large-scale patterns of abiotic dri-vers, though they do not organize theseassumptions in a coherent structure analogous tothe biomes framework in terrestrial ecology.

Dodds et al. (2015) proposed a framework todescribe how global patterns of temperature andprecipitation might constrain stream ecology, ascorrelated with their corresponding terrestrialbiomes. Under the stream biome gradient frame-work, climate was a clear determinant of streamproperties including hydrology (flooding anddrying), geomorphology (drainage density andchannel morphology), canopy cover, seasonality,and many other core abiotic and biotic drivers ofaquatic ecology. These translated to patterns ofdiversity and distributions of animal life historytraits as well as differences in expected ecosys-tem rates (Dodds et al. 2015). It is unclear towhat degree these patterns transfer to other sur-face waters (wetlands and lakes).

Here we explore the idea of biomes forstreams, wetlands, and lakes. We strive to pro-vide a predictive framework of ecological prop-erties across the broadest spatial scales(macroscales). Macrosystem approaches are rele-vant to biogeography, conservation, and humanscales of management (e.g., national laws relateto conservation and environmental protection)and represent a novel way to view ecology oncontinental scales (Heffernan et al. 2014, McClu-ney et al. 2014). We view this as a framework, asthe prior paper Dodds et al. (2015) used the term“concept” but did not provide one simple con-cept, but rather a general predictive approachbased on constraints imposed by climate. Weunderstand that broad generality requires atrade-off with specific local prediction (Levins1966), and that there will be many local excep-tions to the generalizations of the framework. Wehope this framework identifies macroscale pat-terns that freshwater scientists can use to provideexpectations on how systems will vary based onlarge-scale drivers.The first step is providing a definition of biome

that could apply to freshwaters. Our generalworking definition of a biome is a unique spatialregion defined by predictable abiotic and bioticcharacteristics (process, form, and function) ofecological interest driven by climatic factors andothers that vary over broad spatial scales. As withterrestrial biomes, aquatic biomes will not havehard and fast boundaries. Transitions amongbiomes across space occur largely as gradientsrather than as sharp edges. Biomes are not bio-geographically systematic (not zoogeographicregions or ecoregions; sensu Abell et al. 2008).Ecologists define terrestrial biomes based onplant architecture and productivity even thoughthey can contain different species that have con-vergent evolution leading to similar function.Aquatic primary producers may function simi-larly, and we explore this and other potentialways to delineate freshwater biomes. Whittaker(1970) stated biomes could be developed foraquatic systems based on physiognomy but notedthat aquatic communities intergrade with each


SYNTHESIS & INTEGRATION DODDS ET AL.

http://creativecommons.org/licenses/by/3.0/

other in different ways and are less dependent onglobal climatic gradients than are terrestrial habi-tats (presumably because water is less limiting).

We assessed patterns of both abiotic and bioticcharacteristics across global gradients of precipi-tation and temperature (driven by elevation andlatitude) and additional characteristics thatmight be a function of elevation or latitude (e.g.,glaciation and slope). We considered surfacefreshwater habitats, including streams, wetlands,lakes, and reservoirs. Our goal was to assess pat-terns of abiotic factors across global gradientsand link these abiotic patterns to observed bio-logical responses. We chose abiotic factorsknown to be drivers of ecological structure andfunction in freshwater systems. These factorsinclude temperature, light, geological processes,hydrology, chemistry, and habitat density anddistribution. Where data are lacking, we do nottest predictive patterns, but hypothesize poten-tial patterns based on logic. We intend to gener-ate a general predictive framework of testablehypotheses but cannot test every hypothesis.Thus, this paper is a statement of approachrather than a complete application. We use acombination of literature review and data analy-sis (see Appendix S1 for data descriptions). Forsome patterns, we explore representative tran-sects (n = 20) across global gradients (Fig. 1) totest predictions. The precipitation gradient tran-sects were set to cut across landscapes keepingtemperature somewhat constant, while tempera-ture gradients were set along gradients of alti-tude or latitude attempting to keep precipitationroughly constant. In some cases, we find biologi-cal patterns by latitude, altitude, temperature, orprecipitation and then hypothesize potentialcauses. We consider these potential areas forfuture research.

PHYSIOGRAPHY: STREAMS, WETLANDS, ANDLAKES ACROSS ABIOTIC GRADIENTS

Abiotic factors structuring freshwater biotavary across broad scales. We characterize fresh-water systems across abiotic gradients known tobe important drivers of ecosystem structure andfunction. Those abiotic characteristics can alsodrive ecological and evolutionary processes pre-dictably over broad spatial scales. In this section,we evaluate and synthesize spatial gradients of

geomorphic, climatic, hydrologic, and otherphysiographic characteristics of lentic and loticsystems. These sections form the basis of the sub-sequent discussion on how climate can drive bio-tic processes to lead to unique freshwaterbiomes.

TemperatureTemperature controls metabolism of individual

organisms, ecosystem rates, the balance betweenactual and potential evapotranspiration (a deter-minant of runoff to freshwaters), and other physi-cal factors of habitats (e.g., lake stratification,extent and duration of ice cover, dissolved oxy-gen levels). Daily and seasonal water temperatureextremes of freshwater systems vary predictablywith climate (Table 1). In smaller water bodies,canopy cover modulates diurnal swings in tem-perature (create a microclimate, Allen 2016), sogiven a constant amount of energy input, daily,and seasonal water temperatures will likely bemore extreme in grasslands, tundra, and desertsand less extreme in forests (except in deciduousforests during times of no leaf cover).Elevation and latitude control temperature;

tropical systems are warmer (by up to 30°C aver-age temperature), but less variable seasonally.With mean average temperatures of �20°C orless in high latitude systems, water is not liquidfor much of the year. Ice formation exerts multi-ple effects on the ecology and geomorphology ofaquatic environments. For instance, shallowhabitats in high latitudes and altitudes mayfreeze completely during winter, allowing onlyfreeze-tolerant organisms to survive. In areaswith permanent ice cover, summer thawing maylead to ponds or streams on the surface of the icethat form a unique cold habitat. Ice also influ-ences stratification of lakes and river flows. Forexample, intense flooding can occur followingice breakup which influences erosional processes,water quality (e.g., temperature, nutrients),in-channel and floodplain vegetation, and whichcan disturb aquatic organisms such as benthicmacroinvertebrates and fishes (Prowse and Culp2003).The solubility of oxygen in water is relatively

low and declines with increasing water tempera-tures (Dodds and Whiles 2019). This, combinedwith greater metabolic rates at higher tempera-tures (up to a point) and the prevalence of



amictic lakes in tropical climates, means with allthings equal, anoxic conditions will developmore frequently in aquatic habitats of warmerclimates. However, ice and snow cover can alsolead to anoxia, so there will be exceptions to thisgeneral pattern. Anoxia alters biogeochemicalcycling and survival of many freshwater organ-isms. Overall, temperature provides an indirectcontrol of ecosystem processes through its effectson metabolic rates and anoxic conditions, and wewill discuss these implications more fully in thesections on ecosystem rates.

LightLight fuels aquatic production, warms water,

harms organisms in the form of ultraviolet radia-tion, and controls photolysis rates (Kirk 1994).

Latitude, altitude, cloud cover, and air pollutioninfluence solar radiation. Incoming irradiance,photosynthetically active radiance (PAR), andultraviolet (UV) light influence heating, photosyn-thesis, photolysis, and light damage, respectively.Incoming total irradiance energy is the main

force that warms water; it is greatest in tropicalregions and least in high latitude regions (Fig. 2).All five of the warm-cold transects from Fig. 1with varying latitude had a trend of increasingdirect normal irradiance as latitude neared 0°.For example, along transect 3, direct normal irra-diance at 44.5° latitude was 4.9 kWh�m�2�d�1,compared to 3.21 at 67.5° latitude.The influence of elevation on optical atmo-

spheric thinning causes a 50% increase in PARover a 3 km altitude increase (Barry and Chorley

Fig. 1. Twenty transects across varying climate gradients used to investigate patterns across biome gradients.Solid black lines represent precipitation gradients drawn with temperature held relatively constant (10), whiledashed black lines represent temperature gradients driven by altitude or latitude (10). Background biome imagefrom https://commons.wikimedia.org/wiki/File:Vegetation.png.

Table 1. Physical and chemical patterns over climate gradients.

Feature High altitude Low altitude High latitude Low latitude Mesic Xeric

Solar radiation Higher Lower Lower Higher Lower HigherUltraviolet radiation Higher Lower Lower Higher Lower HigherMonthly irradiance minima Lower Higher (near tropics)Water temperature Lower Higher Lower Higher Higher LowerWater temperature extremes No trend No trend Greater Lesser Lesser GreaterNutrients Lower Higher Lower HigherSalinity Lower Higher Lower Higher

Note: Blank cells are where no pattern is predicted, and no trend is indicated where a direct statistical test indicated no trend(see Temperature, Light, and Chemistry).



https://commons.wikimedia.org/wiki/File:Vegetation.png

2003). Incoming PAR peaks at 20° N latituderather than at the equator, due to atmosphericmoisture (Lewis 2011). Cloud cover decreasesPAR and UV radiation (H€ader et al. 2015) as is

evident in desert areas of Australia, NorthernAfrica, and Southwest North America (Fig. 2).Photosynthetically active radiance declines dra-matically between 20° and 70°.

Fig. 2. (a) Global map of average daily global horizontal irradiation in kWh/m2 from SolarGIS 2014 GeoModelSolar. Highlights global trends of light availability in terms of latitude and altitude. (b) Global map of annualmonthly mean ultraviolet-B radiation (UV-B) from 2004 to 2013 (Beckmann et al. 2014); glUV dataset availableat: https://www.ufz.de/gluv/.



https://www.ufz.de/gluv/

High levels of UV radiation cause photolysis,which can lead to cell damage and influencechemical properties of freshwater systems. UVradiation exhibits similar latitudinal trends asother forms of irradiance, with around ten timesmore UV-B in the tropics than in high latituderegions (Fig. 2). Altitude also plays a role, withclear UV maxima in the Himalayan Mountainsalong the Tibetan Plateau and the Andes on thedry coast of South America, where UV radiationlevels are around 6000 J�m�2�d�1, more thandouble the UV flux in neighboring lower altitudeareas.

Additional climate-driven variables can influ-ence the availability of incoming solar radiationonce it reaches freshwater systems. Ice and snowcover influence light availability in rivers, wet-lands, and lakes (Wetzel 2001), so high latitudeand elevation habitats have lower light season-ally. High inorganic sediment loads can intercepta substantial amount of light. Absorption by dis-solved organic carbon (DOC) reduces light avail-ability in freshwaters. In boreal lakes, inputs ofDOC increase with latitude, in turn decreasingthe penetration of PAR and UV radiation in thewater column during summer months (Pienitzand Vincent 2000). Paleolimnological studies ofsubarctic lakes indicate minimum light availabil-ity and smaller UV/PAR ratios occurred duringperiods of maximum vegetation coverage andhigher DOC (Pienitz and Vincent 2000). Finally,in small streams, small wetlands, and forestedwetlands, tree canopy can intercept light. Thus,regions of forest may have more shading thangrassland, desert, or tundra regions.

Geological processesMajor geological processes forming river and

stream networks are somewhat different fromthose forming wetlands and lakes, though someprocesses are shared (e.g., oxbow lakes andriparian wetlands). Rivers and streams form aswater drains to lower topography, driven by tec-tonics and erosion, although glacial outburstphenomena can set river courses at temperateand higher latitudes (e.g., Waitt 1985). Fluvialgeomorphologists classify streams based on mor-phological characteristics such as formative dis-charge, dominant sediment type and load, valleycharacteristics, and channel dimensions and pro-file (Leopold and Wolman 1957, Schumm 1977,

Rosgen 1996), and these characteristics all varywith changes in precipitation and seasonality.Valley characteristics such as confinement,

slope, lithology, and soils, which are functions ofelevation, drive material sorting and channel pat-tern (Table 2). Slope and available sediment con-trol the dominant sediment transport mechanismof streams, which can influence network struc-ture. A global assessment indicates steeperslopes are related to narrower and less sinuousrivers and there are more rivers in areas withlower slopes (Frasson et al. 2019).In general, bedload transport is prevalent in

upland streams of steep slopes while suspendedload transport dominates in lowland rivers(Church 2006). Therefore, transportable materialat higher elevations tends to be cobble- to boul-der-sized rock of colluvial origin, while finermaterials of alluvial, eolian, lacustrine, and/orglacial origin dominate at lower elevations.Channel pattern is expected to evolve fromstraight to braided and/or sinuous (Leopold andWolman 1957), while transportable bed materialsize decreases moving from small, higher eleva-tion streams to large, low elevation rivers(Church 2006). Channel form generally follows aprecipitation gradient; braided channels with ahigh sediment load and width-to-depth ratio areexpected to be more prevalent in drier climatesthan wet ones, holding elevation constant(Dodds et al. 2015). Based on these generaliza-tions, unstable channel forms should be morecommon in regions of low elevations and/orprecipitation.Wetlands occur where shallow water perma-

nently or periodically inundates the land surfaceand are especially common where depth togroundwater is shallow and local geomorphol-ogy allows for positive surface water balance(Fan et al. 2013). Wetlands form from processessuch as glacial scarring, tectonic interaction, suc-cession from small lakes, and sediment inputfrom rivers (Tooth and Viles 2014). Glacialdepressions are associated with wetlands inhigher latitudes and altitudes. Therefore, wet-lands in more arid, lower latitude regions (unin-fluenced by Pleistocene glacial cover) usuallyhave longer formation histories (Tooth andMcCarthy 2007). Arid wetlands are more depen-dent on surface flow as a source of water thanhumid wetlands and therefore should have



longer periods of desiccation and higher levels ofsedimentation. Across global gradients, weexpect increased dependence on local physicaltopography in wetland formation with decreas-ing precipitation, as deeper depressions arerequired to meet the top of the water table.

Major processes that form wetlands and lakesinclude tectonic, volcanic, landslide, glacial, dis-solution, fluvial, eolian, beaver dammed, coastal,and meteoric (Hutchinson 1957). Jackson et al.(2014) followed Hutchinson’s synopsis of lakeformation and provided a modified geomorphicclassification of inland freshwater wetlands. Flu-vial wetlands include stream and river floodplainwetlands, as well as beaver wetlands. Lacustrinewetlands occur at lake fringes and drying lakes.Glacial wetlands include bogs, fens, tundra, pot-holes, and kettles. Groundwater wetlandsinclude dissolution and low-lying coastal plaindepressions. This leaves miscellaneous inlanddepressions such as wind-carved playas, paleo-marine depressions, buffalo wallows, andanthropogenic. The distribution of lake sizes isstill not well understood as a function of climateand topography, though progress is being madeon this front (Steele and Heffernan 2017).

Of these lake and wetland formations, tectonic,volcanic, coastal, dissolution, and meteoric arenot closely related to broad biogeographic pat-terns of climate, but others in Table 3 are eitherdirectly or indirectly. Coastal formations aremainly prevalent at low elevations, and glacialdepressions predominate at high latitudes and/orhigh altitudes. Wind-scoured depressions are

more common in drier areas where vegetativecover is less likely to inhibit eolian erosion. Land-slide lakes are more common in areas with stee-per topography (with high local elevations), andfluvial lakes and wetlands are more abundant inareas with higher densities of large rivers.

HydrologyRegional climate and geology control hydrol-

ogy and maintain all freshwater ecosystems.Direct precipitation, surface runoff, and ground-water are the main sources of water that feed intofreshwaters. The dominant timing and form (rainor snow) of precipitation can greatly affect wateravailability, phenology, and intensity of effects onfreshwater systems (Table 4). For all aquaticenvironments, depth to groundwater is a crucialaspect dictated by climate and is particularlyimportant in dry areas. In high altitude or lati-tude areas, permafrost strongly limits depth togroundwater (Fan et al. 2013).Flow patterns in streams correlate to the terres-

trial biomes in which they are embedded (Doddset al. 2015). The flow regime of streams and riv-ers shifts from ephemeral and flashy to morestable, groundwater-fed, perennial systems asprecipitation increases, though floods still occurin mesic areas (Poff 1996, Ullrich et al. 2010).Snowmelt influences flow in high latitude or alti-tude areas. Mountain ranges may affect climate,inducing a rain-shadow effect on the leewardside of the range. For example, habitats found onthe leeward side of a mountain range may bemore ephemeral or intermittent near the range,

Table 2. Variation of geomorphological attributes of freshwater systems across gradients of precipitation andtemperature driven by elevation and latitude.

Feature High altitude Low altitudeHigh

latitudeLow

latitude Mesic Xeric

River slope High gradient Low gradientDominant transportmechanism

Bedload Suspended

Transportable material Larger FinerTurbidity (concentration) Low High Lower HigherChannel pattern Straight, braided Sinuous, braided More sinuous More braidedSediment load variability Greater† Smaller† Smaller† Greater† Smaller GreaterStream width-to-depth ratio Smaller Greater Smaller GreaterStability of channel form Greater stability Less stability Greater stability Less stabilityFormation history Shorter Longer Shorter Longer Shorter Longer

Note: Blank cells are where no general expected pattern is predicted.†Over the last 1000 yr (see Dearing and Jones 2003).



shifting to more perennial systems at lower ele-vations further away from the range, while onthe windward side, flow regime is expected to beperennial, regardless of elevation.

Wetland character is determined by the domi-nant sources of water (direct precipitation, sur-face flow, and/or groundwater; Brinson 1993).These sources control chemistry and hydrope-riod or hydropattern (i.e., the general seasonalpattern of surface water inundation, Jacksonet al. 2014). Thus, climate and biome gradientswill dictate regional wetland character. Forexample, in humid climates, groundwater

discharge into wetlands contributes to localhydrology, but direct precipitation and surfaceflow dominate (Prigent et al. 2007, Tooth andMcCarthy 2007). In contrast, arid climates havefewer wetlands, which are typically sustained bygroundwater discharge (e.g., spring-associatedponds, pans) or flowing surface water (e.g., flu-vial wetlands). While occasional intense rainsmay cause floods, these events will be ephem-eral, and usually insufficient to create conditionscharacteristic of wetlands such as hydric soilsand obligate wetland biota (Keddy 2000). Aridwetlands are prone to prolonged desiccation

Table 3. Variation of hypothesized attributes related to lake and wetland formation across climate gradients.


Glacial depressions More Less More LessWind-caused depression Less MoreLandslide lakes More LessFluvial lakes and wetlands Less More More LessCoastal/lakeshore Rare More common

Note: Blank cells are where no general expected pattern is predicted.

Table 4. Gradients related to hydrology over climate gradients driven by precipitation, altitude, and latitude.

Feature High altitude Low altitudeHigh

latitudeLow


Flow regime Ephemeral onleeward side,perennial onwindward side,snowmelt

Variable Snowmelt Stormflow More stable,groundwater-fed,perennial

Ephemeraland flashy

Primarywater source(wetlands)

Snowmelt Precipitation Snowmelt Precipitation Precipitation Groundwaterdischarge, flowfrom nearbyhighaltitude areas

Evapotranspiration Lower Higher Lower HigherActual: potentialevapotranspiration

Lower Higher Lower Higher >1 <1

Seasonalvariation ininundatedsurface area

Higher Lower

Mixing andstratification

Mono- ordimictic

Variabledependingon latitude

Mono- ordimictic

Amixis(tropical)

Depth togroundwater

Shallow(permafrost,shallow soil)

Variable Shallow(permafrost,shallow soil)

Variable Shallow Deep

Spatialconnectivity

Higher Lower(closed basins,intermittentriver networks)




(Tooth and McCarthy 2007, Tooth and Viles2014). In semi-arid climates, wetland habitat willbe predominantly associated with seasonal riverflooding rather than precipitation or groundwa-ter. Along altitude gradients, precipitation driveswater tables at low elevations, whereas high-ele-vation wetlands are more temporally stable dueto snowmelt (Millar et al. 2017). We expect simi-lar relationships with increasing water depth inwetlands with decreasing latitude (Jim�enez-Alfaro et al. 2014).

Evapotranspiration is a universally importantprocess of water loss from wetlands, varyingstrongly with climatic conditions (Brinson 1993,Kroes and Brinson 2004), as the balance of poten-tial and actual evapotranspiration rates controlsurface flow and groundwater discharge. Emer-gent and adjacent macrophyte transpiration moststrongly affects wetlands but can influencestreams and small ponds, making the link to ter-restrial biomes especially strong. Evaporationvaries across temperature and humidity gradi-ents, with potential rates being greatest in hot,dry climates, and least important in cool, humidclimates. Terrestrial biome type (e.g., presence orabsence of trees) will control transpiration rates,with direct influences on hydrology. For exam-ple, flooding is not completely controlled by sea-sonal changes in precipitation, but instead can bedriven by changes in evaporation and tree tran-spiration rates, both of which decline in winter,allowing rivers to fill and spill into adjacentfloodplains (Benke et al. 2000). These samechanges also affect how much water accumulateson floodplains from direct rainfall and ground-water discharge.

Precipitation can drive inundation directly orindirectly. In warm regions, monsoonal precipita-tion cycles can directly cause seasonal flooding.In contrast, inundation of boreal wetlands(where snowmelt is a major water source) has lit-tle correlation with the timing of snowfall, butrather the total magnitude and pattern of warm-ing. In humid regions like the Pantanal or Ama-zon Basin of South America, inundationdependence of riverine wetlands on precipitationincreases with distance from the river channel.The Pantanal inundation depends upon a sea-sonal cycle of precipitation, but downstreamareas of wetlands may lag by 6 months fromperiods of peak rainfall (Hamilton 2002). Across

latitudinal gradients, seasonal variation in wet-land surface area increases with distance fromthe equator (Prigent et al. 2007).Lakes (except saline lakes in closed basins) are

less responsive to climatic fluctuations withrespect to hydrology compared to streams andwetlands, but some hydrologic attributes of lakesare linked to climate, elevation, and latitude. Forexample, a key controlling factor of seasonalcycles and biogeochemistry in lakes is the extentand frequency of mixing and stratification(Dodds and Whiles 2019). Temperate to high alti-tude or latitude lakes will not mix when there isice cover and may not stratify in the summer;therefore, mixing may only occur once (mo-nomictic) or twice (dimictic) per year. Tropicallakes are susceptible to biogenically inducedamixis (never mixing). In this case, continuousprimary production in the surface waters resultsin organisms sinking into the profundal zoneand releasing salts on decomposition, leading tohigher density of the deeper waters. The biogenicactivity leads to a relatively permanent stateof stratification. All of the factors controllingstratification could be influenced by climaticfluctuations.

SedimentIncoming sediment size and load affect mor-

phological characteristics of all freshwater sys-tems, as well as biology and water quality. Bedmaterial sediment size tends to decrease withdecreasing elevation (Church 2006). Based on thetransportable sediment available and the domi-nant sediment transport mechanism (Church2006), turbidity is expected to be greater inlowland rivers where suspended load transporta-tion dominates, vs. in steeper streams of higherelevations.Sediment load depends on watershed sources

and sinks, both of which can vary considerablyover space and time. A global study conductedby Dearing and Jones (2003) using 100- to 1000-yr-old sediment records from various lakes andseas found the relative magnitude of changebetween the minimum sediment accumulationrate (Smin) and the maximum rate (Smax) variedalong both elevation and latitudinal gradients.The dimensionless ratio Smax/Smin indicated thegreatest variance in accumulation in regions ofhigh elevations, low latitudes, and/or small



drainages. However, these sediment load gener-alizations developed over temperature gradientsmay vary significantly when assessing differenttime scales.

Assessing sediment concentration and loadsover a precipitation gradient is less complex. Ingeneral, sediment concentration should be higherin semi-arid and arid regions due to more surfaceexposure and, as a result, erosional events. Forexample, rivers in United States desert andMediterranean biomes have roughly 3–12 timesgreater mean sediment concentrations than thosefrom forested regions (Dodds and Whiles 2004).However, sediment load is generally propor-tional to discharge, so load should be greater inmore mesic areas; Table 2 summarizes these sed-iment characteristics.

Characteristics of sediments can also dependupon climate. In dry regions, surface soils havelower organic carbon content than more mesicbiomes (Jobb�agy and Jackson 2000). Thus, sedi-ments entering aquatic systems in drier areas areexpected to have a higher inorganic content. Theorganic content of allochthonous sediment canstrongly influence biota (Whiles and Dodds2002).

ChemistryNutrient concentration trends may vary across

climatic gradients as nutrient ecoregions tend tohave variable expected baseline nutrient concen-trations (Smith et al. 2003). Dodds et al. (2015)hypothesized nutrient concentrations would gen-erally be low in biomes where terrestrial vegeta-tion is dense and moderately higher in desertsdue to an increase in erosion potential. Nitrogenconcentrations in undisturbed tropical water-sheds are about five times greater than in similartemperate watersheds, but this is enigmatic assome also claim higher rates of denitrification tobe characteristic of the tropics (Downing et al.1999).

Climate has no clear relationship with the bal-ance between N and P availability in freshwater.Elser et al. (2007) found little support for latitudi-nal variation in the autotrophic productionresponse to nutrient enrichment. However, wehave relatively few data from the tropics or arcticsystems and comprehensive analyses mayuncover trends with climate. Dodds and Smith(2016) noted considerable variance in total N to

total P ratios at reference concentrations acrossecoregions of the United States, suggesting thatstoichiometric nutrient supply might link tobiome.Turner et al. (2003) documented dissolved sili-

cate to dissolved nitrate ratio is a sensitive indi-cator of aquatic food web health and harmfulalgal blooms. Their analysis of silicate and nitrateconcentrations in the world’s largest rivers (37%of Earth’s watershed area) indicated dissolvedsilicate yield increases with increasing runoff(i.e., following a precipitation gradient) but is notdependent on temperature. Thus, we expect sili-cate yield not to depend on latitude or altitudeexcept as they influence precipitation.Saline lakes are one habitat in which climate

has a strong influence on chemistry, which influ-ences biological and physical properties (Hutchin-son 1957; Table 1). Saline waters are common indry climates because terminal lake formation ispossible when evaporation equals rates of input.Saline lakes can have unique chemical character-istics including high nutrient content, and chem-istry depends upon the geology of the areasfeeding them. Saline waters can demonstrategreater capacity for stratification, as dissolvedsalts have a greater influence on water densitythan temperature, and areas with saline watersare likely to have waters of highly contrastingsalinities. Salinity has a strong influence on thetypes of organisms that can persist.

Distribution, density, and size of habitatsThe density and size of freshwater habitats can

influence species dispersal and diversity, andother aspects of connectivity. Climate gradientsobviously influence density of freshwaters(Table 5). Density is highest in wet areas, depend-ing on geology. High-latitude areas have highdensities of wetlands because evaporation ratesare low and permafrost limits infiltration (Fig. 3).Continental ice sheets also flattened large areaswhile leaving behind many depressions thatbecame lakes and wetlands. This is apparent inboreal and arctic latitudes of the Northern Hemi-sphere (45–75° N), which have the highest densi-ties of lakes (Verpoorter et al. 2014).Even though wetlands are abundant at high

latitudes, wetlands are also widely dispersedthroughout the tropics in lower latitudes (Fig. 3;Mitsch et al. 2009, Fan et al. 2013, Gumbricht



et al. 2017). Wetland density is high in mesicareas and increases with decreasing elevationbecause riparian wetlands follow many large riv-ers (e.g., the Amazon) and flood plains are widerwhere there is little topographic relief (Tooth andViles 2014). In arid climates, river inflows andlocal physical topography (i.e., tectonic depres-sion and windblown sediments) promote surfacewater retention and therefore support wetlandformation.

We used global data for delineated stream net-works (Appendix S2) to assess the distribution ofrivers across all continents except Antarctica. Thenumber of rivers within an area generallydecreased as temperature increased (Fig. 4).Area-corrected densities of flowing waterincreased with distance from the equator.

Perhaps this is related to increased temperaturesand evaporation as well as drier conditions at 30°north and south of the equator driven by Hadleycells, coupled with potential underestimation ofriver density in wetter tropical areas. Thereappear to be slightly more streams per unit areain higher latitude areas of the Northern Hemi-sphere (Fig. 4). Rivers are relatively evenly dis-tributed across the Southern Hemisphere, whilein the Northern Hemisphere, the highest andmost variable densities of streams are foundaround 50–60° N (Fig. 4). Interestingly, we foundno clear trends between the density of rivers andelevation or precipitation at a global scale(Appendix S2). This could be an artifact of ourmethod to extract densities, which does not takeinto account stream size or water volume.

Table 5. Patterns of distribution, density, and habitat size of streams, wetlands, and lakes across climategradients.


Drainage density Lower Higher Higher LowerPerennial stream density Lower Higher No trend No trend Higher LowerWetland and lake density Lower Higher Higher Lower Higher Lower

Note: Blank cells are where no pattern is predicted, and no trend is indicated where a direct statistical test in this paper orone cited in the text indicated no trend (see Distribution, density, and size of habitats).

Fig. 3. Global distribution of different types of wetlands. Spatial data were compiled from multiple sources byLehner and D€oll (2004). Because wetlands can exhibit seasonal variations in extent, boundaries are difficult todelineate. Wetland classifications containing percentages therefore represent the proportion of wetland withinthe area based on extents defined by multiple data sources.



Small streams (1st order) contribute most tothe length of rivers globally, while moderate-sized rivers (5th–9th order) contribute most tostream area (Downing et al. 2012). Bifurcationratios (Leopold 1962, Welcomme 1976) predictthe size distribution of rivers to be similar glob-ally, even across regions with different climates.Downing et al. (2012) compared the size distri-bution of rivers in the United States and Africaand found similar ratios of stream area to landarea between the two continents (0.56% and0.30%, respectively). They concluded that theproportion of area covered by rivers is small (lessthan 1%) regardless of geographic region. Thesecomparisons were made at very broad scales(continents), and they were not specificallydesigned to capture precipitation and tempera-ture gradients.

Verpoorter et al. (2014) found a generaldecreasing trend of lake density, area, andperimeter, from north to south across the globe.Similar to patterns of wetlands, lake densities arehighest in boreal and arctic latitudes of theNorthern Hemisphere (45–75° N, Appendix S2).The low total number of lakes in the SouthernHemisphere in part relates to lower proportionof land mass relative to the Northern

Hemisphere. We found even when correcting forland area, a greater proportion of terrestrial areawas covered by lake habitats in the NorthernHemisphere relative to the Southern Hemi-sphere, although a peak in lake coverage existsjust below the equator (Fig. 5). This high densityof lakes at high latitudes is likely due to the largenumber of glaciated lakes in those regions. Thearea covered by lake habitats also decreases withincreasing elevation. Over half of the global den-sity, perimeter, and area of lakes is located at ele-vations lower than 500 m above sea level(Verpoorter et al. 2014). While lake density andsize decrease with increasing temperature acrosselevation and latitudinal gradients (AppendixS2), our analysis of global lakes indicated thehighest proportion of lake surface area/total landarea was found at intermediate levels of precipi-tation (Fig. 5).Although we classify freshwater habitats into

lakes, rivers, wetlands, these different habitatsinteract with one another to form mosaics offreshwater habitat. The summed connectivity ofthese different habitats may be important forboth ecological and evolutionary phenomena.We combined global datasets of freshwater dis-tributions (Appendix S1) to assess patterns of

Fig. 4. Distribution of number of river segments within 1000-km2 grids distributed globally in relation tolatitude and temperature. Spatial data were obtained from Lehner et al. (2008; rivers) and Fick and Higmans(2017; temperature).



hydrologic connectivity across temperature andprecipitation gradients globally (analysis descri-bed in Appendix S2). We used patch cohesionvalues of small landscapes (10,000 km2) acrossthe globe to compare connectivity of all freshwa-ter habitats. Patch cohesion measures the physi-cal connectedness of freshwater habitat patchesby relating the perimeter of freshwater patches tothe total area of freshwater habitat within eachsub-landscape unit (Gustafson 1998). Higher val-ues indicate more connectivity or aggregation offreshwater habitats within a landscape, whilelower numbers indicate freshwater habitats aredispersed and fragmented across the landscape.Our analysis indicated higher connectivitywithin landscapes at latitudes around 50° andaround the equator (Fig. 6). Connectivity wasgreatest at intermediate elevation and precipita-tion levels. Overall, the large area of lakes andwetlands in the Northern Hemisphere andaround the equator drive connectivity across large(latitudinal) temperature gradients, although wesee the highest connectivity at high temperatures(Fig. 6). In wetter climates, where streams have ahigh drainage density (i.e., channels per unitarea), there is a greater probability that wetlandsand lakes will be linked by flowing water, andwetlands and lakes will also tend to have greater

numbers per unit area (with less distance betweenthem).Glaciation can lead to relatively sharp patterns

of connectivity. Similar to our analysis, Ferguset al. (2017) analyzed lake, wetland, and streamconnectivity at the southern extent of glaciationof Eastern North America. They found areas thathad been covered by the continental ice sheethad more connected lakes and wetlands. Areasimmediately south of the glaciation had less con-nected wetlands and lakes but more connectedstreams and rivers.

BIOTIC CHARACTERISTICS ACROSS BIOMEGRADIENTS: WHAT IS A FRESHWATER BIOME?

Plant architecture and productivity are the keyfeatures of terrestrial biomes. Dominant planttypes (e.g., type of macrophytes or phytoplank-ton) might connect less strongly to aquatic thanto terrestrial habitats across climate gradients aswater limitation is a key determinant of plantproductivity and form in terrestrial environ-ments. Emergent plants in wetlands follow cli-mate patterns (e.g., short stature or mosses athigh altitudes or latitudes), but trends in otherproducers (e.g., periphyton, submerged macro-phytes) may tie less closely to climate as we

Fig. 5. Total surface lake area within 1000-km2 grids distributed across the globe in relation to latitude and pre-cipitation. Lake area data were obtained from the Global Lakes and Wetlands Database (Lehner and D€oll 2004)and temperature data from Fick and Higmans (2017).



discuss below. Thus, any categorization of fresh-water biomes may require consideration of vari-ables in addition to the plant physiognomy thatforms the primary basis of terrestrial biome clas-sification. Such variables could include ecosys-tem rates in addition to productivity, communityarchitecture, or functional types or diversities ofprimary producers and other organisms. Here,we consider how climatic gradients influencepredictability of these potential variables. Weconsider primary producers, invertebrates, andvertebrates in this context as well as other

biological characteristics such as ecosystem rates,food chain length, and trophic state.

Ecosystem ratesGross primary production (GPP), ecosystem

respiration (ER), and their balance, net ecosystemproduction (NEP), are fundamental ecosystemrates that control biogeochemistry as well as sys-tem productivity. The balance between GPP andER indicates the primacy of autochthonous orallochthonous carbon sources and potentialenergy pathways in the system. These can be

Fig. 6. Patch cohesion, a measure of connectivity, of all freshwater habitats across the globe varied across glo-bal gradients of elevation, precipitation, latitude, and temperature. White dots represent individual landscapes(1000 km2), and heat map colors represent the density of landscape exhibiting given values of patch cohesionand latitudes. Higher values of patch cohesion represent greater degree of connectivity between freshwater habi-tat patches. Freshwater spatial data were combined from Lehner and D€oll (2004) and Lehner et al. (2008), and cli-mate data were obtained from Fick and Higmans (2017). See Appendix S2 for geoprocession explanation.



influenced by climate, as well as additionalecosystem processes, such as nutrient uptakeand retention (Table 6).

Gross primary production.—Stream GPP increa-ses with temperature and latitude (Bernot et al.2010). This is likely due to indirect and con-founding variables, including light availability(discussed in Light). The rates of GPP in streamsacross biomes are discussed in more detail else-where (Dodds et al. 2015, Bernhardt et al. 2018).Bernhardt et al. (2018) suggest that position inthe river continuum is potentially a more impor-tant control on metabolism than the biome theriver is embedded in.

In wetlands dominated by emergent vegeta-tion, light or nutrients likely limit GPP, so lati-tude, altitude, and cloud cover could limitproduction of emergent vegetation. Temperatureand length of growing season are also importantdrivers of wetland plant production. Middletonand McKee (2004) considered factors limitingbald cypress production across about 10 degreeslatitude. Cold temperatures and shorter growingseasons limited production at higher latitudeswhile respiration and water loss limit productionat lower latitudes, making the relationshipbetween latitude and production curvilinear.Brinson et al. (1981) found a similar curvilinearrelationship with wetland net biomass produc-tion. Lowest biomass production occurred inwetlands in southern Florida (lowered fertility)and northern Canada (lower temperatures).Northern ombrotrophic (deriving nutrients fromatmospheric moisture) bogs had minimal

nutrient availability or water flow and were low-est in biomass production.At constant temperature, important influences

in biomass production include water availability,water flow, and nutrient availability. Brinsonet al.’s (1981) expansive freshwater wetlandstudy indicated the importance of moisture gra-dients to wetland GPP. In forested wetlands,GPP was higher in floodplain forests (which hadhigh nutrients and water levels) than in shallowcypress ponds (reliant mostly on precipitation).In non-forested wetlands, GPP was greatest inriverine marshes and lowest in stagnant, nutri-ent-poor wetlands. Overall, GPP was greater innon-forested wetlands than in forested wetlands.Unlike most terrestrial systems, bryophytes cancontribute significantly to overall ecosystem pro-ductivity of wetlands (Brinson et al. 1981, Pere-gon et al. 2008). Sphagnum mosses exhibit strongcorrelations with climate gradients; growth andproductivity are positively correlated withincreased temperature and precipitation. Climatevariables affect herb or shrub production lessdirectly, with productivity being influenced moreby water levels. Shrubs can grow better in shal-lower water, while herbs and grasses are moreproductive under deeper water conditions(Thornman and Bayley 1997).Many factors could influence broad biogeo-

graphic patterns of GPP in lakes, but light is akey variable (Lewis 2011). Light interacts withnutrient concentrations to drive the magnitudeand temporal variability of GPP. Across 7500lakes globally, daily GPP rates in Northern

Table 6. Ecosystem rates as they vary over climate gradients driven by precipitation, altitude, and latitude.

FeatureHigh

altitudeLow

altitudeHigh

latitudeLow


Ecosystem respiration Lower Higher Lower Higher Higher LowerGross primaryproduction

Higher Lower

Net ecosystemproduction

Lower Higher(near tropics)

Lower (in small systems,more riparian canopylight interception)

Higher (in small systems,less riparianlight interception)

Wetland primaryproduction

Lower Higher Lower Higher

Sphagnum production Higher LowerWetland netecosystem production

Higher Lower Lower Higher

Wetland pH Lower Higher Lower Higher




Hemisphere temperate lakes increased withincreased surface irradiance, whereas daily GPPin tropical lakes decreased with increased lightavailability due to photoinhibition (Staehr et al.2016). Greater irradiance warmed lakes andweakened thermal stratification in tropical lakesrelative to temperate lakes (Lewis 1996). Warmertemperatures combined with more efficientnutrient cycling and less variable solar irradiancemake primary production twice as high in tropi-cal lakes relative to temperate lakes (Lewis 2011).

Ecosystem respiration and decomposition.—Ecosys-tem respiration can indicate heterotrophic activ-ity in addition to respiration by primaryproducers. Many freshwaters are net hetero-trophic and have more allochthonous than auto-chthonous carbon input (Dodds and Cole 2007).In general, respiration rates respond to tempera-ture similarly across terrestrial, marine, andfreshwater lakes and rivers (Yvon-Durocher et al.2012). Young et al. (2004) proposed leaf decom-position and metabolism as priority functionalecosystem indicators, as they collectively repre-sent the autochthonous and allochthonous basisof freshwater food webs and energy systems.Decomposition has been studied more instreams, but trends are likely to apply to wet-lands and lakes. In general, decomposition ratesof litter in streams increase as latitude and lignincontent drop and temperature, precipitation, andnutrient availability rise (Zhang et al. 2008, Boy-ero et al. 2016). Leaf decomposition rate is alsodriven by litter quality (LeRoy and Marks 2006)and is therefore likely to vary among terrestrialbiomes with different vegetation inputs. Leafdecomposition rates increase by 5–21% with a1–4°C increase in temperature based on broad-scale meta-analysis (Follstad Shah et al. 2017).This increase is less than expected from metabolictheory but leads to substantial expected differ-ences in rates from high latitude to the tropics.Supporting this, Tiegs et al. (2019) report animpressive experiment using constant substratatype to measure decomposition that demonstratedhigher rates of decomposition at lower latitudes.

Ecosystem respiration in streams largelyincreases with increasing temperature anddecreasing altitude (Mulholland et al. 2001,Dodds et al. 2015). In streams, ER appears to belargely unconnected from precipitation gradientsbecause of metabolic compensation, with aquatic

primary producers dominating when canopycover is low (discussed in Dodds et al. 2015).Woody inputs to freshwater systems may also

be important both with respect to respirationrates and as habitat for freshwater animals.Across terrestrial biomes, woody vegetation isassociated with humid climates, so woody debrisinput is greatest in regions of high precipitation.Tree canopy cover, and thus large woody debrisinput, can also vary across gradients of tempera-ture and light availability, but is strongly associ-ated with certain terrestrial biomes. Obviously,large woody debris input to rivers, wetlands,and lakes will be greatest in forested biomes andless in grasslands, tundra, and semi-arid to ariddeserts.Net ecosystem production.—Freshwater meso-

cosm experiments suggest warmer habitats willhave lower net ecosystem production becauserespiration rates are more strongly influenced bytemperature than GPP (Yvon-Durocher et al.2010). This result was verified for streams usinginstream measures of metabolism across a widearray of biomes (Song et al. 2018). The later mea-surements included organisms adapted to thestreams that were investigated, circumventing apotential criticism of the mesocosm experiments.A few studies indicate wetland NEP increases

with elevation. Millar et al. (2017) found NEPwas positively correlated with water table depthand negatively correlated with air temperature.At higher altitudes, wetlands are less dependenton rain, due to increased ice and snowmelt.Additionally, Lovell and Menges (2013) alsofound soil moisture across a modest elevationgradient to be the primary factor controllinggrowth patterns of wetland plants.Any factor that limits light to aquatic habitats

could decrease NEP. For example, greater inor-ganic turbidity in desert areas could limit pro-duction. Cloudy areas, areas with highly tannicwaters (dystrophic), dense canopy cover, andprolonged ice cover can all decrease NEP.Other metabolic rates.—Nitrogen cycling rates in

streams do not vary predictably across climategradients as indicated by broadly geographicallydistributed stream isotopic tracer measurements.Tank et al. (2018) found no relationship betweenammonium uptake rates and temperature forstreams, indicating limitation by other factors.Mulholland et al. (2009) found little evidence for



temperature effects on stream denitrificationrates across temperate biomes.

Turetsky et al. (2014) found many factors influ-enced methane release from wetlands as basedon analysis of data from 71 northern, temperate,and subtropical wetlands. Permanently, wetareas had emission rates influenced by the capac-ities of plants to transport methane from sedi-ments to the atmosphere. In wetlands that dried,which predominate in drier climates, hydro-period and temperature were key controllingfactors.

Lewis (2002) suggests rates of denitrificationare substantially greater in tropical than temper-ate lakes. In lakes that stratify, the greaterhypolimnetic temperatures and prolonged peri-ods of stratification make depletion of dissolvedoxygen more rapid and more likely, leading toconditions that favor denitrification. Presumably,similar arguments would apply to methanogene-sis in lakes, so rates are likely greater in tropicalthan temperate lakes.

Trophic structureSome aspects of food web structure that might

be influenced by climatic gradients includetrophic cascades, food web complexity and sta-bility, and food chain length (Table 7). However,many of these are insufficiently cataloged to gen-eralize across biomes. In very harsh habitats(including high salinity habitats and the freshwa-ters of Antarctica), multicellular animals are rareor absent. In these cases, food chain lengths areshort. Attempts to generalize the importance ofomnivory across broad geographic areas havebeen inconclusive. For example, F€ureder (2007)found decreased omnivory with increasing tem-perature because greater diversity allowed formore specialization, while Gonzalez-Bergonzoniet al. (2012) found increasing omnivory withincreasing temperature because of food limita-tion.

Fern�andez et al. (2017) analyzed ecosystemcomplexity across a latitudinal gradient andquantified complexity using several metrics suchas emergence, self-organization, homeostasis,and autopoiesis (the potential of an ecosystem todevelop organization). They found no significantglobal pattern of any metric of complexity acrossthe latitudinal gradient, but they did observe

complexity driven largely by photoperiod andseasonality, which should relate to latitude.Norman et al. (2017) reviewed isotopic experi-

ments from streams across several continents todescribe the efficiency of nitrogen transfer fromprimary uptake compartments, such as epilithonand detritus, to animals. They found nitrogentransfer rates were greater with open canopy.This indicates that biomes with low riparian veg-etation around streams (and potentially wetlandsand small lakes) should have greater rates ofnitrogen transfer into animal components of theecosystem.Relationships among food chain length and

gradients of temperature, latitude, elevation, orprecipitation are weak in rivers, wetlands, andlakes (Post et al. 2000, Vander Zanden and Fetzer2007). Jardine et al. (2015) found species shiftsfrom tropical to temperate areas of Australia butfound no latitudinal differences in communitystructure in the streams. However, Irons et al.(1994) found a positive correlation betweenstream invertebrate leaf litter processing ratesand latitude, which suggests invertebrates mattermore with respect to nutrient recycling and foodchain length at higher latitudes. Dodds et al.(2015) noted some evidence that food chainlength in streams will increase across a latitudi-nal gradient toward the tropics.Food chain length in lakes is probably driven

more by trophic state than by latitude, altitude,or precipitation. A global synthesis of lake foodchain lengths found weak or no relationshipswith ecosystem size, mean annual air tempera-ture, or latitude (Vander Zanden and Fetzer2007). However, there is ample evidence that cli-mate-dependent factors influence lakes. Thesefactors include changes in phosphorus loading(Vollenweider 1968, Schindler 1977, Jeppesenet al. 2005), mixing regime and stratification,oxygen saturation, and frequency of extremewind events (Mooji et al. 2005, Blenckner et al.2007, Kernan et al. 2009). Additionally, lakesexperience changes in trophic structure drivenby temperature (Gyllstr€om et al. 2005, Bekliogluet al. 2007, Jeppesen et al. 2007, Meerhoff et al.2007a, b) and by complex interactions betweentemperature, nutrients, and physical forces(Jeppesen et al. 2009). Thus, climate gradientsmay be linked to food chain length, but



anthropogenic eutrophication can override thispotential interaction.

Aquatic microbial diversityMicrobes obviously respond to temperature,

light, and nutrients, so aquatic microbial commu-nities should relate to climate gradients. How-ever, global surveys of microbial diversity arelimited given the relatively recent ability to iden-tify the dominant taxonomic units of bacteriaand archaea from the natural environment. Evenglobal surveys of eukaryotic microbes are rare. Insubarctic ponds in Norway and Finland, diatomand cyanobacteria richness followed a unimodaltrend with increased elevation, non-cyanobac-teria richness decreased, while heterotrophic bac-terial diversity was more directly related toelevation and terrestrial productivity (Teittinenet al. 2017). A meta-analysis on freshwater bacte-ria of lakes found pH to be a very influentialenvironmental variable and latitude a weak, butsignificant, variable in influencing bacterial com-munity composition (Newton et al. 2011). Thesefindings, and others (Koopman and Carstens2011, Newton et al. 2011, Antoniades et al.2014), suggest aquatic microbial diversity, at leastto some extent, can be related to terrestrial biodi-versity patterns, and thus should link to thebiomes freshwaters are embedded in. This is apromising area for future research.

Primary producers: distribution and diversityFreshwater macrophytes exhibit complex pat-

terns of diversity across global climate gradients.

We found little comparative information on lati-tudinal patterns of phytoplankton diversity orcommunity structure in lakes and wetlands, orperiphyton in streams. Therefore, our followingdiscussion focuses on macrophytes. Many fresh-water macrophytes have broad geographicranges, but diversity is highest in the Neotropics,intermediate in the Oriental, Nearctic, andAfrotropics, lower in the Palearctic and Aus-tralasia, even lower in the islands of the Pacific,and lowest in Antarctica (Chambers et al. 2008).Of the 412 macrophyte genera, 39% are endemicto a single biogeographic region, with the highestrates of endemism found in the Neotropics andAfrotropics. Grimaldo et al.’s (2017) analysis ofmacrophyte diversity in warm, calcareous riversin Africa and the Americas found latitude wasthe strongest driver of diversity and couldpredict macrophyte richness. Although broadtemperature gradients drive patterns of macro-phyte diversity, local factors can have strongimpacts on diversity, especially the presence ofinvasive species (O’Hare et al. 2012, Grimaldoet al. 2017).Wetland vegetation composition and structure

varies in response to surrounding terrestrialenvironment, water source, salinity, nutrientavailability, and climate conditions. Obviously,forested wetlands (e.g., swamps, boreal bogs) aredominated by perennial woody plants, whereasnon-forested wetlands (e.g., marshes) are domi-nated by annual or perennial emergent orsubmersed plants (Brinson et al. 1981). Gener-ally, dominant wetland species tend to follow

Table 7. Variation of biological properties over climate gradients from literature, data patterns analyzed, orlogical argument.

FeatureHigh

altitudeLow

altitude High latitude Low latitude Mesic Xeric

Canopy cover Lower Higher Lower Higher Higher LowerMacrophytediversity

Lower Higher(Neo and Afro tropics)

Wetland species Less trees Greater woodyvegetation

Greater C:N,less trees

Lesser C:N,more trees

Greaterwoodyvegetation

Greaterherbaceousvegetation

Non-fish vertebratediversity: freshwaterturtle diversity

Lower Higher

Fish species richness Lower Higher Lower Higher Higher LowerPercentageendemic species

No trend No trend No trend No trend

Note: Blank cells are where no pattern is predicted, and no trend is indicated where a direct statistical test indicated no trend(see Patterns of animal diversity across freshwater biomes).



precipitation gradients. Dominant species showedhigh turnover across rainfall gradients in coastalwetlands, with more woody vegetation in wetterareas and graminoid marshes in drier climates(Osland et al. 2014). Wetland plant species arecharacterized by their tolerance for inundation.Woody plants are associated with a higher mois-ture index and seasonally flooded wetlands,while emergent plants are tolerant of prolongedflooding (Garris et al. 2015).

Wetland vegetation also varies with latitude.Collantes et al. (2009) found floristic compositionof Argentinian freshwater wetlands changesalong a latitudinal gradient, due to variations infactors such as soil pH, base cations, and soil C:Nratio. The C:N increased with increasing latitude(most likely due to slow decomposition rates),and pH increased with rainfall (caused by basecation leaching). Studies of rich fens across Eur-ope similarly suggested compositional shifts inspecies with latitude (Jim�enez-Alfaro et al. 2014),with vascular plants responding more stronglyto latitude shifts than bryophytes. Based on priorpublished studies, we predict a transition fromwoody to herbaceous wetland plants from wet todry regions. Wetland net biomass productionand primary production should generally behigher in lower latitude, warmer regions.

Patterns of animal diversity across freshwaterbiomes

There are around 126,000 described freshwateranimal species, making up approximately 9.5%of known global diversity (Balian et al. 2008).This estimate is likely low, as large geographicalgaps in knowledge of species diversity exist,especially across invertebrate groups of CentralAfrica, South America, and South-East Asia. Dueto knowledge gaps in tropical systems, currentestimates of diversity that indicate highest levelsof freshwater diversity in the Palearctic zoogeo-graphic region may be inaccurate. We will focusour discussion of animal diversity patterns onvertebrates and arthropods because these groupshave been most thoroughly described.

A broad consensus exists that aquatic inverte-brate communities in streams, rivers, wetlands,and lakes vary strongly across climatic gradients.This variation occurs among terrestrial biomes,and hypotheses about factors controlling thisvariation are emerging. For stream invertebrate

fauna, assessing latitudinal gradients in diversityhas been a major focus of study, particularlybetween the temperate and tropical zones (e.g.,Boulton et al. 2008). Generally, major inverte-brate groups vary in their diversity responses tolatitude (Table 8). Vinson and Hawkins (2003)observed Ephemeroptera and Trichoptera rich-ness declined with increasing altitude, while Ple-coptera richness increased with altitude, up to2500 m, and then declined. More direct links tobiome gradients for stream invertebrates are alsofound in Vinson and Hawkins (2003). Theygrouped their 493 collections of stream insectsinto biogeographic regions (8) and biomes (9).While both biogeographic region and biomeexplained significant variation in the genericrichness of individual and combined Ephe-meroptera, Plecoptera, and Trichoptera richness,biome tended to account for about twice the vari-ation relative to biogeographic region. In general,forested biomes tend to support a greater rich-ness of Ephemeroptera, Plecoptera, and Tri-choptera taxa than do arid or tundra biomes,with grassland biomes having intermediaterichness.Meta-analyses addressing freshwater wetland

aquatic invertebrates also point toward theimportance of biomes. Batzer and Ruh�ı (2013)compiled global family-level taxa lists from 447individual wetlands, largely from North Americaand Europe. Multivariate community analyses ofthis fauna identified 11 distinct groups of wet-lands, many of which corresponded to biomevariation (e.g., desert depressions, semi-arid (i.e.,grassland) depressions, alpine ponds, northernboreal forest wetlands, Mediterranean ponds).This geographic variation dwarfed macroinverte-brate community variation between temporarilyand permanently flooded habitats, typically con-sidered a major ecological transition for wetlandinvertebrates (Batzer and Wissinger 1996, Well-born et al. 1996, Wissinger 1999). However, tem-porary and permanent wetland subgroupingswere often evident within geographic locations.Ruh�ı et al. (2013) examined wetland invertebrategeneric composition among 225 temporary wet-lands located in the Nearctic and westernPalearctic. Genera in the harsh arid and cold cli-mates tended to concentrate in a few higher taxo-nomic groupings, and in the more benignclimates, genera were spread across many higher



taxa. This suggests harsh climate conditions filtersome groups, but those adapted for the condi-tions can proliferate. For example, the diversityof large branchiopods (fairy shrimps, clamshrimps, tadpole shrimps) tends to be greatest inwetlands with harsh environmental conditions(see Boix and Batzer 2016).

For lake invertebrates, studies of broad geo-graphic variation have focused on crustaceanzooplankton (Cladocera and Copepoda), anddata collected thus far are more ambiguous interms of biome gradients. Dodson (1992)assessed species richness across 66 North Ameri-can lakes of varying latitude, size, depth, andtrophic status. Latitude was not a significant pre-dictor of richness, although lake size and produc-tivity were. However, Gillooly and Dodson(2000) found Cladocera size across WesternHemisphere lakes was greater in the north tem-perate zone than in equatorial or polar regions,suggesting a temperature effect. Havens et al.(2014) confirmed body sizes of assorted crus-tacean zooplankton in lakes declines withincreasing water temperatures associated withlatitudinal change. Beaver et al. (2014) also foundmore small bodied species in warmer waterswhen they assessed 102 western U.S. reservoirsacross three ecoregions.

Vertebrates (other than fishes) dependent onfreshwater habitats to complete at least somepart of their life cycle follow variable relation-ships across temperature and precipitation gradi-ents. Amphibian and freshwater turtle richnesshave positive relationships with precipitationand temperature; richness is greatest near theequator and decreases toward the poles (Buckleyand Jetz 2007, Buhlmann et al. 2009). Individualamphibian groups, however, display complexpatterns of richness. Plethodontid salamander

richness increases with decreasing latitude, prob-ably more related to colonization times followingglacial periods than temperature (Kozak andWiens 2012). Anuran richness exhibits both posi-tive and negative relationships with temperatureacross elevation gradients depending on latitudeand precipitation (Navas 2002). Freshwater turtlediversity typically increases with decreasing ele-vation (Buhlmann et al. 2009). Richness of birdsthat specialize on riverine habitats is greatest atmoderate-to-low elevations, though there ispotentially a hotspot in high elevations of south-central–South-East Asia where numerous rangesoverlap (Buckton and Ormerod 2002).Some relationships with fish diversity and cli-

mate gradients are clear. Antarctic freshwaterhabitats are mostly too harsh to support fish.However, their presence may also be limited bydispersal. Intermittent habitats and saline habi-tats characteristic of drier areas are often devoidof fishes.Abell et al. (2008) delineated 426 freshwater

ecoregions combining river-, wetland-, and lake-dwelling species across the globe and observedboth the total number and density of fish species(total number/ecoregion area) were generallyhighest at low latitudes. We further assessed fishspecies richness and endemism along transects ofclimate gradients (Fig. 1) using the same dataset(see Appendix S1 for data description, andAppendix S3 for analysis details) to assessbroad-scale patterns of freshwater fish diversity.Our analyses generally agreed with Abell et al.

(2008), but we found mixed responses for somemetrics across climate gradients. Specifically, fishspecies richness significantly increased withincreases in minimum annual temperature (sim-ple linear regression: F1,46 = 12.6, R2 = 0.21, P <0.001), and log richness significantly increased

Table 8. Summary of literature on diversity of some stream invertebrate groups as a function of latitude.

Group High latitude Low latitude

Amphipoda More diverse Less diverse (V€ain€ol€a et al. 2007)Coleoptera Less diverse More diverse (Brown 1981, J€ach and Balke 2008)Ephemeroptera Peaks at 30° S and 40°N Peak at 10° N (Vinson and Hawkins 2003) less diverse in general

(Pearson and Boyero 2009)Hemiptera Less diverse More diverse (Polhemus and Polhemus 2008)Odonata Less diverse More diverse (Pearson and Boyero 2009)Plecoptera Peaks at 30° S and 40°N Less diverse (Vinson and Hawkins 2003, Pearson and Boyero 2009)Trichoptera Peaks at 30° S and 40°N Lower (Vinson and Hawkins 2003) or no pattern (Pearson and Boyero 2009)



as mean annual precipitation increased (F1,29 =65.3, R2 = 0.69, P < 0.0001; Fig. 7). Percentage ofendemic species showed no obvious trends withgradients of either temperature or precipitation(Fig. 7). We did not generate regression modelsfor this relationship due to the lack of linear pat-tern and the relatively large number of zeroes(Appendix S3).

The absence of any obvious relationshipbetween endemism and temperature and precip-itation suggests factors other than climate maybe more important in determining distributionsof endemic species. The historical availability ofextinction refugia or rates of allopatric speciationcould be more important than current climategradients for endemic species distributions glob-ally (Collen et al. 2014). Future analyses of diver-sity in rivers, lakes, and wetlands separately, or

data compiled at finer spatial scales may providemore insight into these relationships.

Variation in life historyMany factors influence life history of aquatic

organisms, but some (e.g., intermittence, meantemperature, and seasonality) are more likelythan others to vary across biome gradients. Ephe-meral habitats have clear impact on life history,as organisms that inhabit them must have desic-cation resistant stages or the ability to retreat torefugia during dry periods. Such adaptations arerarer with more complex animal body architec-ture and with greater size. Hydrology plays alarge role in the life history patterns for freshwa-ter shrimp (Hancock and Bunn 1997) and manyother freshwater taxa (Lytle and Poff 2004). How-ever, there is much variation within and among

Fig. 7. Bi-plots of fish richness and percentage endemism per ecoregion across global precipitation and tem-perature transects (Fig. 1) obtained from freshwater ecoregions developed by Abell et al. (2008). We used linearregression to test for trends in species richness (number of species). Richness, which was log transformed acrossprecipitation transects to improve normality, exhibited a significant positive relationship with both temperature(F1,46 = 12.6, R2 = 0.21, P < 0.001) and precipitation (F1,29 = 65.3, R2 = 0.69, P < 0.0001), while the proportion ofendemic species per ecoregion did not exhibit any clear linear trends.



groups, and both abiotic and biotic factors willdrive variation in life history patterns acrosstaxa.

Temperature is important to invertebrate lifehistory, with warmer temperatures increasingprimary productivity and leading to more foodsources for growth by invertebrates (Hurynand Wallace 2000). Temperature, latitude, andphotoperiod affected the abundance of variousspecies of immature zooplankton. Higher tem-peratures advanced phenology of all groups andcopepod hatch rates were highest at midlati-tudes, during periods of high temperature andlong days. Rotifers had greater numbers ofhatches at lower latitudes, although hatchsuccess was greater for shorter days (Jones andGilbert 2016).

Wissinger (1999) diagrammed how hydrope-riod variation may control biota in North Ameri-can wetlands, though his results probably applyto other parts of the world. Wetland hydroperiodvariation in permanence, duration, predictability,phenology, and harshness (of both wet and dryphases) all were predicted to affect biota relatedin large part to invertebrate life history strategies.Wetland characteristics of certain hydroperiodswere often defined by biome type (e.g., desertpools, northern peatlands, California marshes).

Patterns of fish life history strategies (Wine-miller 2005, Mims et al. 2010) vary in response tolocal environmental variables, and these patternslikely translate into patterns across global climategradients. Mims et al. (2010) observed oppor-tunistic life history strategies (small bodied, earlymaturation, low juvenile survivorship) weremost prominent in the southeast portion of theUnited States, periodic strategists (large body,high fecundity, late maturation, low juvenile sur-vivorship) in the northwest portion, and equilib-rium strategists (moderate maturation age, lowfecundity, high juvenile survivorship) in thenorthern part of the country. This study sug-gested a latitudinal gradient, with equilibriumstrategists being more dominant at higher lati-tudes than periodic strategists. Using a similardataset, Mims and Olden (2012) found oppor-tunistic strategists were associated with variableflows, while equilibrium strategists were associ-ated with predictable flows, and additionallyperiodic strategists occurred in areas with highflow seasonality. This same pattern was detected

in Australian fish assemblages (Olden and Ken-nard 2010). Bunn and Arthington (2002) foundmany fishes from highly variable flow regimeshave evolved strategies that allow for strongrecruitment despite disturbances, includingfishes which spawn in months with low andstable flows and avoid months with high andunpredictable flows. As flow regimes are drivenby climatic variability, this suggests that biomescan constrain fish life history strategy.

PREDICTIVE ECOLOGY, THE FRESHWATER BIOMEGRADIENT FRAMEWORK, AND ANTHROPOGENICIMPACTS

The frameworkEven before Alexander von Humboldt pub-

lished the concept of life zones in the geographyof plants (von Humboldt and Bonpland 1807),humanity had words for grasslands, tundra,deserts, and forests. Perhaps humans grasp theconcept of terrestrial biomes more readily as ourcolloquial vocabulary had classified them beforethe idea of biomes was formalized scientifically.However, we view terrestrial biomes as scientificconstructs meant to facilitate generalizations andcomparisons, and we assert that they actuallyoccur across gradients of abiotic factors ratherthan representing discreet ecosystem states.Our review and analyses throughout this

paper suggest many—but not all—characteristicsof lakes, streams, and wetlands are linked to theterrestrial biome in which the freshwater systemis embedded. Still, there are certain pitfalls inusing terrestrial biomes to delineate freshwaterbiomes. Classifying freshwater systems based offterrestrial biomes alone is problematic because itignores heterogeneity both within and amongterrestrial biomes, and because some aquaticecosystems traverse traditional terrestrial biomedelineations.One of the most important shortcomings is

geology, a factor not considered with respect totraditional terrestrial biomes. The geology awatershed is embedded in has strong effects onhydrology and chemistry of freshwaters in addi-tion to geomorphology. An additional problemarises when considering classifications assumingaltitude and latitude as proxies for temperatureto delineate biomes. In the terrestrial biomeframework, tundra would occur at high altitude



and latitude. However, high-elevation tundra atlower latitudes, compared to Arctic tundra,would have substantial slope (which Arctic tun-dra may or may not), less pronounced annualvariation in day length, substantially greaterPAR and UV influx, and lower levels of dissolvedoxygen saturation.

We think that freshwater systems, being dis-tributed across abiotic gradients used to delin-eate terrestrial biomes, can be characterizedfurther given specific combinations of abiotic fac-tors. This is the core of a biome gradientapproach, rather than attempting to draw pre-scribed lines on a map delineating specificregions. Thus, there is merit in classifying fresh-water biomes based on combinations of factorsthat determine the surrounding terrestrial biome(altitude, latitude, and precipitation). Further-more, there is value in refining predictions basedon that information to create a conceptual frame-work specific to freshwater biomes.

This framework could then be used to predicthow a system might behave, what system vul-nerabilities to anthropogenic pressures exist, andhow we might restore or protect ecological func-tions. Essentially, we can create a broad baselineexpectation based on the climate a habitat isfound in. One could consider different freshwa-ter types (rivers, lakes, and wetlands) as biomesin their own regard, but we view these types asmore similar to habitat variance within terrestrialbiomes. For example, a mountainous, coniferousforested area may have meadows, wetlands,rocky areas, and shrub-dominated regions, butecologists would still consider the area as a singleterrestrial biome. The relative distribution of thedifferent types may vary with freshwater biome(e.g., deep, low-conductivity lakes may be rela-tively rare in very dry habitats, and epheme-ral streams and wetlands may be relativelycommon).

In the remainder of this section, we use thefreshwater biome perspective to hypothesizehow abiotic and biotic factors of aquatic ecosys-tems might change with increasing anthro-pogenic impact. We explore how the intensity ofthese impacts varies predictably across latitudi-nal, altitudinal, and precipitation gradients inconjunction with human population densities.This exercise is intended to illustrate some of the

potential strengths of a biome gradient frame-work as applied to continental waters.

Climate changeThe potential synergistic effects between cli-

mate change and other anthropogenic pressureson freshwater systems (e.g., fragmentation andhabitat loss; Markovic et al. 2017) are quicklybeing realized. Global warming and associatedclimate change will have pervasive effects atmacrosystem scales on structure and function ofall freshwater communities; lakes and streamsare important sentinels of global change (Wil-liamson et al. 2008). Conservation and manage-ment could be facilitated by understanding howthese changes could alter the expected freshwaterbiome type in a region. Arctic and high-elevationlakes will experience more warming (MountainResearch Initiative 2015, Kraemer et al. 2017).Warming could lead to increased species diver-sity in eutrophic systems and decreased diversityin oligotrophic systems (Binzer et al. 2015).Jeppesen et al. (2009) reported that warmer lakesare more sensitive to external changes such asincreased nutrient loading or water-level fluctua-tions, suggesting that temperate lakes undergo-ing warming may have heightened sensitivity toeutrophication. Marino et al. (2018) analyzedempirical studies of how experimental warmingstrengthened or weakened trophic cascades.They found that warming strengthened top-down control in colder environments but weak-ened it in warm environments. We previouslydiscussed how warming will shift system meta-bolism by decreasing NEP.A changing climate will also affect precipita-

tion amount, type, duration, intensity, and tim-ing. This will influence hydrology and alterrequirements for land and water management(Goudie 2006). A shift in hydrology can impactstream morphology, leading to excessive erosionand/or deposition until a new dynamic equilib-rium is reached (Lane 1955). These kinds ofmorphological adjustments can have seriousimplications for the biota of many streams, wet-lands and lakes (Cluer and Thorne 2014). Goudie(2006) hypothesized that river stability will beimpacted the most in cold, tropical, and aridregions of the globe. Furthermore, these effectson river stability could be compounded by



human impacts that disrupt flood rhythms (Jar-dine et al. 2015).

Wetlands are particularly vulnerable to climatechanges in terms of the quality and quantity oftheir water sources. This vulnerability will bepredominantly driven by hydropattern alter-ations from changing precipitation and tempera-ture regimes. Temperature increase coulddecrease water levels and increase vegetationcover in semi-permanent wetlands (Erwin 2009).Macrophyte growth is predicted to increase withclimate warming due to higher temperatures andreductions in water tables (Jeppesen et al. 2009).

Global warming has led to longer ice-free sea-sons in lakes and rivers around the world (Mag-nuson et al. 2000). Earlier ice-off associated withwarming will lead to large shifts in lake ecologybecause many physical and chemical parametersin these lakes will change (Preston et al. 2016).Climate change will likely trigger shifts in thephenology of lakes (see Heino et al. 2009) bychanging the timing of stratification patterns andduration of ice cover; it will also likely influencethe biotic characteristics of zooplankton commu-nities (Jones and Gilbert 2016). These changescould alter diversity, abundances, and phenologyof zooplankters within lakes, and responses maydiverge across latitudes due to differences inoptimal hatching cues and life history strategies.We can potentially use information from areas oflower latitudes to predict future conditions atany current latitude.

Land use changes and reservoirsLand use changes have widespread influences

on streams, wetlands, and lakes. Streams receivemore pollutants and have altered hydrology,wetlands are drained and converted to agricul-tural or urban uses, and lakes receiving watersfrom impaired streams will be degraded as well(Dodds and Whiles 2019). Read et al. (2015)found lake water quality is heavily driven bylocal factors, especially land use within thewatershed.

Changes in land use, particularly riparianmodification, alter inputs of terrestrial vegetationinto freshwaters. Large woody debris can be asignificant allochthonous source of organic car-bon in freshwater ecosystems, and anthro-pogenic alterations to large woody debris inputshave had lasting negative impacts on these

ecosystems (Brooker 1985, Gurnell and Sweet1998). The relative amount of large woody debrisnaturally occurring in systems is a function ofthe biome in which they occur, and restoration ofthese inputs should take into account the biomecontext.Dams impact flowing waters across the globe,

greatly influencing the hydrology, channel stabil-ity, and ecosystem functioning of rivers andstreams (Cluer and Thorne 2014, Fox et al. 2016).We assessed the number of reservoirs across gra-dients of temperature, latitude, elevation, andprecipitation, focusing on the difference betweenlake and reservoir surface coverage within agiven area. Similar to natural lakes, most reser-voirs are located below 500 m above sea level(Appendix S2). More reservoirs occur in theNorthern Hemisphere, with the highest densityof reservoirs occurring around 50° N (AppendixS2). This differs from densities of natural lakes,which have higher densities between 60° and 70°N (Lehner and D€oll 2004, Verpoorter et al. 2014).The distribution of lakes across northern mid-latitudes corresponds to higher densities of reser-voirs in areas with moderate-to-high meanannual temperatures. Across gradients of precip-itation, reservoirs have high densities in rela-tively dry landscapes, with the highest density ofreservoirs occurring in areas receiving an aver-age of only 500–1000 mm of precipitation peryear (Appendix S2). We also compared the rela-tive amount of area covered by natural lakes andreservoirs across the same global gradients(Fig. 8). Reservoir surface area is only higherthan natural lake area at low elevations, thoughlake area is typically still higher. Reservoir areaexceeds lake area at relatively high levels of pre-cipitation, but lake area is typically still higher,and there are no clear spikes in reservoir/lakearea across temperature and latitudinal gradi-ents. Lake area exceeds reservoir area followingsimilar patterns of global lake surface area, withlakes exceeding reservoir area in northern mid-latitudes, and just below the equator.

Water abstractionHighest levels of groundwater abstraction

occur at midlatitudes (Wada et al. 2010).Although these latitudes generally have moresurface water, large drier areas such as the GreatPlains of the United States have been converted



to crop production based on groundwaterabstraction. These drier areas, such as northeastPakistan, northwest India, northeast China, thecentral and western USA, Yemen, and southeastSpain, have high groundwater depletion rates(Wada et al. 2010), and surface waters are disap-pearing from these areas. Where data are avail-able, threats to human water security followsimilar trends across latitudinal and precipitationgradients, though high human population densi-ties in areas with relatively high precipitation canstill face significant water scarcity threats(V€or€osmarty et al. 2010). Global water abstrac-tion, paired with global climate change and habi-tat loss, will likely have a profound impact onfreshwater systems, especially regarding nutrientdynamics and concentrations, water quality,salinity, and trophic structure (Jeppesen et al.2015).

Aquatic pollutionHeavy metals, pharmaceuticals, microplastics,

acid precipitation, nutrients, and other pollutantsare common impacts on freshwater habitats; onecan detect aquatic pollutants such as traceorganic compounds in the most remote areas ofthe Arctic and Antarctic. The impact of volatilepersistent organochlorine compounds increaseswith increasing latitude and altitude due to con-densation (Wania and Mackay 1993). The con-centrations of many pollutants may varypredictably or unpredictably across gradients infreshwater biomes. A study of mercury andmethylmercury concentrations in fish tissue fromacross the western USA and Canada found noconsistent latitudinal geospatial patterns at thecontinental scale (Eagles-Smith et al. 2016). How-ever, their results suggested a modest correlationbetween methylmercury concentration and

Fig. 8. The difference between reservoir areas relative to lake area within 1000-km2 areas as an index of rela-tive human impact on water surface area. Negative values indicate larger area of lakes, while positive valuesindicate larger area of reservoirs. Reservoir and lake area data were obtained from the Global Lakes andWetlands Database (Lehner and D€oll 2004) and climate data from Fick and Higmans (2017).



precipitation levels. For African fishes, latitude,trophic level, and mass were all positive predic-tors of mercury (Hg) concentration (Hanna et al.2014). In a global review of Hg biomagnificationin aquatic food webs, Lavoie et al. (2013)reported a positive relationship between Hgtrophic magnification and latitude; they hypoth-esized that the trend reflected decreasing temper-atures from south to north, which would in turndrive lower biomagnification rates due to differ-ences in Hg assimilation and excretion associatedwith temperature. Overall, continental-scaleanalyses of freshwater pollution patterns arelacking and warranted in a time of unprece-dented anthropogenic change.

CONCLUSION

Our conceptual framework is not so much asingle hypothesis as a logical way to characterizepatterns in freshwaters at scales relevant to evo-lutionary processes that gave rise to freshwaterbiodiversity, to the scales at which regulatoryunits interact with the environment, and the cur-rent scope of anthropogenic impacts on ecosys-tems that provide vital ecosystem services(Dodds et al. 2013). Many of our observations inthis paper are obvious (e.g., freshwaters in drierareas will be sparser and prone to drying), but abroad framework including the ramifications ofbroad-scale gradients from basic geomorphologyand hydrology to ecosystem and communityresponses has not been approached, as far as weknow. While much has been published on somecharacteristics of freshwaters across continentalscales, there is much more to learn. For example,far less is known about tropical waters than tem-perate ones with respect to diversity and ecosys-tem rates. Another key issue is mappingbiodiversity at the appropriate scale (McMana-may et al. 2018). Many biotic characteristics (e.g.,food chain length, food web connectivity, rates ofproduction and decomposition, endemism, andomnivory) have been poorly studied acrossbroad spatial scales. Characterization of broadpatterns requires substantially more data to findpatterns in inherently noisy ecological data. Asresearchers in developing countries expand theirscope of study, and government agencies becomemore comprehensive in their monitoring pro-grams while making their data freely available,

we will be able to further test predictions madeunder the freshwater biome gradient framework.As we refine this framework, it will allow betterpredictive ability at macrosystem scales based onthe resulting expected ecological baselines andhow they will be altered in a changing world.

ACKNOWLEDGMENTS

We thank Patricia Chambers, Michael Pace, PriscillaMoley, and Henry Camarillo for developing the con-cepts presented in this paper. Thanks to two anony-mous reviewers who greatly improved the paper. Wealso acknowledge the World Wildlife Fund Inc. andThe Nature Conservancy for sharing data. US NSFMacrosystems grant no. 1442595 supported WKD,AES, and FT. We also received support from the NSFKonza LTER award 1440484 and the NSF Kansas EPS-CoR EPS-0903806. This is publication #19-326-J fromthe Kansas Agricultural Experiment Station.

LITERATURE CITED

Abell, R., et al. 2008. Freshwater ecoregions of theworld: a new map of biogeographic units forfreshwater biodiversity conservation. BioScience58:403–414.

Allen, D. C. 2016. Microclimate modification by ripar-ian vegetation affects the structure and resourcelimitation of arthropod communities. Ecosphere 7:e01200.

Antoniades, D., M. S. V. Douglas, N. Michelutti, and J.P. Smol. 2014. Determining diatom ecotones andtheir relationship to terrestrial ecoregion designa-tions in the central Canadian Arctic Islands. Jour-nal of Phycology 50:610–623.

Balian, E. V., H. Segers, C. L�ev�eque, and K. Martens.2008. The freshwater animal diversity assessment:an overview of the results. Hydrobiologia 595:627–637.

Barry, R. G., and R. J. Chorley. 2003. Atmosphere,weather and climate. Eighth edition. Routledge,London, UK.

Batzer, D. P., and A. Ruh�ı. 2013. Is there a core set oforganisms that structure macroinvertebrate assem-blages in freshwater wetlands? Freshwater Biology58:1647–1659.

Batzer, D. P., and S. A. Wissinger. 1996. Ecology ofinsect communities in nontidal wetlands. AnnualReview of Entomology 41:75–100.

Beaver, J. R., C. E. Tausz, T. R. Renicker, G. C. Holdren,D. M. Hosler, E. E. Manis, K. C. Scotese, C. E. Tea-cher, B. T. Vitanye, and R. M. Davidson. 2014. Thelate summer crustacean zooplankton in western



U.S.A. reservoirs reflects ecoregion, temperature,and latitude. Freshwater Biology 59:1173–1186.

Beckmann, M., T. V�aclav�ık, A. M. Manceur, L. �Sprtov�a,H. von Wehrden, E. Welk, and A. F. Cord. 2014.glUV: a global UV-B radiation dataset for macroe-cological studies. Methods in Ecology and Evolu-tion 5:372–383.

Beklioglu, M., S. Romo, I. Kagalou, X. Quintana, andE. B�ecares. 2007. State of the art in the functioningof shallow Mediterranean lakes: workshop conclu-sions. Hydrobiologia 584:317–326.

Benke, A. C., I. Chaubey, G. M. Ward, and E. L. Dunn.2000. Flood pulse dynamics of an unregulated riverfloodplain in the southeastern US coastal plain.Ecology 81:2730–2741.

Bernhardt, E. S., et al. 2018. The metabolic regimes offlowing waters. Limnology and Oceanography 63:S99–S118.

Bernot, M. J., et al. 2010. Interregional comparison ofland-use effects on stream metabolism. FreshwaterBiology 55:1874–1890.

Binzer, A., C. Guill, B. C. Rall, and U. Brose. 2015.Interactive effects of warming, eutrophication andsize structure: impacts on biodiversity and food-web structure. Global Change Biology 22:220–227.

Blenckner, T., et al. 2007. Large-scale climatic signa-tures in lakes across Europe: a meta-analysis. Glo-bal Change Biology 13:1314–1326.

Boix, D., and D. P. Batzer. 2016. Invertebrate assem-blages and their ecological controls across theworld’s freshwater wetlands. Pages 601–639 in D.Batzer and D. Boix, editors. Invertebrates in fresh-water wetlands: an international perspective ontheir ecology. Springer, New York, New York,USA.

Boulton, A. J., L. Boyero, A. P. Covich, M. Dobson, S.Lake, and R. Pearson. 2008. Are tropical streamsecologically different from temperate streams?Pages 257–284 in D. Dudgeon, editor. Tropicalstream ecology. Elsevier, New York, New York,USA.

Boyero, L., R. G. Pearson, C. Hui, M. O. Gessner, J.P�erez, M. A. Alexandrou, M. A. Grac�a, B. J. Cardi-nale, R. J. Albari~no, and M. Arunachalam. 2016.Biotic and abiotic variables influencing plant litterbreakdown in streams: a global study. Proceedingsof the Royal Society B 283:20152664.

Brinson, M. M. 1993. Changes in the functioning ofwetlands along environmental gradients. Wetlands13:65–74.

Brinson, M. M., A. E. Lugo, and S. Brown. 1981. Pri-mary productivity, decomposition and consumeractivity in freshwater wetlands. Annual Review ofEcological Systems 12:123–161.

Brooker, M. P. 1985. The impact of river channeliza-tion: IV the ecological effects of channelization.Geographic Journal 151:63–69.

Brown, H. P. 1981. A distributional survey of the worldgenera of aquatic dyropoid beetles (Coleoptera:Dryopidae, Elmidae, and Psephenidae sens.lat.).Pan-Pacific Entomologist 57:133–148.

Buckley, L. B., and W. Jetz. 2007. Environmental andhistorical constraints on global patterns of amphib-ian richness. Proceedings of the Royal Society B274:1167–1173.

Buckton, S. T., and S. J. Ormerod. 2002. Global patternsof diversity among the specialist birds of riverinelandscapes. Freshwater Biology 47:695–709.

Buhlmann, K. A., T. S. B. Akre, J. B. Iverson, D. Karap-atakis, R. A. Mittermeier, A. Georges, A. G. J. Rho-din, P. P. van Dijk, and J. W. Gibbons. 2009. Aglobal analysis of tortoise and freshwater turtle dis-tributions with identification of priority conserva-tion areas. Chelonian Conservation and Biology8:116–149.

Bunn, S. E., and A. H. Arthington. 2002. Basic princi-ples and ecological consequences of altered flowregimes for aquatic biodiversity. EnvironmentalManagement 30:492–507.

Chambers, P. A., P. Lacoul, K. J. Murphy, and S. M.Thomaz. 2008. Global diversity of aquatic macro-phytes in freshwater. Hydrobiologia 595:9–26.

Church, M. 2006. Bed material transport and the mor-phology of alluvial river channels. Annual Reviewof Earth and Planetary Sciences 34:325–354.

Cluer, B., and C. Thorne. 2014. A stream evolutionmodel integrating habitat and ecosystem benefits.River Research and Applications 30:135–154.

Collantes, M. B., J. Anchorena, S. Stoffella, C. Escart�ın,and R. Rauber. 2009. Wetlands of the MagellanicSteppe (Tierra del Fuego, Argentina). Folia Geob-otanica 44:227–245.

Collen, B., F. Whitton, E. E. Dyer, J. E. M. Baillie, N.Cumberlidge, W. R. T. Darwall, C. Pollock, N. I.Richman, A. Soulsby, and M. B€ohm. 2014. Globalpatterns of freshwater species diversity, threat andendemism. Global Ecology and Biogeography23:40–51.

Dearing, J. A., and R. T. Jones. 2003. Coupling tempo-ral and spatial dimensions of global sediment fluxthrough lake and marine sediment records. Globaland Planetary Change 39:147–168.

Dodds, W. K., and J. J. Cole. 2007. Expanding the con-cept of trophic state in aquatic ecosystems: it’s notjust the autotrophs. Aquatic Sciences 69:427–439.

Dodds, W. K., K. B. Gido, M. R. Whiles, M. D. Daniels,and B. P. Grudzinski. 2015. The stream biome gra-dient concept: factors controlling lotic systems



across broad biogeographic scales. FreshwaterScience 34:1–19.

Dodds, W. K., J. S. Perkin, and J. E. Gerken. 2013.Human impact on freshwater ecosystem services: aglobal perspective. Environmental Science & Tech-nology 47:9061–9068.

Dodds, W., and V. H. Smith. 2016. Nitrogen, phospho-rus, and eutrophication in streams. Inland Waters6:155–164.

Dodds, W. K., and M. R. Whiles. 2004. Quality andquantity of suspended particles in rivers: conti-nent-scale patterns in the United States. Environ-mental Management 33:355–367.

Dodds, W. K., and M. R. Whiles. 2019. Freshwater ecol-ogy: concepts and environmental applications oflimnology. Third edition. Academic Press, Cam-bridge, Massachusetts, USA.

Dodson, S. 1992. Predicting crustacean zooplanktonspecies richness. Limnology and Oceanography37:848–856.

Downing, J. A., J. J. Cole, C. M. Duarte, J. J. Middel-burg, J. M. Melack, Y. T. Prairie, P. Kortelainen, R.G. Striegl, W. H. McDowell, and L. J. Tranvik. 2012.Global abundance and size distribution of streamsand rivers. Inland Waters 2:229–236.

Downing, J. A., M. McClain, R. Twilley, J. Melack, J.Elser, N. Rabalais, W. Lewis, R. Turner, J. Corredor,and D. Soto. 1999. The impact of accelerating land-use change on the N-cycle of tropical aquaticecosystems: current conditions and projectedchanges. Biogeochemistry 46:109–148.

Eagles-Smith, C. A., et al. 2016. Spatial and temporalpatterns of mercury concentration in freshwaterfish across the Western United States and Canada.Science of the Total Environment 568:1171–1184.

Elser, J. J., M. Bracken, E. E. Cleland, D. S. Gruner, W.S. Harpole, H. Hillebrand, J. T. Ngai, E. W. Sea-bloom, J. B. Shurin, and J. E. Smith. 2007. Globalanalysis of nitrogen and phosphorus limitation ofprimary producers in freshwater, marine and ter-restrial ecosystems. Ecology Letters 10:1135–1142.

Erwin, K. L. 2009. Wetlands and global climate change:the role of wetland restoration in a changingworld. Wetlands Ecology Management 17:71–84.

Fan, Y., H. Li, and G. Miguez-Macho. 2013. Global pat-terns of groundwater table depth. Science 339:940–943.

Fergus, C. E., J.-F. Lapierre, S. K. Oliver, N. K. Skaff, K.S. Cheruvelil, K. Webster, C. Scott, and P. Soranno.2017. The freshwater landscape: lake, wetland, andstream abundance and connectivity at macroscales.Ecosphere 8:e01911.

Fern�andez, N., J. Aguilar, C. A. Pi~na-Garc�ıa, and C.Gershenson. 2017. Complexity of lakes in a latitu-dinal gradient. Ecological Complexity 31:1–20.

Fick, S. E., and R. J. Higmans. 2017. WorldClim 2: new1-km spatial resolution climate surfaces for globalland areas. International Journal of Climatology37:4302–4315.

Follstad Shah, J. J., et al. 2017. Global synthesis of thetemperature sensitivity of leaf litter breakdown instream and rivers. Global Change Biology 23:3064–3075.

Fox, G. A., A. Sheshukov, R. Cruse, R. L. Kolar, L.Guertault, K. R. Gesch, and R. C. Dutnell. 2016.Reservoir sedimentation and upstream sedimentsources: perspectives and future research needs onstreambank and gully erosion. EnvironmentalManagement 57:945–955.

Frasson, R. P. D. M., T. M. Pavelsky, M. A. Fonstad, M.T. Durand, G. H. Allen, G. Schumann, C. Lion, R.E. Beighley, and X. Yang. 2019. Global relationshipsbetween river width, slope, catchment area, mean-der wavelength, sinuosity, and discharge. Geo-physical Research Letters 46:3252–3262.

F€ureder, L. 2007. Life at the edge: habitat conditionand bottom fauna of alpine running waters. Inter-national Review of Hydrobiology 92:491–513.

Garris, H. W., R. Mitchell, L. H. Fraser, and L. R. Bar-rett. 2015. Forecasting climate change impacts onthe distribution of wetland habitat in the Midwest-ern United States. Global Change Biology 21:766–776.

Gillooly, J. F., and S. Dodson. 2000. Latitudinal pat-terns in the size distribution and seasonal dynam-ics of new world freshwater cladocerans.Limnology and Oceanography 45:22–30.

Gonzalez-Bergonzoni, I., M. Meerhoff, T. A. Davidson,F. Teixeira-de Mello, A. Baattrup-Pedersen, and E.Jeppesen. 2012. Meta-analysis shows a consistentand strong latitudinal pattern in fish omnivoryacross ecosystems. Ecosystems 15:492–503.

Goudie, A. S. 2006. Global warming and fluvial geo-morphology. Geomorphology 79:384–394.

Grimaldo, T. J., M. T. O’Hare, M. P. Kennedy, T. A.Davidson, J. Bonilla-Barbosa, B. Santamar�ıa-Ara�uz,L. Gettys, S. Varandas Martins, S. M. Thomaz, andK. J. Murphy. 2017. Environmental drivers of fresh-water macrophyte diversity and community com-position in calcareous warm-water rivers ofAmerica and Africa. Freshwater Biology 62:1511–1527.

Gumbricht, T., R. M. Roman-Cuesta, L. Verchot, M.Herold, F. Wittmann, E. Householder, N. Herold,and D. Murdiyarso. 2017. An expert system modelfor mapping tropical wetlands and peatlandsreveals South America as the largest contributor.Global Change Biology 23:3581–3599.

Gurnell, A. M., and R. Sweet. 1998. The distribution oflarge woody debris accumulations and pools in



relation to woodland stream management in asmall, low-gradient stream. Earth Surface Pro-cesses and Landforms 23:1101–1121.

Gustafson, E. J. 1998. Quantifying landscape spatialpattern: What is the state of the art? Ecosystems1:143–156.

Gyllstr€om, M., et al. 2005. The role of climate in shap-ing zooplankton communities of shallow lakes.Limnology and Oceanography 50:2008–2021.

H€ader, D. P., C. E. Williamson, S. �A. W€angberg, M.Rautio, K. C. Rose, K. Gao, E. W. Helbling, R. P.Sinhah, and R. Worresti. 2015. Effects of UV radia-tion on aquatic ecosystems and interactions withother environmental factors. Photochemical & Pho-tobiological Sciences 14:108–126.

Hamilton, S. K. 2002. Hydrological controls of ecologi-cal structure and function in the Pantanal wetland(Brazil). The ecohydrology of South American Riv-ers and Wetlands. International Association ofHydrological Sciences, Special Publication 6:133–158.

Hancock, M. A., and S. E. Bunn. 1997. Populationdynamics and life history of Paratya australiensisKemp, 1917 (Decapoda: Atyidae) in upland rain-forest streams, south-east Queensland. Marine andFreshwater Research 48:361–369.

Hanna, D. E. L., D. G. Buck, and L. J. Chapman. 2014.Effects of habitat on mercury concentrations in fish:a case study of Nile perch (Lates niloticus) in LakeNabugabo, Uganda. Ecotoxicology 25:178–191.

Havens, K. E., et al. 2014. Temperature effects on bodysize of freshwater crustacean zooplankton fromGreenland to the tropics. Hydrobiologia 743:27–35.

Heffernan, J. B., P. A. Soranno, M. J. Angilletta, L. B.Buckley, D. S. Gruner, T. H. Keitt, J. R. Kellner, J. S.Kominoski, A. V. Rocha, and J. Xiao. 2014.Macrosystems ecology: understanding ecologicalpatterns and processes at continental scales. Fron-tiers in Ecology and the Environment 12:5–14.

Heino, J., R. Virkkala, and H. Toivonen. 2009. Climatechange and freshwater biodiversity: detected pat-terns, future trends and adaptations in northernregions. Biological Reviews 84:39–54.

Holdridge, L. R. 1947. Determination of world plantformations from simple climatic data. Science105:367–368.

Huryn, A. D., and J. B. Wallace. 2000. Life history andproduction of stream insects. Annual Review ofEntomology 45:83–110.

Hutchinson, G. E. 1957. A treatise on limnology, vol I.Geography, physics and chemistry. Wiley, NewYork, New York, USA.

Irons, J. G., M. W. Oswood, R. J. Stout, and C. M. Prin-gle. 1994. Latitudinal patterns in leaf litter

breakdown: Is temperature really important?Freshwater Biology 32:401–411.

J€ach, M. A., and M. Balke. 2008. Global diversity ofwater beetles (Coleoptera) in freshwater. Hydrobi-ologia 595:419–442.

Jackson, C. R., J. A. Thompson, and R. K. Kolka. 2014.Wetland soils, hydrology, and geomorphology.Pages 23–60 in D. P. Batzer and R. R. Sharitz, edi-tors. Ecology of freshwater and estuarine wetlands.UC Press, Berkeley, California, USA.

Jardine, T. D., et al. 2015. Does flood rhythm driveecosystem responses in tropical riverscapes? Ecol-ogy 96:684–692.

Jeppesen, E., M. Meerhoff, B. A. Jacobsen, R. S.Hansen, M. Søndergaard, J. P. Jensen, T. L. Lau-ridsen, N. Mazzeo, and C. W. C. Branco. 2007.Restoration of shallow lakes by nutrient controland biomanipulation – the successful strategy var-ies with lake size and climate. Hydrobiologia581:269–285.

Jeppesen, E., et al. 2005. Lake responses to reducednutrient loading – an analysis of contemporarylong-term data from 35 case studies. FreshwaterBiology 50:1747–1771.

Jeppesen, E., et al. 2009. Climate change effects on run-off, catchment phosphorus loading and lake eco-logical state, and potential adaptations. Journal ofEnvironmental Quality 38:1930–1941.

Jeppesen, E., et al. 2015. Ecological impacts of globalwarming and water abstraction on lakes and reser-voirs due to changes in water level and relatedchanges in salinity. Hydrobiologia 750:201–227.

Jim�enez-Alfaro, B., et al. 2014. Biogeographic patternsof base-rich fen vegetation across Europe. AppliedVegetation Science 17:367–380.

Jobb�agy, E. G., and R. B. Jackson. 2000. The verticaldistribution of soil organic carbon and its relationto climate and vegetation. Ecological Applications10:423–436.

Jones, N. T., and B. Gilbert. 2016. Changing climatecues differentially alter zooplankton dormancydynamics across latitudes. Journal of Animal Ecol-ogy 85:559–569.

Keddy, P. A. 2000. Wetland ecology: principles andconservation. Cambridge University Press, Cam-bridge, UK.

Kernan, M., M. Ventura, P. Bitu�s�ık, A. Brancelj, G.Clarke, G. Velle, G. G. Raddum, E. Stuchl�ık, and J.Catalan. 2009. Regionalisation of remote Europeanmountain lake ecosystems according to their biota:environmental versus geographical patterns. Fresh-water Biology 54:2470–2493.

Kirk, J. T. 1994. Light and photosynthesis in aquaticecosystems. Cambridge University Press.



Koopman, M., and B. Carstens. 2011. The microbialphyllogeography of the carnivorous plant Sarrace-nia alata. Microbial Ecology 61:750–758.

Kozak, K. H., and J. J. Wiens. 2012. Phylogeny, ecology,and the origins of climate – richness relationships.Ecology 93:167–181.

Kraemer, B. M., et al. 2017. Global patterns in lakeecosystem responses to warming based on the tem-perature dependence of metabolism. GlobalChange Biology 23:1881–1890.

Kroes, D. E., and M. M. Brinson. 2004. Occurrence ofriverine wetlands on floodplains along a climaticgradient. Wetlands 24:167–177.

Lane, E. W. 1955. The importance of fluvial morphol-ogy in hydraulic engineering. American Society ofCivil Engineering 81:1–17.

Lavoie, R. A., T. D. Jardine, M. M. Chumchal, K. A.Kidd, and L. M. Campbell. 2013. Biomagnificationof mercury in aquatic food webs: a worldwidemeta-analysis. Environmental Science and Technol-ogy 47:13385–13394.

Lehner, B., and P. D€oll. 2004. Development and valida-tion of a global database of lakes, reservoirs andwetlands. Journal of Hydrology 296:1–22.

Lehner, B., K. Verdin, and A. Jarvis. 2008. New globalhydrography derived from spaceborne elevationdata. Eos Transactions 89:93–94.

Leopold, L. B. 1962. Rivers. American Scientist 50:511–537.

Leopold, L. B., and M. G. Wolman. 1957. River channelpatterns – braided, meandering and straight. Pro-fessional Paper No. 282B. United States GeologicalSurvey, Washington, D. C., USA.

LeRoy, C. J., and J. C. Marks. 2006. Litter quality,stream characteristics and litter diversity influencedecomposition rates and macroinvertebrates.Freshwater Biology 51:605–617.

Levins, R. 1966. The strategy of model building inpopulation biology. American Scientist 54:421–431.

Lewis Jr., W. M.. 1996. Tropical lakes: how latitudemakes a difference. Perspectives in Tropical Lim-nology 43–64.

Lewis Jr., W. M.. 2002. Causes for the high frequencyof nitrogen limitation in tropical lakes. Interna-tionale Vereinigung f€ur Theoretische und Ange-wandte Limnologie: Verhandlungen 28:210–213.

Lewis Jr., W. M.. 2011. Global primary production oflakes: 19th Baldi Memorial Lecture. Inland Waters1:1–28.

Lovell, J. T., and E. Menges. 2013. Dominant speciesresponses to drought in seasonal wetlands: evi-dence from reciprocal transplants across a moisturegradient. Journal of the Torrey Botanical Society140:157–169.

Lytle, D. A., and N. L. Poff. 2004. Adaptation to natu-ral flow regimes. Trends in Ecology and Evolution19:94–100.

Magnuson, J. J., et al. 2000. Historical trends in lakeand river ice cover in the Northern Hemisphere.Science 289:1743–1746.

Marino, N. D. A. C., G. Q. Romero, and V. F. Farjalla.2018. Geographical and experimental contextsmodulate the effect of warming on top-down con-trol: a meta-analysis. Ecology Letters 21:455–466.

Markovic, D., S. F. Carrizo, O. K€archer, A. Walz, and J.N. W. David. 2017. Vulnerability of Europeanfreshwater catchments to climate change. GlobalChange Biology 23:3567–3580.

McCluney, K. E., N. L. Poff, M. A. Palmer, J. H. Thorp,G. C. Poole, B. S. Williams, M. R. Williams, and J. S.Baron. 2014. Riverine macrosystems ecology: sensi-tivity, resistance, and resilience of whole riverbasins with human alterations. Frontiers in Ecol-ogy and the Environment 12:48–58.

McManamay, R. A., N. A. Griffiths, C. R. DeRolph,and B. M. Pracheil. 2018. A synopsis of global map-ping of freshwater habitats and biodiversity: impli-cations for conservation. Pages 57–87 in L.Hufnagel, editor. Pure and applied biogeography.IntechOpen, London, UK.

Meerhoff, M., J. M. Clemente, F. Teixeira de Mello, C.Iglesias, A. R. Pedersen, and E. Jeppesen. 2007a.Can warm climate-related structure of littoralpredator assemblies weaken the clear water statein shallow lakes? Global Change Biology 13:1888–1897.

Meerhoff, M., C. Iglesias, F. Teixeira de Mello, J. M.Clemente, E. Jensen, T. L. Lauridsen, and E. Jeppe-sen. 2007b. Effects of habitat complexity on com-munity structure and predator avoidance behaviourof littoral zooplankton in temperate versus sub-tropical shallow lakes. Freshwater Biology 52:1009–1021.

Middleton, B. A., and K. L. McKee. 2004. Use of a lati-tudinal gradient in bald cypress (Taxodium dis-tichum) production to examine physiologicalcontrols of biotic boundaries and potentialresponses to environmental change. Global Ecol-ogy and Biogeography 13:247–258.

Millar, D. J., D. J. Cooper, K. A. Dwire, R. M. Hubbard,and J. Fischer. 2017. Mountain peatlands rangefrom CO2 sinks at high elevations to sources at lowelevations: implications for a changing climate.Ecosystems 20:416–432.

Mims, M. C., and J. D. Olden. 2012. Life history theorypredicts streamflow effects on fish assemblageresponse to hydrologic regimes. Ecology 93:35–45.

Mims, M. C., J. D. Olden, Z. R. Shattuck, and N. L.Poff. 2010. Life history trait diversity of native



freshwater fishes in North America. Ecology ofFreshwater Fish 19:390–400.

Mitsch, W. J., J. G. Gosselink, L. Zhang, and C. J.Anderson. 2009. Wetland ecosystems. Wiley,Chichester, UK.

Mooji, W. M., et al. 2005. The impact of climate changeon lakes in the Netherlands: a review. AquaticEcology 39:381–400.

Mountain Research Initiative. 2015. Elevation-depen-dent warming in mountain regions of the world.Nature Climate Change 5:424.

Mulholland, P. J., et al. 2001. Inter-biome comparisonof factors controlling stream metabolism. Freshwa-ter Biology 46:1503–1517.

Mulholland, P. J., et al. 2009. Nitrate removal in streamecosystems measured by 15N addition experi-ments: denitrification. Limnology and Oceanogra-phy 54:666–680.

Navas, C. A. 2002. Herpetological diversity alongAndean elevational gradients: Links with phys-iological ecology and evolutionary physiology.Comparative Biochemistry and Physiology Part A:Molecular and Integrative Physiology 133:469–485.

Newton, R. J., S. E. Jones, A. Eiler, K. D. McMahon,and S. Bertilsson. 2011. A guide to the natural his-tory of freshwater lake bacteria. Microbiology andMolecular Biology Reviews 75:14–49.

Norman, B. C., et al. 2017. Drivers of nitrogen transferin stream food webs across continents. Ecology98:3044–3055.

O’Hare, M. T., I. D. Gunn, D. S. Chapman, B. J. Dudley,and B. V. Purse. 2012. Impacts of space, local envi-ronment and habitat connectivity on macrophytecommunities in conservation lakes. Diversity &Distributions 18:603–614.

Olden, J. D., and M. J. Kennard. 2010. Intercontinentalcomparison of fish life history strategies along agradient of hydrologic variability. Pages 83–107 inK. B. Gido and D. A. Jackson, editors. Communityecology of stream fishes: concepts, approaches, andtechniques. American Fisheries Society, Bethesda,Maryland, USA.

Osland, M. J., N. Enwright, and C. L. Stagg. 2014.Freshwater availability and coastal wetland foun-dation species: ecological transitions along a rain-fall gradient. Ecology 95:2789–2802.

Pearson, R. G., and L. Boyero. 2009. Gradients inregional diversity of freshwater taxa. Journal ofthe North American Benthological Society 28:504–514.

Peregon, A., S. Maksyutov, N. P. Kosykh, and N. P.Mironycheva-Tokareva. 2008. Map-based inven-tory of wetland biomass and net primary produc-tion in western Siberia. Journal of GeophysicalResearch 113:G01007.

Pienitz, R., and F. Vincent. 2000. Effect of climatechange relative to ozone depletion on UV exposurein subarctic lakes. Nature 404:484–487.

Poff, N. L. 1996. A hydrogeography of unregulatedstreams in the United States and an examination ofscale-dependence in some hydrological descrip-tors. Freshwater Biology 36:71–91.

Polhemus, J. T., and D. A. Polhemus. 2008. Globaldiversity of true bugs (Heteroptera; Insecta) infreshwater. Hydrobiologia 595:379–391.

Post, D. M., M. L. Pace, and N. G. Hairston Jr. 2000.Ecosystem size determines food-chain length inlakes. Nature 405:1047.

Preston, D. L., N. Caine, D. M. McKnight, M. W. Wil-liams, K. Hell, M. P. Miller, S. J. Hart, and P. T. J.Johnson. 2016. Climate regulates alpine lake icecover phenology and aquatic ecosystem structure.Geophysical Research Letters 43:5353–5360.

Prigent, C., F. Papa, F. Aires, W. B. Rossow, and E. Mat-thews. 2007. Global inundation dynamics inferredfrom multiple satellite observations, 1993–2000.Journal of Geophysical Research 112:D12107.

Prowse, T. D., and J. M. Culp. 2003. Ice breakup: aneglected factor in river ecology. Canadian Journalof Civil Engineering 30:128–144.

Read, E. K., et al. 2015. The importance of lake-specificcharacteristics for water quality across the conti-nental United States. Ecological Applications25:943–955.

Rosgen, D. L. 1996. Applied river morphology. Wild-land Hydrology, Pagosa Springs, Colorado, USA.

Ruh�ı, A., D. Boix, S. Gasc�on, J. Sala, and D. P. Batzer.2013. Functional and phylogenetic relatedness intemporary wetland invertebrates: current macroe-cological patterns and implications for future cli-mate change scenarios. PLoS ONE 8:e81739.

Schindler, D. W. 1977. Evolution of phosphorus limita-tion in lakes. Science 195:260–262.

Schumm, S. A. 1977. The fluvial system. Volume 338.Wiley, New York, New York, USA.

Smith, R. A., R. B. Alexander, and G. E. Schwarz. 2003.Natural background concentrations of nutrients instreams and rivers of the conterminous UnitedStates. Environmental Science and Technology37:2039–3047.

Song, C., W. K. Dodds, J. R€uegg, A. Argerich, C. L.Baker, W. B. Bowden, M. M. Douglas, K. J. Farrell,M. B. Flinn, and E. A. Garcia. 2018. Continental-scale decrease in net primary productivity instreams due to climate warming. Nature Geo-science 11:415–420.

Staehr, P. A., L. S. Brighenti, M. Honti, J. Christensen,and K. C. Rose. 2016. Inland Waters 6:593–607.

Steele, M. K., and J. B. Heffernan. 2017. Land use andtopography bend and break fractal rules of water



body size-distributions. Limnology and Oceanog-raphy Letters 2:71–80.

Tank, J. L., et al. 2018. Partitioning assimilatory nitro-gen uptake in streams: an analysis of stable isotopetracer additions across continents. EcologicalMonographs 88:120–138.

Teittinen, A., J. Wang, S. Str€omg�ard, and J. Soininen.2017. Local and geographical factors jointly driveelevational patterns in three microbial groupsacross subarctic ponds. Global Ecology and Bio-geography 26:973–982.

Thornman, M. N., and S. E. Bayley. 1997. Above-ground plant production and nutrient content ofthe vegetation in six peatlands in Alberta, Canada.Plant Ecology 131:1–16.

Tiegs, S. D., et al. 2019. Global patterns and drivers ofecosystem functioning in rivers and riparian zones.Science Advances 5:eaav0486.

Tooth, S., and T. S. McCarthy. 2007. Wetlands in dry-lands: geomorphological and sedimentologicalcharacteristics, with emphasis on examples fromsouthern Africa. Progress in Physical Geography31:3–41.

Tooth, S., and H. Viles. 2014. 10 reasons why geomor-phology is important. British Society for Geomor-phology. https://geomorphology.org.uk/sites/default/files/10_reasons_full.pdf

Turetsky, M. R., et al. 2014. A synthesis of methaneemissions from 71 northern, temperate, and sub-tropical wetlands. Global Change Biology 20:2183–2197.

Turner, R. E., N. N. Rabalais, D. Justic, and Q.Dortch. 2003. Global patterns of dissolved N,P and Si in large rivers. Biogeochemistry 64:297–317.

Ullrich, C., B. Lehner, B. Wickel, and T. Sickley. 2010.World of rivers (supplement to National Geo-graphic). National Geographic Society, Washing-ton, D.C., USA. https://www.nationalgeographic.org/hires/world-rivers/

V€ain€ol€a, R., J. D. S. Witt, M. Grabowski, J. H. Bradbury,K. Jazdzewski, and B. Sket. 2007. Global diversityof amphipods (Amphipoda; Crustacea) in freshwa-ter. Pages 241–255 in E. V. Balian, C. L�eveque, H.Segers, and K. Martens, editors. Freshwater animaldiversity assessment. Developments in hydrobiol-ogy. Volume 198. Springer, Dordrecht, The Nether-lands.

Vander Zanden, M. J., and W. W. Fetzer. 2007. Globalpatterns of aquatic food chain length. Oikos116:1378–1388.

Verpoorter, C., T. Kutser, D. A. Seekell, and L. J. Tran-vik. 2014. A global inventory of lakes based onhigh-resolution satellite imagery. GeophysicalResearch Letters 41:6396–6402.

Vinson, M. R., and C. P. Hawkins. 2003. Broad-scalegeographical patterns in local stream insect generarichness. Ecography 26:751–767.

Vollenweider, R. A. 1968. Scientific fundamentals oflake and stream eutrophication, with particular ref-erence to phosphorus and nitrogen as eutrophica-tion factors. Technical Report DAS/DSI/68.27.OECD, Paris, France.

von Humboldt, A., and A. Bonpland. 1807. Essay onthe geography of plants. Pages 49–145 in S. T. Jack-son, editor. 2009 Essay on the geography of plants.University of Chicago Press, Chicago, Illinois,USA.

V€or€osmarty, C. J., et al. 2010. Global threats to humanwater security and river biodiversity. Nature467:555–561.

Wada, Y., L. P. H. Van Beek, C. M. Van Kempen, J. W.T. M. Reckman, S. Vasak, and M. F. P. Bierkens.2010. Global depletion of groundwater resources.Geophysical Research Letters 37:1–5.

Waitt, R. B. 1985. Case for periodic, colossal j€okulh-laups from Pleistocene glacial Lake Missoula. GSABulletin 96:1271–1286.

Wania, F., and D. Mackay. 1993. Global fractionationand cold condensation of low volatility organochlo-rine compounds in polar regions. Ambio 22:10–18.

Welcomme, R. L. 1976. Some general and theoreticalconsiderations on the fish yield of African rivers.Journal of Fish Biology 8:351–364.

Wellborn, G. A., D. K. Skelly, and E. E. Werner. 1996.Mechanisms creating community structure acrossa freshwater habitat gradient. Annual Review ofEcology and Systematics 27:337–363.

Wetzel, R. G. 2001. Limnology: lake and river ecosys-tems. Elsevier, London, UK.

Whiles, M. R., and W. K. Dodds. 2002. Relationshipsbetween stream size, suspended particles, and fil-ter-feeding macroinvertebrates in a Great Plainsdrainage network. Journal of Environmental Qual-ity 31:1589–1600.

Whittaker, R. H. 1970. Communities and ecosystems.Macmillan, New York, New York, USA.

Williamson, C. E., W. Dodds, T. K. Kratz, and M. A.Palmer. 2008. Lakes and streams as sentinels ofenvironmental change in terrestrial and atmo-spheric processes. Frontiers in Ecology and theEnvironment 6:247–254.

Winemiller, K. O. 2005. Life history strategies, popula-tion regulation, and implications for fisheries man-agement. Canadian Journal of Fisheries andAquatic Sciences 52:872–885.

Wissinger, S. A. 1999. Ecology of wetland inverte-brates: synthesis and applications for conservationand management. Pages 1043–1086 in D. P. Batzer,



https://geomorphology.org.uk/sites/default/files/10_reasons_full.pdf

https://geomorphology.org.uk/sites/default/files/10_reasons_full.pdf

https://www.nationalgeographic.org/hires/world-rivers/

https://www.nationalgeographic.org/hires/world-rivers/

R. D. Rader, and S. A. Wissinger, editors. Inverte-brates in freshwater wetlands of North America:ecology and management. John Wiley and Sons,New York, New York, USA.

Young, R. G., C. D. Matthaei, and C. R. Townsend.2004. Organic matter breakdown and ecosystemmetabolism: functional indicators for assessingriver ecosystem health. Freshwater Science 27:605–625.

Yvon-Durocher, G., J. M. Caffrey, A. Cescatti, M. Dos-sena, P. del Giorgio, J. M. Gasol, J. M. Montoya, J.Pumpanen, P. A. Staehr, and M. Trimmer. 2012.

Reconciling the temperature dependence of respi-ration across timescales and ecosystem types. Nat-ure 487:472.

Yvon-Durocher, G., J. I. Jones, M. Trimmer, G. Wood-ward, and M. Montoya Jose. 2010. Warming altersthe metabolic balance of ecosystems. PhilosophicalTransactions of the Royal Society B: BiologicalSciences 365:2117–2126.

Zhang, D., D. Hui, Y. Luo, and G. Zhou. 2008. Rates oflitter decomposition in terrestrial ecosystems: glo-bal patterns and controlling factors. Journal ofPlant Ecology 1:85–93.

SUPPORTING INFORMATION

Additional Supporting Information may be found online at: http://onlinelibrary.wiley.com/doi/10.1002/ecs2.2786/full



http://onlinelibrary.wiley.com/doi/10.1002/ecs2.2786/full

http://onlinelibrary.wiley.com/doi/10.1002/ecs2.2786/full

the freshwater biome gradient framework: predicting ......ways to delineate freshwater biomes....

Documents