* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Williamson, Craig E., Jennifer A. Brentrup, Jing Zhang, William H
Economics of global warming wikipedia , lookup
Politics of global warming wikipedia , lookup
Climate change adaptation wikipedia , lookup
Climate governance wikipedia , lookup
Climatic Research Unit documents wikipedia , lookup
Global warming wikipedia , lookup
General circulation model wikipedia , lookup
Climate engineering wikipedia , lookup
Media coverage of global warming wikipedia , lookup
Climate change in Tuvalu wikipedia , lookup
Citizens' Climate Lobby wikipedia , lookup
Climate sensitivity wikipedia , lookup
Public opinion on global warming wikipedia , lookup
Climate change and agriculture wikipedia , lookup
Scientific opinion on climate change wikipedia , lookup
Effects of global warming wikipedia , lookup
Instrumental temperature record wikipedia , lookup
Climate change feedback wikipedia , lookup
Effects of global warming on human health wikipedia , lookup
Attribution of recent climate change wikipedia , lookup
Climate change in Saskatchewan wikipedia , lookup
Climate change in the United States wikipedia , lookup
Global Energy and Water Cycle Experiment wikipedia , lookup
Climate change and poverty wikipedia , lookup
Surveys of scientists' views on climate change wikipedia , lookup
Effects of global warming on humans wikipedia , lookup
Climate change, industry and society wikipedia , lookup
Limnol. Oceanogr., 59(3), 2014, 840–850 2014, by the Association for the Sciences of Limnology and Oceanography, Inc. doi:10.4319/lo.2014.59.3.0840 E Lakes as sensors in the landscape: Optical metrics as scalable sentinel responses to climate change Craig E. Williamson,1,* Jennifer A. Brentrup,1 Jing Zhang,2 William H. Renwick,3 Bruce R. Hargreaves,4 Lesley B. Knoll,5 Erin P. Overholt,1 and Kevin C. Rose 6 1 Department of Biology, Miami University, Oxford, Ohio of Statistics, Miami University, Oxford, Ohio 3 Department of Geography, Miami University, Oxford, Ohio 4 Department of Earth and Environmental Sciences, Lehigh University, Bethlehem, Pennsylvania 5 Lacawac Sanctuary, Lake Ariel, Pennsylvania 6 Smithsonian Environmental Research Center, Edgewater, Maryland 2 Department Abstract As the lowest point in the surrounding landscape, lakes act as sensors in the landscape to provide insights into the response of both terrestrial and aquatic ecosystems to climate change. Here a novel suite of climate forcing optical indices (CFOI) from lakes across North America is found to respond to changes in air temperature, precipitation, and solar radiation at timescales ranging from a single storm event to seasonal changes to longerterm interdecadal trends with regression r2 values ranging from 0.73 to 0.89. These indices are based on two optical metrics of dissolved organic carbon (DOC) quality: DOC specific absorbance (a*320) and spectral slope (S275–295), where the ratio a*320 to S275–295 gives a composite climate forcing index. These indices of DOC quality are more responsive to climate forcing than is DOC concentration. A similar relationship between the component indices a*320 and S275–295 is observed across a wide range of lake types. A conceptual model is used to examine the similarities and differences in DOC-related mechanisms and ecological consequences due to increased temperature vs. precipitation. While both warmer and wetter conditions increase thermal stratification, these two types of climate forcing will have opposite effects on water transparency as well as many ecological consequences, including oxygen depletion, the balance between autotrophy and heterotrophy, and depth distributions of phytoplankton and zooplankton. In recent decades, trends of increasing DOC concentration have provided one of the strongest signals of change in inland aquatic ecosystems (Williamson et al. 2009a), with as much as a doubling of DOC concentrations in lakes in some regions of North America and Europe (Monteith et al. 2007). The causes of increasing DOC are a subject of active debate. While earlier work (Monteith et al. 2007) seemed to be converging on a central role for recovery from anthropogenic acidification, more recent work (Zhang et al. 2010; Couture et al. 2012) suggests that climate plays an equally or perhaps more important role, even in systems in which recovery from acidification has been observed. Furthermore, climate change effects on terrestrial ecosystems are predicted to increase future DOC concentrations in boreal lakes by as much as 65% (Larsen et al. 2011). A clearer understanding of these increasing DOC concentrations is critical because DOC is one of the most fundamental regulators of aquatic ecosystem structure and function (Williamson et al. 1999; Couture et al. 2012). In most inland waters the primary source of DOC is from the surrounding terrestrial landscape, though phytoplankton and macrophytes also produce DOC. Due largely to its ability to absorb light (evidenced by color), terrestrially derived DOC plays an important role in determining water transparency, mixing depth, and oxygen depletion in deeper waters, as well as the processing and bioavailability of nutrients and toxic compounds (Williamson et al. 1999). Increases in DOC can also alter the Climate change is driving us toward an overall warmer and wetter world, with more frequent and severe droughts as well as more intense precipitation and floods. Future climate scenarios point to dry regions and seasons getting drier and wet regions and seasons getting wetter (IPCC 2012, 2013), yet deciphering the effects of the temperature vs. precipitation components of a changing climate is notoriously difficult. Long-term climate trends are clouded by variability that creates noise in the system (Clark et al. 2010). To add to the challenge, climate change is manifested not only in long-term trends but also in extreme events and geographic variation across the landscape (IPCC 2012). For example, while precipitation has increased by 7% on average across the United States in the past 100 yr, the amount of rainfall in extreme precipitation events (the top 1% of all daily events) has increased by 20% (Karl et al. 2009). Some regions, such as the northeastern United States, have experienced up to a 67% increase in extreme precipitation events, while in other regions, such as the southwestern United States, this increase has been only 9% (Karl et al. 2009). These heavy rainfall events can lead to more intense runoff into lakes, which can in turn increase terrestrially derived dissolved organic carbon (DOC) concentrations that substantially alter thermal structure and ecosystem metabolism of lake ecosystems (Klug et al. 2012; Sadro and Melack 2012). * Corresponding author: [email protected] 840 Lakes as climate sentinels response of aquatic food webs to climate change (Hansson et al. 2012; Read and Rose 2013). Insights into the color and quality as well as the concentration of these climatedriven changes in DOC can also help resolve important questions about how these changes in DOC alter food web subsidies, or inhibit them through shading due to absorption of photosynthetically active radiation (PAR; Jones et al. 2012). Strategies for developing a more quantitative and informative framework to provide insights into the contrasting effects of warmer vs. wetter conditions on terrestrial and aquatic ecosystems are thus essential. One of the most promising approaches to deciphering the ecological effects of climate change involves the use of the sentinel responses of ecosystems themselves, as these may provide the most sensitive indicators of both drivers and responses. For example, in the world’s oceans, ecosystems amplify climate variation and provide stronger signals of regime shifts than do actual physical climate measurements (Taylor et al. 2002; Hays et al. 2005). While continental landscapes are more complex than the pelagic zones of oceans, these landscapes also have embedded sensors: lake ecosystems that are highly responsive to climate change (Williamson et al. 2009a,b). Lake responses to climate change include physical, chemical, and biological signals that are readily measured over multiple timescales (Adrian et al. 2009), yet to date there is no basic framework with which to take advantage of the power of lakes as climate sentinels. There is a need to identify which responses are the most valuable as clear and informative sentinels of alteration of both terrestrial and aquatic systems by climate change. Lakes are sensors in the landscape that provide many types of sentinel responses, defined here as physical, chemical, and biological signals of climate-driven ecosystem change that are both informative and quantifiable. Sentinel responses have important ecological consequences, defined here as responses that involve living organisms that inform us about how climate change alters ecosystem structure and function. Here we use this simple framework to develop and apply a suite of DOC-related climate forcing optical indices (CFOI) for assessing the sentinel responses of lakes to temperature and precipitation. We also examine the response of these indices to solar radiation, which changes as a function of cloud cover, one of the greatest unknowns in predicting future climate change (IPCC 2013). We examine changes across multiple timescales to evaluate whether short-term extreme events ranging from single storm events to periods of high or low drought can inform us about ecosystem-level responses over longer timescales. Sentinel responses that share common mechanisms and thus respond similarly across multiple timescales or space scales can be referred to as scalable sentinel responses as a result of their greater potential to predict long-term responses based on shortterm data. We then present a conceptual model of some of the potential sentinel responses to warmer vs. wetter climates on lakes and their ecological consequences. While these two components of climate change both induce similar increases in thermal stratification, they have opposite effects on water transparency and result in opposite ecological consequences. 841 Our approach springs from recent progress in absorption and fluorescence spectroscopy that has created new insights into the role of DOC in aquatic ecosystems. These techniques are based on the fact that DOC strongly and selectively absorbs shorter wavelength solar ultraviolet (UV) radiation (Williamson et al. 1996), making UVrelated metrics much more sensitive sentinels of environmental change than is DOC concentration (Williamson et al. 1999). UV and PAR transparency have been shown to be valuable sentinels of climate change in a wide variety of lakes ranging from boreal lakes in Canada (Gunn et al. 2001) to Arctic and Antarctic lakes (Vincent et al. 1998). More recent studies of UV and PAR transparency in a wide variety of subalpine and alpine lakes have shown that UV is more sensitive than PAR to environmental change (Rose et al. 2009a,b). Prior work (Fee et al. 1996; Snucins and Gunn 2000) has suggested that small, clear-water lakes are likely to be the most responsive systems to changes in DOC concentration and climate as a result of the greater importance of DOC and transparency vs. wind in controlling mixing depth in smaller lakes and the low background ‘noise’ levels of DOC in clear-water lakes. Here we show that shallow turbid reservoirs can be similarly responsive when considering DOC quality. UVbased optical metrics have also been used to demonstrate that dust from the Sahara Desert may contribute a major portion of the DOC to high alpine lakes in the Sierra Nevada of Spain (Mladenov et al. 2009). CFOI—The suite of CFOI that we propose here builds on these prior advances in UV absorbance spectroscopy and is based on two optical metrics of DOC quality: a*320, the DOC-specific absorbance at 320 nm (m2 (g C)21), which is the dissolved absorption coefficient at 320 nm (ad320, m21) divided by DOC concentration (mg C L21), and S275–295, the spectral slope, the slope of the linear relationship between the natural log of the absorption coefficient and wavelength in the 275–295-nm UV range (nm21). The ratio of these two indices can be used as a novel third, composite climate forcing index (CF 5 a*320 : S275–295, nm m2 (g C) 21). Increases in S275–295 are driven largely by DOC photobleaching (Helms et al. 2008), making these metrics informative indicators of increased exposure to sunlight. Increases in a*320 are driven by inputs of fresh terrestrial DOC, which decrease transparency by introducing non-photobleached (or less-photobleached) DOC (Hargreaves 2003; Helms et al. 2008; SanClements et al. 2012) that is often less biolabile. The contrasting and interrelated responses of a*320 and S275–295 to the balance between precipitation and temperature are largely independent of DOC concentration (Hargreaves 2003; Helms et al. 2008; Loiselle et al. 2009). These two metrics are inversely related because of the generally low prior exposure of terrestrially derived DOC to sunlight and the loss of UV absorbance that occurs during photobleaching. This complementary relationship of the two metrics provides a more robust and informative response index to climate change than could either metric alone, and this relationship also has the potential to more clearly differentiate mechanisms that underlie DOC source and processing. 842 Williamson et al. three different types of lakes were used to assess the value of the CFOI relationship over time intervals ranging from a single extreme event to months, years, and decades. Methods Fig. 1. Conceptual model of the climate forcing optical index (CFOI) relationship. CFOI curve (dashed line) showing the hypothesized response of DOC-specific absorbance (a*320) and spectral slope (S275–295) to warmer, drier (e.g., drought with high sunlight exposure, low terrestrial DOC inputs) vs. cooler, wetter (e.g., high precipitation with low sunlight exposure, high terrestrial DOC inputs) regimes. Here we present a series of hypotheses predicting how this suite of indices will respond to climate change. First of all, as a result of the complementary responses of a*320 vs. S275–295 to climate forcing, we predict that plotting these two component indices against each other will provide a robust relationship across a wide variety of lakes located in different climate regions across North America (Fig. 1). While myriad other environmental conditions may alter the position of a given lake on this CFOI curve, we predict that lakes will exhibit the potential to respond to climate change by moving along this curve. We hypothesize that for a given lake, warmer, drier conditions such as those associated with heat waves and droughts will lead to a decrease in CF as a*320 decreases as a result of a decrease in the flux of DOC from the surrounding watershed and S275–295 increases due to higher photobleaching (Fig. 1). In contrast, we hypothesize that cooler, wetter conditions associated with high precipitation events will lead to an increase in CF because S275–295 will decrease as the result of cloudy conditions and less photobleaching and a*320 will increase as the result of a hydrologic balance that favors increased influx of more light-absorbing terrestrially derived DOC (Fig. 1). Thus, we predict that CF will decrease with increasing temperature or exposure to solar radiation and increase with precipitation. We also hypothesize that conditions that are warmer and wetter, or colder and drier, will fall in between these extremes depending on the relative balance of allochthonous DOC flux and photobleaching. We tested these hypotheses and predictions in a suite of lakes across broad spatial and temporal scales. Single samples in 35 lakes from highly contrasting climate zones in different geographic regions of North America were used to test how robust the CFOI relationship was among diverse lakes. Multiple samples in Lake study sites—The 35 lakes sampled from across North America are all small lakes (all but one measured , 3.25 km2) that range from sinkholes in arid regions of New Mexico to seepage and surface runoff lakes of glacial origin in heavily forested regions of the central and northern United States, to artificial impoundments in heavily agricultural landscapes in Texas and Ohio, to subalpine and alpine lakes at and above tree line in the U.S. Sierra Nevada and Beartooth Mountains and Canadian Rockies. Data for these lakes are from just one sampling during the ice-free season. Of the three more intensively studied lakes, Acton Lake is a small (surface area 5 2.40 km2), shallow (maximum depth 5 8 m), hypereutrophic (mean summer chlorophyll 5 59 mg L21, DOC 5 3.7 mg L21) Ohio reservoir with a short hydraulic retention time (29.2 d), embedded in an agricultural landscape. Lakes Giles and Lacawac are both small (surface areas 5 0.48 and 0.21 km2, respectively), natural glacial lakes located in heavily forested, largely undisturbed catchments in northeastern Pennsylvania. Giles is a deeper (maximum depth 5 23 m), oligotrophic (mean summer chlorophyll 5 0.8 mg L21, DOC 5 1.3 mg L21) lake with a hydraulic retention time of 5.6 yr. Lacawac is a moderately deep (maximum depth 5 13 m), mesotrophic, and somewhat dystrophic (mean summer chlorophyll 5 3.8 mg L21, DOC 5 5.1 mg L21) lake with a hydraulic retention time of 3.4 yr. Data and analysis—Weekly water samples were taken during the spring and summer from Acton Lake over the course of 2 yr with contrasting weather conditions to examine the CFOI response, enabling us to capture a major storm event and seasonal transitions during 2008 from the cool wet spring (April and May average daily air temperature [ADT] 5 13.8uC, average daily precipitation [ADP] 5 3.8 mm) to the warm dry summer (July–August ADT 5 22.6uC, ADP 5 1.1 mm). These same data were also used to assess interannual variation in a warmer, drier year (2007, May–August ADT 5 22.3uC, ADP 5 1.4 mm) vs. a cooler, wetter year (2008, May–August ADT 5 20.6uC, ADP 5 3.2 mm). July samples from Lake Giles and nearby Lake Lacawac were used to look at the CFOI response to interdecadal variations in rainfall during years for which data were available (1993, 1996, 1997, 1999, 2000, 2005, 2006, 2008, 2010, and 2011) for the drier 1990s (May–July ADP 5 2.3 mm) and the wetter 2000s (May– July ADP 5 3.6 mm). Temperatures were similar during May–July for the two decades (ADT 5 18.5uC for the 1990s, 18.3uC for the 2000s). The meteorological data (ADT, ADP) are for the 75 d prior to the mid-July sampling dates for each of these years. For Acton Lake, meteorological data were obtained from the Environmental Protection Agency Clean Air Status and Trends Network (CASTNET) database for the OXF122 site (Oxford, Ohio) located about 5 km from the lake. For Lacawac and Giles Lakes as climate sentinels Lakes, meteorological data were obtained from automated instruments maintained at the Lacawac Sanctuary Field Station by one of the authors (BRH). Water samples were collected from the epilimnion and filtered using ashed Whatman GF/F filters. Filtered samples were analyzed for DOC using a Shimadzu Total Organic Carbon Analyzer (TOC-5000 or TOC-VCPH). A Milli-Q deionized water blank was subtracted from all DOC measurements, and values were calibrated using a diluted certified DOC standard (Aqua Solutions, 50 mg L21 potassium biphthalate). Dissolved absorbance properties were measured with a Shimadzu UV and visible spectrophotometer (UV-160U, UV-1601, or UV-1650 PC). Corrected dissolved absorbance values were calculated by subtracting a Milli-Q water blank and the average absorption from 775 to 800 nm from raw absorbance values. Naperian dissolved absorption coefficients were calculated as follows: 2:303|D r where D is the decadal optical density value from the spectrophotometer and r (units: m) is the path length of the quartz cuvette. The DOC-specific absorption coefficients (a*320) were calculated by dividing ad (320 nm, m21) by the DOC concentration (mg C L21). Spectral slopes (nm21) over the 275–295-nm range (S275–295) were calculated using linear regression to estimate the slope of the relationship between ln ad and wavelength, expressed according to convention as a positive number. For the regression analyses that examined the timing and strength of the response variables within the three primary study lakes, all meteorological data were averaged over the preceding time interval (0–15 d, 0–30 d, 0–45 d, 0–60 d, and 0–75 d). The CFOI variables and DOC concentration were the response variables and the weather drivers were the predictor variables. The response variables were log transformed to linearize the data. The data are from July samples for Lacawac and Giles and from the weekly samples during the 2007–2008 period in Acton. A simple linear regression was performed to model the relationship between the CFOI and DOC response variables and weather drivers ranging over different time intervals. Significance was determined at an a level of 0.05. All linear regressions were run using R (R Development Core Team 2011 versions 0.94.110 and 0.97.248). Scatter plots (not shown) were used to visualize the relationship between sentinel responses of the lakes and climate drivers. ad ~ Results All 35 lakes from climate regions ranging from the warm, dry southwest to the colder, wetter northeast United States fell along the same line on the plot of a*320 vs. S275–295, indicating a high degree of conformity of these diverse lakes to the CFOI relationship (r2 5 0.88, Fig. 2A). Furthermore, two very different types of lakes showed similar responses to temperature and precipitation across multiple timescales (Fig. 2B–D). Acton Lake (shallow, hypereutrophic reservoir) showed responses that follow 843 the CFOI curve for a single extreme precipitation event (arrow in Fig. 2B), for seasonal changes from spring to summer (r2 5 0.73, Fig. 2B), and for interannual changes from a warm, dry to a cool, wet year (r2 5 0.89, Fig. 2C). The CF index (mean 6 standard error [SE]) was 92 6 4.1 during the warmer, drier year (2007) and 155 6 12.7 in the cooler, wetter year (2008). The transition from the cooler, wetter spring to the warmer, drier summer in 2008 decreased the CF index from the 134–190 range in May to values of 94–95 in August. The extreme precipitation event in early June 2008 (over 109 mm of rain in 2 d) drove the CF index up to a value of 335. Lake Giles (deeper, oligotrophic, natural lake) similarly showed responses that conformed to the CFOI relationship on an interdecadal timescale (r2 5 0.80, Fig. 2D). The CF index was 11 6 1.4 for the drier 1990s and almost three times as high (31 6 6.0) for the wetter 2000s. For Lake Lacawac, a dystrophic, natural glacial lake in the same forested region as Giles, the CF value was 94 6 16.9 for the 1990s and 140 6 10.6 for the 2000s. Following an extreme precipitation event in 2006 (over 200 mm in 4 d from June 25 through 28), the CF values were driven up to 58 in Giles and 170 in Lacawac during July. We used linear regression to test how rapidly and strongly a*320, S275–295, CF, and DOC concentration in the same three lakes responded to average daily air temperature, precipitation, and solar radiation integrated over timescales of 15, 30, 45, 60, and 75 d prior to sampling (to give a start time close to ice-out). The r2 values were used to evaluate the strength of each sentinel response variable to each climate driver as well as the timing of the maximum response. For all three lakes the strongest CFOI responses were to precipitation (r2 5 0.62–0.78), with less pronounced responses to air temperature (only one significant response, a*320 in Lacawac, r2 5 0.41) and solar radiation (all three CFOI response variables in Lacawac, r2 5 0.58–0.61, and S275–295 in Giles, r2 5 0.39). The coefficients of determination for DOC concentration were generally lower than for the CFOI variables, with significant responses to air temperature only in Lake Giles (r2 5 0.48, Table 1) and to precipitation in Lacawac (r2 5 0.52) and Acton (r2 5 0.14). All CFOI variables showed significant responses to precipitation in all three lakes within 30–75 d, with peak responses generally at 60–75 d. Acton responded more rapidly to precipitation than did the other two lakes, with a significant response to all three CFOI variables within as few as 15 d (Table 1). Significant responses of the CFOI indices to solar radiation were observed between 30 and 75 d, with peak responses in Giles in as few as 45 d, in Lacawac at 60–75 d, and with no significant responses in Acton. Significant responses of DOC concentration were observed within 30–75 d, with peak responses to air temperature after 60 d in Giles and significant responses to precipitation after 60 and 30 d in Lacawac and Acton, respectively (Table 1). Discussion Sentinel responses that are effective over multiple timescales are of great interest because they may permit 844 Williamson et al. Fig. 2. Sentinel responses of climate forcing optical index (CFOI) variables in (A) 35 diverse lakes across North America (r2 5 0.88, y 5 21.0 e2129 x), an agricultural reservoir following (B) an extreme rain event as well as seasonal variation in climate (r2 5 0.73, y 5 10.9 e293.9 x), and (C) interannual variation among a warm, dry (2007) and a cool, wet (2008) year (r2 5 0.89, y 5 10.2 e291.4 x), and (D) interdecadal variation of a glacial lake in a forested watershed during the drier 1990s vs. the wetter 2000s (r2 5 0.80, y 5 10.2 e2117 x). Note the similar CFOI responses (move to lower right during hot, dry regimes) across different timescales and lake types. (A) Letters represent geographic region of the United States (E 5 East, MW 5 Midwest, N 5 North, W 5 West, S 5 South), R 5 reservoir, CR 5 Canadian Rocky Mountains. (B–D) Dashed lines are the regression line from (A). new insights into the ecological consequences of both extreme events and long-term trends in climate change by examining short-term responses. But such responses are especially valuable if the underlying mechanisms are the same and if they respond in a similar fashion across timescales so that they comprise scalable sentinel responses. The optical metrics of DOC quality reported here share the same mechanisms and show similar responses across multiple timescales from single storm events to seasonal, interannual, and interdecadal timescales and thus meet the criteria of being scalable sentinel responses. These indices are highly responsive to variations in precipitation and, to a lesser extent, to changes in solar radiation and temperature, extending prior work (Schindler et al. 1996; Snucins and Gunn 2000) demonstrating decreases in DOC concentration in response to drought. The data presented here show that responses of DOC concentration are sensitive across a wide range of timescales, from weeks to decades, and, more importantly, demonstrate how weather and climate alter DOC quality with important implications for both carbon loss from terrestrial ecosystems and lake ecosystem structure and function. These indices are particularly sensitive to climate change as a result of the fact that DOC quality is more sensitive to climate change than is DOC concentration alone and as a result of the high responsiveness of both component indices to solar radiation and precipitation in complementary ways. For example, drought conditions are usually characterized by high solar radiation and low precipitation, while storm events are characterized by low solar radiation and high precipitation. The variations in timing and magnitude of the CFOI responses observed here highlight the ways in which lakes Lakes as climate sentinels 845 Table 1. Time interval (d) for maximum (top number) coefficient of determination (r2) for response of DOC concentration (mg L21) and three climate forcing optical index variables (a*320 5 DOC-specific absorbance at 320 nm, S275–295 5 275–295-nm spectral slope, CF 5 climate forcing index, the ratio of a*320 to S275–295) to average daily precipitation (mm d21), temperature (uC), and solar radiation (kJ m22 d21). All time intervals with significant regressions are given in parentheses when there is more than one. Data for Lakes Giles and Lacawac were only available for 10 of the years from 1993 to 2011 (see Methods). Dashes indicate no time interval shown because r2 is so low (, 0.1). [DOC] Days Lake Giles, 1993–2011 (July) Precipitation Air temperature Solar radiation 60 60 30 Lake Lacawac, 1993–2011 (July) Precipitation 60(30,60,75) Air temperature — Solar radiation 15 a*320 r2 0.3 0.48 0.12 0.52 ,0.1 0.32 Acton Reservoir, 2007–2008 (May–August) Precipitation 30 0.14 Air temperature — ,0.1 Solar radiation — ,0.1 S275–295 Days r2 Days 60(30,60,75) — — 0.62 ,0.1 ,0.1 75(30–75) 15 45 0.78 0.41 0.61 0.75 ,0.1 ,0.1 45(30–75) 15 60(45–75) 75(15–75) — — CF r2 Days r2 0.68 0.28 0.39 60(30–75) — 60 0.66 ,0.1 0.14 75(30–75) 15 75(30–75) 0.76 0.28 0.58 75(30–75) 15 60(45–75) 75(15–75) — — 0.74 ,0.1 ,0.1 75(15–75) — — 0.76 0.37 0.61 0.77 ,0.1 ,0.1 * Statistically significant values (p , 0.05) in bold type; significant negative correlations in bold italics. with different types of land use and catchment characteristics may vary in the timing and magnitude of their responses to climate forcing. For example, the more rapid response of Acton to precipitation (compared to the other two lakes; Table 1) is likely due to its shallow nature and short retention time, allowing tributary water with high allochthonous inputs to rapidly dominate the lake after a rainfall event. The high sensitivity of Lacawac to solar radiation (in comparison to the other two lakes; Table 1) is likely due to the substantial portion of its watershed that comprises a bog that provides highly photolabile DOC to the lake, but it may also be related to iron-bound DOC from anoxic sediments (Maloney et al. 2005). Similarly, the short retention time combined with high inputs of inorganic sediments from the surrounding, highly disturbed agricultural watershed (Knoll et al. 2013) may contribute to the lack of a response to photobleaching in Acton (Table 1). While previous work (Snucins and Gunn 2000) has suggested that higher transparency lakes are more sensitive sentinels of climate-induced changes in DOC, we found that both the CFOI variables and DOC concentration were as responsive, if not more responsive, to precipitation in Lacawac than they were in the higher transparency Lake Giles (Table 1). We attribute these patterns to DOC quality being more responsive to precipitation than is DOC concentration alone as a response metric. The data reported here also suggest that both DOC concentration and quality are less sensitive to changes in temperature than they are to changes in precipitation and photobleaching. This is consistent with the findings of prior studies (Zhang et al. 2010) that have suggested that precipitation and photobleaching are the two most important processes controlling differences in the quality and concentration of DOC in lakes and their response to climate change. Changes in a*320 and S275–295 may inform us about DOC sources and lake ecosystem responses to altered DOC quality (Hargreaves 2003) that may underlie recently observed DOC trends. Terrestrially derived DOC has a generally lower S275–295 and higher a*320 than does pelagic source DOC; it also has a higher molecular weight and is more highly colored and more photolabile but less biolabile than pelagic source DOC (Helms et al. 2008). The balance between photolability and biolability will in turn largely determine whether a given concentration of DOC will depress photosynthesis through shading or serve as a resource subsidy for aquatic food webs (Jones et al. 2012). Changes in a*320 may also be indicative of conditions favorable to methane production and can lead to higher d13C stable isotope ratios through photobleaching (Osburn et al. 2001). In this prior study of our two Pennsylvania study lakes the d13C values diverged between the surface and deep waters by as much as 3–4% during periods of thermal stratification (Osburn et al. 2001). This is likely due to the conversion of higher molecular weight, terrestrially derived DOC to lower molecular weight compounds in the surface waters (Osburn et al. 2001). There are also deepwater and bottom sediment processes in the anoxic region of Lacawac that could contribute to these differences (Maloney et al. 2005). This difference in the stable isotope signatures in shallow vs. deep waters has important implications for the ongoing debates about the relative importance of allochthonous carbon to aquatic food webs (Jones et al. 2012) because hypolimnetic DOC has a similar optical and isotopic signature to terrestrially derived DOC. To further explore the relative response strength of the CF index to precipitation vs. temperature we ranked the years in our interannual data set in Lake Giles to identify the warmest year (2006), the wettest year (also 2006), and the years with the highest drought (1993, lowest Palmer Drought Severity Index, PDSI) and with the lowest drought (2011, highest PDSI) and compared the differences in their CF values. The logic was that the drought indices 846 Williamson et al. would reveal the years during which temperature, evaporation, and precipitation interacted to give us the warmest, driest year (1993) and the coolest, wettest year (2011). The Giles CF index values ranged from 7 in the most severe drought year (1993) to 58 in the warmest and wettest year (2006). These were the lowest and highest points on the CFOI curve, respectively (Fig. 2D). The least severe drought year (2011) and the coolest, driest year (1997) had intermediate values of 23 and 14, respectively. This suggests that our initial predictions that a higher CF value will be related to colder, wetter conditions (Fig. 1) did not hold in Lake Giles for that year at least. This pattern is consistent with the results that emerged in our regression analyses— precipitation is more important than temperature in driving the observed changes in DOC quality (Table 1). As with any indices of change, the CFOI relationship is sensitive to other potential drivers of environmental change, such as wind speed, land use, land management, and acid deposition. Thus, the CFOI metrics must be used cautiously with these potential other drivers in mind. For example, wind speed has shown long-term trends in recent decades (Pryor et al. 2009), and wind speed influences the depth of the surface mixed layer in larger lakes and consequently both photobleaching rates (decreasing a* and increasing S275–295) as well as advection of DOC from deeper waters (potentially increasing a* and decreasing S275–295). In spite of the shallower mixing depths that have been observed in these lakes and anticipated higher photobleaching rates, the response to higher precipitation in recent years was fully consistent with the CFOI relationship (Figs. 1, 2). Changes in land use patterns similarly have the potential to influence the CFOI relationship. The amount of wetlands within a watershed is an important determinant of DOC concentration in lakes, particularly in alpine and subalpine regions (Winn et al. 2009). In agricultural regions, storm events may alter DOC quality and concentration in downstream systems in ways that differ from those seen in well-forested watersheds (Caverly et al. 2013). Thus, draining of wetlands or conversion to (or recovery from) agriculture must be considered when applying the CFOI metrics. Changes in land use in the watersheds of Lacawac and Giles have been minimal. Lake Lacawac is located within the Lacawac Sanctuary, and preservation of the intact watershed is a high priority, as it is in the nearby privately owned property surrounding Lake Giles. Thus, land use change does not explain the long-term trends that we observed. The very similar CFOI responses observed in Acton Lake (highly agricultural watershed) and the Pocono lakes (preserved forested watersheds) suggest that the CFOI metrics are robust across lakes with a wide variety of land use patterns. One of the leading hypotheses proposed to explain the observed increases in DOC in both Europe and North America is recovery from acidification, with reductions in anthropogenic sulfate deposition (Monteith et al. 2007). Recovery from acidification has been clearly demonstrated to cause an increase in DOC concentration (Williamson et al. 1996; Tanentzap et al. 2008). Drought and acidification can also interact to alter the UV transparency of lakes (Schindler et al. 1997). In one lake with extreme exposure to acid deposition, drought induced a threefold increase in UV transparency (Yan et al. 1996). The processes that control DOC optical quality and thus potentially influence the CFOI relationship are also pH dependent, so the nature of the response depends on the pH range. In the pH range of 4 to 7, the loss rate of total organic carbon due to photobleaching increases with decreasing pH (Gennings et al. 2001), but there is little change in DOC-specific absorbance (similar to our a*320, but absorbance was measured at 440 nm; Pace et al. 2012). Above a pH of 7, photobleaching induces a more rapid loss of DOC-specific absorbance with increasing pH. Of our two primary study lakes, only Lake Giles (pH 5.4–6.1) is likely to be influenced by recovery from acid precipitation, because Acton has a pH of 7.6–9.1 and a very high buffering capacity (, 3000 meq L21). While in Giles we observed no significant relationship between CF and pH (p 5 0.15), the trend was positive. We suspect that pH may play some role in the observed interdecadal response of the CFOI because the pH in the lake increased from 5.4 in 1993 to 6.3 in 2012. In Acton we observed a highly significant inverse relationship between CF and pH (r2 5 0.63, p , 0.01). Both of these patterns are consistent with what has been previously observed with the effects of pH on photobleaching (Pace et al. 2012). However, the relationship between pH and CF could also be driven by the fact that high runoff events carry a higher portion of water that has not passed through the calcic bedrock, while at times of low flow stream water contains a greater portion of groundwater flow. The relationship between DOC and pH is complex because influx of DOC during rain events can alter pH, and the response will depend on the nature of the soils in the catchment as well as on the hydrology of a particular lake. In spite of these caveats about pH and DOC quality, the similarity of the responses of the CFOI across lakes of very differing pH and multiple timescales remains compelling and argues against acid deposition playing a dominant role in the patterns that we observed. The relationship between the CFOI response and pH remains intriguing, however, and needs more investigation. On a more general level, the debate over the role of pH and other drivers of the observed trends in DOC concentration has yet to be resolved. While the relative importance of different drivers is likely to vary regionally, climate clearly plays an important role. In Europe, analysis of some of the largest data sets has shown climate to be highly correlated with DOC concentration (Weyhenmeyer and Karlsson 2009). Climate change is also predicted to increase future DOC concentrations in boreal lakes by as much as 65% as a result of warming increasing terrestrial vegetation (Larsen et al. 2011). A wide range of studies have clearly demonstrated the link between precipitation and increases in DOC concentrations during recovery from drought (Schindler et al. 1996), wetter climates in the springtime (Pace and Cole 2002), and heavy rainfall associated with global oscillations such as the Atlantic Multidecadal Oscillation (Gaiser et al. 2009). Even in regions where recovery from acidification has been a clear contributor to regional increases in DOC, precipitation and incident solar radiation appear to provide similar, if not better, explanatory power than pH for observed trends of Lakes as climate sentinels 847 increasing DOC and decreasing transparency (Gunn et al. 2001; Zhang et al. 2010; Couture et al. 2012). Similarly, in 58 Adirondack lakes with pH values between 4.2 and 7.0 there was no significant explanatory power of pH in terms of predicting variation in PAR, UV transparency, DOC concentration, or DOC-specific absorbance at 440 nm among lakes within a single year (Bukaveckas and Robbins-Forbes 2000). Our work extends this prior work by demonstrating the central importance of precipitation and incident solar radiation in driving changes in DOC quality, as well as in demonstrating a smaller role for temperature. Fig. 3. Conceptual model showing the response of lake ecosystems to warmer vs. wetter climate regimes. These two extremes of climate forcing both increase thermal stratification but have opposite effects on water transparency and ecological consequences, signaled by changes in the climate forcing optical indices, including DOC-specific absorbance (a*320), spectral slope (S275–295), and overall climate forcing, the ratio of a*320 to S275–295 (CF). Warmer climate regimes will lead to greater water transparency (reducing a*320, increasing S275–295) and consequently more autotrophic, nutrient-limited conditions where the ecological consequences include less (2) oxygen depletion and lower (2) zooplankton food quality but positive (+) effects on all other listed ecological consequences (see + and 2 in left column). Wetter climate regimes will increase terrestrial DOC inputs to lakes, reducing water transparency and consequently leading to more heterotrophic, light-limited conditions (increasing a*320, reducing S275–295) where the effects on the ecological consequences are the opposite effects of those of warming (see + and 2 in right column). Conceptual model—The effects of climate change on lakes have been demonstrated repeatedly. Yet there remains an urgent need to identify key sentinel responses to develop a more coherent framework with which to address the contrasting effects of temperature and precipitation components of climate change and their ecological consequences for lake ecosystems. What signals are ecosystems giving us that can help to reveal their responses to these two primary components of climate change? Here we present a conceptual model within the context of climate-driven changes in DOC quality and concentration and discuss past findings in order to better understand the sentinel responses of lakes to these two very different, but interrelated, climate drivers (Fig. 3). In spite of the contrasting effects of warmer vs. wetter conditions on water transparency, both higher temperatures and increases in precipitation will increase the strength of thermal stratification of lakes (Fig. 3). Modeling efforts have long suggested that warmer air temperatures will lead to both shallower and warmer surface mixed layers and longer and stronger periods of thermal stratification (De Stasio et al. 1996). These modeling efforts are supported by extensive empirical data. For example, a compilation of long-term data from lakes across the Northern Hemisphere showed trends of increasing surface-water temperatures in 11 out of 16 lakes (Adrian et al. 2009). Data from both remote sensing and actual in-lake measurements indicate that the water temperatures of large lakes are warming at almost twice the rate of air temperatures in the same regions (Schneider et al. 2009). Comparison of two lakes in Switzerland following the European heat wave of 2003 showed warmer surface-water temperatures but also thermal stratification and Schmidt stability values that exceeded the values for the prior 1956– 2002 period by 4–5 standard deviations (Jankowski et al. 2006). Analysis of 35 yr of data (1955–1992) in the 26-kmlong South Bay of Lake Huron revealed that periods with warmer air temperatures and higher solar radiation were characterized by warmer surface waters and shallower thermoclines (King et al. 1997). A study of 86 small boreal lakes during a cool year and two warm years found that warmer air temperatures were related to warmer surfacewater temperatures, shallower mixed layers, and stronger thermal stratification (Snucins and Gunn 2000). In this study the deep waters of clear lakes warmed while those of higher DOC, more colored lakes either did not change or even showed a cooling during the warmer years. Lake 848 Williamson et al. cooling has actually been observed during periods of unchanging air temperatures when increases in DOC concentration decrease water transparency (Tanentzap et al. 2008). Higher precipitation increases loading of terrestrial DOC to inland waters, which reduces water transparency, leading to more absorption of sunlight in the surface waters, shallower mixed layer depths, and stronger thermal stratification (Read and Rose 2013). Our own work reported above shows that increases in precipitation increase not only DOC concentration but also the DOC-specific absorbance (a*320). These increases in a*320 further enhance the amount of solar radiation absorbed in the surface waters of lakes and decrease the heating of deeper strata, increasing the strength of thermal stratification beyond the effects of DOC concentration alone. Thus, both warmer and wetter conditions combined will lead to particularly strong increases in thermal stratification, consistent with our observation that the warmest and wettest year from our multiyear data set also had the highest CF value. During a 20 yr period with severe drought involving trends of increasing air temperature (2uC) and decreasing precipitation, small boreal lakes exhibited a trend of increasing mixing depth (Schindler et al. 1996). This increase in mixing depth was associated with an increase in wind speed. Consistent with prior studies on the effects of warming on lakes and our conceptual model (Fig. 3), in this study we did observe an increase in transparency (Secchi) due to decreases in DOC as well as increases in volume-corrected lake temperature. Volume-corrected lake temperature and air temperature increased by about the same amount (0.108uC yr21 and 0.107uC yr21, respectively) over the 20 yr period from 1969 to 1988. More information on DOC quality obtained through the suite of CFOI variables proposed here would further elucidate the importance of DOC source and processing that likely underlies these observed sentinel responses. Optical differences in DOC quality have important ecological consequences related to water transparency and thermal stratification (Fig. 3). One of the most important ecological consequences is that periods of warmer surface waters and stronger thermal stratification can both depress primary production (O’Reilly et al. 2003) and favor toxinproducing cyanobacteria (Carey et al. 2012). Periods of higher precipitation that increase DOC can similarly favor cyanobacteria and shift the trophic structure of pelagic communities (Bastidas Navarro and Modenutti 2012). In Lake Zurich, which provides drinking water for 1.5 million people, increased thermal stratification has led to an increase in the abundance of the toxic cyanobacterium Planktothrix rubescens. This is a non–nitrogen fixing species that in contrast to many cyanobacteria shows optimal growth under colder conditions and low light (Posch et al. 2012). Deep annual holomixis to depths that burst buoyancy-regulating gas vacuoles is the primary source of mortality for this inedible cyanobacterium. Stronger thermal stratification reduces mixing to the damaging deeper depths, thereby permitting this species to grow throughout the year and increasing its population size in recent decades (Posch et al. 2012). Other ecological consequences of increased DOC and a*320 and reduced photobleaching and S275–295 include a shallower compensation depth, more oxygen depletion in deep waters, shallower chlorophyll maximum and zooplankton distributions, reduced amplitude of zooplankton vertical migration, and alteration of the refuge from visual predators as well as timing and strength of critical phenological events such as the clear-water phase (Fig. 3). Shallower vertical distributions of zooplankton related to decreasing UV transparency following even a single extreme storm event in 2006 have been demonstrated (Rose et al. 2012). In contrast, drought-induced increases in lake water residence time and incident sunlight will increase DOC photobleaching with consequent increases in S275–295 and decreases in a*320 and corresponding increases in water transparency and mixing depth (Figs. 1, 3). Associated increases in DOC biolability, water transparency, and light : nutrient ratios lead to decreases in food quality for zooplankton and fish by increasing the carbon : phosphorus ratio (Dickman et al. 2008). The ecological consequences of increases in DOC combined with warming temperatures have been demonstrated to be lake-specific depending on food chain length (Hansson et al. 2012). Warmer water temperatures can also cause shifts in zooplankton community structure, favoring species that either have a higher physiological tolerance for warmer temperatures or that have higher growth rates at warmer temperatures (Wojtal-Frankiewicz 2012). Increased DOC loading also increases the cost of water treatment by increasing the production of carcinogenic disinfection by-products in drinking water supplies (Beggs and Summers 2011) and reducing water disinfection of parasites and pathogens by sunlight (Overholt et al. 2012). The dependence of these noxious chemical reactions on DOC quality is unknown but will be altered by climate change (Figs. 1–3). Recent advances in optical and other sensor technologies make the ‘‘lakes as sentinels’’ approach particularly amenable to use in ecological observatory networks (EONS) such as the Global Lake Ecological Observatory Network (GLEON) and the U.S. National Ecological Observatory Network (NEON). These EONS are playing a key role in establishing both the broadscale networking of sites as well as the computational and analytical infrastructure necessary for deciphering macrosystem-scale ecological processes. Yet EONS are also producing massive amounts of data that are a challenge to manage. One strategy for ‘‘staying afloat in the sensor deluge’’ (Porter et al. 2012) is to use targeted approaches involving specific indices, such as the CFOI metrics, coupled with high-frequency data to enable new insights into ecosystem-level responses to climate change. Acknowledgments We thank Wiebke Boeing, Janet Fischer, Evelyn Gaiser, Weston Nowlin, John Melack, Mark Olson, Jasmine Saros, and Hilary Swain for help and logistics in collecting water samples that contributed to the 35 lake data set and the late Robert E. Moeller for contributing to earlier discussions regarding indices to look at the shifting balance between hydraulic flushing and photobleaching. Comments from two anonymous reviewers improved an earlier version of this manuscript. We also thank the GLEON for providing forums at their meetings for presenting and discussing some of the ideas presented here with the broader limnological community as well as for travel support for some of the coauthors who attended these meetings. This work was supported Lakes as climate sentinels in part by National Science Foundation grants from the Division of Environmental Biology (DEB)-0734277, DEB Integrated Research Challenges in Environmental Biology-0552283, and the Division of Graduate Education-0903560. References ADRIAN, R., AND OTHERS. 2009. Lakes as sentinels of climate change. Limnol. Oceanogr. 54: 2283–2297, doi:10.4319/ lo.2009.54.6_part_2.2283 BASTIDAS NAVARRO, M., AND B. MODENUTTI. 2012. Precipitation patterns, dissolved organic matter and changes in the plankton assemblage in Lake Escondido (Patagonia, Argentina). Hydrobiologia 691: 189–202, doi:10.1007/s10750-0121073-5 BEGGS, K. M. H., AND R. S. SUMMERS. 2011. Character and chlorine reactivity of dissolved organic matter from a mountain pine beetle impacted watershed. Environ. Sci. Technol. 45: 5717–5724, doi:10.1021/es1042436 BUKAVECKAS, P. A., AND M. ROBBINS-FORBES. 2000. Role of dissolved organic carbon in the attenuation of photosynthetically active and ultraviolet radiation in Adirondack lakes. Freshw. Biol. 43: 339–354, doi:10.1046/j.1365-2427.2000.00518.x CAREY, C. C., B. W. IBELINGS, E. P. HOFFMAN, D. P. HAMILTON, AND J. D. BROOKES. 2012. Eco-physiological adaptations that favour freshwater cyanobacteria in a changing climate. Water Res. 46: 1394–1407, doi:10.1016/j.watres.2011.12.016 CAVERLY, E., J. M. KASTE, G. S. HANCOCK, AND R. M. CHAMBERS. 2013. Dissolved and particulate organic carbon fluxes from an agricultural watershed during consecutive tropical storms. Geophys. Res. Lett. 40: 5147–5152, doi:10.1002/grl.50982 CLARK, J. M., AND OTHERS. 2010. The importance of the relationship between scale and process in understanding long-term DOC dynamics. Sci. Total Environ. 408: 2768–2775, doi:10.1016/j.scitotenv.2010.02.046 COUTURE, S., D. HOULE, AND C. GAGNON. 2012. Increases of dissolved organic carbon in temperate and boreal lakes in Quebec, Canada. Environ. Sci. Pollut. Res. 19: 361–371, doi:10.1007/s11356-011-0565-6 DE STASIO, B. T., JR, D. K. HILL, J. M. KLEINHANS, N. P. NIBBELINK, AND J. J. MAGNUSON. 1996. Potential effects of global climate change on small north temperate lakes: Physics, fishes, and plankton. Limnol. Oceanogr. 41: 1136–1149, doi:10.4319/lo.1996.41.5.1136 DICKMAN, E. M., J. M. NEWELL, M. J. GONZÁLEZ, AND M. J. VANNI. 2008. Light, nutrients, and food-chain length constrain planktonic energy transfer efficiency across multiple trophic levels. Proc. Natl. Acad. Sci. USA 105: 18408–18412, doi:10.1073/pnas.0805566105 FEE, E. J., R. E. HECKY, S. E. M. KASIAN, AND D. R. CRUIKSHANK. 1996. Effects of lake size, water clarity, and climatic variability on mixing depths in Canadian Shield lakes. Limnol. Oceanogr. 41: 912–920, doi:10.4319/lo.1996.41.5.0912 GAISER, E. E., N. D. DEYRUP, R. W. BACHMANN, L. E. BATTOE, AND H. M. SWAIN. 2009. Multidecadal climate oscillations detected in a transparency record from a subtropical Florida Lake. Limnol. Oceanogr. 54: 2228–2232, doi:10.4319/lo.2009. 54.6.2228 GENNINGS, C., L. A. MOLOT, AND P. J. DILLON. 2001. Enhanced photochemical loss of organic carbon in acidic waters. Biogeochemistry 52: 339–354, doi:10.1023/A:1006499713835 GUNN, J. M., E. SNUCINS, N. D. YAN, AND M. T. ARTS. 2001. Use of water clarity to monitor the effects of climate change and other stressors on oligotrophic lakes. Environ. Monit. Assess. 67: 69–88, doi:10.1023/A:1006435721636 849 HANSSON, L. A., AND OTHERS. 2012. Food-chain length alters community responses to global change in aquatic systems. Nat. Clim. Change 3: 228–233, doi:10.1038/nclimate1689 HARGREAVES, B. R. 2003. Water column optics and penetration of UVR, p. 59–105. In E. W. Helbling and H. E. Zagarese [eds.], UV effects in aquatic organisms and ecosystems. Comprehensive series in photochemistry & photobiology. Royal Society of Chemistry. HAYS, G. C., A. J. RICHARDSON, AND C. ROBINSON. 2005. Climate change and marine plankton. Trends Ecol. Evol. 20: 337–344, doi:10.1016/j.tree.2005.03.004 HELMS, J. R., A. STUBBINS, J. D. RITCHIE, E. C. MINOR, D. J. KIEBER, AND K. MOPPER. 2008. Absorption spectral slopes and slope ratios as indicators of molecular weight, source, and photobleaching of chromophoric dissolved organic matter. Limnol. Oceanogr. 53: 955–969, doi:10.4319/lo.2008.53.3.0955 INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE (IPCC). 2012. Managing the risks of extreme events and disasters to advance climate change adaptation, p. 1–582. In C. B. Field and others [eds.], A special report of Working Groups I and II of the Intergovernmental Panel on Climate Change. Cambridge Univ. Press. INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE (IPCC) 2013. Summary for policymakers: Climate change 2013—the physical science basis, p. 1–27. In T. F. Stocker and others [eds.], Working Group 1 contribution to the IPCC fifth assessment report of the Intergovernmental Panel on Climate Change. Cambridge Univ. Press. JANKOWSKI, T., D. M. LIVINGSTONE, H. BUHRER, R. FORSTER, AND P. NIEDERHAUSER. 2006. Consequences of the 2003 European heat wave for lake temperature profiles, thermal stability, and hypolimnetic oxygen depletion: Implications for a warmer world. Limnol. Oceanogr. 51: 815–819, doi:10.4319/lo.2006.51.2.0815 JONES, S. E., C. T. SOLOMON, AND B. WEIDEL. 2012. Subsidy or subtraction: How do terrestrial inputs influence consumer production in lakes? Freshw. Rev. 5: 37–49. KARL, T. R., J. M. MELILLO, AND T. C. PETERSON [EDS]. 2009. Global climate change impacts in the United States. Cambridge Univ. Press. KING, J. R., B. J. SHUTER, AND A. P. ZIMMERMAN. 1997. The response of the thermal stratification of South Bay (Lake Huron) to climatic variability. Can. J. Fish. Aquat. Sci. 54: 1873–1882, doi:10.1139/f97-093 KLUG, J. L., AND OTHERS. 2012. Ecosystem effects of a tropical cyclone on a network of lakes in northeastern North America. Environ. Sci. Technol. 46: 11693–11701, doi:10.1021/es302063v KNOLL, L. B., M. J. VANNI, W. RENWICK, E. DITTMANN, AND J. A. GEPHART. 2013. Temperate reservoirs are large carbon sinks and small CO2 sources: Results from high-resolution carbon budgets. Glob. Biogeochem. Cycles 27: 52–64, doi:10.1002/gbc.20020 LARSEN, S., T. ANDERSEN, AND D. HESSEN. 2011. Climate change predicted to cause severe increase of organic carbon in lakes. Glob. Change Biol. 17: 1186–1192, doi:10.1111/j.13652486.2010.02257.x LOISELLE, S. A., L. BRACCHINI, A. M. DATTILO, M. RICCI, A. TOGNAZZI, A. COZAR, AND C. ROSSI. 2009. Optical characterization of chromophoric dissolved organic matter using wavelength distribution of absorption spectral slopes. Limnol. Oceanogr. 54: 590–597, doi:10.4319/lo.2009.54.2.0590 MALONEY, K. O., D. P. MORRIS, C. O. MOSES, AND C. L. OSBURN. 2005. The role of iron and dissolved organic carbon in the absorption of ultraviolet radiation in humic lake water. Biogeochemistry 75: 393–407, doi:10.1007/s10533-005-1675-3 MLADENOV, N., J. LÓPEZ-RAMOS, D. M. MCKNIGHT, AND I. RECHE. 2009. Alpine lake optical properties as sentinels of dust deposition and global change. Limnol. Oceanogr. 54: 2386–2400, doi:10.4319/lo.2009.54.6_part_2.2386 850 Williamson et al. MONTEITH, D. T., AND OTHERS. 2007. Dissolved organic carbon trends resulting from changes in atmospheric deposition chemistry. Nature 450: 537–541, doi:10.1038/nature06316 O’REILLY, C. M., S. R. ALIN, P.-D. PLISNIER, A. S. COHEN, AND B. A. MCKEE. 2003. Climate change decreases aquatic ecosystem productivity of Lake Tanganyika, Africa. Nature 424: 766–768, doi:10.1038/nature01833 OSBURN, C. L., D. P. MORRIS, K. A. THORN, AND R. E. MOELLER. 2001. Chemical and optical changes in freshwater dissolved organic matter exposed to solar radiation. Biogeochemistry 54: 251–278, doi:10.1023/A:1010657428418 OVERHOLT, E. P., S. H. HALL, C. E. WILLIAMSON, C. K. MEIKLE, M. A. DUFFY, AND C. E. CÁCERES. 2012. Solar radiation decreases parasitism in Daphnia. Ecol. Lett. 15: 47–54, doi:10.1111/j.1461-0248.2011.01707.x PACE, M. L., AND J. J. COLE. 2002. Synchronous variation of dissolved organic carbon and color in lakes. Limnol. Oceanogr. 47: 333–342, doi:10.4319/lo.2002.47.2.0333 ———, I. RECHE, J. J. COLE, A. FERNÁNDEZ-BARBERO, I. P. MAZUECOS, AND Y. PRAIRIE. 2012. pH change induces shifts in the size and light absorption of dissolved organic matter. Biogeochemistry 108: 109–118, doi:10.1007/s10533-011-9576-0 PORTER, J. H., P. C. HANSON, AND C. C. LIN. 2012. Staying afloat in the sensor data deluge. Trends Ecol. Evol. 27: 121–129, doi:10.1016/j.tree.2011.11.009 POSCH, T., O. KOSTER, M. M. SALCHER, AND J. PERNTHALER. 2012. Harmful filamentous cyanobacteria favoured by reduced water turnover with lake warming. Nat. Clim. Change 2: 809–813, doi:10.1038/nclimate1581 PRYOR, S. C., AND OTHERS. 2009. Wind speed trends over the contiguous United States. J. Geophys. Res 114: D14105, doi:10.1029/2008JD011416 R DEVELOPMENT CORE TEAM. 2011. R: A language and environment for statistical computing [Internet]. Accessed 03 June 2013. Available from http://www.r-project.org/ READ, J., AND K. C. ROSE. 2013. Physical responses of small temperate lakes to variation in dissolved organic carbon concentrations. Limnol. Oceanogr. 58: 921–931. ROSE, K. C., C. E. WILLIAMSON, J. M. FISCHER, S. J. CONNELLY, M. OLSON, A. J. TUCKER, AND D. A. NOE. 2012. The role of UV and fish in regulating the vertical distribution of Daphnia. Limnol. Oceanogr. 57: 1867–1876, doi:10.4319/lo.2012.57.6.1867 ———, ———, J. E. SAROS, R. SOMMARUGA, AND J. M. FISCHER. 2009a. Differences in UV transparency and thermal structure between alpine and subalpine lakes: Implications for organisms. Photochem. Photobiol. Sci. 8: 1244–1256, doi:10.1039/b905616e ———, ———, S. G. SCHLADOW, M. WINDER, AND J. T. ORIS. 2009b. Patterns of spatial and temporal variability of UV transparency in Lake Tahoe, California-Nevada. J. Geophys. Res. Biogeosci 114: G00D03, doi:10.1029/2008JG000816 SADRO, S., AND J. MELACK. 2012. The effect of an extreme rain event on the biogeochemistry and ecosystem metabolism of an oligotrophic high-elevation lake. Arct. Antarct. Alp. Res. 44: 222–231, doi:10.1657/1938-4246-44.2.222 SANCLEMENTS, M. D., G. P. OELSNER, D. M. MCKNIGHT, J. L. STODDARD, AND S. J. NELSON. 2012. New insights into the source of decadal increases of dissolved organic matter in acid-sensitive lakes of the Northeastern United States. Environ. Sci. Technol. 46: 3212–3219, doi:10.1021/es204321x SCHINDLER, D. W., P. J. CURTIS, S. E. BAYLEY, B. R. PARKER, K. G. BEATY, AND M. P. STAINTON. 1997. Climate-induced changes in the dissolved organic carbon budget of boreal lakes. Biogeochemistry 36: 9–28, doi:10.1023/A:1005792014547 ———, AND OTHERS. 1996. The effects of climatic warming on the properties of boreal lakes and streams at the Experimental Lakes Area, northwestern Ontario. Limnol. Oceanogr. 41: 1004–1017, doi:10.4319/lo.1996.41.5.1004 SCHNEIDER, P., S. J. HOOK, R. G. RADOCINSKI, G. K. CORLETT, G. C. HULLEY, S. G. SCHLADOW, AND T. E. STEISSBERG. 2009. Satellite observations indicate rapid warming trend for lakes in California and Nevada. Geophys. Res. Lett 36: L22402, doi:10.1029/2009GL040846 SNUCINS, E., AND J. GUNN. 2000. Interannual variation in the thermal structure of clear and colored lakes. Limnol. Oceanogr. 45: 1639–1646, doi:10.4319/lo.2000.45.7.1639 TANENTZAP, A. J., AND OTHERS. 2008. Cooling lakes while the world warms: Effects of forest re-growth and increased dissolved organic matter on the thermal regime of a temperate urban lake. Limnol. Oceanogr. 53: 404–410, doi:10.4319/lo.2008.53.1.0404 TAYLOR, A. H., J. I. ALLEN, AND P. A. CLARK. 2002. Extraction of a weak climate signal by an ecosystem. Nature 416: 629–632, doi:10.1038/416629a VINCENT, W. F., I. LAURION, AND R. PIENITZ. 1998. Arctic and Antarctic lakes as optical indicators of global change. Ann. Glaciol. 27: 691–696. WEYHENMEYER, G. A., AND J. KARLSSON. 2009. Nonlinear response of dissolved organic carbon concentrations in boreal lakes to increasing temperatures. Limnol. Oceanogr. 54: 2513–2519, doi:10.4319/lo.2009.54.6_part_2.2513 WILLIAMSON, C. E., D. P. MORRIS, M. L. PACE, AND O. G. OLSON. 1999. Dissolved organic carbon and nutrients as regulators of lake ecosystems: Resurrection of a more integrated paradigm. Limnol. Oceanogr. 44: 795–803, doi:10.4319/lo.1999.44.3_part_2.0795 ———, J. E. SAROS, AND D. W. SCHINDLER. 2009a. Sentinels of change. Science 323: 887–888, doi:10.1126/science.1169443 ———, ———, W. F. VINCENT, AND J. P. SMOL. 2009b. Lakes and reservoirs as sentinels, integrators, and regulators of climate change. Limnol. Oceanogr. 54: 2273–2282, doi:10.4319/ lo.2009.54.6_part_2.2273 ———, R. S. STEMBERGER, D. P. MORRIS, T. M. FROST, AND S. G. PAULSEN. 1996. Ultraviolet radiation in North American lakes: Attenuation estimates from DOC measurements and implications for plankton communities. Limnol. Oceanogr. 41: 1024–1034, doi:10.4319/lo.1996.41.5.1024 WINN, N., C. E. WILLIAMSON, R. ABBITT, K. ROSE, W. RENWICK, M. HENRY, AND J. SAROS. 2009. Modeling dissolved organic carbon in subalpine and alpine lakes with GIS and remote sensing. Landsc. Ecol. 24: 807–816, doi:10.1007/s10980-009-9359-3 WOJTAL-FRANKIEWICZ, A. 2012. The effects of global warming on Daphnia spp. population dynamics: A review. Aquat. Ecol. 46: 37–53, doi:10.1007/s10452-011-9380-x YAN, N. D., W. KELLER, N. M. SCULLY, D. R. S. LEAN, AND P. J. DILLON. 1996. Increased UV-B penetration in a lake owing to drought-induced acidification. Nature 381: 141–143, doi:10.1038/ 381141a0 ZHANG, J., AND OTHERS. 2010. Long-term patterns of dissolved organic carbon in lakes across eastern Canada: Evidence of a pronounced climate effect. Limnol. Oceanogr. 55: 30–42, doi:10.4319/lo.2010.55.1.0030 Associate editor: Dariusz Stramski Received: 02 October 2013 Accepted: 27 January 2014 Amended: 04 February 2014