Ecosystem responses to climate change at a Low Arctic and a High Arctic long-term research site
Long-term measurements of ecological effects of warming are often not statistically significant because of annual variability or signal noise. These are reduced in indicators that filter or reduce the noise around the signal and allow effects of climate warming to emerge. In this way, certain indicators act as medium pass filters integrating the signal over years-to-decades. In the Alaskan Arctic, the 25-year record of warming of air temperature revealed no significant trend, yet environmental and ecological changes prove that warming is affecting the ecosystem. The useful indicators are deep permafrost temperatures, vegetation and shrub biomass, satellite measures of canopy reflectance (NDVI), and chemical measures of soil weathering. In contrast, the 18-year record in the Greenland Arctic revealed an extremely high summer air-warming of 1.3 °C/decade; the cover of some plant species increased while the cover of others decreased. Useful indicators of change are NDVI and the active layer thickness.
KeywordsAlaska Toolik Climate change Ecological effects Greenland Zackenberg Medium pass filter Vegetation
Climate warming in the Arctic, substantial over recent decades and well-documented in IPCC reports (IPCC 2001, 2013), is reflected in changes in a wide range of environmental and ecological measures. These illustrate convincingly that the Arctic is undergoing a system-wide response (ACIA 2005; Hinzman et al. 2005). The changing measures range from physical state variables, such as air temperature, permafrost temperature (Romanovsky et al. 2010), or the depth of seasonal thaw (Goulden et al. 1998), to changes in ecological processes, such as plant growth, which can result in changes in the state of ecosystem components such as plant biomass or changes in ecosystem structure (Chapin et al. 2000; Sturm et al. 2001; Epstein et al. 2004). In spite of the large number of environmental and ecological measurements made over recent decades, it has proven difficult to discover statistically significant trends in these measurements. This difficulty is caused by the high annual and seasonal variability of warming in the air temperature and the complexity of biological interactions.
Ecological settings for Toolik and Zackenberg research sites
Toolik field station
Inland, Northern Alaska 68o38′N, 149o43′W, 719 m altitude
Coast, Northeast Greenland 74o30′N, 21o30′W, 0 m altitude
Rolling foothills, Continuous permafrost (200 m), annual temperature −8 °C, summer (mid-June to mid-August) 9 °C, annual precipitation 312 mm
Mountain valley, Continuous permafrost (estimated 200–400 m), annual temperature −8 °C, summer (3 months) 4.5 °C, annual precipitation 261 mm
Tussock tundra (sedges, evergreen and deciduous shrubs, forbs, mosses, and lichens). Low shrubs, birches, and willows grow between tussocks and along water tracks and stream banks. Low Arctic
Central valley floor dominated by Ericaceous evergreen (Cassiope tetragona), by heaths and arctic willow (Salix arctica)j, and by snow-beds, grasslands, and fens. This High Arctic ecosystem has relatively low biodiversity and low species redundancy
LTER (Long Term Ecological Research), ITEX (International Tundra Experiment), NOAA’s Arctic Program, CALM (Circumpolar Active Layer Monitoring), and the TFS environmental monitoring program
BioBasis programme of NERI, Danish Environmental Protection Agency, CALM (Circumpolar Active Layer Monitoring), ECOGLOBE (Aarhus University), INTERACT, World Wildlife Fund, GeoBasis, NARP
Both sites are underlain by hundreds of meters of continuous permafrost and have similar average annual temperatures of ~ −8 °C. Summers, however, are shorter and cooler at Zackenberg (4.5 °C) than at Toolik (9 °C). The short and cool summers of the Zackenberg valley restrict the number of vascular plant species in the dominant moist heath tundra so this High Arctic site has a relatively low biodiversity (Callaghan 2005; Schmidt et al. 2012). In contrast, the rolling uplands at the Low Arctic Toolik site are dominated by dwarf-shrub heath-tussock tundra and have many more plant species. Bliss (1997) surveyed the North American Arctic, including Greenland, and reported that the High Arctic has 300 species, mostly herbaceous forms, while the Low Arctic has 700 species, including a number of woody species such as birch and willow.
Materials and methods
Environmental and ecological monitoring at Toolik and Zackenberg
Environmental and ecological variables measured over the long-term at Toolik and Zackenberg sites
Environmental and ecological variables
Air temperature, precipitation, wind speed and direction, and growing season dates for 1989–2010 are in Cherry et al. (2014)
Summer depths of thaw for July and August in the Tussock Watershed, 1989–2010, are in Kling et al. (2014)
Net primary production aboveground for moist acidic tundra from 6 harvests 1989–2000 and point-frame data (4 harvests 1989–2008) are in Shaver et al. (2014)
Climatic norms for river basin (1989–2010) and discharge and temperature (1972–2010) are in Bowden et al. (2014)
Primary production and respiration (1984–1998), epilithic chlorophyll (1983–2010), bryophyte cover (1992–2006), benthic insect taxa (1984–1998), and grayling growth (1985–2005) data are in Bowden et al. (2014)
Physics and chemistry
Epilimnion temperature (July, 1985–2007) and summer alkalinity (1975–2011) data are in Luecke et al. (2014)
Chlorophyll (July, 1985–2010) data are in Luecke et al. (2014)
Temperature, 1991–2005, wind direction and speed (1985–2005), and precipitation, 1997–2005, are given in Hansen et al. (2008). Data are available at Greenland Ecosystem Monitoring (http://www.data.g-e-m.dk)
The summer thaw depth progression from June 1 to September 7 at ZEROCALM-2, 1996–2005, is given in Christiansen et al. (2008)
Plant communities and production
Plant communities were analyzed (1997, 2008) in relation to summer temperature and spring snow cover. Five replicate plots in eight plant communities were sampled (Schmidt et al. 2012). NDVI measures (Tagesson et al. 2012) gave gross primary production at the peak of the growing season from 1992 to 2008
Variations and trends in biotic and abiotic ecosystem compartments
Precipitation, temperature, and snow depth measured hourly (1996–2010). Abundance of 6 plant species, 6 taxa of arthropods, 4 species of birds, and 3 mammals measured weekly and seasonally (Mortensen et al. 2014). At 2 lakes, temperature, ice cover, and nutrients were measured (1997–2005) as well as volume of phytoplankton and abundance of zooplankton (Christoffersen et al. 2008)
Methods for data from Toolik
As a part of the international CALM program (Circumpolar Active Layer Monitoring described in Brown et al. 2000), summer thaw depth of the active layer in moist acidic tundra at Toolik was measured using steel probes at 96 individual sites within a 200 × 900 m grid. At each site, three measurements were averaged, and a grand average of all sites was calculated for each of two dates in summers from 1990 to 2011. Additional information on thawing the soil came for measures of alkalinity in Toolik Lake. Alkalinity was determined by potentiometric titration (Kling et al. 1992, 2000) and was averaged across depth and season to provide an annual estimate. Keller et al. (2010) measured strontium isotope ratios (87Sr/86Sr), which decrease with depth in soils at the Arctic LTER, to estimate the increasing depth of water flow within the soil.
Using the point-frame method described by Walker (1996), Gould and Mercado-Díaz (in Shaver et al. 2014) monitored the response of plant communities to ambient climate in 155 permanent plots. Measurements were made at 5- to 7-year intervals since 1989 in two 1 km2 grids set up by Walker et al. (1989) at Toolik Lake and nearby Imnavait Creek. This monitoring was a part of the International Tundra Experiment (ITEX).
Guay et al. (2014) analyzed satellite data to determine annual dynamics of normalized-difference vegetation index (NDVI), a measure of plant productivity, which is also highly correlated with aboveground biomass in arctic systems (Boelman et al. 2003; Raynolds et al. 2012). The NDVI data were derived from the GIMMS-AVHRR times series, version 3 g (Pinzón and Tucker 2014), with a 0.07o (8 km) spatial resolution. We analyzed the GIMMS-3 g dataset across the years 1982–2014 for a 40-km (20 km radius) area surrounding the Toolik Field Station. Seasonal periods of NDVI trends through time were consistent with the seasonal periods used to assess trends in air temperature (see legend for Fig. 3).
Climate trends: Arctic, North Slope of Alaska, Toolik, and Zackenberg
Over the entire Arctic, the average SAT for the past century increased by approximately 0.09 °C per decade; since the mid 1960s that rate has increased to 0.4 °C per decade (ACIA 2005). The North Slope of Alaska has warmed even faster than the rest of the Arctic during the past few decades; Shulski and Wendler (2007) report an increase of more than 3 °C over the past 60 years or 0.5 °C per decade. The coastal town of Barrow, some 310 km northwest of the Toolik site, has warmed significantly (p < 0.01) over the last 60 years with a temperature increase of ~3.1 °C or 0.5 °C per decade (Fig. 2) (Alaska Climate Research Center 2015).
In contrast to the Arctic and North Slope trends, a linear trend analysis of the Toolik datasets revealed no significant trend (p > 0.05) in the ~25-year record of SAT from 1989 to 2010 (Cherry et al. 2014) or in SAT from 1989 to 2014 (Fig. 2). This inability to detect a significant trend (p > 0.05) for these dates also occurred for the Barrow record for the same short period (Fig. 2). The lack of significant warming is also apparent in a closer analysis of the Toolik record for winter, spring, summer, and fall (Fig. 3).
In contrast, the Zackenberg annual air temperatures and the summer temperatures (Figs. 2, 3) show a significant (p < 0.01) warming. Schmidt et al. (2012) report that over the 1997–2008 period, the measured average summer temperature increased dramatically resulting in an increase of between 1.8 and 2.7 °C per decade (p < 0.01), while precipitation data show no significant trends for annual averages or for summer months. To extend the Zackenberg climate database, Hansen et al. (2008) used data from a nearby meteorological station (established in 1958) and from elsewhere in Greenland to create a dataset and calculate a long-term increase in average annual temperature for the period 1901–2005 of 1.39 °C (p < 0.01) and for 1991–2005 of 2.25 °C (p < 0.01); they mention that these trends are similar to trends from other studies along the east coast of Greenland.
From Zackenberg, there are no permafrost temperature data below 1.3 m (Christiansen et al. 2008).
Changes in depth of active layer thaw
Direct measure of depth of thaw with steel probes
The Zackenberg data, in contrast, show a significant increase (p < 0.01) in the maximum depth of thaw in a 10-year record at ZEROCALM-1 (Christiansen et al. 2008) which varied slightly from 60 to 65 cm in the first 5 years and then increased steadily from 60 to 79 cm over the last 5 years in response to the significant increase in summer temperatures (Fig. 3).
Indirect measures of depth of thaw: Chemical measures of soil weathering
The weathering and water movement in the soil that led to both the increase in alkalinity and the decrease in strontium isotope ratios also integrate the chemical signal over several years. This integration occurs because some of the alkalinity that is produced in one year remains in the soil water at the end of the summer and is not released until the thaw of the active layer the next summer. For example, Everett et al. (1996) measured the Ca2+ in soil water for 22 days in August and found an average of 31.4 µEq L−1 in overland flow (n = 3), 79.8 at 20 cm depth (n = 21), and 112 µEq L−1 at 40 cm (n = 21). Rainfall each fall ensured that the active layer was saturated at the beginning of each winter (Hinzman et al. 1996). The next spring, most of the runoff from the watershed occurred from snowmelt in the spring as surficial runoff when the active layer was still frozen (Woo and Steer 1983). The ions that are a part of the soil water are not released until the thaw depth deepens later in the summer (Cornwell 1992).
At Zackenberg (Christiansen et al. 2008), twenty lakes showed no change in chemical conductivity when monitored twice (1997 and 2003). Two of these lakes also showed no changes when monitored every year from 1997 to 2003. It is not known if weathering of the previously frozen soil would show alkalinity and isotopic changes in the Zackenberg stream and lake watersheds in the same way as soils at Toolik.
Relative species abundance and composition of tundra vegetation
The ITEX protocol was also used twice at Zackenberg to measure changes in the eight dominant plant communities from 1997 to 2008 (Schmidt et al. 2012). Each community had four replicate sampling plots. In contrast to the Toolik results, there were significant reductions of up to 55% in the cover of grasses and lichens across all plant communities. Yet, some species and groups, including the willow (Salix arctica), exhibited only minor changes during this period. The interpretations suggested for Zackenberg by Schmidt et al. (2012) for point-frame analysis and Campioli et al. (2013) for heating experiments are that some of the reductions may be due to the lower sensitivity of High Arctic plant communities to warming than those in the Low Arctic or High Arctic communities could even be resistant to climate change. However, a complicating factor was reduced availability of water during the summers caused by deepening of the active layer. In addition, there was little sign of the marked expansion of shrubs found in most of the Low Arctic (Walker et al. 2006) but musk oxen grazing (Forchhammer et al. 2005) and the relatively short period of observations might make it difficult to measure any expansion.
NDVI measures of plant biomass
NDVI for the Toolik region in northern Alaska
The NDVI for the Toolik region has also been analyzed at much finer scales by Raynolds et al. (2013) who used six scenes from Landsat 4 or more-recent sensors (1985–2007) showing the annual peak NDVI as measured at a 30-m pixel resolution over an 823 km2 area. They analyzed changes in 14 types of vegetation and found that nearly all the patches showed either no increase or a small increase in NDVI; in fact, sizeable increases in NDVI were found only in tussock tundra, non-tussock-sedge tundra, and acidic dwarf-shrub tundra, the latter making up only 5% of the pixels. Thus, the increase in NDVI evident at a coarser scale (Fig. 9) was also present at the finer scale but was heterogeneously distributed. Further comparisons between the AVHRR (Fig. 9) and the Landsat values (Raynolds et al. 2013) are difficult because NDVI values measured with different sensors and at different levels of resolution and types of rectification may be quite different (Goetz 1997).
The changes in NDVI (Fig. 9) indicate a regional increase in vegetation photosynthetic activity and aboveground plant biomass. The plot measurements of plant and leaf biomass at the Toolik site (Fig. 8) indicate that this biomass increase is largely the result of increased growth by deciduous shrubs (e.g., dwarf birch, willows, and alder) in response to multi-year warming, but this response is shared with graminoids and forbs. Several researchers attribute the slow increase in biomass to a slow increase in the availability of N to plants (Shaver et al. 1992, 2014; Pearce et al. 2015; Jiang et al. 2015). It is well known through warming and fertilization experiments that the N supply strongly limits plant growth in northern Alaska and that warming increases the microbial mineralization of organic nitrogen in the soil, the major source of N to plants in the tundra.
NDVI for the Zackenberg region in Greenland
At Zackenberg (Tagesson et al. 2012), the annual maximum NDVI increased from 0.35 to 0.61 between 1992 and 2004, an increase of 74%, before dipping to 0.49 in 2005 and returning to 0.57 in 2007 and 2008. The authors suggest that this dip in the NDVI could have been caused by a one-year change in the satellite sensor. In any case, the increase in NDVI at Zackenberg is consistent with other studies of NDVI trends in the Arctic, including the Toolik data (Fig. 9), that interpret the trend as an increase in both greening and plant photosynthesis (see also Jia et al. 2003; Verbyla 2008). However, the rates of change of the Zackenberg NDVI are much higher (0.02/year) than rates in other studies (0.003–0.006/year). Tagesson et al. (2012) suggested the high rates may be due to the almost complete plant cover in the 1.4 km2 study area as compared with other arctic study areas that typically include a large fraction of gravel, rock, and water.
As with the permafrost temperatures at 20 m, vegetation responses to temperature are integrated over several years and thereby damp out the hourly, daily, and seasonal variation in the noisy air temperature signal. Plants accumulate and retain biomass and nutrients from year-to-year and decade-to-decade. Thus, the effects of an unusually cold or warm summer are damped out but the long-term trend in temperature is preserved in long-lived tissues such as rhizomes, woody stems, and branches. Results of small but persistent changes, like the warming signal, accumulate over decades and become detectable because these long-lived tissues integrate effects over many years and are not reset to a base level each year.
Ecosystem and biota temporal trends, Zackenberg
As noted in the description of Zackenberg ecosystem research, one type of long-term study is the detailed analysis of changes across the entire aquatic and terrestrial food webs. In contrast, at the Toolik site, the long-term studies are only of selected parts of the entire ecosystem, for example, the point-frame measures of changes in the plant community types described above.
From 1996 to 2010, studies were made of trends and temporal variations across the whole aquatic and terrestrial ecosystem. In the terrestrial ecosystem, 6 plant species, 6 taxa of arthropods, 4 species of birds, and 3 species of mammals were measured (Mortensen et al. 2014). Plants were sampled weekly for such factors as abundance, flowering dates, and emergence, arthropods were sampled in traps, and mammals by census. Of the biotic variables, 39% exhibited significant linear trends; of the significant trends in the abundance of terrestrial biota, 12 had a positive slope and 22 had a negative slope. The Zackenberg data also revealed (Høye et al. 2013) that warmer summer temperatures led to a shortening of the season of plant flowering and likely to a decline in the abundance of the insects that visit these plant flowers for basking, nectar and pollen feeding, mating, and ovipositioning. For the aquatic ecosystems, both the species and abundance of phytoplankton and zooplankton responded strongly to water temperatures that varied from 4 to 11 °C from one summer to the next. One lake was dominated by chrysophytes in a warm summer (95% of abundance) and by dinophytes (96% of abundance) in a cold summer.
Filtering long-term data to determine environmental trends
Low Arctic Toolik
Toolik and Zackenberg long-term monitoring results
Useful for long-term series (decades) of moderate warming
Useful for shorter-term series(<decades) in case of intense warming
No significant trend because of short record and high annual variability
Significant trend related to large annual warming
Active layer thickness
Is reset every year so closely related to summer temperature
No long-term significant trend in thickness
Changes in surface water alkalinity are evidence of deeper penetration of water into soil previously permanently frozen
Doubling of alkalinity in lakes is evidence that water is penetrating deeper into soil as active layer increases
No change in surface water chemistry
Permafrost temperature at 20 m
As temperature signal moves deeper into the soil, the high-frequency noise is filtered out
Increase of 0.8 °C over long-term
No measures made at depths >1.2 m
Plant and animal abundance with point-frame and species counts
Abundance of individual terrestrial and aquatic species or of communities
Shrubs, graminoids increase
Some species increase, most decrease but grasses and lichens decrease
NDVI satellite measure of peak greenness at annual or seasonal periods
Estimate of peak annual or seasonal biomass or photosynthesis. Integrates over various areas
Steady increase 1982–2014
Increase 1992–2004, instrument changes to 2008
The permafrost record of temperature at 20 m depth in the soil (Fig. 4) is an excellent example of naturally occurring moving average or running mean, and no manipulation of the data is needed. As Lachenbruch and Marshall (1986) point out, as the temperature signal moves deeper into the soil, the high-frequency noise is progressively filtered out. However, as previously noted, some of the change in snow depth will also affect soil temperatures so that permafrost temperature change reflects the temperature of the upper layers of the soil and not just the air temperature during climate change (Stieglitz et al. 2003).
Another way that the natural environmental measures act as filters to reduce the variability from year to year is by accumulation of the effects. That is, some of the environmental effects of one year are stored and affect the data for the next year or years. The plants that are most useful for tracing warming (Figs. 8, 9) have large amounts of biomass in their stems, branches, and roots; these tissues persist from year to year such that small warming-induced increments in biomass accumulate over time. The plants also translocate and store nutrients and carbohydrates over winter for use in leaf production the next spring, so that resources for growth that are acquired in one-year affect subsequent year’s growth. In much the same way, the alkalinity that is added to soil water as a product of weathering is not completely flushed out at the end of the summer and may persist in the soil for years before it is transported to lakes (Fig. 6).
A part of understanding the concept of environmental and ecological filters is the notion that some measurements are not useful as indicators of change because they are reset to the same level every year. For example, the depth of thaw of the soil at Toolik is measured at its summer peak (Table 2; Fig. 5) but is reset to zero centimeters each winter when the entire active layer is frozen. Similarly, summer temperatures in lakes, chlorophyll content of lake plankton, or temperature of the Kuparuk River are all reset at the start of each year.
High Arctic Zackenberg
These insights about environmental filters developed from the Low Arctic site at Toolik do not really agree with the results of long-term environmental studies at the High Arctic site at Zackenberg. Out of all the ecological changes that occurred at this High Arctic site only one, the change in canopy reflectance or NDVI, appeared to be the result of integration of a climate change effect from one year to another. This integration, which occurs because plants retain biomass, nutrients, and carbohydrates from one year to another, decreases signal noise and acts as a medium pass filter.
A possible reason for the difference between these sites is the consistent and high rate of warming occurring during the period of research at Zackenberg when the summer average temperatures increased between 1.8 and 2.7 °C/decade (p < 0.01) (Schmidt et al. 2012). At Toolik, there was no significant trend of summer and of annual average temperatures and at least summer temperatures were more variable (Fig. 3). However, at Zackenberg, the dramatic rise in the summer temperatures has increased the thickness of the active layer, increased the overall plant biomass or productivity as measured by NDVI, and increased numbers or biomass of some of the biota while decreasing the numbers or biomass of others. For example, at Zackenberg, there was a large reduction of the biomass of grasses and lichens but no increase in shrubs. Evidently some plants and plant communities are sensitive to change while others have a high degree of resistance.
An overall warming trend does occur in northern Alaska where air temperatures at Barrow, the only permanent town, have warmed ~3.1 °C in the last 65 years. However, at the Toolik site, the 25-year record of annual mean air temperatures revealed no significant trend due to the high variability from year to year and the relatively short period of record. In spite of the temperature variability and the absence of a warming trend, three of the Toolik measurements did produce evidence of change: a warming of permafrost temperatures at a depth of 20 m, an increase of plant biomass as measured on the ground and by means of satellite NDVI, and a change in surface water chemistry indicating an increase in weathering of previously frozen soil. These indicators have the common feature of integrating effects of the climate signal on multi-year to decadal time scales. They act as a medium pass filter that reduces the signal variability yet allows the effects of long-term warming to emerge within the 20- to 30-year dataset. Based on the indicators that have passed through the medium pass filter, we conclude that there has been a measureable response of the Toolik ecosystem to arctic warming, even though there has been no statistically significant warming trend in the annual air temperature.
The concept of the medium pass filter, developed from the Low Arctic Toolik dataset, has only limited application for analyzing results from the High Arctic at Zackenberg in northern Greenland. At this site, the summer temperatures are increasing at an unusually high rate. Some results of the environmental measurements, such as the thickness of the active layer and abundance of some plants, increase in parallel with the warming temperatures. Others, such as biomass of grasses and lichens, have dramatically decreased. The NDVI measure does increase in a similar manner in the High Arctic and Low Arctic sites, and it is likely that there is a medium pass filter at work in which some integration is achieved because of year-to-year carryover and reuse of plant biomass, nutrients, and carbohydrates. It would be interesting to see results of temperature measures at 20 m depth where the soil acts to filter out the year-to-year variability.
We thank Amy Jacobs, at the University of Alaska, Fairbanks, for assistance with climate statistics and Bonnie Kwiatkowski of the Ecosystems Center, Woods Hole, Massachusetts, for help with figures. The Toolik research was supported in part by NSF Grants DEB 0207150, DEB 1026843, ARC 1107701, and ARC 1504006.
- ACIA. 2005. Arctic climate impact assessment. Cambridge: Cambridge University Press.Google Scholar
- Alaska Climate Research Center. Retrieved 1 April 2015, from http://climate.gi.alaska.edu.
- Bliss, L.C. 1997. Arctic ecosystems of North America. In Polar and alpine tundra, ed. F.E. Wiegolawski, 551–683. Amsterdam: Elsevier.Google Scholar
- Bowden, W.B., B.J. Peterson, L.A. Deegan, A.D. Huryn, J.P. Bensted, H. Golden, M. Kendrick, S.M. Parker, et al. 2014. Ecology of streams of the Toolik region. In Alaska’s changing arctic: Ecological consequences for tundra, streams, and lakes, ed. J.E. Hobbie, and G.W. Kling, 173–258. New York: Oxford University Press.CrossRefGoogle Scholar
- Callaghan, T.V. 2005. Arctic tundra and polar desert ecosystems. In Arctic climate impact assessment, ed. C. Symon, 243–353. Cambridge: Cambridge University Press.Google Scholar
- Chatfield, C. 2004. The analysis of time series: An introduction, 6th ed. Boca Raton: CFC Press.Google Scholar
- Cherry, J.E., S.J. Déry, Y. Cheng, M. Stieglitz, A.S. Jacobs, and F. Pan. 2014. Climate and hydrometeorology of the Toolik Lake region and the Kuparuk River Basin: Past, present, and future. In Alaska’s changing arctic: Ecological consequences for tundra, streams, and lakes, ed. J.E. Hobbie, and G.W. Kling, 21–60. New York: Oxford University Press.CrossRefGoogle Scholar
- Elmendorf, S.C., G.H.R. Henry, R.D. Hollister, R.G. Bjork, N. Boulanger-Lapointe, E.J. Cooper, J.H.C. Cornelissen, T.A. Day, et al. 2012. Plot-scale evidence of tundra vegetation change and links to recent summer warming. Nature Climate Change 2: 453–457. doi:10.1038/nclimate1465.CrossRefGoogle Scholar
- Hobbie, J.E., G.R. Shaver, J. Laundre, K. Slavik, L.A. Deegan, J. O‘Brien, S. Oberbauer, and S. MacIntyre. 2003. Climate forcing at the arctic LTER site. In Climate variability and ecosystem response at long-term ecological research (LTER) sites, ed. D. Greenland, 74–91. New York: Oxford University Press.Google Scholar
- IPCC. 2001. In Climate change 2001: Synthesis report. A contribution of working groups I, II, and III to the third assessment report of the intergovernmental panel on climate change, ed. R.T. Watson, Core Writing Team. Cambridge and New York: Cambridge University Press.Google Scholar
- IPCC. 2013. Summary for policymakers. In Climate change 2013: The physical science basis. Contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change, ed. T.F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex, and P.M. Midgley. Cambridge and New York: Cambridge University Press.Google Scholar
- Kling, G.W., H.E. Adams, N.D. Bettez, W.B. Bowden, B.C. Crump, A.E. Giblin, K.E. Judd, K. Keller, et al. 2014. Land–water interactions. In Alaska’s changing arctic: Ecological consequences for tundra, streams, and lakes, ed. J.E. Hobbie, and G.W. Kling, 143–172. New York: Oxford University Press.CrossRefGoogle Scholar
- Luecke, C., A.E. Giblin, N.D. Bettez, G.D. Burkart, B.C. Crump, A.M. Evans, C. Gettel, S. MacIntyre, et al. 2014. The response of lakes near the Arctic LTER to environmental change. In Alaska’s changing Arctic: Ecological consequences for tundra, streams, and lakes, ed. J.E. Hobbie, and G.W. Kling, 238–286. New York: Oxford University Press.CrossRefGoogle Scholar
- Mercado-Díaz, J.A. 2011. Plant community responses of the Alaskan Arctic tundra to environmental and experimental changes in climate. MSc Thesis. University of Puerto Rico, Río Piedras.Google Scholar
- Shaver, G.S., J.A. Laundre, M.S. Bret-Harte, F.S. Chapin, J.A. Mercado-Díaz, A.E. Giblin, L. Gough, W.A. Gould, et al. 2014. Terrestrial ecosystems at Toolik Lake, Alaska. In Alaska’s changing Arctic: Ecological consequences for tundra, streams, and lakes, ed. J.E. Hobbie, and G.W. Kling, 90–142. New York: Oxford University Press.CrossRefGoogle Scholar
- Shiklomanov, N.I., D.A. Streletskiy, F.E. Nelson, R.D. Hollister, V.E. Romanovsky, C.E. Tweedie, J.G. Bockheim, and J. Brown. 2010. Decadal variations of active-layer thickness in moisture-controlled landscapes, Barrow, Alaska. Journal of Geophysical Research 115: G00I04. doi:10.1029/2009JG001248.CrossRefGoogle Scholar
- Shulski, M., and G. Wendler. 2007. The climate of Alaska. Fairbanks: University of Alaska Press.Google Scholar
- Tagesson, T., M. Mastepanov, M.P. Tamstorf, L. Eklundh, P. Schubert, A. Ekberg, C. Sigsgaard, T.R. Christensen, et al. 2012. High-resolution satellites reveal an increase in peak growing season gross primary production in a high-Arctic wet tundra ecosystem 1992-2008. International Journal of Applied Earth Observation and Geoinformation 18: 407–416.CrossRefGoogle Scholar
- Walker, D.A., E. Binnian, B.M. Evans, N.D. Lederer, E. Nordstrand, and P.J. Webber. 1989. Terrain, vegetation and landscape evolution of the R4D research site, Brooks Range foothills, Alaska. Holarctic Ecology 12: 238–261.Google Scholar
- Walker, M.D. 1996. Community baseline measurements for ITEX studies. In ITEX manual, 2nd ed, ed. U. Molou, and P. Molgaard, 39–41. Copenhagen: Danish Polar Centre.Google Scholar
- Walker, M.D., C.H. Wahren, R.D. Hollister, G.H.R. Henry, L.E. Ahlquist, J.M. Alatalo, M.S. Bret-Harte, M.P. Calef, et al. 2006. Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences of the United States of America 103: 1342–1346.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.