SEARCH

SEARCH BY CITATION

Keywords:

  • adaptive phenotypic plasticity;
  • Arabidopsis thaliana;
  • artificial selection;
  • Brassicaceae;
  • flowering time;
  • photoperiod;
  • red-to-far-red ratio (R : FR)

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. References

Covariation between light quality- and photoperiod-mediated phenotypic plasticity was investigated using Arabidopsis thaliana. Three episodes of artificial selection were imposed on an index that quantified the plastic response to reduced red to far-red ratios (R : FR), with higher index values indicating greater plasticity. Relative to control lines, two high plasticity (HP) lines showed 1.6- and 2.4-fold increases in the index; low plasticity (LP) lines showed 1.4- and 1.1-fold decreases. A factorial experiment combining high and low R : FR conditions with long and short photoperiods assessed indirect consequences of selection on plasticity. Despite divergent R : FR-mediated plasticities in HP vs. LP lines, all four lines showed increases in photoperiod-mediated responses and decreases in mean leaf number. Complex relationships among trait means, plasticities and underlying mechanisms caution against generalizing about the genetic architecture of plastic traits. Partially independent developmental and evolutionary responses to R : FR and photoperiod are somewhat unsurprising, given this species’ cosmopolitan nature.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. References

Plants and other organisms respond developmentally, via phenotypic plasticity, to many different environmental stimuli. Populations may also respond evolutionarily, via natural selection, to multiple environmental pressures simultaneously. Constraints on the adaptive evolution of traits and their associated plasticities may reflect details of the underlying physiological mechanisms and biochemical pathways involved in detection of and plastic responses to multiple stimuli (e.g. Dorn et al., 2000; Bochdanovits et al., 2003; Davidowitz et al., 2003). Although it is widely recognized that such constraints can either impede or facilitate adaptive evolution, we often lack information about the genetic architecture of plasticity (Windig et al., 2004). Artificial selection studies are useful for dissecting the quantitative genetic architecture of traits and their associated plastiticies (Scheiner, 2002; Bochdanovits & de Jong, 2003; Bochdanovits et al., 2003; Brakefield, 2003). This is a particularly appealing approach when working with model organisms, where a wealth of information about physiological and genetic mechanisms is available.

In this paper, we report an artificial selection study conducted with Arabidopsis thaliana (Brassicaceae), motivated by our interest in four constraints. First, and very fundamentally, natural populations may not harbour significant genetic variation for plasticity. This would result in the inability of a population to respond to selection targeting the phenotypic plasticity of a trait. Second, a genetic correlation between the mean of a trait and its phenotypic plasticity can result in plasticity evolving as a correlated, indirect response to direct selection on the trait itself. Alternatively, as discussed in this paper, selection targeting the plasticity of a trait might indirectly shift the mean of the trait (reviewed by Scheiner, 2002). In general, such correlations must be characterized empirically rather than assumed (Bradshaw, 1965; Schlichting, 1986). A third issue is whether the phenotypic plasticity of a trait evoked by a given stimulus may condition whether and how other traits respond plastically to the same stimulus (Relyea, 2001; Weinig & Delph, 2001; Weinig, 2002; Bochdanovits et al., 2003; Davidowitz et al., 2003). A fourth and final concern is especially important because all organisms, including plants, experience variation in multiple environmental stimuli simultaneously. The phenotypic plasticity of a particular trait induced by one stimulus may affect that trait's plasticity to other stimuli (Bell & Lechowicz, 1994; DeWitt et al., 2000; Dixon et al., 2001; Davidowitz et al., 2003; Relyea, 2003,2004).

To investigate these four constraints, the research described in this paper takes advantage of the model annual A. thaliana and focuses on the timing of the transition from vegetative growth to reproduction, often referred to as ‘flowering time’ or ‘bolting time’. Flowering-time traits in Arabidopsis are known to show plastic responses to many environmental factors (e.g. temperature regimes, plant nutrition: Putterill et al., 2004), and the magnitude of these plastic responses can vary considerably depending on the genotypes considered (Pigliucci, 2002,2003; Pigliucci et al., 2003). Here, we narrow our focus to two light cues – red to far-red ratio (R : FR) and photoperiod – and to the plasticity of the number of rosette leaves at bolting. The number of rosette leaves at bolting is an indication of the developmental stage at which the transition to reproductive maturity occurs. With some exceptions, it is correlated with the chronological timing of bolting (measured in days) and with the developmental and chronological timing of flowering per se. For these two light stimuli, the directions of plastic responses in Arabidopsis are predictable, with reduced R : FR conditions generally accelerating bolting and short photoperiods delaying it (Koornneef et al., 1991; Mitchell-Olds, 1996; Pigliucci, 2002).

Using a sample of Arabidopsis genotypes derived from a naturally variable population, we used short-term artificial selection to create lines with either increased or decreased R : FR-mediated plasticity. These lines were then used to conduct a factorial experiment exposing all lines to combinations of high and reduced R : FR and long and short photoperiods. Data from this experiment are used to address four questions about the evolution of bolting time and its plasticity to two different light stimuli. (1) Was there a direct response to selection? This would confirm that a wild population harboured heritable variation for the R : FR-mediated plasticity of the focal trait, indicating the potential for shade-avoidance responses to evolve if natural selection regimes are heterogeneous at appropriate spatial or temporal scales (Schmitt et al., 2003; Huber et al., 2004). (2) Did changes in the focal plasticity indirectly select for a change in the associated trait's mean? If not, results would support the suggestion that the plasticity of a trait and the trait itself can evolve as independent characters (Bradshaw, 1965; Schlichting, 1986; Scheiner, 2002). (3) Do changes in the focal plasticity indirectly select for a change in the R : FR-mediated plasticity of a related trait, the number of days between germination and bolting? A strong indirect response to selection would corroborate the genetic and functional integration of not only these two traits but also their corresponding plasticities (Mitchell-Olds, 1996). (4) Do direct changes in the focal, R : FR-mediated plasticity of a trait also indirectly alter the photoperiod-mediated plasticity of that trait? Such covariation of R : FR- and photoperiod-mediated plasticities would be consistent with substantial overlap in the actions of relevant photoreceptors (and downstream signal transduction pathways) and may influence whether the two plasticities can evolve independently.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. References

Plant materials

The study population was derived from a wild population of A. thaliana (Brassicaceae), a highly selfing temperature annual. It was originally collected in Kendalville, Michigan, USA. A bulk sample from the Kendalville population has been maintained by Lehle Seed Inc. (Round Rock, TX, USA), and a previous study in our lab had established a total of 68 full-sib maternal families from this bulk collection (Camara et al., 2000). All 68 of these were used as the base population for artificial selection. The 68 random families in the base population are referred to as ‘founding genotypes’. Beyond this base generation, our artificial selection protocol results in maternal families that are not necessarily random, and we therefore refer to them as ‘sublines’. Also, given the almost completely selfing mating system (Abbott & Gomes, 1989), we chose not to impose outcrossing artificially. This departure from some previous studies (e.g. Ward et al., 2000; Fischer et al., 2004) mirrors the choice made by other researchers (e.g. Mauricio, 1998). The realism and potential biases introduced by this choice are discussed later in this report.

Growing conditions

All studies were conducted in an air-conditioned, artificially lit (see below) walk-in growth room, with high plant densities. Seeds were imbibed in the dark for 48 h on Petri plates containing an agar-based medium of MS Basal Salts (Sigma, St. Louis, MO, USA). Plates were then randomly assigned to light treatments. After 1 week, seedlings were transplanted into small pots in 96-pot greenhouse flat inserts, with each pot containing about 120 mL of standard greenhouse growth medium (Fafard #2, Agawam, MA, USA) and returned to their assigned light treatment. Flats were bottom watered twice weekly. Temperature was maintained at 22–24 °C.

Characterizing the base population

We grew 68 founding genotypes in two light treatments: control and reduced R : FR, both with an 18L/6D photoperiod. In the control treatment we used a bank of seven standard fluorescent tubes (F20T12 Sylvania, Danvers, MA, USA) interspersed with four 25 W incandescent bulbs. In the reduced R : FR, three of the four standard tubes were replaced with specialized tubes that emit far-red wavelengths (F20S-FR74 Toshiba Ltd., Tokyo, Japan), with standard and specialized tubes alternated and again interspersed with four 25 W incandescent bulbs. The R : FR ratios were approximately 1.0 and 0.5 in the control and reduced treatments, respectively. Quantities of photosynthetically active radiation was approximately the same in both treatments. There were seven replicate flats per treatment, with one individual per genotype in each flat (n = 952). In this and in follow-up experiments, flats were rotated frequently to minimize any effects of patchiness in the light treatments.

All plants were checked at least every other day for bolting, indicated by the appearance of ∼2 mm of an elongating inflorescence. We recorded the number of days between bolting and initial exposure of seeds to light; we also counted the number of rosette leaves. For each founding genotype, the leaf number data was used to calculate a mean plasticity index defined as the proportional change in mean leaf number when plants are exposed to reduced R : FR conditions [=1− (genotype mean in low R : FR/genotype mean in high R : FR)]. This index was used as the basis for selection.

To evaluate whether this base population harboured genetic variation for the plasticity of either bolting time trait (i.e. leaf number at bolting or days to bolting), we analysed data using SAS PROC MIXED and restricted maximum likelihood (REML) (SAS Institute Inc., 1999). We conducted separate analyses for the two traits. Our mixed models included two levels of the R : FR treatment [fixed effect, tested with an F-value and Satterthwaite approximated degrees of freedom(d.f.)], genotype, and the genotype-by-R : FR treatment interaction (random effects, tested with Wald's Z). Results of these tests did not differ from tests involving sequential comparison of models with individual random effects included vs. excluded (Littell et al., 1996). If a population harbours genetic variation for a trait (expressed within environments or averaged across environments), there should be a significant variance component for the genotype effect. A broad-sense heritability (H2) can be estimated by the ratio of this variance component to total variance. If a population harbours genetic variation for plasticity, there should be a significant variance component for the genotype-by-R : FR treatment interaction term; a broad-sense heritability for the plasticity of a trait can be calculated as the ratio of this variance component to total variance (Scheiner, 1993; Pigliucci, 2001; van Kleunen et al., 2002). Broad-sense heritabilities are often considered to be inflated in comparison to narrow-sense heritabilities, but are perhaps more appropriate and relevant here for two reasons: we are working with a naturally selfing plant and our selection protocol targets plasticity, which is an attribute of a group of individuals with similar genotype (e.g. inbred families, clones) rather than individuals.

Establishing the lines

Six lines were initiated from the base population of 68 founding genotypes: two ‘high plasticity’ (HP1, HP2) and two ‘low plasticity’ (LP1, LP2) selection lines and two control lines (C1, C2). We used a line-sorting procedure that narrowed the number of founding genotypes gradually, across three episodes of selection, rather than all at once. This procedure maintained a large number of sublines in each line across several generations, avoided any artificial disruption of this species’ highly selfing mating system, and sampled multiple maternal plants of the same genotype in an effort to represent any potential within-family variation.

Using the plasticity index for each genotype, we identified the top (or bottom) 34 genotypes in the base population. For the 17 genotypes in the top (or bottom) quartile, three randomly selected maternal plants (51 maternal plants total) were chosen to contribute seed for three sublines in the next generation. For 17 genotypes in the next highest (or next lowest) quartile, one randomly selected maternal plant contributed seed for just one subline in the next generation. (To minimize heterogeneity of maternal effects, only maternal plants grown under high R : FR conditions contributed seed from one generation to the next.) To create control lines, we randomly chose 17 founding genotypes and drew seed from three randomly selected maternal plants and another 17 founding genotypes from which we drew seed from just a single randomly selected maternal plant. Selection procedures for the HP, LP and C lines were repeated to produce a second line of each type, drawing from the same 68 founding genotypes in the base population, but not necessarily from the same maternal plants (6 lines × 68 sublines = 408 sublines). Since the same founding genotypes may have contributed seed to both selection lines, we were careful to track the founding genotype of each subline throughout the study. Also, replicate lines are not completely independent with respect to the initial episode of selection.

During the subsequent episodes of selection, however, sublines within all six lines were handled independently. This entailed growing all 408 sublines for a second and a third generation in the high and low R : FR treatments in the growth room, with three individuals per subline per treatment (n = 2448). Individual plants were scored for leaf number and the R : FR plasticity index was calculated for each subline in each generation. Between the second and third generation, this index was used to rank and select sublines for each HP and LP line; sublines were randomly selected for each control line. A final episode of selection occurred after the third generation and prior to the follow-up experiment for evaluating the selection lines. The R : FR plasticity index was used to rank sublines and the top (or bottom) 18 sublines were selected within each line, with one randomly selected maternal plant per subline contributing seed. This shift in the selection protocol was logistically necessary because the experiment to evaluate selection lines involved four rather than just two light treatments and our growth room could not accommodate multiple individuals from all 408 sublines in all four environments.

For all six lines, selection differentials, S were calculated for the base, second and third generations of the selection experiment as the difference between the mean of selected parents and the mean of all individuals in the parental generation before selection was imposed. Means were weighted since sublines differed in the number of progeny contributed (see above). The response to selection, R was calculated as the difference between the current generation's mean as compared to the previous generation's mean. For the four selection lines, S and R were standardized by dividing each line's differential and response by the differential and response for the C1 and C2 lines averaged. Standardized responses to selection were plotted against standardized selection differentials. Finally, cumulative responses and differentials across the three episodes of selection were calculated and used to estimate realized heritabilities, h2 = R/S.

Loss of sublines; expected and actual sorting of sublines

In the base population study, 42 of 952 plants (4.4%) failed to germinate or died as seedlings. No sublines were lost. In the next two generations, lack of germination or early mortality sometimes resulted in loss of sublines. After the second generation, three sublines were lost from the HP2 line, two each from HP1 and LP2 and one each from C1, C2 and LP1. After the third generation, one additional subline went extinct in LP1.

Our study is clearly a departure from experimental evolution studies conducted with out-crossing species, because Arabidopsis has a highly selfing mating system and we did not experimentally impose an out-crossed mating scheme (given its artificiality for this species). Potential limitations and biases introduced by using a selfing species are discussed later in this report. Such an approach also entails selection via line sorting, and here we briefly sketch its potential impact given the protocols described above. At one extreme, the final 18 sublines in any given selection or control line could represent 18 founding genotypes (i.e. of a total of 68 in the base population). At the other extreme, the three episodes of selection described could result in representation by a minimum of two founding genotypes. Also, since replicate lines were derived from a common pool of founding genotypes, ‘sharing’ of founders was possible and is expected to be more prevalent in selection as compared to control lines. To examine the extent to which these potential line-sorting artifacts are actually present in our lines, we tracked each subline's founding genotype in all six lines.

Follow-up experiment: evaluating lines

We conducted a factorial experiment that exposed all sublines to high and low R : FR conditions and to both 12L/12D (‘short’) and 18L/6D (‘long’) treatments. In the short day treatment, 12 h of the high or low R : FR lights described above were used for the short treatment (see Characterizing the base population). In the long day treatment, these conditions were augmented with 6 h of dim incandescent light (i.e. low R : FR at the end of the day). A light-proof curtain shielded plants in the short photoperiod treatment from this extended period of dim light. The ecological realism of these conditions represents a compromise between two other experimental objectives. First, we aimed at equalizing the quantity of PAR not only between contrasting R : FR conditions, but also between short and long photoperiods. Second, we chose ‘short’ and ‘long’ photoperiod that would not unduly prolong the experiment and such that the ‘short’ treatment elicited a significant delay in bolting time.

The experiment involved a total of 7 × 18 = 126 sublines (n = 2106), with the seventh line composed of 18 sublines randomly drawn directly from the base population (saved as seed). If fewer than three plants per treatment were available for estimating a subline's trait mean and plasticities, the subline was omitted from the analysis (four from C2; two each from HP2, LP1 and LP2; one from C1). Plants grown from saved seed from the base population behaved anomalously, bolting with many fewer leaves compared to all other lines and expressing different phenotypes than in the base population study. This line was exposed to unique seed storage, and studies with other Arabidopsis genotypes have demonstrated that seed age, stratification and maternal environment effects can affect life-history traits such as flowering time (Shaw et al., 2000; Munir et al., 2001). Although we included data from the base line in our analyses, this paper compares the HP and LP selection lines to the average of the controls lines (C).

Data were analysed using two approaches. First, with data from individual plants, we performed a mixed model analysis using SAS PROC MIXED with REML. Leaf number data were square root transformed to improve normality. F-values with Satterthwaite approximated d.f. were used to test the fixed effects of selection line, R : FR treatment, photoperiod treatment and two- and three-way interaction terms. Variance components were calculated and tested for significance with Wald's Z statistic for four random effects, all nested within selection line: subline, subline-by-R : FR treatment, subline-by-photoperiod treatment and subline-by-R : FR treatment-by-photoperiod treatment.

Data from the follow-up experiment were also used to calculate each subline's R : FR-mediated plasticity index for leaf number (as described above), a comparable R : FR-mediated plasticity index for days to bolting and a photoperiod-mediated plasticity index. This latter index was defined as the proportional acceleration in mean leaf number when plants were exposed to long photoperiods [=1 − (subline mean for trait in long photoperiod/subline mean for trait in short photoperiod)].

For these three indices and for leaf number in high R : FR conditions, we examined differentiation among selection lines by conducting univariate one-way anovas and two orthogonal contrasts: between the two control lines and the four selection lines and between the two HP and two LP lines. Formal tests for unequal variance among selection lines (Levene's test) found heteroscedasticity for one of the plasticity indices (R : FR plasticity index for leaf number) and for leaf number and we could not successfully homogenize variances using transformations of either variable. In the case of leaf number, this is most likely because there is a bimodal distribution in the L1 and L2 lines. However, one-way anovas are robust to violations of heteroscedasticity and normality assumptions, particularly when samples sizes are well balanced, as in this case (Zar, 1999, p. 185). For simplicity, we present parametric analyses of untransformed data, with the caveat that deviations from normality potentially compromise single d.f. planned contrasts (Sokal & Rohlf, 1995).

To evaluate indirect responses to selection, we examined genetic correlations, estimated as Pearson product-moment correlations based on subline (i.e. family) means within HP, LP and C lines. We focused specifically on correlations between the plasticity subject to direct selection (i.e. the R : FR-mediated plasticity index for leaf number) and (1) the mean of that trait, (2) the R : FR-mediated plasticity index for days to bolting and (3) the photoperiod-mediated plasticity index for leaf number. This method of estimating genetic correlations has been criticized as biased (Lynch & Walsh, 1998), but the approach has been explored and supported by simulation studies (Roff & Preziosi, 1994; Windig, 1997) and employed for studies with Arabidopsis (Mauricio, 1998) and with the clonal plant Ranunculus reptans (van Kleunen et al., 2002). It is an appropriate approach here because plasticity indices cannot be scored on individuals, but require estimation from subline means (i.e. family means).

To supplement our examination of genetic correlations, we used an ancova approach that essentially asks whether direct and indirect responses to selection are linearly interdependent. The ancova models were modifications of the among-line anova. In all models, the plasticity index for leaf number was the dependent variable. Each of the three ancova models added one of the following continuous covariates and its interaction with the main effect of line: (1) mean leaf number, (2) R : FR-mediated plasticity index for days to bolting and (3) photoperiod-mediated plasticity index for leaf number. F-values were used to evaluate the significance of the covariate effect and the covariate-by-line interaction term (Willis, 1996).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. References

Base population

Averaging across all plants, bolting occurred with only 9.0 ± 2.4 leaves and after 21.4 ± 4.2 days in low R : FR, as compared with 13.8 ± 5.1 leaves and after 29.7 ± 7.0 days in high R : FR [means ± standard deviation (SD)]. The R : FR treatment had a statistically significant effect for both traits (Table 1). The mean plasticity index for leaf number at bolting was 0.32, with the value of the index ranging as low as 0.075 and as high as 0.641. There was variation among sublines for the plasticities associated with both traits and for trait means, indicated by the significant subline-by-treatment interaction term and the significant main effect of subline (Table 1).

Table 1.  Variation in the base population for rosette leaf number at bolting and the timing of bolting as affected by maternal genotype and red to far-red ratio (R : FR).
 Rosette leaf number at boltingDays from germination to bolting
Variance componentH2F or ZVariance componentH2F or Z
  1. H2 refers to broad-sense heritability estimates for the traits and for the trait plasticities (see text for explanation).

  2. For F-values, d.f. = 1.67.

  3. * * *indicates significance at the P < 0.001 level.

R : FR treatment 155*** 300***
Genotype7.40.454.16***17.70.524.65***
R : FR × Genotype4.50.275.01***6.40.184.57***
Residual error4.5  9.9  

Follow-up experiment

In the follow-up study, averaging across all plants in all conditions, bolting occurred with as few as 3 and as many as 62 rosette leaves (mean = 24.1) and as early as 14 and as late as 139 days (mean = 49 days). Our mixed model anovas indicate that both environmental and genetic factors, as well as complex interactions among them, contributed significantly to this variation (Table 2). The greatest developmental and chronological delay (i.e. most leaves at bolting, most days before bolting) occurred in high R : FR and short photoperiod; the greatest acceleration occurred in low R : FR and long photoperiod. Averaging across all sublines, the main effects of both the R : FR Treatment and the Photoperiod Treatment are both significant but in an additive manner (i.e. there is not a significant R : FR Treatment-by-Photoperiod Treatment interaction term). Differentiation for R : FR- and photoperiod-mediated plasticity among selection lines is supported by the statistical significance of the selection line by R : FR interaction term (for leaf number) and the selection line by photoperiod interaction term (for both traits). Variation among sublines within selection lines for both R : FR- and photoperiod-mediated plasticities of both traits were detected, as indicated by significant three-way interaction terms (subline × R : FR × photoperiod).

Table 2.  Variation in rosette leaf number at bolting and the timing of bolting in the follow-up experiment associated with subline, selection line, red to far-red ratio (R : FR) and photoperiod.
Source of variationRosette leaf number at boltingDays from germination to bolting
Variance componentd.f.F or ZVariance componentd.f.F or Z
  1. For F-values, degrees of freedom are Satterthwaite approximated.

  2. * * *indicates effects significant at the P < 0.001 level; **P < 0.01; *P < 0.05

  3. n.s.indicates effects for which P > 0.05.

Selection line6.1188.53***6.1177.07***
R : FR treatment1.97136***1.19652.6***
Photoperiod treatment1.108929***1.112387***
R : FR × Photoperiod1.1020.01n.s.1.1950.02n.s.
Selection line × R : FR6.973.00**6.1961.77n.s.
Selection line × Photoperiod6.10812.16***6.1126.02***
Selection line × R:FR × Photoperiod6.1020.65n.s.6.1950.87n.s.
Subline (selection line)47.1517.09***125.416.32***
Subline × R : FR0.466410.38n.s.01
Subline × Photoperiod0.433010.35n.s.14.2111.9*
Subline × R : FR × Photoperiod7.21214.26***46.1816.61***
Residual error17.24  79.71  

Expected and observed line sorting

By tracking the founding genotypes, we found that the number of founders surviving in the C lines (12 in C1, 14 in C2, 5 shared) differed little from the number surviving in the LP lines (10 in LP1, 11 in LP2, 4 shared; nonsignificant test of homogeneity, inline image = 0.72.). The HP lines narrowed to fewer founders (five in HP1, six in HP2, three shared), but differences between control and HP lines in the impact of line sorting were only marginally significant (inline image = 7.61, 0.10 > P > 0.05).

Direct response to selection: differentiation among lines for R : FR-mediated plasticity

Since the study of the base population detected statistically significant variation among genotypes and for the treatment-by-genotype interaction term, a short-term response to selection was expected. Indeed, the R : FR-mediated plasticity index for leaf number decreased in both LP lines and increased in both HP lines, with the HP2 line showing the greatest response to selection. The plasticity of leaf number varied across generations, however, with all lines showing lower plasticity indices in the different environment used for the follow-up study (fourth generation) as compared to the second and third generations (Table 3). Responses to selection for the HP1, HP2, LP1 and LP2 lines were therefore standardized with respect to the average of the C1 and C2 lines. These responses were then plotted against cumulative selection differentials (Fig. 1) and also used to calculate realized heritabilities. Realized heritability estimates based on the response of the HP2 selection line were comparable to the broad-sense heritability estimated in the base population (Table 1), but estimates based on the other three selection lines were smaller.

Table 3.  The red to far-red (R : FR) ratio-mediated plasticity of rosette leaf number at bolting, as quantified by plasticity indices (±se) (see text for details).
 Selection lineBOverall
HP1HP2C1C2LP1LP2
  1. HP, high plasticity; LP, low plasticity, C, control lines.

Generation 20.38 ± 0.0210.39 ± 0.0190.35 ± 0.0190.34 ± 0.0210.32 ± 0.0210.35 ± 0.0160.35 ± 0.009
Generation 30.39 ± 0.0220.40 ± 0.0320.32 ± 0.0160.31 ± 0.0220.32 ± 0.0190.30 ± 0.0280.34 ± 0.010
Follow-up Study0.22 ± 0.0350.32 ± 0.0410.13 ± 0.0590.14 ± 0.0370.10 ± 0.0260.12 ± 0.0580.21 ± 0.0460.17 ± 0.017
image

Figure 1. Direct responses to selection standardized relative to the mean of the C1 and C2 control lines. Artificial selection targeted the plasticity index for the red to far-red (R : FR) ratio-mediated plasticity of rosette leaf number at bolting. For each selection line, h2 is a realized heritability estimate calculated using the cumulative selection differential and the standardized response to selection.

Download figure to PowerPoint

Differentiation among lines for the R : FR-mediated plasticity index (for leaf number) was significant, based on our one-way anova (Fig. 2a). A contrast of the mean of the two HP lines vs. the mean of the two LP lines was highly significant (F1,108 = 18.2, P < 0.0001). Standard errors around the means of the plasticity indices were smaller in the two LP lines as compared to the HP and C lines (Fig. 2a), even though both the LP1 and LP2 lines showed a strongly bimodal distribution for leaf number at bolting (averaged across treatments, or in either treatment; see additional details below). With the method used for calculating the plasticity index, low values are theoretically possible with any number of leaves, as our results illustrate.

image

Figure 2. Means (±1 se) for the selection and control lines for (a) the red to far-red (R : FR) ratio-mediated plasticity index for leaf number at bolting, (b) mean number of leaves in high R : FR and long photoperiods conditions, (c) the R : FR-mediated plasticity index for days to bolting and (d) the photoperiod plasticity index for leaf number at bolting. A one-way anova found significant differentiation among selection lines for the R : FR-mediated plasticity index for leaf number (F6,108 = 3.95, P < 0.0013). The text provides additional details of statistical tests focusing on specific contrasts among selection and control lines and ancovas.

Download figure to PowerPoint

Indirect responses to selection

There was significant differentiation among lines for mean leaf number (Fig. 2b; F6,108 = 7.07, P < 0.001), with higher means in the C lines as compared to HP and LP selection lines (statistically significant orthogonal contrast of the two C lines vs. all four selection lines, F1,108 = 26.2, P < 0.0001). Overall (i.e. across all lines) and in the control lines, the correlation of mean leaf number and its R:FR-mediated plasticity was positive. Within HP and LP lines, this correlation was near zero (Table 4). This trait was more variable in the LP lines than in other lines (Fig. 2b); for this trait, both the LP1 and LP2 lines showed a strongly bimodal distribution (data not shown). Specifically, some sublines bolted with as few as four leaves in both high and low R : FR environments; others had as many as 25–35 leaves in both environments. This bimodality most likely accounts for the results of our Levene's test, which indicated significant heterogeneity of variance for this trait among lines (F6,108 = 13.9, P < 0.0001) and for the failure of efforts to homogenize variance by transforming this variable. Results of anovas and contrasts for this trait should be interpreted cautiously.

Table 4.  Correlations of plasticity indices and trait means, approximated with subline means from the follow-up study.
 Mean leaf number (high R : FR and long photoperiod)R : FR-mediated plasticity index (days to bolting)Photoperiod-mediated plasticity index (leaf number)
  1. R : FR, red to far-red ratio.

  2. Bold indicates significance at the P < 0.01 level.

  3. Sample sizes: for high plasticity (HP), n = 34; for Control, n = 31; for low plasticity (LP), n = 32.

R : FR-mediated plasticity index (leaf number)HP −0.08HP0.58HP −0.08
Control0.52Control0.93Control0.47
Low0.04LP0.58LP −0.08
Mean leaf number (high R :FR and long photoperiod)  HP0.03HP0.77
  Control0.47Control0.79
  Low0.02LP0.90
R : FR-mediated plasticity index (days to bolting)    HP0.07
    Control0.44
    LP −0.03

For each selection line's R : FR plasticity index for days to bolting, the pattern of variation was quite similar to what was observed for the R : FR plasticity index for leaf number, which was the target of selection. Specifically, R : FR plasticity of days to bolting decreased slightly in both LP lines and increased in both HP lines, with the greatest change occurring in the HP2 line (Fig. 2c). There was a very strong positive correlation of these two plasticity indices overall (r = 0.85) and within the C and HP selection lines; a somewhat weaker but still positive correlation was detected in LP selection lines (Table 4).

The photoperiod plasticity index for leaf number (Fig. 2d) was higher in the HP lines as compared to the two C lines, which themselves differed slightly (but not significantly) for this index. In the LP lines, this index was also high relative to the C lines. A contrast of the mean photoperiod plasticity index of the two C lines vs. the mean of all four selection lines confirmed the significance of this differentiation (F1,108 = 14.19, P < 0.0003).

To assess whether direct and indirect responses to selection were interdependent, we used ancova models that were modifications of the among-line anova presented above. In all models, the dependent variable is always the R : FR-mediated plasticity index for leaf number and ‘line’ is main effect, with different covariates and interaction terms added to investigate indirect responses to selection (see Methods). When the covariate was the R : FR-mediated plasticity index for days to bolting, both the selection line main effect and the selection-line-by-covariate interaction term were strongly significant, suggesting that the R : FR-mediated plasticities of these two traits are not independent genetically. When the covariate was mean leaf number, the interaction term and the covariate term were both significant, but only modestly so. For the photoperiod-mediated plasticity index for leaf number, the covariate was nonsignificant regardless of whether the nonsignificant interaction term was included in the model (Table 5). These latter two results suggest that leaf number at bolting, and the R : FR- and photoperiod-mediated plasticities of leaf number are either genetically independent or related in a quite complicated rather than a simple, linear fashion.

Table 5.  Analysis of covariance tables.
Continuous covariateSourced.f.FP-value
  1. In all models, the red to far-red (R : FR) ratio-mediated plasticity index of leaf number is the dependent variable and selection line is the fixed main effect. The interaction between the continuous covariate and selection line was only included when significant.

Mean leaf number in high R : FRSelection line62.610.0217
Covariate14.390.0387
Selection line × Covariate12.40.0332
Error101  
R : FR-mediated plasticity of days to flowerSelection line65.610.0001
Covariate12090.0001
Selection line × Covariate15.530.0001
Error101  
Photoperiod-mediated responsiveness of leaf numberSelection line64.580.0004
Covariate13.730.0561
Error107  

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. References

Because we detected significant genotype-by-environment interactions and among genotype variation in the base population, we were not surprised to observe responses – admittedly modest – to artificial selection. Like other experimental Arabidopsis populations (Callahan et al., 2005), the Kendalville population harbored a genetic correlation between leaf number at bolting and days to bolting, consistent with other studies showing that this constraint is maintained, even in the face of mutation and selection in combination (Pigliucci, 2003). Our data lend additional support to the idea that this correlation reflects tight functional and physiological integration of these traits and, very likely, substantial overlap in their genetic basis (Mitchell-Olds, 1996). Our most striking result, a change in the photoperiod-mediated plasticity of leaf number, was apparently mediated by a corresponding shift in mean leaf number rather than a direct response due to overlap in the photoreceptors that are jointly involved in R : FR and photoperiod signal detection and transduction.

Direct responses to selection: what limits them?

The assumption that A. thaliana is a highly selfing species is supported by a recent study of pollination biology (Hoffmann et al., 2003), several published estimates of mating system parameters (1–2%Snape & Lawrence, 1971; <0.3%Abbott & Gomes, 1989) and several surveys of population genetic structure reporting extremely low heterozygosity (Bergelson et al., 1998a,b; Clauss et al., 2002) and rather limited within-population genetic diversity, usually with just one or a few genotypes detectable (Kuittinen et al., 1997; Breyne et al., 1999; Clauss et al., 2002). Yet Jørgensen & Mauricio's (2004) recent survey of North American populations based on AFLP markers detected both among- and within-population genetic variation, with the latter actually accounting for a greater proportion of overall variation. This finding was based on intensive sampling within six populations, three of them in the vicinity of Lake Michigan and rather close to the native site of our Kendalville population. Such results suggest that multiple founding events create considerable genetic variability within natural populations of even highly selfing plants. This is consistent with our finding that a population can harbour genetic variation and respond to short-term selection, even if the protocol involves line sorting without any cross-pollination and recombination among genotypes.

The responses to selection observed in our study, while statistically significant, were somewhat modest. For example, the two LP lines showed quite limited response to selection and the impact of line sorting did not reduce the number of founding genotypes in these lines (as compared to randomly selected control lines). Indeed, only the HP2 line, which showed a larger response to selection (as compared to the HP1 line), provided a realized heritability estimate comparable to the broad-sense heritability estimated in the base population. Although this may reflect the tendency for broad-sense heritabilities to be inflated (e.g. due to maternal effects), Scheiner & Lyman's (1991) study of temperature-induced plasticity in Drosophila melanogaster also found that narrow-sense heritability estimates exceeded realized heritabilities.

That there was an asymmetrical response to selection, and that replicate HP lines differed in their responses to selection may reflect two sources of variation. First, our protocol of drawing seed from three rather than a single maternal plant in some selected sublines may have introduced maternal microenvironment effects, which can inflate estimates of within-population genetic variance. This inflation may have been greater in the HP2 than in the other line. Second, different founding genotypes were represented. For example, in the HP lines a total of only eight founding sublines survived, two unique to HP1, three unique to HP2 and three shared. Without greater replication of selection lines, it is not possible to gauge the relative contributions of these two sources of variation.

Several other experiments, including some involving out-crossing organisms, have also documented limited direct responses to selection on plasticity (Scheiner & Lyman, 1989,1991; Wijngaarden & Brakefield, 2001; van Kleunen et al., 2002). Interestingly, Scheiner & Lyman (1991) noted that they could not reduce plasticity to less than zero (i.e. to reverse the pattern of plasticity). Our results also suggest that it may be easier to select for increased plasticity than for reduced or reversed plasticity. Interestingly, this is consistent with Botto & Smith's (2002) survey of R : FR-mediated plasticity in 157 A. thaliana accessions, which found a marked skew toward higher than average plasticity. Asymmetrical responses to selection are a common feature of artificial selection studies, particularly if selected traits influence fitness (Hill & Caballero, 1992), but rigorously testing for and properly interpreting them requires greater line replication, longer-term selection and more sophisticated mating designs (e.g. Mackay et al., 1994).

Did trait means change in response to selection on plasticity?

Several artificial selection studies have imposed direct selection on a target trait and examined the evolution of plasticity as an indirect response (see also Brumpton et al., 1977; Falconer, 1990; Scheiner, 2002; van Kleunen et al., 2002). Our study involved direct selection only on a plasticity index, without imposing corresponding selection on the trait itself in other selection lines. Instead, we examined the evolution of trait means as an indirect response to selection on a plasticity index.

The correlation coefficient between mean leaf number and its R : FR-mediated plasticity in the base population study, r = +0.71 (n = 68, P < 0.001), as well as in the control populations, were quite similar to a highly significant correlation of r = +0.66 found in a study of Scandinavian accessions of Arabidopsis (Pigliucci et al., 2003). Assuming simple linear relationship between this trait and its plasticity, an increased leaf number is predicted for HP lines and a decrease in LP lines. Yet using a relativized plasticity index also might have the reverse effect by introducing a bias: plants with a larger number of leaves will not necessarily have a high index. For example, in a subline with 20 leaves, a decrease of five leaves in response to low R : FR conditions means an index of only 0.25 while for a subline with 15 leaves, a decrease of five leaves means an index of 0.33. Indeed, Botto & Smith's (2002) survey found that accessions with many leaves generally had limited plasticity to R : FR signals.

Contrary to either expectation, mean leaf number had neither a direct linear nor an inverse linear relationship with the R : FR-mediated plasticity index. Instead, our study found that selection increased mean leaf number in both HP and LP lines. While supporting the contention that the plasticity of a trait and the trait itself can be partially genetically independent (Bradshaw, 1965; Schlichting, 1986), this result may also indicate a more complex functional constraint. Namely, there is only one way to have greater plasticity to reduced R : FR: bolt with many leaves in high R : FR and very few in low R : FR. In contrast, a genotype can have a zero (or very low) plasticity index with any number of leaves, provided leaf number is maintained across environments. The LP1 and LP2 lines, which both showed a slight decrease in the mean of the plasticity index, also showed a decrease in the index's variability, even though both selection lines continued to harbour both many-leaved and few-leaved sublines after several rounds of selection.

Plasticity of flowering time traits, including leaf number at bolting, is very extensive in Arabidopsis and plants’ flowering habits respond to gradations of light signals in a graded fashion (Putterill et al., 2004). This is an important consideration, since our follow-up study involved a shift in conditions: just 12 h of full-intensity light augmented by 6 h of dim illumination as compared to the 18 h at full illumination during previous generations. This shift probably accounts for an increased mean leaf number at bolting and correspondingly lower plasticity indices in the follow-up study (Table 3). While the inclusion of control lines makes it possible to adjust for the effect of this environmental variation, an assumption underlying all of our interpretations is that the contrasting environments between the follow-up study and the base population study would not change the relative ranking of HP, LP and C lines.

Was there a correlation between the two flowering time traits (and their plasticities)?

In Arabidopsis, the developmental and chronological timing of flowering (i.e. number of rosette leaves at bolting and days to bolting) tend to be strongly and linearly related in many populations and environments. Using a recombinant inbred population, Mitchell-Olds (1996) documented a strong positive genetic correlation between reproductive age (i.e. time of flowering in days) and reproductive size (i.e. leaf number at flowering), mapped variation for these two traits to quantitative trait loci (QTLs) and demonstrated that plants flowering in fewer days but with more leaves were favoured by selection. While available field studies indicate that natural selection may sometimes oppose rather than reinforce the strong correlation between these two bolting time traits (e.g. Callahan & Pigliucci, 2002) we unfortunately lack information about natural selection regimes in the Kendalville habitat, making it difficult to determine whether the correlation we detected reflects an adaptive responses to past selection pressures or, instead, a genetic constraint that has been maintained in the face of selection.

Are changes in R : FR-mediated plasticity associated with altered responsiveness to photoperiod?

Such a prediction is consistent with recent progress in understanding how overlapping photoreceptors are involved in regulating R : FR- and photoperiod-mediated plasticities. Briefly, several flowering genes, notably FLOWERING LOCUS T (FT), are complexly regulated by the light sensitivity of upstream gene products. In particular, the transcription factor CONSTANS (CO) has its own transcription regulated by both cryptochrome and phytochrome photoreceptors (Simpson & Dean, 2002; Klejnot & Lin, 2004). Both types of photoreceptors play roles in entraining the circadian clock, which regulates the diurnal rhythmicity of CO mRNA levels. In addition, degradation of CO protein in the proteasome is promoted by phytochrome B but inhibited by phytochrome A and cryptochrome. Therefore, not only photoperiod but also the R : FR of natural light conditions can jointly affect CO protein levels and associated flowering time (Klejnot & Lin, 2004; Valverde et al., 2004).

This model of R : FR- and photoperiod-mediated responses does not account for variation among genotypes in mean leaf number. As discussed above, reduced R : FR-mediated plasticity can occur in two ways: by producing either very many or very few leaves in both high and low R : FR conditions. The same is true for a reduction in photoperiod-mediated plasticity of flowering time: genotypes with any number of leaves can show reduced plasticity to photoperiod, while genotypes with few leaves in long photoperiods have the greatest potential for showing a photoperiod response. (A. thaliana generally responds to short, noninductive conditions by increasing leaf number.) This functional assessment is consistent with genetic correlations and indirect responses to selection documented in our study. Based on subline means, R : FR-mediated plasticity and mean leaf number are not correlated. As an indirect response to artificial selection, both HP lines and both LP lines declined in leaf number relative to the two C lines. With this decrease in mean leaf number, there was also an increase in photoperiod response. This reflects the negative correlation between photoperiod-mediated plasticity and leaf number, which is significant overall and also within HP, LP and C lines. In sum, it appears that mean leaf number and R : FR plasticity may be able to evolve in a partially independent manner, but that the same may not be true for mean leaf number and photoperiod plasticity.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. References

To the extent possible, a population's quantitative genetic architecture should be interpreted integratively, taking into account both details about underlying genetic mechanisms (Windig et al., 2004) and the past and current environmental variability of natural environments (e.g. Huber et al., 2004; Stinchcombe et al., 2004). With respect to genetic mechanisms, our results suggest that a complete understanding of the relationship between R : FR- and photoperiod-mediated responses requires considering variation in leaf number, which itself may reflect ecotypic differentiation, environmental conditions or some combination of these two. With respect to the natural environments, we unfortunately lack specific information about the Kendalville natural habitat and cannot speculate about whether the specific pattern of co-variation found in our base population or our selection lines would indeed be selectively advantageous in that natural habitat. Yet it is obvious that a plant population at a given latitude experiences photoperiod as a predictable signal, while canopy shade and R : FR signals are likely to vary more idiosyncratically (e.g. Scheiner & Callahan, 1999; Callahan & Pigliucci, 2002). Given the cosmopolitan nature of Arabidopsis, it is perhaps unsurprising that responses to photoperiod and R : FR are partially independent both developmentally and evolutionarily.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. References