JXB Advance Access originally published online on December 21, 2005
Journal of Experimental Botany 2006 57(2):267-281; doi:10.1093/jxb/erj029
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RESEARCH PAPER |
Phenotypic plasticity and growth temperature: understanding interspecific variability
1Department of Biology, University of York, PO Box 373, York YO1 5YW, UK
2Ecosystem Dynamics, Research School of Biological Sciences, The Australian National University, GPO Box 475, Canberra, ACT 2601 Australia
3Department of Plant Ecophysiology, Utrecht University, PO Box 80084, 3508 TB Utrecht, The Netherlands
* To whom correspondence should be addressed. Fax: +44 (0)1904 328505. E-mail: OKA1{at}york.ac.uk
Received 27 May 2005; Accepted 25 October 2005
 |
Abstract |
|---|
 Top Abstract IntroductionPhenotypic responses to growth...Systematic differences among...Temperature and relationship...ConclusionsReferences |
|---|
The subject of this review is the impact of long-term changes in temperature on plant growth and its underlying components. The discussion highlights the extent to which thermal acclimation of metabolism is intrinsically linked to the plasticity of a range of biochemical and morphological traits. The fact that there is often a trade-off between temperature-mediated changes in net assimilation rates (NAR) and biomass allocation [in particular the specific leaf area (SLA)] when plants are grown at different temperatures is also highlighted. Also discussed is the role of temperature-mediated changes in photosynthesis and respiration in determining NAR values. It is shown that in comparisons that do not take phylogeny into account, fast-growing species exhibit greater temperature-dependent changes in RGR, SLA, and NAR than slow-growing plants. For RGR and NAR, such trends are maintained within phylogenetically independent contrasts (i.e. species adapted to more-favourable habitats consistently exhibit greater temperature-mediated changes than their congeneric counterparts adapted to less-favourable habitats). By contrast, SLA was not consistently more thermally plastic in species from favourable habitats. Interestingly, biomass allocation between leaves and roots was consistently more plastic in slow-growing species within individual phylogenetically independent contrasts, when plants were grown under contrasting temperatures. Finally, how interspecific variations in NAR account for an increasing proportion of variability in RGR as growth temperatures decrease is highlighted. Conversely, SLA played a more dominant role in determining interspecific variability in RGR at higher growth temperatures; thus, the importance of SLA in determining interspecific variation in RGR could potentially increase if annual mean temperatures increase in the future.
Key words: Acclimation, biomass allocation, growth analysis, net assimilation rate, plasticity, photosynthesis, respiration, specific leaf area, temperature
|
Introduction |
|---|
|
TopAbstractIntroductionPhenotypic responses to growth...Systematic differences among...Temperature and relationship...ConclusionsReferences |
|---|
Being sessile organisms, plants experience large temporal variations in temperature; moreover, large spatial variations in temperature are commonly experienced (Larcher, 2004
). Variations in temperature affect metabolic processes that contribute to biosynthesis and cellular maintenance (e.g. photosynthesis and respiration; Berry and Raison, 1981). Biomass allocation is also temperature sensitive, with long-term exposure to low temperatures resulting in plants exhibiting a reduced investment in the shoot, and leaves that are thicker and display less leaf area per unit mass than their warm-grown counterparts (Fig. 1; Potter and Jones, 1977; Woodward, 1979a; Friend and Woodward, 1990; Williams and Black, 1993; Körner, 1999; Bruhn et al., 2000; Loveys et al., 2002). Collectively, such responses result in the relative growth rate (RGR, rate of increase in dry mass per unit starting mass and time) of individual plants being temperature dependent (Blackmann et al., 1955; Potter and Jones, 1977; Ackerly et al., 1992; Loveys et al., 2002; Hendrickson et al., 2004; Storkey, 2004). Given that the temperatures experienced by plants are changing as a result of global climate change (Hulme, 1999; Schimel et al., 2001), predicting the response of plant growth to temperature is fundamental to projecting the impact of global change on the biosphere (Long and Hutchin, 1991).
|
50 d following initial 4 weeks of growth at 25/20 °C. (B, C) Transverse sections of representative warm-grown (B) and cold-developed leaves (C). Note the increased overall leaf thickness, increased number of cell layers, and decreased cell size in cold-developed leaves. Scale bars=100 µm. Photograph in (A) courtesy of Peter Gorsuch.This paper reviews the impacts of long-term changes in temperature on plant growth and its underlying components. First, the extent to which individual traits are altered following long-term exposure to contrasting temperatures is considered; the paper focuses on the extent to which some traits exhibit plastic responses to growth temperature, and the degree to which plasticity of those traits contributes to thermal acclimation of metabolism, and rates of growth in plants experiencing contrasting growth temperatures; highlighted is the tight coupling that exists between phenotypic plasticity and metabolic acclimation. The impacts of growth temperature on the underlying components of RGR [e.g. the net assimilation rate (NAR), specific leaf area (SLA), and leaf mass ratio (LMR)] are discussed. As part of this section, published data sets (Loveys et al., 2002
, 2003b) are re-analysed to illustrate the changes to the components of plant growth that can occur when plants are grown at 18 °C, 23 °C, and 28 °C. The importance of respiratory responses to temperature in determining the response of RGR to growth temperature is highlighted. An assessment is then made whether there are systematic differences among contrasting species (e.g. inherently fast- versus slow-growing species) in thermal plasticity of growth-related traits. Here, the data sets of Loveys et al. (2002, 2003a, b) that previously did not take phylogeny into account, but which nevertheless contain several sets of phylogenetically independent contrasts (PICs) (Westoby, 2002) along gradients of environmental favourability (altitude, rainfall, nutrient availability) have been used again. Finally, the impact of interspecific differences in thermal plasticity on the relationship between RGR and its underlying components is considered in a wide range of contrasting species (in the absence and presence of phylogenetic correction). It is shown that irrespective of whether phylogeny is taken into account or not, changes in growth temperature alter the relationship between RGR and its underlying components (the roles of SLA and NAR in determining interspecific variability in RGR increased and decreased, respectively, at higher growth temperatures). |
Phenotypic responses to growth temperature |
|---|
|
TopAbstractIntroductionPhenotypic responses to growth...Systematic differences among...Temperature and relationship...ConclusionsReferences |
|---|
Plasticity and acclimation
The impact of long-term changes in temperature on plant growth depends, in part, on the ‘phenotypic plasticity’ of traits underpinning the growth response when plants experience a change in growth temperature. For the purposes of this review, ‘phenotypic plasticity’ is defined as the capacity for traits within an individual to exhibit a diverse range of phenotypes in response to growth at different temperatures (based on previous definitions by Bradshaw, 1965
; Piersma and Drent, 2003; Strand and Weisner, 2004). A good example of phenotypic plasticity to growth temperature is the contrasting leaf structure/morphology exhibited by warm- and cold-grown plants (Fig. 1). In fully expanded, mature leaves, such contrasting phenotypes are largely fixed irrespective of the growth temperature that the plant subsequently experiences [i.e. such plasticity is irreversible; Piersma and Drent (2003) defined such cases as ‘developmental plasticity’]. By contrast, reversibility is possible in traits that exhibit ‘phenotypic flexibility’ (Piersma and Drent, 2003). Phenotypic flexibility encompasses molecular and biochemical changes that contribute to changes in the capacity of physiological processes (see next section) within fully expanded tissues that experience changes in growth temperature. In subsequent sections, examples of both forms of phenotypic plasticity are provided.
An important consequence of the wide range of changes associated with temperature-mediated phenotypic plasticity is that the capacity of key metabolic processes often differs between plant tissues with a cold- and warm-grown phenotype. For example, cold-grown plants typically exhibit greater photosynthetic and respiratory capacities than their warm-grown counterparts. As a result, in situ rates of metabolism are often similar in plants experiencing contrasting growth temperatures, when measured at their respective growth temperatures (i.e. metabolism ‘acclimates’; Körner and Larcher, 1988; Stitt and Hurry, 2002; Atkin and Tjoelker, 2003). Acclimation of metabolism can result in ratios of respiration to photosynthesis (R:P) being similar in plants experiencing contrasting growth temperatures (Gifford, 1995; Ziska and Bunce, 1998; Dewar et al., 1999; Loveys et al., 2002, 2003a; Atkin et al. 2005a, b). Ultimately, the changes associated with phenotypic plasticity can also lead to RGR remaining relatively constant across contrasting growth temperatures (see below).
Although acclimation of photosynthesis and respiration to a new growth temperature can occur in pre-existing leaves formed at the previous growth temperature (Pisek et al., 1973; Loveys et al., 2003a) (an example of phenotypic flexibility), data from studies using short-lived herbaceous species suggest that full acclimation to a new growth temperature requires that leaves need to be formed at the new growth temperature (Hurry et al., 1995; Strand et al., 1997; Loveys et al., 2003a; Talts et al., 2004; Atkin et al., 2005b); thus, the changes associated with developmental plasticity result in greater apparent acclimation of metabolism than do the changes associated with phenotypic flexibility. In addition to increased thickness (Fig. 1), cold-developed leaves exhibit increased levels of membrane lipid unsaturation (Berry and Raison, 1981), contain higher nitrogen concentrations (Ryan, 1995; Tjoelker et al., 1999b), have higher transcript and activity levels of photosynthetic and sucrose synthesis enzymes (Hurry et al., 2001), and decreased potential for inorganic phosphate-mediated feedback inhibition of photosynthesis (Strand et al., 1997, 1999) than their warm-grown counterparts. Underpinning many of the acclimation-related changes in metabolism are changes in expression of several cold-responsive genes; these include CBF/DREB1, the family of COR genes, and the HOS genes (Thomashow, 2001; Zarka et al., 2003; Knight et al., 2004). Collectively, such responses enable photosynthesis to operate at similar rates in the cold compared with leaves grown and measured at warmer temperatures (Hurry et al., 2001).
Like photosynthesis, acclimation of respiration to contrasting growth temperatures is closely linked to changes in development and chemical composition (Atkin and Tjoelker, 2003; Atkin et al., 2005a, b). Cold-developed plants often exhibit enhanced respiratory rates both at low and high temperatures (Atkin et al., 2000a; Talts et al., 2004), reflecting an increase in both the density of mitochondria per unit tissue mass and an increase in potential rates of respiratory activity per unit mitochondrial volume (Klikoff, 1966, 1968; Miroslavov and Kravkina, 1991; Armstrong et al., 2005). The abundance of individual proteins/protein complexes and the capacity of terminal oxidases (e.g. alternative and cytochrome oxidase) are also altered in tissues developed at contrasting temperatures (Ribas Carbó et al., 2000; Kurimoto et al., 2004b; Armstrong et al., 2005). Importantly, the available data suggest that pre-existing leaves and roots of short-lived species are unable to alter their structure (i.e. anatomy/morphology), chemical composition, and enzymatic capacity to the degree required for full acclimation of respiration to occur; examples are provided in Loveys et al. (2003a). Further work is needed, however, to assess whether the same is true for slower-growing, longer-lived species.
Underlying components of RGR: responses to temperature
One way of assessing the effects of long-term exposure on plants is to quantify the impact of temperature on RGR and its underlying components [i.e. NAR (the increase in plant mass per unit leaf area and time), SLA (the ratio of leaf area to leaf dry mass), and LMR (ratio of leaf mass to plant mass)]. NAR largely reflects the rate of whole-plant photosynthetic carbon gain minus carbon released by respiration (Lambers and Poorter, 1992). RGR is related to these parameters according to:
 | (1) |
; Berry and Björkman, 1980; Atkin et al., 2005a) according to:
 | (2) |
 | (3) |
Recently, a growth analysis was conducted on large Arabidopsis thaliana (Col-0) plants shifted from 25 °C to 5 °C (OK Atkin, LJ Atkinson, unpublished results); RGR decreased from 75 mg g–1 d–1 to near 0 mg g–1 d–1, due to an immediate drop in the NAR (brought about by a large decline in Pa). However, RGR subsequently recovered to 30 mg g–1 d–1 [due to metabolic changes within pre-existing tissues and alterations in the phenotype of newly formed, cold-developed tissues (Fig. 1), which collectively resulted in a decrease in the SLA and an increase in NAR]. Similar homeostatic adjustments in RGR are reported for plants grown over a wide range of temperatures (Tjoelker et al., 1998b; see Table 1), due to temperature-mediated changes in NAR, LMR, and/or SLA. Because long-term changes in temperature often result in adjustments in all parameters underpinning the RGR, the linkage between NAR and RGR is often less clear in plants exposed to contrasting temperatures for extended periods than it is in plants experiencing short-term changes in temperature (Blackmann et al., 1955; Stirling et al., 1998; Bruhn et al., 2000). To illustrate further how long-term differences in growth temperature impact on RGR and its underlying components (see equation 1), the mean values for 16 contrasting species grown at 18 °C, 23 °C, and 28 °C, using data reported by Loveys et al. (2002, 2003b), were calculated for several growth-related traits. Table 2 shows that growth at 18 °C and 23 °C had little impact on the average RGR of all species and that average RGR was substantially reduced when plants were grown at 28 °C. Between 18 °C and 23 °C, RGR remained relatively constant due to decreases in NAR being largely offset by increases in SLA. Above 23 °C, RGR declined largely due to decreases in NAR.
|
|
An example of plastic changes in biomass allocation are temperature-dependent changes in the relative investment in above- and below-ground organs, particularly when leaves and roots experience contrasting growth temperatures (Kummerrow and Ellis, 1984
; Clarkson et al., 1986; DeLucia et al., 1992; Körner, 1999; Weih and Karlsson, 2001; Cunningham and Read, 2003) (for examples of root:shoot ratios, see Table 1). However, when roots and leaves experience equal temperatures in a controlled environment, in which plants are grown hydroponically, other reports have found LMR to be temperature insensitive (Tjoelker et al., 1998b; Loveys et al., 2002). An example is shown in Table 2, which provides mean values for 16 contrasting species grown at 18 °C, 23 °C, and 28 °C, using LMR data reported by Loveys et al. (2002). It seems that temperature-mediated changes in LMR are more likely in cases where leaf and roots experience contrasting temperatures.
How plastic is biomass allocation within leaves (i.e. SLA) in plants grown under contrasting temperature regimes? Analysis of the Loveys et al. (2002) data set demonstrates that average SLA values were lowest at 18 °C, with 23 °C and 28 °C exhibiting relatively similar average SLA values; average SLA at 18 °C was 31% lower than that at 28 °C (Table 2). Similar temperature-dependent changes in SLA were reported by Williams and Black (1993), when they grew Pennisetum setaceum (a grass introduced to Hawaii) at two temperatures (Table 1). Moreover, Woodward (1979a, b) found that SLA of both alpine and lowland plants decreased with decreasing growth temperature (Table 1). By contrast, Hovenden (2001) found no difference in SLA values between southern beech seedlings grown at 23 °C and 18 °C. Tjoelker et al. (1998b) also reported no systematic temperature-induced change in SLA values of five deciduous and evergreen tree species grown at five temperatures (Table 1). The impact of temperature on SLA therefore depends on which species is being investigated and the temperature regimes at which the plants are grown.
The response of NAR to long-term differences in growth temperature is highly variable in published studies. For example, in a comparison of four rainforest canopy species grown under controlled conditions (14/6 °C, 19/11 °C, 22/14 °C, 25/17 °C, and 30/22 °C day/night temperatures), Cunningham and Read (2003) found that NAR increased with increasing growth temperature, whereas Tjoelker et al. (1998b) reported that NAR values were often similar in tree seedlings grown at five temperatures (Table 1). NAR also exhibited little systematic change with growth temperature in several other species (Stirling et al., 1998; Bruhn et al., 2000; Kurimoto et al., 2004a). By contrast, both Bednarz and van Iersel (2001) and Loveys et al. (2002) reported that NAR decreased with increasing growth temperature (see Table 2 for a summary of the data of Loveys et al., 2002). In the study by Loveys et al. (2002), average NAR values decreased as growth temperatures increased, being 75% higher at 18 °C than at 28 °C (Table 2).
Growth temperature-mediated variations in NAR
What factors might explain the variable response of NAR to growth temperature? As shown in equation 2, NAR is equal to Pa minus Ra divided by CC, with Ra being influenced by both biomass allocation and rates of Rm (equation 3). Given that temperature-sensitive changes in CC are unlikely to account for temperature-induced changes in NAR either in the short- or long-term [rather than being lower at low growth temperatures, Loveys et al. (2002) found that CC values were actually highest in plants grown at the lowest growth temperature (Table 2), particularly in fast-growing species], changes in NAR following long-term exposure to contrasting temperatures must reflect responses of Pa and/or Ra. In the absence of metabolic acclimation or plastic changes in biomass allocation, increases in temperature are expected to decrease NAR (and, conversely, NAR will increase with decreasing temperature), due to respiration being more temperature sensitive than photosynthesis in short-term experiments (Berry and Björkman, 1980). In such cases, the ratio of respiration to photosynthesis increases with increasing temperature (Gifford, 1995; Ziska et al., 1998; Dewar et al., 1999; Atkin et al., 2000b). Thus, the fact that NAR is often similar or even greater in plants grown at high temperatures (compared with cool-grown plants; Tjoelker et al., 1998b; Cunningham and Read, 2003) suggests that in those studies, Pa and/or Ra must have adjusted to long-term changes in temperature (Tjoelker et al., 1998a, 1999a). Such adjustment could reflect acclimation of the underlying photosynthetic and respiratory machinery, or plastic changes in biomass allocation (e.g. equation 3). By contrast, less adjustment of Pa and/or Ra is likely in species where NAR decreased markedly with increasing growth temperature. Thus, variations in the temperature response of NAR are likely to reflect the degree to which Pa and Ra adjust to long-term changes in temperature.
Underpinning variations in the degree of acclimation of area-based fluxes could be either changes in rates per unit mass and/or the ratio of leaf area to mass relationships (e.g. as driven by changes in SLA). Theoretically, decreases in SLA (underpinned by increases in leaf thickness) could lead to an increase in NAR due to increases in capacity of the photosynthetic machinery per unit leaf area that result from greater numbers of mesophyll cells and/or abundance of chloroplasts per unit leaf area (Lambers and Poorter, 1992; Shipley, 2000). Examples of temperature-mediated changes in shoot phenotype and leaf structure are provided in Fig. 1, which shows shoots of warm-grown, cold-treated (i.e. warm-grown plants shifted to chilling temperatures for several days) and cold-developed (i.e. leaves produced following a total of 50–60 d chilling treatment) plants, as well as transverse leaf sections of Arabidopsis thaliana that developed at 25/20 °C or 5 °C (Armstrong et al., 2005). However, realizing gains from additional photosynthetic capacity requires that leaves be exposed to high irradiance; in leaves exposed to low irradiance, decreases in SLA are unlikely to result in concomitant increases in NAR.
To what extent are lower NAR values in high temperature-grown plants due to increased respiratory losses (i.e. high daily Ra values)? Although Loveys et al. (2002) did not measure whole-plant gas exchange in the 16 species shown in Table 2, they did conduct measurements using a sub-set of six species. Averaged across the six species, daily Ra values were nearly identical in plants grown at 18 °C and 23 °C (102 and 109 mmol CO2 m–2 d–1), increasing by nearly 30% to 134 mmol CO2 m–2 d–1 in plants grown at 28 °C (Loveys et al., 2002). In the absence of changes in Pa or CC, such changes in Ra alone should decrease NAR by nearly 1.0 g m–2 d–1 (this calculation was based on the assumption that SLA, LMR, Pa, and CC values remained constant at 24 m2 kg–1, 0.65 
540 mmol CO2 m–2 d–1, and 32 mmol C g–1, respectively). Given that NAR was >3.0 g m–2 d–1 lower at 28 °C than 23 °C (Table 2), this suggests that temperature-mediated reductions in NAR at the high growth temperature must have also been due, in part, to reductions in Pa; indeed, when averaged across the six species, daily whole-shoot Pa values were lower at 28 °C (540 mmol CO2 m–2 d–1) than at 23 °C (583 mmol CO2 m–2 d–1). Thus, alterations in Pa and Ra contribute to differences in NAR at contrasting growth temperatures. In turn, variations in Ra largely reflect temperature-mediated changes in Rm values rather than changes in biomass allocation (see Loveys et al., 2002).
Temperature-mediated changes in RGR: what underlying factor is more important?
What underlying factors play the dominant role in determining variations in RGR when plants are grown at different temperatures? Although there is often a negative interplay between NAR and SLA, changes in one parameter are often not matched by concomitant changes in the other. An example is shown in Table 2, where increases in NAR at low growth temperature were not fully matched by decreases in SLA; in such cases, temperature-mediated variations in NAR play a more dominant role in determining RGR at each growth temperature than do variations in SLA. This is further illustrated in Fig. 2, where the extent to which RGR remained constant over the 18–23 °C interval (i.e. RGR18/RGR23, RGR at 18 °C divided by RGR at 23 °C) for individual species was plotted against the difference in NAR values at 18 °C and 23 °C (raw data was extracted from Loveys et al., 2002); the ratio RGR18/RGR23 was greatest in plants that exhibited the greatest increase in NAR at 18 °C compared with 23 °C. In a study assessing the impact of temperature on growth of three wheat cultivars and rice, Kurimoto et al. (2004a) also found that temperature-mediated changes in NAR played a prominent role in determining the variations in RGR in plants grown under controlled-environment conditions. Similarly, in a study of 18 annual weed and two crop species, Storkey (2004) found that variations in RGR were largely determined by the NAR in the autumn in plants grown outdoors. However, in spring, variations in the ratio of leaf area to plant mass largely determined the RGR (Storkey, 2004). The linkage between NAR and RGR may thus be dependent on developmental stage or on other factors associated with seasonal changes in the environment.
|
Taken together, this section highlights how, in response to lower growth temperatures, SLA values decrease with a potential negative effect on RGR. However, acting against this decline in SLA is an increase in the NAR (at least for plants grown under sub-saturating irradiance). Further work is needed to assess whether growth temperature has similar effects on SLA and NAR in plants experiencing higher irradiances and/or lower temperature regimes than those used by Loveys et al. (2002)
. |
Systematic differences among species in their response to growth temperature |
|---|
|
TopAbstractIntroductionPhenotypic responses to growth...Systematic differences among...Temperature and relationship...ConclusionsReferences |
|---|
In the previous section, the extent to which some of the underlying components of RGR exhibit plastic responses when grown at different temperatures, and how acclimation of metabolism is underpinned by plasticity of a range of morphological and biochemical traits was discussed. In this section, consideration is given to whether there are systematic differences among contrasting species (e.g. inherently fast- versus slow-growing species) in the response to growth temperature.
Acclimation of photosynthesis and respiration
Studies investigating the temperature responses of photosynthesis and respiration have found that there is considerable variability in the degree of photosynthetic and respiratory acclimation among species, with some species fully acclimating while others are incapable of even partial acclimation (Berry and Raison, 1981; Larigauderie and Körner, 1995; Xiong et al., 2000; Atkin et al., 2005a, b). Tjoelker et al. (1999a) suggested that broad-leaved tree species exhibited a lower degree of acclimation of leaf respiration than selected conifer species; they suggested that it may be possible to predict degrees of acclimation/homeostasis of respiration using structural and/or functional traits. Moreover, in a comparison of thermal acclimation of photosynthesis and leaf respiration in two lowland and one alpine Plantago species, Atkin et al. (2005a) found that only the fast-growing lowland species exhibited significant metabolic acclimation when grown at three different temperatures (13 °C, 20 °C, and 27 °C); moreover, only the lowland species exhibit plasticity of leaf structure in response to growth temperature. Collectively, such results suggest that inherently fast-growing species may indeed be more morphologically plastic and capable of exhibiting greater acclimation/homeostasis of metabolic flux than their slow-growing counterparts. By contrast, other studies have not found systematic differences in the degree of acclimation of respiration. For example, Larigauderie and Körner (1995) found no systematic differences in thermal acclimation of respiration among alpine and lowland plant species. Moreover, in a comparison of 16 fast- and slow-growing species, Loveys et al. (2003a) found no systematic difference in respiratory homeostasis of either roots or leaves, when homeostasis/acclimation ratios were plotted against RGR values. Thus, it may be premature to conclude that all fast-growing species are more plastic and capable of exhibiting greater thermal acclimation of respiration than their slow-growing counterparts.
Case study: comparisons of plasticity in contrasting species across all taxa
To assess further whether there are systematic differences in response to growth temperature among the fast- and slow-growing species used by Loveys et al. (2002), the absolute difference between the maximum and minimum value of a trait (irrespective at which growth temperature the maximum and minimum values occurred) was calculated (Fig. 3). In comparisons of all taxa (i.e. not taking into account phylogeny), each max–min value against the maximum RGR exhibited by that species across the three growth temperatures was then plotted (Fig. 3). No systematic differences were found in the max–min values of leaf mass ratio (LMR), stem mass ratio (StMR), or root mass ratio (RMR) among fast- and slow-growing species in comparisons of all taxa (data not shown). However, there was a significant positive correlation between max–min values of RGR, SLA, and NAR, and the maximum RGR value exhibited by each species across the three growth temperatures, with the slope being greatest in the NARmax–NARmin versus RGRmax plot (Fig. 3B). Thus, although there is no systematic difference among species in how biomass allocation between leaves, stems, and roots respond to growth temperature, fast-growers do exhibit greater absolute changes in RGR, SLA, and NAR than their slow-growing counterparts. Such absolute differences are relevant when considering how growth temperature alters the competitive ability of species (e.g. large absolute induced changes in leaf area may enable a plant to out-compete its neighbours for light).
|
Importantly, no systematic difference in plasticity of SLA among the fast- and slow-growing species was found when the relative ability of an individual species to alter traits (e.g. biomass allocation) was assessed on being challenged with different growth temperatures (Fig. 4). The relative ability to alter SLA was calculated using the percentage coefficient of variation approach (%CV, where CV=100xstandard deviation of individual temperature treatment means)/grand mean of the temperature treatment means; Ryser and Eek, 2000
; Bloor and Grubb, 2004). Similarly, no systematic differences were found among fast- and slow-growing species in terms of relative change in RGR or NAR brought about by differences in growth temperature (not shown). Thus, while growth temperature does have a greater absolute effect on growth parameters of fast-growing species, contrasting species do not appear to differ in plasticity assessed using scales that are relative.
|
The data shown in Fig. 3 were for comparisons where phylogeny was not taken into account. Here, it is asked whether those trends across all species (irrespective of phylogeny) also exist within PICs. Each PIC contained a fast- and a slow-growing species characteristic of environmentally favourable and unfavourable habitats, respectively (three nutrient contrasts, two rainfall contrasts, and two altitudinal contrasts). PIC patterns were interpreted in a qualitative manner (Wright and Westoby, 1999
), as significance testing would have required greater replication than was available from the data sets of Loveys et al. (where data on seven PICs were available); consequently, only majority trends among PICs (five out of seven PICs, six out of seven PICs, or seven out of seven PICs) were considered indicative of a consistent relationship (Wright and Westoby, 1999). Figure 5 shows PIC stick graphs (Westoby et al., 1998) of absolute differences between the maximum and minimum values of a trait (irrespective of which growth temperature the maximum and minimum values occurred) for plant species representative of favourable and unfavourable habitats. Temperature-induced differences in RGR were generally greater in the species characteristic of favourable habitats (in six out of the seven contrasts; Fig. 5A). Only in the genus Acacia did the species from the unfavourable habitat (A. aneura) respond more dynamically to temperature than the species from a more favourable habitat (A. melanoxylon). Similarly, temperature-induced differences in NAR were consistently higher in plants from favourable habitats (Fig. 5B). However, by contrast to the general trends across all species (Fig. 3), species from favourable habitats did not consistently exhibit greater temperature-induced differences in SLA (differences in SLA were only greater in the species from favourable habitats in four out of the seven contrasts; Fig. 5C). Thus, while temperature-induced changes in RGR and NAR were consistently different in species from contrasting habitats, no such conclusion can be reached for SLA (i.e. adaptation to stressful habitats is not consistently associated with a decrease in thermal plasticity of leaf area to mass relationships). This result is surprising, as it is often assumed that fast-growing species with high SLA values are more plastic than their slow-growing counterparts (Lambers et al., 1998). While this is true when phylogeny is not taken into account (Fig. 3), it is not necessarily the case within individual PICs (Fig. 5C).
|
Another surprising outcome of the present PIC analysis was that, by contrast to the general trends across all species (see above), consistent differences in plasticity of LMR and RMR to temperature were found within individual PICs (Fig. 6). In the case of LMR, five out of the seven contrasts exhibited greater plasticity of LMR in species from unfavourable habitats. For RMR, the trend was highly consistent, with species from unfavourable habitats exhibiting greater plasticity than their favourable habitat counterparts in seven out of the seven contrasts. Thus, even though temperature has relatively little absolute effect on LMR and RMR (Loveys et al., 2002
), and across all taxa there is no systematic difference in plasticity of LMR and RMR among fast- and slow-growing species (see above), consistent differences in plasticity are found within most clades. This suggests that adaptation to stressful, unfavourable habitats is associated with increased plasticity of biomass allocation between leaves and roots in response to differences in growth temperature. One possible explanation for this result is that the nutrient uptake potential of a given unit of root mass varies with growth temperature (e.g. decreasing at lower growth temperatures). To overcome limitations in nutrient uptake at unfavourable temperatures, a plant could either increase the uptake capacity of a given mass of roots, or increase allocation of biomass to roots. It is suggested that in species characteristic of unfavourable habitats, greater reliance is placed on changes in biomass allocation to roots than is the case for species characteristic of more favourable habitats. Further research is needed to determine whether this hypothesis is correct.
|
|
Temperature and relationship between RGR and its underlying components |
|---|
|
TopAbstractIntroductionPhenotypic responses to growth...Systematic differences among...Temperature and relationship...ConclusionsReferences |
|---|
Trends across all taxa, irrespective of phylogeny
Most of the studies investigating the relationship between inherent differences in RGR and its underlying components have been conducted using plants grown under controlled environmental conditions at a single temperature (Lambers and Poorter, 1992
; Poorter and Garnier, 1999). Although both Garnier and Freijsen (1994) and Poorter and De Jong (1999) found that changes in growth environment do not alter the ranking of leaf traits (e.g. SLA) across species, one consequence of interspecific differences in thermal plasticity (see previous section) is that the relationship between RGR and its underlying components in contrasting species varies with growth temperature (Loveys et al., 2002). Loveys et al. (2002) found that the importance of SLA for explaining variations in RGR was reduced as growth temperature decreased (Fig. 7). Conversely, variations in NAR played a greater role in determining interspecific differences in RGR in plants grown at 18 °C than in plants grown at 23 °C and 28 °C (Fig. 7). Loveys et al. (2002) found that the increasing importance of NAR in determining interspecific differences in RGR as growth temperatures decreased was due to interspecific differences in the degree of thermal acclimation of whole plant respiration (Ra) (with the degree of homeostasis of Ra being greatest in inherently slow-growing species). Rates of whole plant respiration per unit plant mass (Rm) were also more homeostatic (across growth temperatures) in the slow-growing species (Loveys et al., 2002). In a subsequent study, Loveys et al. (2003a) found no systematic differences in the degree of acclimation exhibited by mature leaves of the fast- and slow-growing species; it was speculated that the apparent contradiction between the two studies (Loveys et al., 2002, 2003a) reflected the fact that Loveys et al. (2002) measured R rates in whole shoots (i.e. developing and mature leaves), whereas Loveys et al. (2003a) focused on the response of mature leaves only. If the degree of acclimation potential of developing leaves differs from that of mature leaves, then it may be that responses seen at the whole shoot level will differ from that of mature leaves alone. To date, no published studies have assessed the extent to which growth temperature affects rates of R in developing and mature leaves.
|
Another factor that could contribute to the discrepancy between Loveys et al. (2002)
and Loveys et al. (2003a) could be the impact of temperature on respiratory flux in species that differ inherently in rates of respiration. Fast-growing species typically respire at a higher rate than their slow-growing counterparts (Poorter et al., 1991; Atkin et al., 1996; Loveys et al., 2003a). In the absence of acclimation and assuming equal Q10 values in fast- and slow-growing species (Loveys et al., 2003a), exposure to a new growth temperature will have a greater absolute effect on respiration in the fast-growing species. Conversely, changes in growth temperature would have less effect on respiration in inherently slow-growing species. Taken together, the above studies demonstrate that changes in growth temperature do alter the relationship between RGR and its underlying components in comparisons of contrasting species, that variations in NAR play a greater role in determining interspecific differences in RGR at low growth temperature, and, finally, that the greater role of NAR at lower growth temperatures is due to rates of whole-plant R being more homeostatic across growth temperatures in slow-growing species than their fast-growing counterparts. Greater homeostasis of whole plant R in the slow-growing species (compared with fast-growing species) may reflect interspecific differences in the ratio of mature and immature tissues, as well as differences in specific rates of R and subsequent temperature responses of R.
Case study: responses to growth temperature in PICs
Do the trends observed in Loveys et al. (2002) across all species (irrespective of phylogeny) also exist within PICs? Here, a reanalysis from a PIC perspective of the effect of temperature on relationships between RGR and its underlying components is given. Figure 8 shows correlated change graphs (Wright and Westoby, 1999) of dSLA and dNAR values plotted against dRGR values for each PIC. Values were calculated by subtracting the attribute value for the favourable habitat from that of the unfavourable environment within each PIC. Positively correlated change would cause a point to fall in either upper-left or lower-right quadrants. For clarity of presentation, data for plants grown at 23 °C have not been shown. The correlated divergence between SLA and RGR was more consistent at 28 °C than at 18 °C (Fig. 8A), as shown by seven out of seven points (at 28 °C) versus four out of seven (at 18 °C) points falling in the top-right/lower-left quadrants (Fig. 8A). Moreover, for a given divergence in RGR, divergences in SLA were greater at 28 °C than at 18 °C. For correlated divergence between NAR and RGR, six out of seven points fell in the top-right/lower-right quadrants (Fig. 8B), with divergence in NAR generally being greater (for a given divergence in RGR) at 18 °C than at 28 °C. Thus, the SLA versus RGR and NAR versus RGR trends observed in Loveys et al. (2002) across all species (irrespective of phylogeny) are held within PICs but also changes in growth temperature alter the roles that SLA and NAR play in determining variations in RGR. At high growth temperatures, variations in RGR are strongly correlated with variations in SLA, irrespective of phylogeny. However, this relationship is less consistent at lower growth temperatures, suggesting that divergences in leaf structure may become less important in determining divergences in growth rate as growth temperatures decrease. Conversely, NAR plays a greater role in determining RGR at lower growth temperatures. Such results cast some doubt on the generality of SLA being an important determinant of RGR of plants growing under cool field conditions and raise the prospect that divergences in NAR may play a greater role than previously thought in determining differences in RGR exhibited by closely related fast- and slow-growing species from contrasting environments.
|
|
Conclusions |
|---|
|
TopAbstractIntroductionPhenotypic responses to growth...Systematic differences among...Temperature and relationship...ConclusionsReferences |
|---|
In this review, how some plants, when challenged with contrasting growth temperatures, maintain relatively similar growth rates, is highlighted. Underpinning the response of RGR to growth is plasticity of a wide range of biochemical, anatomical, and morphological traits; collectively, such plastic responses contribute to homeostasis of metabolic activity, and potentially to maintenance of RGR in plants experiencing contrasting thermal regimes. Thus, acclimation of metabolism is intrinsically linked to plasticity of a range of biochemical and morphological traits. It has also been shown how plants often exhibit a trade-off between temperature-mediated changes in NAR and biomass allocation (in particular, changes in leaf structure that contribute to changes in SLA) when grown at different temperatures. The present analysis suggests that temperature-mediated changes in respiration play a pivotal role in determining NAR and thus RGR at different temperatures, particularly in plants grown at sub-saturating irradiance. It has also been shown that in comparisons that do not take phylogeny into account, fast-growing species exhibit greater temperature-dependent changes in RGR, SLA, and NAR than slow-growing plants (but with no systematic difference in the effect of temperature on biomass allocation between leaves, stems, and roots). Such trends are maintained with PICs for RGR and NAR [i.e. species adapted to more favourable habitats consistently exhibit greater temperature-mediated changes (and thus less homeostasis) in RGR and NAR than their congeneric counterparts adapted to less favourable habitats]. By contrast, no consistent trend in plasticity of SLA within PICs was found, suggesting that adaptation to contrasting habitats is not associated with systematic changes in plasticity of leaf structure. The present analysis has revealed the surprising fact that biomass allocation between leaves and roots is consistently more plastic in slow-growing species within each PIC. Finally, the review has highlighted how interspecific differences in plasticity/homeostasis alter the relationship between RGR and its underlying components, regardless of whether phylogeny is taken into account or not; it has been shown that variations in NAR account for an increasing proportion of variability in RGR as growth temperatures decrease. Conversely, the importance of SLA in determining interspecific variation in RGR could potentially increase if annual mean temperatures increase in the future.
|
Acknowledgements |
|---|
This work was supported by grants from NERC in the UK (GR3/11898, NERC/A/S/2001/01186, NERC/B/S/2001/00875 to OKA), the Cooperative Research Centre for Greenhouse Accounting, Australia (BRL), the Nordic Academy of Advanced Studies (NorFA) Temperature Stress Network, and a Daphne Jackson Fellowship funded by the NERC (to LJA). We acknowledge the generous support provided by the Society for Experimental Biology and the Journal of Experimental Botany towards the running of the Plant Frontiers meeting in Sheffield.
|
Footnotes |
|---|
Abbreviations: CC, carbon concentration; CV, coefficient of variation; P, net photosynthesis; PIC, phylogenetically independent contrast; NAR, net assimilation rate; RGR, relative growth rate; RMR, root mass ratio; LMR, leaf mass ratio; SLA, specific leaf area; R, respiration.
|
References |
|---|
|
TopAbstractIntroductionPhenotypic responses to growth...Systematic differences among...Temperature and relationship...ConclusionsReferences |
|---|
Ackerly DD, Coleman JS, Morse SR, Bazzaz FA. 1992. CO2 and temperature effects on leaf area production in two annual plant species. Ecology 73, 1260–1269.[CrossRef]
Armstrong AF, Logan DC, O'Toole P, Tobin AK, Atkin OK. 2005. Heterogeneity of plant mitochondrial responses underpinning respiratory acclimation to the cold in Arabidopsis thaliana leaves. Plant, Cell and Environment (in press).
Atkin OK, Botman B, Lambers H. 1996. The causes of inherently slow growth in alpine plants: an analysis based on the underlying carbon economies of alpine and lowland Poa species. Functional Ecology 10, 698–707.[CrossRef]
Atkin OK, Bruhn D, Hurry VM, Tjoelker MG. 2005b. The hot and the cold: unraveling the variable response of plant respiration to temperature. Functional Plant Biology 32, 87–105.[CrossRef]
Atkin OK, Evans JR, Ball MC, Lambers H, Pons TL. 2000b. Leaf respiration of snow gum in the light and dark: interactions between temperature and irradiance. Plant Physiology 122, 915–923.
Atkin OK, Holly C, Ball MC. 2000a. Acclimation of snow gum (Eucalyptus pauciflora) leaf respiration to seasonal and diurnal variations in temperature: the importance of changes in the capacity and temperature sensitivity of respiration. Plant, Cell and Environment 23, 15–26.[CrossRef]
Atkin OK, Scheurwater I, Pons TL. 2005a. High thermal acclimation potential of both photosynthesis and respiration in two lowland Plantago species in contrast to an alpine congeneric. Global Change Biology (in press).
Atkin OK, Tjoelker MG. 2003. Thermal acclimation and the dynamic response of plant respiration to temperature. Trends in Plant Science 8, 343–351.[CrossRef][ISI][Medline]
Bednarz CW, van Iersel MW. 2001. Temperature response of whole-plant CO2 exchange rates of four upland cotton cultivars differing in leaf shape and leaf pubescence. Communications in Soil Science and Plant Analysis 32, 2485–2501.[CrossRef]
Berry JA, Björkman O. 1980. Photosynthetic response and adaptation to temperature in higher plants. Annual Review of Plant Physiology 31, 491–543.[ISI]
Berry JA, Raison JK. 1981. Responses of macrophytes to temperature. In: Lange OL, Nobel PS, Osmond CB, Zeigler H, eds. Physiological plant ecology. I. Responses to the physical environment. Berlin: Springer-Verlag, 277–338.
Blackmann GE, Black JN, Kemp AW. 1955. Physiological and ecological studies in the analysis of plant environment. X. An analysis of the effects of seasonal variation in daylight and temperature on growth of Helianthus annuus in the vegetative phase. Annals of Botany 19, 527–548.
Bloor JMG, Grubb PJ. 2004. Morphological plasticity of shade-tolerant tropical rainforest tree seedlings exposed to light changes. Functional Ecology 18, 337–348.[CrossRef]
Bradshaw AD. 1965. Evolutionary significance of phenotypic plasticity in plants. Advances in Genetics 13, 115–155.
Bruhn D, Leverenz JW, Saxe H. 2000. Effects of tree size and temperature on relative growth rate and its components of Fagus sylvatica seedlings exposed to two partial pressures of atmospheric [CO2]. New Phytologist 146, 415–425.[CrossRef]