Plant and Soil

, Volume 424, Issue 1–2, pp 145–156 | Cite as

Autumnal warming does not change root phenology in two contrasting vegetation types of subarctic tundra

  • Sarah SchwiegerEmail author
  • Jürgen Kreyling
  • Ann Milbau
  • Gesche Blume-Werry
Regular Article


Background and aims

Root phenology is important in controlling carbon and nutrient fluxes in terrestrial ecosystems, yet, remains largely unexplored, especially in the Arctic. We compared below- and aboveground phenology and ending of the growing season in two contrasting vegetation types of subarctic tundra: heath and meadow, and their response to experimental warming in autumn.


Root phenology was measured in-situ with minirhizotrons and compared with aboveground phenology assessed with repeat digital photography.


The end of the growing season, both below- and aboveground, was similar in meadow and heath and the belowground growing season ended later than aboveground in the two vegetation types. Root growth was higher and less equally distributed over time in meadow compared to heath. The warming treatment increased air and soil temperature by 0.5 °C and slightly increased aboveground greenness, but did not affect root growth or prolong the below- and aboveground growing season in either of the vegetation types.


These results imply that vegetation types differ in root dynamics and suggest that other factors than temperature control autumnal root growth in these ecosystems. Further investigations of root phenology will help to identify those drivers, in which including responses of functionally contrasting vegetation types will help to estimate how climate change affects belowground processes and their roles in ecosystem function.


Belowground Climate change Fine roots Plant phenology Root growth Subarctic tundra 





carbon dioxide


Growing Degree Days








open top chambers




plant functional type


photosynthetic photon flux density


warming treatment


Plant phenology (i.e. the timing of recurring life history events) has been shown to be sensitive to changes in temperature and is often considered as one of the primary indicators of climate change (Linderholm 2006; Barichivich et al. 2013; Richardson et al. 2013; Keenan and Richardson 2015; Radville et al. 2016a). Northern ecosystems (> 45°N) experienced and will experience substantially greater than average climate warming with annual average temperatures having increased almost twice the rate of those of the rest of the world (ACIA 2004). As a consequence, changes in phenology and growing season length are expected to be more pronounced at high latitudes and altitudes, indeed, an intensive greening connected with longer growing seasons and greater photosynthetic activity is observed in northern areas (Zhou et al. 2001; Marchand et al. 2004; Körner 2007; Piao et al. 2008; Pau et al. 2011).

Until now, growing season length and phenology are mainly assessed aboveground (but see Steinaker and Wilson 2008; Blume-Werry et al. 2015; Sloan et al. 2016). However, over 80% of plant biomass in northern ecosystems is located belowground (Pendall et al. 2004; Sloan et al. 2013; Iversen et al. 2015; Radville et al. 2016b). In fact, belowground processes play a key role in ecosystem function, and in particular, fine roots (defined as ≤2 mm in diameter) fulfil important functions by controlling a dominant flux of carbon from plants into soils, and mediating potential uptake and cycling of nutrients and water in terrestrial ecosystems (Pendall et al. 2004; McCormack et al. 2013; Iversen et al. 2015). This is of especial importance in tundra ecosystems, as they contain large stores of soil carbon and therefore have the potential to provide a large positive feedback to climate change through its release (Mack et al. 2004). Despite the importance of fine roots in biogeochemical cycles, little is known about what controls fine root growth and function in cold climate ecosystems. Consequently, we have a very poor knowledge about how a future warmer climate will affect fine root processes, such as growth, mortality, and phenology. Gaining knowledge of belowground phenological responses to climate change is crucial to understand the widespread consequences for trophic interactions, ecosystem functions, and the ability to predict the shape of future plant communities (Pau et al. 2011; McCormack et al. 2013).

It is known that rates of root growth and root mortality generally increase with increasing soil temperatures (Pregitzer et al. 2000; Pendall et al. 2004; McCormack et al. 2013). Changes in temperature accompanied with longer periods with temperatures favourable for root growth will thus most likely alter timing and rates of root turnover and thus be responsible for changes in soil carbon storage, water and nutrient fluxes (McCormack et al. 2013). Thereby, the ‘shoulder seasons’ (i.e. (late) spring and (early) autumn) will be of special importance because most warming happens outside the growing season (Post et al. 2009; Ernakovich et al. 2014).

For instance, surface air temperatures have increased by 0.98 °C in autumn compared to an increase of about 0.29 °C in summer over the last hundred years at arctic latitudes (65°-90°N; ACIA 2004; IPCC 2013; Johannessen et al. 2016). Both, the sensitivity of roots for temperature and the rising autumn temperatures support the idea that rising temperatures potentially contribute to a prolonging of the belowground growing season. In fact, belowground production has been shown to be asynchronous with aboveground production in tundra ecosystems with substantial root growth occurring after aboveground senescence (Blume-Werry et al. 2015; Radville et al. 2016b; Sloan et al. 2016). However, whether this becomes even more pronounced with warming temperatures is unclear, as there are still fundamental knowledge gaps on how root dynamics (production, lifespan and turnover) in tundra ecosystems respond to changing environmental conditions, such as warming (Iversen et al. 2015).

Subarctic tundra in Fennoscandia is characterized by two co-dominant vegetation types, heath and meadow (Björk et al. 2007). The heath is dominated by woody dwarf-shrubs and occurs on acidic, less fertile soils, with limited biomass production (Björk et al. 2007; Sundqvist et al. 2011). In contrast, the meadow, mainly comprised of herbs, graminoids and sedges, represents a more productive system with higher pH and lower C:N ratios than heath (Björk et al. 2007). It has been shown that those two contrasting vegetation types not only differ in their current state but also respond differently to changing temperatures, both below- and aboveground, where meadow was more responsive to changes in elevations, and therefore temperature, than heath (ACIA 2004; Sundqvist et al. 2011). For example, meadow showed higher turnover rates of PFTs with decreasing temperature, while on the other hand a decrease in temperature caused a stronger decrease in mineralisation rates and plant nutrient availability in the heath compared to the meadow (Sundqvist et al. 2011). Consequently, effects of temperature on community and ecosystem properties in arctic tundra, such as biodiversity, biomass productivity and nutrient availability, may be determined by the type of dominant vegetation present (Sundqvist et al. 2011). However, few studies thus far have addressed differences in root growth patterns among different vegetation types of the same area, or their responses to increases in temperature. Therefore, how changes in plant community composition manifest belowground, and the consequences for ecosystem function are still largely left unanswered (Iversen et al. 2015). Sloan et al. (2016) found considerable differences in timing of root production among contrasting plant communities in a subarctic region. Compared to woody plant communities, sedge-dominated vegetation showed earlier cessation of growth (Sloan et al. 2016). It is likely that vegetation types differ in their root phenology, but they might also differ in their response to warmer autumn temperatures, for instance because the role of photoperiod as a control of root cessation differed between arctic species (Radville et al. 2016a, b).

In this study, we compared belowground autumn phenology of heath and meadow, i.e. the two dominant vegetation types in the subarctic tundra in northern Sweden. Further, we tested the consequences of higher autumnal temperatures by increasing soil and air temperature in the field using horticultural fleece covers. In-situ measurements of root phenology were performed using the minirhizotron technique and compared with aboveground measurements of senescence using repeat digital photography for estimates of greenness.

Specifically, we hypothesized that:
  1. (i)

    below- and aboveground autumn phenology differ between meadow and heath due to differences in community composition

  2. (ii)

    autumnal warming will prolong the below- and aboveground growing season in both vegetation types

  3. (iii)

    the prolongation of the growing season with warming will be more pronounced in the meadow than in the heath vegetation, as meadow has previously been shown to be more responsive to environmental changes.


Material and methods

Study area

The study took place at Mount Nuolja close to the Abisko Scientific Research Station (Abisko Naturvetenskapliga Station, ANS) in northern Sweden (68°21′N, 18°49′E). The region has a subarctic climate, characterised by long, usually very cold winters with an average temperature of −11.3 °C and an average precipitation of 8.7 mm in January. The area has short, cool to mild summers with an average July temperature of 10.4 °C and a precipitation of 93.8 mm (average of 30 year mean temperatures of 1971–2000; ANS weather data). Plots were located above the tree line, at approximately 900 m a.s.l. The two vegetation types, heath (H) and meadow (M) are interspersed at the study site.

A survey of plant functional types (PFTs) was conducted in the end of July, 2015 using the method of Braun-Blanquet (Braun-Blanquet 1964). Plant species were classified into the following PFTs as described by Chapin et al. (1996): forbs (herbaceous flowering plants, e.g. Ranunculus nivalis, Bistorta vivipara, Viola biflora, Trollius europaeus), evergreen shrubs (e.g. Empetrum hermaphroditum, Vaccinium vitis-idaea), deciduous shrubs (e.g. Salix reticulata, Vaccinium myrtillus, Vaccinium uliginosum, Betula nana), graminoids (e.g. Carex spp.) and bryophytes (Table S1).

Warming treatment

In total, the study site consisted of 15 paired plots (1 m × 1 m) each with a control (C) and a warming treatment (W). Eight plots were placed in heath vegetation, while seven plots were located in meadow vegetation, because the previously installed minirhizotron tubes in the eighth meadow plot was not accessible and partially destroyed and could not be used for the experiment. The heath and meadow plots were distributed randomly across the field site. In the warming treatment, the vegetation was covered with two layers of a horticultural fleece (1 m × 1 m, polypropylene, 17 g m−2, HP Johannesson Trading AB) to create an insulation. The use of fleece was chosen because other remote methods, such as open top chambers (OTCs), usually do not increase air temperatures significantly at the end of the growing season from August onwards at this latitude (e.g., Marion et al. 1997; Dorrepaal et al. 2004). To measure the amount of shading caused by the fleece with that of commonly used OTCs and control plots, the photosynthetic photon flux density (PPFD) in μmol m−2 s−1 was measured with a PAR-sensor (QSO-S PAR Photon Flux sensor; Decagon Devices Inc., Pullman, WA, USA). Light intensity was measured on a clear day, on a partly cloudy day and on an overcast rainy day under ambient conditions, with two layers of fleece, and two types of OTCs (type 1: transparent polycarbonate, 3.0 mm thickness, Sun-Lite HP®, Solar Components Corp., Manchester, NH, USA; type 2: fibreglass 1.14 mm thickness, ePlastics®, Ridout Plastics Co. Inc., San Diego, CA, USA). The PAR measurements showed that the PPFD measured under the two layers of fleece was on average 14% less than the values of two types of open top chambers under clear sky conditions, but only a slight difference was measured under (partly-) cloudy conditions (see Fig. S1).

Air and soil temperatures

To measure soil and air temperature, iButtons from the type Thermochron (DS1920 and DS1921G iButton from Maxim) were used. iButtons were placed at 5 cm depth in the soil in each control and warming plot (N = 30). Additionally, iButtons were placed under small radiation shields at ca. 5 cm above ground (directly under the fleece in warming plots) and randomly distributed in one meadow and two heath plots in both the control and the warming treatment (N = 6). The radiation shields were manufactured by 3D LAB (Abisko, Sweden) using a 3D–Printer (M2 from MakerGear; MakerGear™ LLC, Beachwood, OH, USA). Temperature data was collected every hour for both soil (beginning of August until mid-October) and air (mid-August until mid-October).

To better describe the relationship between temperature and root growth, Growing Degree Days (GDD) were estimated with the following equation:
$$ GDD=\left[\frac{\left( Tmax+ Tmin\right)}{2}\right]- Tbase $$
where Tmax is the daily maximum soil temperature, Tmin is the daily minimum soil temperature, and Tbase is the base temperature below which the process of interest, in this case root growth, does not occur (McMaster and Wilhelm 1997). Tbase was set to 2 °C and 5 °C, as root growth is commonly believed to cease at soil temperatures below 5 °C, but has also been observed at temperatures between 0 °C and 5 °C, in some species of arctic ecosystems (Shaver and Billings 1977; Bell and Bliss 1978; Chapin 1978, 1983). In case the mean of Tmax and Tmin was below the value of Tbase the GDD was set to zero.

Aboveground phenology

To estimate the start of aboveground plant senescence, pictures of the aboveground vegetation were taken with a digital camera (Canon EOS 350 D; Canon Inc., Tokyo, Japan) from each plot and treatment every 10 days on the same days that belowground phenology was measured from the beginning of August until the middle of September. A diffusion tent (Cubelite®, Lastolite™ Professional, 90 × 90 cm; Leicestershire, UK) was used to avoid reflections or shading. A grey-card was used to adjust white-balance and colour and hold them constant between plots and sampling events. Downward-looking digital photographs were made of each plot, always of the same permanently marked area (40 × 40 cm), the same height (1 m) and with the same aperture. Greenness was defined by calculating the percentage of green pixels in each plot using Adobe Photoshop CS5 (Adobe Systems Inc., San Jose, CA, USA). Greenness increase within each sampling interval was used as aboveground production, while a loss of greenness (the difference in greenness in % between the sampling events became negative) was considered as the start of aboveground senescence and thus the end of aboveground-growing season. Lichens and bryophytes were excluded in the greenness measurement.

Belowground phenology

To estimate root growth and productivity in-situ, minirhizotrons (Bartz Technology Corporation, Carpinteria, CA, USA) were used. These provide a non-destructive method for directly studying roots (Johnson et al. 2001). In each plot two transparent minirhizotron observation tubes (cellulose acetate butyrate, 90 cm long, 28.5 mm internal diameter) were installed horizontally in the soil at approximately 10 cm depth, where the majority of fine roots are situated (Iversen et al. 2015). A camera system (BTC-100×, Bartz Technology Corporation, Carpinteria, CA, USA), which can be inserted into the tubes, was used for taking pictures of the roots. Both ends of each tube were covered with a cap to exclude light and prevent soil, mesofauna and roots from entering the tubes. Tubes were installed in 2011 giving roots enough time to recolonize (Johnson et al. 2001).

Sampling was done every 10 days between the 4th of August and the 14th of September in 2015. An additional sampling was conducted on the12th of October in 2015 to examine if root growth continued after the expected growing season. At every sampling round, 44 images (frame size: 14 mm × 18 mm) per tube were collected. Data analysis (i.e. fine root production and mortality) was performed using the program Rootfly, version 2.0.2 (Birchfield & Wells, Clemson University, Clemson, SC, USA). The root length (in mm) as well as the diameter (in mm) of all existing roots were measured on the images. Root production was defined as the sum of the length of new roots formed and any increase in length of existing roots since the previous sampling event (Johnson et al. 2001).

Statistical analysis

To test for differences in root growth (in mm) and greenness (in %) between the two vegetation types (H, M), the two treatments (W, C) and their interactions, a linear mixed-effects model ANOVA (R packages ‘lme4’ and ‘lmerTest’) was used with treatment and vegetation as interacting explanatory variables and block (i.e. pairs of control and warming) as a random factor. To estimate differences in root growth and greenness for the vegetation types and the treatments over time, a repeated measure ANOVA was conducted with the same interacting explanatory variables, but adding time as a random factor. To test the effectiveness of the warming treatment for air and soil temperatures a linear mixed-effects model ANOVA was used with time as random factor. In the same way it was tested for significant differences in soil and air temperatures between the two vegetation types. Statistical significance was accepted when the probability of the result assuming null hypothesis (P) is less than 0.05.

All data was analysed graphically (R packages ‘sciplot’ and ‘ggplot2’) to test the assumptions of normality and homogeneity of variance. Thus, the residuals were plotted with normal q-q-plots and residual vs. fitted plots of the model. The data was square root-transformed if necessary to fit the assumptions. A simple linear least squares regression and a Pearson correlation were used to test if root growth can be predicted by GDD. Due to the fact that the sampling intervals were not constant over the study, the data was normalised by dividing root growth and GDD for the single sample rounds by the amount of days it included. All statistical analyses were performed using R version 3.0.2 (R Core Team 2013, Vienna, Austria).


Phenology of vegetation below- and aboveground

Root growth varied significantly over time, but the pattern depended on the vegetation type (ANOVA interaction vegetation x time: P = 0.003). In both meadow and heath, root growth continued after the onset of aboveground senescence (Fig. 2b, d). Meadow vegetation showed a peak in root production shortly after the decline in greenness, whereas the root phenology in the heath vegetation was more evenly distributed and showed no peak during the time of measurements (Fig. 2c, d).

Aboveground greenness averaged over the sampling period was approximately 3% higher in the heath than in the meadow (F1,68 = 4.2, P = 0.045), whereas total root growth was significantly higher in the meadow with 206.55 mm compared to 143.95 mm in the heath (F1,69 = 16.4, P < 0.001). The greenness of aboveground vegetation varied significantly over time (F1,3 = 9.2, P = 0.05) and declined after a peak in mid-August for all treatments and vegetation types (ANOVA interaction vegetation x time: P = 0.086; treatment x time: P = 0.096; Fig. 2a b), indicating the onset of aboveground senescence in mid-August.

The warming treatment showed a significant effect on the greenness of the aboveground vegetation (F1,72 = 9.3, P = 0.0032) with an approximately 4.4% higher greenness on the warm plots than on the control plots. However, the warming treatment showed no significant effect on total root growth over the whole sampling period, averaged across the 15 paired plots (F1,73 = 0.2, P = 0.657).

The two vegetation types did not differ significantly in their responsiveness to the warming treatment, neither belowground (ANOVA interaction treatment x vegetation: P = 0.970) nor aboveground (ANOVA interaction treatment x vegetation: P = 0.257).

Simple linear least squares regression revealed a positive relationship between GDD and root growth, with fine root growth increasing with increasing GDD (Fig. 3). GDD significantly predicted root growth and explained a significant proportion of variance in root growth values (R2 = 0.51, F1,148 = 153, P < 0.001).

Warming treatment

Air temperature

The warming treatment increased maximum air temperatures by 1.7 °C (Table 1, P < 0.001), did not change mean air temperatures (P = 0.445), and reduced minimum air temperatures by 0.2 °C (P < 0.001). The treatment effect varied over time (Table 1, P < 0.001), with the highest variation in the meadow (Fig. 1a, b), in which temperature differences between control and warming ranged from −2.3 °C to 2.9 °C, compared to −1.5 °C to 2.7 °C in the heath plots. Air temperatures dropped below freezing repeatedly during September and in the beginning of October (Fig. 1a, b).
Table 1

Results of the repeated measure ANOVA for air- and soil temperature with treatment as interacting explanatory variable and time as random factor

temperature in °C


treatment (df = 1)

treatment x time (df = 1)







Air temperature




< 0.001



< 0.001




< 0.001



< 0.001







< 0.001

Soil temperature




< 0.001



< 0.001







< 0.001




< 0.001



< 0.001

Values in bold indicate significant differences (P ≤ 0.05) in air- and soil temperature between the treatments warming and control over time. MS = mean squares

Fig. 1

Air (a, b) and soil temperatures (c, d) for meadow (a, c) and heath (b, d) for the warming treatment (continuous lines) and the control (dashed lines) over the time of measurement from mid-August to mid-October in 2015. Values are daily means of the hourly collected temperature data. Air temperatures were averaged across one meadow and two heath plots, each containing two sensors for each control (N = 3) and warming treatment (N = 3). Values for soil temperature were averaged across 15 paired plots (8 heath, 7 meadow), each containing two sensors for each control (N = 15) and warming treatment (N = 15)

Soil temperature

Soil temperatures were increased less by the warming treatment than air temperatures, but the effect was more consistent: maximum soil temperatures increased by 0.53 °C (Table 1, P < 0.001), mean soil temperatures by 0.46 °C (P < 0.001), and minimum soil temperatures by 0.14 °C (P = 0.002). Again, the effect of the warming treatment on soil temperature varied over time (Table 1, P < 0.001), with temperatures in control exceeding those of the warming plots in the beginning and in the end of the measurement (Fig. 1c, d). Soil temperature difference between control and warming in heath ranged from - 0.4 °C to 1.5 °C and - 0.3 °C to 1.3 °C for meadow vegetation, therefore showing similar variation in soil temperature for both vegetation types (Fig. 1c, d). Soil temperatures dropped below freezing in the beginning of October in heath vegetation, while soil temperatures below 0 °C were not measured in the meadow vegetation, but were approaching 0 °C in the beginning of October (Fig. 1c, d). In general, vegetation types differed significantly in soil temperatures with higher temperatures in the meadow (mean = 7.4 °C, SE = 0.27, N = 148) than in the heath (mean = 6.8 °C, SE = 0.26, N = 148; ANOVA: F1,2 = 48.9, P < 0.001).


Autumn phenology of two contrasting vegetation types differs below- but not aboveground

Root growth was significantly higher in meadow than heath, the former being a more fertile and more productive system than heath. According to Blume-Werry et al. (2017) root growth in subarctic tundra can be twice as high in the meadow compared to the heath, showing higher fine root turnover in the meadow. In line with our hypothesis, we found that root phenology in our study differed between the vegetation types; a peak in root production shortly after the decline in greenness was found in the meadow vegetation, whereas the root phenology in the heath vegetation was more evenly distributed and showed no peak during the time of measurement (Fig. 2c, d). Sloan et al. (2016) similarly found considerable differences in the timing of root production between woody plant and sedge-dominated vegetation in subarctic tundra, with sedge communities showing root production early in the growing season, whereas in woody plant communities the majority of root production occurred late in the growing season. Since environmental conditions, such as photoperiod and soil temperature did not differ in their study amongst the studied communities, it could be suggested that differences in the timing of root production relate rather to fundamental inherent differences between PFTs (Sloan et al. 2016). Our study supports this suggestion.
Fig. 2

Aboveground greenness in % (empty dots, a, b) and belowground production in mm (full dots, c, d) over the sampling period. Greenness represents proportion of living aboveground vegetation in % and was estimated on every sampling event except the last (majority of plots covered in snow). Loss of greenness was considered as onset of aboveground senescence and is indicated by an arrow. Belowground production is expressed as total root growth in mm within each sampling interval of 10 days except the last of 30 days. Values represent means across control (dashed lines, circles, N = 15) and warming treatment (continuous lines, squares, N = 15) for the two vegetation types, meadow (a, c, N = 7) and heath (b, d, N = 8). Error bars are ± SE

Contrary to our hypothesis, aboveground phenology of the two vegetation types showed similar dynamics over the sampling period. The onset of aboveground senescence was estimated for both vegetation types at the same date in mid-August with a greenness of less than 10% left in the second week of September (Fig. 2a, b). This result also corresponds with that of Blume-Werry et al. (2017) and Sloan et al. (2016) that the timing of leaf production does not vary substantially across contrasting subarctic plant communities.

Interestingly, root growth was still detected after mid-August, till the end of our measurements (Fig. 2c, d), thus showing an asynchrony for below- and aboveground phenology as previously described in tundra ecosystems (Blume-Werry et al. 2015; Radville et al. 2016b; Sloan et al. 2016). Living roots at the end of the growing season may contribute to ecosystem carbon and nutrient fluxes through respiration and the release of soluble exudates, feeding soil microbes (Iversen et al. 2015). Even though measured root growth for meadow and heath at the end of the study was only 4.7% and 2.6% of the total growth, any longer uptake and cycling of nutrients as well as ongoing carbon fluxes from plants to the soil may be of importance in these nutrient poor systems with a very short summer season.

A possible reason why roots remain active after aboveground senescence could be that through thermal buffering soils remain warm during the autumn (Steinaker and Wilson 2008) and, thus, roots insulated by the soil do not face the risk of freezing as quickly as aboveground plant-parts (Blume-Werry et al. 2015). However, if this is the case, one might expect higher root growth in warmer soils during autumn which we did not observe in either vegetation type. Nutrient availability in autumn could be increased through a decline in soil microbial populations and therefore plants may have their main access to nutrients in autumn and before the microbial populations recovery in spring (Chapin 1978; Jonasson et al. 1999). Allocation to root growth later in the growing season could also facilitate early season nutrient uptake and early root elongation (Sloan et al. 2016). Concerning in this matter is that previous models trying to predict the effects of climate change on plant growth, carbon and nutrient cycling in arctic ecosystems generally link the phenology of root production with the phenology of leaves and therefore underestimate root-mediated soil processes and their potential feedback to climate change (Iversen et al. 2015).

Heath and meadow are the two dominant vegetation types in subarctic tundra that differ strongly in plant species composition and are dominated by different PFTs (Table S1). For instance, the meadow vegetation showed a higher proportion of forbs and sedges (e.g. Ranunculus nivalis, Viola biflora, Carex spp.), whereas the heath plots were comprised of evergreen shrubs (e.g. Empetrum hermaphroditum and Vaccinium vitis-idaea). These fundamental differences between the vegetation types are most likely the reason for the significantly higher greenness observed in the heath plots than the meadow plots in this study (Fig. 2b). The higher presence of evergreens on the heath plots that do not senesce is likely to have influenced our greenness measurements. Therefore, the higher greenness values in heath do in this case not necessarily translate in a higher plant production in heath compared to meadow vegetation. However, the comparison within the vegetation types between the treatments warming and control can be used as an indicator for higher productivity.

No effect of warming on below- and aboveground autumn phenology

In both vegetation types, a significantly higher greenness was found in response to warming, often related to an increased photosynthetic activity and productivity (Zhou et al. 2001; Keenan et al. 2014), but also determined by the composition of plant functional types in the two contrasting vegetation types. Zhou et al. (2001) reported a 12% increase in the greenness of vegetation in Europe over less than two decades (1982–1998), coupled with an increase of the active growing season by about 18 days due to an earlier spring and delayed autumn. Increasing plant growth connected with greener biomass has the potential to alter the ability of ecosystems for long-term storage of CO2 and provide a negative feedback to climate warming (Piao et al. 2008). However, despite the significant differences in greening intensity, warming did not change the onset of aboveground senescence in mid-August, showing no prolongation of the growing season due to warming in either of the two vegetation types (Fig. 2a, b). This contrasts with other studies that found that warming treatments increased the cover of green vegetation and delayed autumn senescence (Marchand et al. 2004; Linderholm 2006).

Besides temperature it has been suggested that the timing of autumn senescence is correlated with the timing of spring budburst (Keenan and Richardson 2015). In this case, the impact of climate change on spring phenology is the crucial factor limiting the potential response of autumn phenology. Because of a higher sensitivity of autumn phenology to photoperiod acting as stronger constrain than temperature increase, autumn senescence may, in general, be less sensitive to temperature variability than spring has shown to be (Barichivich et al. 2013). Considering the focus on autumnal warming in this study and the short-term observation over just one autumn it is possible that a warming effect on the growing season length was much smaller than detected in other studies. It is also possible that aboveground phenology was changed, but only minimally, and we were thus not able to detect an effect on growing season length with our 10-day interval measurements.

Similar to aboveground, we found no evidence that autumnal warming prolonged the belowground growing season of plants in either of the two vegetation types. However, contrary to aboveground, the warming treatment did not increase belowground productivity, as root growth was not significantly higher in the warming plots (Fig. 2c, d). Therefore, we have similar results as Blume-Werry et al. (2017) who also found no response of root growth to their warming treatment caused by earlier snowmelt. As mentioned above, it is possible that the aboveground growing season was prolonged but less than our sampling interval. This is, however, not the case belowground as root growth is an integrative measure (i.e. any additional root growth can be observed in the next root image) and a prolongation of the growing season would at least have led to higher root growth. This implies that the temperature enhancement in this study was either not strong enough to achieve an effect, or that short-term temperature increases simply have no effect on autumnal root growth in the Arctic. Thus, even though the simple linear regression revealed a positve relationship between soil temperature and root growth and GDD explained a significant proportion of variance in root growth values, 50% of the variance in root growth must be explained by other variables (R2 = 0.51, F1,148 = 153, P < 0.001, Fig. 3). Given the adaptation of tundra plants to low temperatures allowing them to grow roots in cold, arctic soils (Chapin 1974; Shaver and Billings 1977; Bell and Bliss 1978; Iversen et al. 2015), it is most likely that other environmental factors, such as moisture, light limitation and the decline in carbohydrate availability, play important roles in controlling amounts and phenology of root growth in the Arctic (Chapin 1983; Barichivich et al. 2013; Radville et al. 2016a, b). For example, soil moisture is often referred to as one of the main drivers of root growth (Iversen et al. 2015; Radville et al. 2016b), but water availability in tundra ecosystems is rarely limited enough to reduce annual productivity (Bell and Bliss 1978). Root growth is also likely correlated with light availability (Iversen et al. 2015), as root growth of some tundra plants has been shown to slow down or even stop at day lengths below 15 h (Shaver and Billings 1977; Billings et al. 1978). Rates of root growth have been found to increase with increasing temperatures in other ecosystems, provided that soil moisture and nutrient availability were not limiting (Pregitzer et al. 2000). But in Arctic tundra, plant production is strongly limited by availability of soil nitrogen (N) and soil phosphorous (P) or both (Chapin 1974, 1983; Henry and Molau 1997; Jonasson et al. 1999; Iversen et al. 2015). Although the factors that govern root growth in arctic ecosystems remain unclear, it is possible that controls on root growth may change over time and throughout the year with soil temperature being more likely a driver of root initiation than root cessation (Radville et al. 2016a).
Fig. 3

The square-root of root growth per day (mm) over the Growing Degree Days per day with a base temperature of 2 °C

Contrary to our expectations that meadow vegetation responds stronger to warming, we found no evidence that meadow and heath differed in their responsiveness to the warming treatment, neither belowground, nor aboveground. This result also corresponds with that of Blume-Werry et al. (2017) that the overall above- and belowground responses to warmer soil temperatures caused by earlier snowmelt did not differ between meadow and heath in a subarctic region.

Although the vegetation types differed strongly in root production, their productivity was not affected by warming. Aboveground, warming increased greenness, but this increase did not differ between the vegetation types. This finding differs from studies at the start or at the peak of the growing season, when contrasting plant communities differ in their responsiveness to warming (Sloan et al. 2016). Our warming treatment with an increase in soil and air temperature of approximately 0.5 °C compared to ambient temperatures (Fig. 1) is consistent with the increase of surface air temperature in the past 40 years over the Arctic, which was an average of 0.4 °C per decade (Elmendorf et al. 2012). In addition, this is of similar effect size as OTCs used for autumnal warming with a range between 0.2 to 0.3 °C for air temperature and 0.4 to 0.5 °C for soil temperature from August till September (Dorrepaal et al. 2004, 2009). It is unlikely that a shading effect (reduction in incoming radiation compared to ambient conditions) could have evened out positive effects of the warmer temperatures due to our previous measurement of the PPFD (Fig. S1) and the higher greenness of aboveground vegetation in the warming plots (Fig. 2a, b).

Occasions when temperatures on the controls exceeded those on the warming treatment are likely caused by an albedo effect of the white fleece. Inversion of the warming treatment was mainly observed during extremely sunny periods (Fig. S3). Higher incoming radiation during sunny days warmed the surface of the controls and kept them warm during nights. Periods with high sun radiation in October showed no increase of air temperature in the controls over the warming treatment because the plots were already covered in snow, resulting in similar albedo effects for both treatments.


We found a significant difference in the belowground autumn phenology of the two contrasting vegetation types, heath and meadow, as well as significantly higher root growth in the more productive meadow vegetation compared to the heath. These differences are likely to arise from differences in the composition of plant functional types and their different growth dynamics over the growing season. In contrast, aboveground phenology and growing season length of the two vegetation types were similar over the sampling period. In both vegetation types, the belowground growing period was longer than the aboveground growing season. This demonstrates a decoupling in shoot and root phenology and should be considered in models predicting future carbon and nutrient fluxes in the face of a changing climate.

Furthermore, even though warming increased aboveground greenness it did not prolong the above- or belowground growing season in either of the two vegetation types implying that other factors (e.g. soil moisture, photoperiod, nutrients or spring budburst) may act as stronger controls on autumnal shoot and root growth than the temperature increase simulated in this experiment. The dynamics of roots remain some of the least understood aspects of plant function in the Arctic, but our and future research will contribute to fill the knowledge gaps considering the critical role of fine root turnover in regulating ecosystem carbon and nutrient balance.



This study was partly funded by the Kempe Foundation, Stiftelsen Oscar och Lili Lamms Minne, and the Humboldt-Ritter-Penck Foundation of the Berlin Geographical Society (Gesellschaft für Erdkunde zu Berlin). We thank Ellen Dorrepaal, Jacob Eckstein, Lea Fink, Laurenz Teuber and the staff of the Abisko Scientific Research Station for support.

Supplementary material

11104_2017_3343_MOESM1_ESM.eps (79 kb)
ESM 1 (EPS 79 kb)
11104_2017_3343_Fig5_ESM.gif (16 kb)

High resolution image (GIF 16 kb)

11104_2017_3343_MOESM2_ESM.eps (947 kb)
ESM 2 (EPS 947 kb)
11104_2017_3343_Fig6_ESM.gif (81 kb)

High resolution image (GIF 80 kb)

11104_2017_3343_MOESM3_ESM.eps (623 kb)
ESM 3 (EPS 622 kb)
11104_2017_3343_Fig7_ESM.gif (35 kb)

High resolution image (GIF 35 kb)

11104_2017_3343_MOESM4_ESM.eps (353 kb)
ESM 4 (EPS 353 kb)
11104_2017_3343_Fig8_ESM.gif (27 kb)

High resolution image (GIF 26 kb)

11104_2017_3343_MOESM5_ESM.docx (25 kb)
ESM 5 (DOCX 25 kb)
11104_2017_3343_MOESM6_ESM.docx (19 kb)
ESM 6 (DOCX 19 kb)


  1. ACIA, Arctic Climate Impact Assessment (2004) Impacts of a warming arctic: arctic climate impact assessment. Cambridge University Press, CambridgeGoogle Scholar
  2. Barichivich J, Briffa KR, Myneni RB, Osborn TJ, Melvin TM, Ciais P, Piao S, Tucker C (2013) Large-scale variations in the vegetation growing season and annual cycle of atmospheric CO2 at high northern latitudes from 1950 to 2011. Glob Chang Biol 19:3167–3183. doi: 10.1111/gcb.12283 CrossRefPubMedGoogle Scholar
  3. Bell KL, Bliss LC (1978) Root growth in a polar semidesert environment. Can J Bot 56:2470–2490CrossRefGoogle Scholar
  4. Billings WD, Peterson KM, Shaver GR (1978) Growth, turnover, and respiration rates of roots and tillers in tundra graminoids, vegetation and production ecology of an Alaskan Arctic tundra. Springer, pp 415–434Google Scholar
  5. Björk RG, Klemedtsson L, Molau U, Harndorf J, Ödman A, Giesler R (2007) Linkages between N turnover and plant community structure in a tundra landscape. Plant Soil 294:247–261. doi: 10.1007/s11104-007-9250-4 CrossRefGoogle Scholar
  6. Blume-Werry G, Wilson SD, Kreyling J, Milbau A (2015) The hidden season: growing season is 50% longer below than above ground along an arctic elevation gradient. New Phytol 209:978–986CrossRefPubMedGoogle Scholar
  7. Blume-Werry G, Jansson R, Milbau A (2017) Root phenology unresponsive to earlier snowmelt despite advanced above-ground phenology in two subarctic plant communities. Funct Ecol 31:1493–1502. doi: 10.1111/1365-2435.12853 CrossRefGoogle Scholar
  8. Braun-Blanquet J (1964) Pflanzensoziologie. Grundzüge der Vegetationskunde, Springer, BerlinCrossRefGoogle Scholar
  9. Chapin FS (1974) Morphological and physiological mechanisms of temperature compensation in phosphate absorption along a latitudinal gradient. Ecology 55:1180–1198CrossRefGoogle Scholar
  10. Chapin FS (1978) Phosphate uptake and nutrient utilization by barrow tundra vegetation, vegetation and production ecology of an Alaskan arctic tundra. Springer, pp 483–507Google Scholar
  11. Chapin FS (1983) Direct and indirect effects of temperature on arctic plants. Polar Biol 2:47–52. doi: 10.1007/BF00258285 CrossRefGoogle Scholar
  12. Chapin FS, Bret-Harte MS, Hobbie SE, Zhong H (1996) Plant functional types as predictors of transient responses of arctic vegetation to global change. J Veg Sci 7:347–358CrossRefGoogle Scholar
  13. Dorrepaal E, Aerts R, Cornelissen JHC, Callaghan TV, van Logtestijn RSP (2004) Summer warming and increased winter snow cover affect Sphagnum Fuscum growth, structure and production in a sub-arctic bog. Glob Chang Biol 10:93–104CrossRefGoogle Scholar
  14. Dorrepaal E, Toet S, van Logtestijn RSP, Swart E, van de Weg MJ, Callaghan TV, Aerts R (2009) Carbon respiration from subsurface peat accelerated by climate warming in the subarctic. Nature 460:616–619. doi: 10.1038/nature08216 CrossRefGoogle Scholar
  15. Elmendorf SC, Henry GHR, Hollister RD, Björk RG, Bjorkman AD, Callaghan TV, Collier LS, Cooper EJ, Cornelissen JHC, Day TA, Fosaa AM, Gould WA, Grétarsdóttir J, Harte J, Hermanutz L, Hik DS, Hofgaard A, Jarrad F, Jónsdóttir IS, Keuper F, Klanderud K, Klein JA, Koh S, Kudo G, Lang SI, Loewen V, May JL, Mercado J, Michelsen A, Molau U, Myers-Smith IH, Oberbauer SF, Pieper S, Post E, Rixen C, Robinson CH, Schmidt NM, Shaver GR, Stenström A, Tolvanen A, Totland O, Troxler T, Wahren C-H, Webber PJ, Welker JM, Wookey PA (2012) Global assessment of experimental climate warming on tundra vegetation: heterogeneity over space and time. Ecol Lett 15:164–175. doi: 10.1111/j.1461-0248.2011.01716.x CrossRefPubMedGoogle Scholar
  16. Ernakovich JG, Hopping KA, Berdanier AB, Simpson RT, Kachergis EJ, Steltzer H, Wallenstein MD (2014) Predicted responses of arctic and alpine ecosystems to altered seasonality under climate change. Glob Chang Biol 20:3256–3269. doi: 10.1111/gcb.12568 CrossRefPubMedGoogle Scholar
  17. Henry GHR, Molau U (1997) Tundra plants and climate change. The international tundra Expemriment (ITEX). Glob Chang Biol 3:1–9. doi: 10.1111/j.1365-2486.1997.gcb132.x CrossRefGoogle Scholar
  18. IPCC (2013) Summary for policymakers. SPM. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley P (eds) Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, pp 1–30Google Scholar
  19. Iversen CM, Sloan VL, Sullivan PF, Euskirchen ES, McGuire AD, Norby RJ, Walker AP, Warren JM, Wullschleger SD (2015) The unseen iceberg: plant roots in arctic tundra. New Phytol 205:34–58. doi: 10.1111/nph.13003 CrossRefPubMedGoogle Scholar
  20. Johannessen OM, Kuzmina SI, Bobylev LP, Miles MW (2016) Surface air temperature variability and trends in the Arctic: new amplification assessment and regionalisation. Tellus A 68Google Scholar
  21. Johnson MG, Tingey DT, Phillips DL, Storm MJ (2001) Advancing fine root research with minirhizotrons. Environ Exp Bot 45:263–289. doi: 10.1016/S0098-8472(01)00077-6 CrossRefPubMedGoogle Scholar
  22. Jonasson S, Michelsen A, Schmidt IK, Nielsen EV (1999) Responses in microbes and plants to changes temperature, nutrient, and light regimes in the arctic. Ecology 80(6):1828–1843Google Scholar
  23. Keenan TF, Richardson AD (2015) The timing of autumn senescence is affected by the timing of spring phenology: implications for predictive models. Glob Chang Biol 21:2634–2641CrossRefGoogle Scholar
  24. Keenan TF, Darby B, Felts E, Sonnentag O, Friedl MA, Hufkens K, O’Keefe J, Klosterman S, Munger JW, Toomey M, others (2014) Tracking forest phenology and seasonal physiology using digital repeat photography: a critical assessment. Ecol Appl 24:1478–1489CrossRefPubMedGoogle Scholar
  25. Körner C (2007) The use of 'altitude' in ecological research. Trends Ecol Evol 22:569–574. doi: 10.1016/j.tree.2007.09.006 CrossRefPubMedGoogle Scholar
  26. Linderholm HW (2006) Growing season changes in the last century. Agric For Meteorol 137:1–14. doi: 10.1016/j.agrformet.2006.03.006 CrossRefGoogle Scholar
  27. Mack MC, Schuur EAG, Bret-Harte MS, Shaver GR, Chapin FS (2004) Ecosystem carbon storage in arctic tundra reduced by long-term nutrient fertilization. Nature 431:440–443. doi: 10.1038/nature02887 CrossRefPubMedGoogle Scholar
  28. Marchand FL, Nijs I, Heuer M, Mertens S, Kockelbergh F, Pontailler J-Y, Impens I, Beyens L (2004) Climate warming postpones senescence in high Arctic tundra. Arct Antarct Alp Res 36:390–394CrossRefGoogle Scholar
  29. Marion GM, Henry GH, Freckman DW, Johnstone J, Jones G, Jones MH, Levesque E, Molau U, Mølgaard P, Parsons AN, others (1997) Open-top designs for manipulating field temperature in high-latitude ecosystems. Glob Chang Biol 3:20–32Google Scholar
  30. McCormack LM, Eissenstat DM, Prasad AM, Smithwick EAH (2013) Regional scale patterns of fine root lifespan and turnover under current and future climate. Glob Chang Biol 19(6):1697–1708Google Scholar
  31. McMaster GS, Wilhelm WW (1997) Growing degree-days: one equation, two interpretations. Agric For Meteorol 87:291–300CrossRefGoogle Scholar
  32. Pau S, Wolkovich EM, Cook BI, Davies TJ, Kraft NJB, Bolmgren K, Betancourt JL, Cleland EE (2011) Predicting phenology by integrating ecology, evolution and climate science. Glob Chang Biol 17:3633–3643. doi: 10.1111/j.1365-2486.2011.02515.x CrossRefGoogle Scholar
  33. Pendall E, Bridgham S, Hanson PJ, Hungate B, Kicklighter DW, Johnson DW, Law BE, Luo Y, Megonigal JP, Olsrud M, Ryan MG, Wan S (2004) Below-ground process responses to elevated CO2 and temperature. A discussion of observations, measurement methods, and models. New Phytol 162:311–322. doi: 10.1111/j.1469-8137.2004.01053.x CrossRefGoogle Scholar
  34. Piao S, Ciais P, Friedlingstein P, Peylin P, Reichstein M, Luyssaert S, Margolis H, Fang J, Barr A, Chen A, Grelle A, Hollinger DY, Laurila T, Lindroth A, Richardson AD, Vesala T (2008) Net carbon dioxide losses of northern ecosystems in response to autumn warming. Nature 451:49–52. doi: 10.1038/nature06444 CrossRefPubMedGoogle Scholar
  35. Post E, Forchhammer MC, Bret-Harte MS, Callaghan TV, Christensen TR, Elberling B, Fox AD, Gilg O, Hik DS, Høye TT, others (2009) Ecological dynamics across the Arctic associated with recent climate change. Science 325:1355–1358Google Scholar
  36. Pregitzer KS, King JS, Burton AJ, Brown SE (2000) Responses of tree fine roots to temperature. New Phytol 147:105–115CrossRefGoogle Scholar
  37. Radville L, McCormack ML, Post E, Eissenstat DM (2016a) Root phenology in a changing climate. Journal of experimental botany:erw062. doi: 10.1093/jxb/erw062
  38. Radville L, Post E, Eissenstat DM (2016b) Root phenology in an Arctic shrub-graminoid community: the effects of long-term warming and herbivore exclusion. Climate Change Responses 3:4. doi: 10.1186/s40665-016-0017-0 CrossRefGoogle Scholar
  39. Richardson AD, Keenan TF, Migliavacca M, Ryu Y, Sonnentag O, Toomey M (2013) Climate change, phenology, and phenological control of vegetation feedbacks to the climate system. Agric For Meteorol 169:156–173. doi: 10.1016/j.agrformet.2012.09.012 CrossRefGoogle Scholar
  40. Shaver GR, Billings WD (1977) Effects of daylength and temperature on root elongation in tundra graminoids. Oecologia 28:57–65. doi: 10.1007/BF00346836 CrossRefPubMedGoogle Scholar
  41. Sloan VL, Fletcher BJ, Press MC, Williams M, Phoenix GK (2013) Leaf and fine root carbon stocks and turnover are coupled across Arctic ecosystems. Glob Chang Biol 19:3668–3676. doi: 10.1111/gcb.12322 CrossRefPubMedGoogle Scholar
  42. Sloan VL, Fletcher BJ, Phoenix GK, Bardgett R (2016) Contrasting synchrony in root and leaf phenology across multiple sub-Arctic plant communities. J Ecol 104:239–248. doi: 10.1111/1365-2745.12506 CrossRefGoogle Scholar
  43. Steinaker DF, Wilson SD (2008) Phenology of fine roots and leaves in forest and grassland. J Ecol 96:1222–1229. doi: 10.1111/j.1365-2745.2008.01439.x CrossRefGoogle Scholar
  44. Sundqvist MK, Giesler R, Graae BJ, Wallander H, Fogelberg E, Wardle DA (2011) Interactive effects of vegetation type and elevation on aboveground and belowground properties in a subarctic tundra. Oikos 120:128–142. doi: 10.1111/j.1600-0706.2010.18811.x CrossRefGoogle Scholar
  45. Zhou L, Tucker CJ, Kaufmann RK, Slayback D, Shabanov NV, Myneni RB (2001) Variations in northern vegetation activity inferred from satellite data of vegetation index during 1981 to 1999. J Geophys Res: Atmos 106:20069–20083CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  1. 1.Experimental Plant Ecology, Institute of Botany and Landscape EcologyErnst-Moritz-Arndt-University GreifswaldGreifswaldGermany
  2. 2.Department of Biodiversity and Natural EnvironmentResearch Institute for Nature and Forest INBOBrusselsBelgium
  3. 3.Climate Impacts Research Centre, Department of Ecology and Environmental ScienceUmeå UniversityAbiskoSweden

Personalised recommendations