Oecologia DOI 10.1007/s00442-014-3055-y
Global change ecology - Original research
Linking belowground and aboveground phenology in two boreal forests in Northeast China Enzai Du · Jingyun Fang
Received: 13 May 2014 / Accepted: 14 August 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract The functional equilibrium between roots and shoots suggests an intrinsic linkage between belowground and aboveground phenology. However, much less understanding of belowground phenology hinders integrating belowground and aboveground phenology. We measured root respiration (Ra) as a surrogate for root phenology and integrated it with observed leaf phenology and radial growth in a birch (Betula platyphylla)–aspen (Populus davidiana) forest and an adjacent larch (Larix gmelinii) forest in Northeast China. A log-normal model successfully described the seasonal variations of Ra and indicated the initiation, termination and peak date of root phenology. Both root phenology and leaf phenology were highly specific, with a later onset, earlier termination, and shorter period of growing season for the pioneer tree species (birch and aspen) than the dominant tree species (larch). Root phenology showed later initiation, later peak and later termination dates than leaf phenology. An asynchronous correlation of Ra and radial growth was identified with a time lag of approximately 1 month, indicating aprioritization of shoot growth. Furthermore, we found that Ra was strongly correlated with soil temperature and air temperature, while radial growth was only significantly correlated with air temperature, implying a down-regulating effect of temperature. Our results indicate different phenologies between
Communicated by Russell Monson. E. Du (*) · J. Fang (*) Department of Ecology, College of Urban and Environmental Sciences and Key Laboratory for Earth Surface Processes of the Ministry of Education, Peking University, Beijing 100871, China e-mail: [email protected] J. Fang e-mail: [email protected]
pioneer and dominant species and support a down-regulation hypothesis of plant phenology which can be helpful in understanding forest dynamics in the context of climate change. Keywords Belowground-aboveground linkages · Leaf phenology · Root respiration · Root phenology · Plant phenology down-regulation
Introduction Plant phenology has strong relevance for primary production, nutrient cycling and plant competition (Jackson et al. 2001; Nord and Lynch 2009; Chuine 2010; Richardson et al. 2010) and is sensitive to climate change (Chmielewski and Rötzer 2001; Gordo and Sanz 2010). Most studies on plant phenology have focused on leaf phenology because of its direct role in photosynthetic production (Menzel 2002), while much less is known about root phenology (Steinaker and Wilson 2008). Root phenology regulates seasonal variations of root activity, which plays a vital role in belowground processes such as nutrient and water uptake, soil C and nutrient cycling, plant–microbe interactions and plant competition (Harris 1977; Burke and Raynal 1994; Côté et al. 1998; Nord and Lynch 2009; Kuzyakov and Gavrichkova 2010). The functional equilibrium between root and shoot growth (Wilson 1988) suggests an intrinsic linkage between belowground and aboveground phenology. Integrating the two can provide a useful insight into belowground-aboveground linkages and improve our understanding of vegetation responses to climate change. Root phenology has been indicated by seasonal variations of root growth (Palacio and Montserrat-Martí 2007), but measurements of root growth suffer from multiple
13
Oecologia
potential biases and limitations (Milchunas 2009). For instance, the traditional method of sequential coring is destructive and labor intensive, and constrains continuous measurements of root growth. Mini-rhizotrons can provide a more precise measure of root dynamics with little disturbance, but these instruments are expensive, which may limit their wide application and large numbers of replicates in field studies (Johnson et al. 2001; Hendrick and Pregitzer 1992). However, root respiration (Ra) provides a non-destructive indicator of root phenology because it is an integrative proxy of root activities including root growth, uptake and transport of ions and maintenance processes (Veen 1981). Ra starts to increase as the growing season begins and decreases to a maintenance level again in the dormant season (Desrochers et al. 2002). The initiation and termination date of the belowground growing season can thus be easily speculated upon seasonal variations of Ra. This proxy method has the advantage of being applicable to existing global data sets of Ra (Bond-Lamberty and Thomson 2010). Previous studies have found strong regulating effects of air temperature on aboveground phenology (e.g., Körner and Basler 2010; Jeong et al. 2011) and soil temperature on seasonal root dynamics (e.g., Tryon and Chapin 1983; Tierney et al. 2003; Quan et al. 2010). Rates of new tissue formation of shoots in cold-adapted plants generally are very low at temperatures below 5 °C (Körner 2012). Due to the faster warming of air compared to soil, leaves may sense and respond to temperature changes earlier than roots at the beginning of the growing season. Evidence also indicates that root activity of trees is driven by carbohydrates from photosynthetic production in leaves with a time lag of several days (Högberg et al. 2001; Kuzyakov and Gavrichkova 2010). Additionally, biochemical signals (e.g., hormones) from leaves may regulate the initiation, cessation and strength of root growth (Kossuth and Ross 1987; Hoad 1995; Aloni 2001, 2013). Based on the above analysis, we proposed a down-regulation hypothesis of plant phenology to link belowground and aboveground phenology. Our down-regulation hypothesis predicts that belowground phenology lags behind aboveground phenology. Previous studies have found decoupling and asynchrony of belowground and aboveground phenology in some forest and grassland ecosystems (Palacio and Montserrat-Martí 2007; Steinaker et al. 2010), but the down-regulation hypothesis of phenology has never been synthesized before. Aboveground phenology, belowground phenology and their linkages are likely to vary among species and vegetation types (Palacio and Montserrat-Martí 2007; Steinaker and Wilson 2008). In the Great Khingan Mountains in Northeast China, pristine boreal forest is solely dominated by larch (Larix gmelinii Ruppr.) and the secondary forest usually consists of pioneer species such as birch (Betula
13
platyphylla Suk.) and aspen (Populus davidiana Dode.). Significant climate warming has occurred in Northeast China since the 1960s (Piao et al. 2010), while precipitation has been decreasing since the 1980s (Ding et al. 2008). Linking belowground and aboveground phenology and detecting phenological differences between the pioneer species and dominant species should have implications for future dynamics of the boreal forest in the context of climate change. During early May to mid-October 2012, we simultaneously measured Ra, leaf phenology, radial growth, and climate factors (soil temperature, soil moisture and air temperature) in a birch-aspen forest and an adjacent larch forest in the Great Khingan Mountains. Root phenology was indicated by a proxy model using Ra and was compared with leaf phenology. Temporal variations and environmental regulators of Ra and radial growth were also compared to gain more information on linkages between belowground and aboveground phenology. Our objectives were to: (1) compare key events of belowground phenology and aboveground phenology between the pioneer and dominant species, and (2) test whether the linkages between belowground phenology and aboveground phenology support the down-regulation hypothesis.
Materials and methods Site description The site is located at the National Field Research Station of Daxing’anling Forest Ecosystem (50°56′, 121°30′E) in Great Khingan Mountains, Northeast China. The annual mean temperature and annual mean precipitation are −5.4 °C and 450–550 mm (50–70 % in July and August), respectively, defining a cold and dry climate. This study was conducted in a birch (Betula platyphylla Suk.)–aspen (Populus davidiana Dode.) (birch vs. aspen 1:3) forest (87 years old) and an adjacent larch (Larix gmelinii Ruppr.)
Table 1 Characteristics of the birch-aspen forest and the larch forest in this study Items
Birch-aspen forest
Larch forest
Forest age (years) Stand density (stems ha−1) Mean diameter at breast height (cm) Mean height (m) Litter production (Mg ha−1) Litter storage (Mg ha−1)
87 1,983 ± 279 12.3 ± 3.2
85 1,816 ± 95 9.8 ± 3.6
12.9 ± 2.8 1.94 ± 0.09 13.35 ± 1.85
11.2 ± 2.9 1.96 ± 0.13 15.45 ± 1.63
Mor layer C:N ratio
18.39 ± 0.13
17.45 ± 0.27
Oecologia
forest (85 years old). Basic information on the two forests is given in Table 1. Leaf phenology Three replicate plots with an area of 10 × 20 m2 were established for each forest type in spring 2011. Leaf phenological events were observed by a well-trained observer from May to October 2012, including the dates of budburst, full unfolding of leaves, leaf senescence and defoliation. Budburst was considered to have occurred in the spring when the first emerging leaves or needles were clearly visible; foliage fully unfolded when the leaves and needles had been fully expanded and elongated; leaf senescence began in the autumn when yellow leaves or needles appeared at the apical portion of branches in the lower and middle part of the canopy; defoliation was completed when all the leaves and needles dropped onto the forest floor. Frequent observations (3–5 days) were conducted to catch each phenological event. The length of the growing season for the aboveground phenology differed from that for root phenology, and was defined as the duration between budburst and leaf senescence. Radial growth Trees with diameter at breast height (DBH) >5 cm were labeled and dendrometer bands were installed in each plot in spring 2011. The window length of the dendrometer band was recorded at the beginning of each month from May to October 2012. The change of window length between each two measurements was the monthly circumference growth, which was further used to calculate the radial growth. The relative increment of DBH (RID; ‰) was calculated to indicate the monthly radial growth of each tree according to the following equation,
RID = 1000 × DBH/DBH′
(1)
soil CO2 flux system (LI-COR, Lincoln, NE) between 9:00 and 12:00 a.m. to avoid the influences of diurnal variations. Simultaneously, soil temperature and soil moisture at 5-cm soil depth were measured using an auxiliary soil temperature probe (Omega Engineering, USA) and a Theta probe (Delta-T Devices, Cambridge, UK). We measured Rs and Rh frequently (2–3 times per month) and in total there were 16 measurements from May to October 2012. One EM50 data logger (Decagon Devices, USA) for each forest was installed to record the soil temperature and moisture at a depth of 5 cm and the open-air temperature every half hour. Modeling respiration‑indicated root phenology In boreal forest, root growth generally occurs during May to September with a single peak of the growth rate during the growing season (Tryon and Chapin 1983). Mini-rhizotron investigation has shown a unimodal seasonal pattern of root growth for larch, birch and aspen trees (Jiang et al. 2010; Quan et al. 2010). Based on the evidence that root growth and Ra of the targeting plant species both show a unimodal seasonal pattern, we proposed a log-normal model using Ra as a surrogate for root phenology to detect the initiation, peak and termination for root growth. Ra (μmol m−2 s−1), calculated as the difference between plot mean Rs and Rh, was fitted to the following equation:
Ra = R0 + a × e−0.5×[ln(t/t0 )/b]
2
(2)
where t indicated day of year, R0 was the maintenance respiration during the dormant season, t0 indicated the date of peak Ra (equivalent to R0 + a), and a and b were parameters. Then the initiation and termination dates of root growth were both obtained by a threshold-based method. Roots were assumed to be dormant when the growth respiration (Ra−R0) was negligible at a threshold of: (3)
(Ra − R0 )/R0 ≤ 0.05
where ΔDBH (mm) was the DBH increment during the current month, and DBH′ (mm) was the DBH at the beginning of the current month. The mean RID were calculated for each plot to indicate the rate of radial growth at a plot scale.
Combining Eqs. 2 and 3, the initiation and termination dates for root growth were thus derived from the following equation:
Soil respiration
Parameters
Birch-aspen forest
Larch forest
In each plot, three collars were inserted randomly into the forest floor to measure soil respiration (Rs) and two collars were inserted to measure heterotrophic respiration (Rh) using a trenching method (Du et al. 2013). Trenches were dug around a 1.5 × 1.5-m2 area and roots were prevented from recolonizing the trenched interior area by a polyethylene sheet. Rs and Rh were measured using a LI-8100A automated
R0
0.29 ± 0.16*
0.23 ± 0.10*
Table 2 Parameters of the log-normal model fit for the birch-aspen forest and the larch forest in Northeast China
a
***
1.72 ± 0.19***
***
0.11 ± 0.01***
2.27 ± 0.33
b
0.08 ± 0.01
t0
***
203 ± 2
207 ± 2***
R0 Maintenance respiration during the dormant season, t0 date of peak root respiration * p
Global change ecology - Original research
Linking belowground and aboveground phenology in two boreal forests in Northeast China Enzai Du · Jingyun Fang
Received: 13 May 2014 / Accepted: 14 August 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract The functional equilibrium between roots and shoots suggests an intrinsic linkage between belowground and aboveground phenology. However, much less understanding of belowground phenology hinders integrating belowground and aboveground phenology. We measured root respiration (Ra) as a surrogate for root phenology and integrated it with observed leaf phenology and radial growth in a birch (Betula platyphylla)–aspen (Populus davidiana) forest and an adjacent larch (Larix gmelinii) forest in Northeast China. A log-normal model successfully described the seasonal variations of Ra and indicated the initiation, termination and peak date of root phenology. Both root phenology and leaf phenology were highly specific, with a later onset, earlier termination, and shorter period of growing season for the pioneer tree species (birch and aspen) than the dominant tree species (larch). Root phenology showed later initiation, later peak and later termination dates than leaf phenology. An asynchronous correlation of Ra and radial growth was identified with a time lag of approximately 1 month, indicating aprioritization of shoot growth. Furthermore, we found that Ra was strongly correlated with soil temperature and air temperature, while radial growth was only significantly correlated with air temperature, implying a down-regulating effect of temperature. Our results indicate different phenologies between
Communicated by Russell Monson. E. Du (*) · J. Fang (*) Department of Ecology, College of Urban and Environmental Sciences and Key Laboratory for Earth Surface Processes of the Ministry of Education, Peking University, Beijing 100871, China e-mail: [email protected] J. Fang e-mail: [email protected]
pioneer and dominant species and support a down-regulation hypothesis of plant phenology which can be helpful in understanding forest dynamics in the context of climate change. Keywords Belowground-aboveground linkages · Leaf phenology · Root respiration · Root phenology · Plant phenology down-regulation
Introduction Plant phenology has strong relevance for primary production, nutrient cycling and plant competition (Jackson et al. 2001; Nord and Lynch 2009; Chuine 2010; Richardson et al. 2010) and is sensitive to climate change (Chmielewski and Rötzer 2001; Gordo and Sanz 2010). Most studies on plant phenology have focused on leaf phenology because of its direct role in photosynthetic production (Menzel 2002), while much less is known about root phenology (Steinaker and Wilson 2008). Root phenology regulates seasonal variations of root activity, which plays a vital role in belowground processes such as nutrient and water uptake, soil C and nutrient cycling, plant–microbe interactions and plant competition (Harris 1977; Burke and Raynal 1994; Côté et al. 1998; Nord and Lynch 2009; Kuzyakov and Gavrichkova 2010). The functional equilibrium between root and shoot growth (Wilson 1988) suggests an intrinsic linkage between belowground and aboveground phenology. Integrating the two can provide a useful insight into belowground-aboveground linkages and improve our understanding of vegetation responses to climate change. Root phenology has been indicated by seasonal variations of root growth (Palacio and Montserrat-Martí 2007), but measurements of root growth suffer from multiple
13
Oecologia
potential biases and limitations (Milchunas 2009). For instance, the traditional method of sequential coring is destructive and labor intensive, and constrains continuous measurements of root growth. Mini-rhizotrons can provide a more precise measure of root dynamics with little disturbance, but these instruments are expensive, which may limit their wide application and large numbers of replicates in field studies (Johnson et al. 2001; Hendrick and Pregitzer 1992). However, root respiration (Ra) provides a non-destructive indicator of root phenology because it is an integrative proxy of root activities including root growth, uptake and transport of ions and maintenance processes (Veen 1981). Ra starts to increase as the growing season begins and decreases to a maintenance level again in the dormant season (Desrochers et al. 2002). The initiation and termination date of the belowground growing season can thus be easily speculated upon seasonal variations of Ra. This proxy method has the advantage of being applicable to existing global data sets of Ra (Bond-Lamberty and Thomson 2010). Previous studies have found strong regulating effects of air temperature on aboveground phenology (e.g., Körner and Basler 2010; Jeong et al. 2011) and soil temperature on seasonal root dynamics (e.g., Tryon and Chapin 1983; Tierney et al. 2003; Quan et al. 2010). Rates of new tissue formation of shoots in cold-adapted plants generally are very low at temperatures below 5 °C (Körner 2012). Due to the faster warming of air compared to soil, leaves may sense and respond to temperature changes earlier than roots at the beginning of the growing season. Evidence also indicates that root activity of trees is driven by carbohydrates from photosynthetic production in leaves with a time lag of several days (Högberg et al. 2001; Kuzyakov and Gavrichkova 2010). Additionally, biochemical signals (e.g., hormones) from leaves may regulate the initiation, cessation and strength of root growth (Kossuth and Ross 1987; Hoad 1995; Aloni 2001, 2013). Based on the above analysis, we proposed a down-regulation hypothesis of plant phenology to link belowground and aboveground phenology. Our down-regulation hypothesis predicts that belowground phenology lags behind aboveground phenology. Previous studies have found decoupling and asynchrony of belowground and aboveground phenology in some forest and grassland ecosystems (Palacio and Montserrat-Martí 2007; Steinaker et al. 2010), but the down-regulation hypothesis of phenology has never been synthesized before. Aboveground phenology, belowground phenology and their linkages are likely to vary among species and vegetation types (Palacio and Montserrat-Martí 2007; Steinaker and Wilson 2008). In the Great Khingan Mountains in Northeast China, pristine boreal forest is solely dominated by larch (Larix gmelinii Ruppr.) and the secondary forest usually consists of pioneer species such as birch (Betula
13
platyphylla Suk.) and aspen (Populus davidiana Dode.). Significant climate warming has occurred in Northeast China since the 1960s (Piao et al. 2010), while precipitation has been decreasing since the 1980s (Ding et al. 2008). Linking belowground and aboveground phenology and detecting phenological differences between the pioneer species and dominant species should have implications for future dynamics of the boreal forest in the context of climate change. During early May to mid-October 2012, we simultaneously measured Ra, leaf phenology, radial growth, and climate factors (soil temperature, soil moisture and air temperature) in a birch-aspen forest and an adjacent larch forest in the Great Khingan Mountains. Root phenology was indicated by a proxy model using Ra and was compared with leaf phenology. Temporal variations and environmental regulators of Ra and radial growth were also compared to gain more information on linkages between belowground and aboveground phenology. Our objectives were to: (1) compare key events of belowground phenology and aboveground phenology between the pioneer and dominant species, and (2) test whether the linkages between belowground phenology and aboveground phenology support the down-regulation hypothesis.
Materials and methods Site description The site is located at the National Field Research Station of Daxing’anling Forest Ecosystem (50°56′, 121°30′E) in Great Khingan Mountains, Northeast China. The annual mean temperature and annual mean precipitation are −5.4 °C and 450–550 mm (50–70 % in July and August), respectively, defining a cold and dry climate. This study was conducted in a birch (Betula platyphylla Suk.)–aspen (Populus davidiana Dode.) (birch vs. aspen 1:3) forest (87 years old) and an adjacent larch (Larix gmelinii Ruppr.)
Table 1 Characteristics of the birch-aspen forest and the larch forest in this study Items
Birch-aspen forest
Larch forest
Forest age (years) Stand density (stems ha−1) Mean diameter at breast height (cm) Mean height (m) Litter production (Mg ha−1) Litter storage (Mg ha−1)
87 1,983 ± 279 12.3 ± 3.2
85 1,816 ± 95 9.8 ± 3.6
12.9 ± 2.8 1.94 ± 0.09 13.35 ± 1.85
11.2 ± 2.9 1.96 ± 0.13 15.45 ± 1.63
Mor layer C:N ratio
18.39 ± 0.13
17.45 ± 0.27
Oecologia
forest (85 years old). Basic information on the two forests is given in Table 1. Leaf phenology Three replicate plots with an area of 10 × 20 m2 were established for each forest type in spring 2011. Leaf phenological events were observed by a well-trained observer from May to October 2012, including the dates of budburst, full unfolding of leaves, leaf senescence and defoliation. Budburst was considered to have occurred in the spring when the first emerging leaves or needles were clearly visible; foliage fully unfolded when the leaves and needles had been fully expanded and elongated; leaf senescence began in the autumn when yellow leaves or needles appeared at the apical portion of branches in the lower and middle part of the canopy; defoliation was completed when all the leaves and needles dropped onto the forest floor. Frequent observations (3–5 days) were conducted to catch each phenological event. The length of the growing season for the aboveground phenology differed from that for root phenology, and was defined as the duration between budburst and leaf senescence. Radial growth Trees with diameter at breast height (DBH) >5 cm were labeled and dendrometer bands were installed in each plot in spring 2011. The window length of the dendrometer band was recorded at the beginning of each month from May to October 2012. The change of window length between each two measurements was the monthly circumference growth, which was further used to calculate the radial growth. The relative increment of DBH (RID; ‰) was calculated to indicate the monthly radial growth of each tree according to the following equation,
RID = 1000 × DBH/DBH′
(1)
soil CO2 flux system (LI-COR, Lincoln, NE) between 9:00 and 12:00 a.m. to avoid the influences of diurnal variations. Simultaneously, soil temperature and soil moisture at 5-cm soil depth were measured using an auxiliary soil temperature probe (Omega Engineering, USA) and a Theta probe (Delta-T Devices, Cambridge, UK). We measured Rs and Rh frequently (2–3 times per month) and in total there were 16 measurements from May to October 2012. One EM50 data logger (Decagon Devices, USA) for each forest was installed to record the soil temperature and moisture at a depth of 5 cm and the open-air temperature every half hour. Modeling respiration‑indicated root phenology In boreal forest, root growth generally occurs during May to September with a single peak of the growth rate during the growing season (Tryon and Chapin 1983). Mini-rhizotron investigation has shown a unimodal seasonal pattern of root growth for larch, birch and aspen trees (Jiang et al. 2010; Quan et al. 2010). Based on the evidence that root growth and Ra of the targeting plant species both show a unimodal seasonal pattern, we proposed a log-normal model using Ra as a surrogate for root phenology to detect the initiation, peak and termination for root growth. Ra (μmol m−2 s−1), calculated as the difference between plot mean Rs and Rh, was fitted to the following equation:
Ra = R0 + a × e−0.5×[ln(t/t0 )/b]
2
(2)
where t indicated day of year, R0 was the maintenance respiration during the dormant season, t0 indicated the date of peak Ra (equivalent to R0 + a), and a and b were parameters. Then the initiation and termination dates of root growth were both obtained by a threshold-based method. Roots were assumed to be dormant when the growth respiration (Ra−R0) was negligible at a threshold of: (3)
(Ra − R0 )/R0 ≤ 0.05
where ΔDBH (mm) was the DBH increment during the current month, and DBH′ (mm) was the DBH at the beginning of the current month. The mean RID were calculated for each plot to indicate the rate of radial growth at a plot scale.
Combining Eqs. 2 and 3, the initiation and termination dates for root growth were thus derived from the following equation:
Soil respiration
Parameters
Birch-aspen forest
Larch forest
In each plot, three collars were inserted randomly into the forest floor to measure soil respiration (Rs) and two collars were inserted to measure heterotrophic respiration (Rh) using a trenching method (Du et al. 2013). Trenches were dug around a 1.5 × 1.5-m2 area and roots were prevented from recolonizing the trenched interior area by a polyethylene sheet. Rs and Rh were measured using a LI-8100A automated
R0
0.29 ± 0.16*
0.23 ± 0.10*
Table 2 Parameters of the log-normal model fit for the birch-aspen forest and the larch forest in Northeast China
a
***
1.72 ± 0.19***
***
0.11 ± 0.01***
2.27 ± 0.33
b
0.08 ± 0.01
t0
***
203 ± 2
207 ± 2***
R0 Maintenance respiration during the dormant season, t0 date of peak root respiration * p
