Introduction

Plant community composition and functioning are strongly influenced by interactions between individuals. Because of plant’s mobility restricted to seed dispersal and clonal growth, success of individual plant is highly dependent on its neighbours (Pacala and Levin 1996; Law and Dieckmann 2000). Plants perceive neighbours’ proximity as reduction in the ratio of red (R) to far-red (FR) radiation via the phytochrome system. The signal triggers a variety of responses known as shade avoidance (Smith and Whitelam 1997). This covers elongation of stems and petioles (Ballaré et al. 1991; Huber et al. 1998) decreased branching (Morgan and Smith 1979; Novoplansky et al. 1994; Niva et al. 2006) and tillering (Casal et al. 1985), more steep orientation of leaves (Ballaré and Scopel 1997) and acceleration of flowering (Halliday et al. 1994). Elongation of vertically growing organs has been interpreted as adaptive placing of the assimilative organs into better radiation conditions in the upper canopy layers (Schmitt and Wulff 1993). High level of plasticity in vertical growth responses has been assigned to high spatio-temporal predictability of the light supply in the vertical plane (Huber et al. 1998; Stuefer 1996).

Despite lower predictability of radiation supply in the horizontal direction, there are some examples of selective placement of plant organs into canopy gaps: bending stems towards free space (Ballaré et al. 1995; Novoplansky et al. 1990), cumulating plant’s stolons and leaves in sites with removed vegetation (Macek and Lepš 2003), growth away from vegetation patches (Sampaio et al. 2004), asymmetric development (lower density of nodes and stolons) at the interface between plants cultivated in pairs (Solangaarachchi and Harper 1989), selective placing of ramets outside of vegetation patches (Evans and Cain 1995) or turning leaves away from neighbours (Maddonni et al. 2002). Shading individual plant parts (one leaf of the pair) results in local reduction of cytokinin concentration (Pons et al. 2001). Cytokinin is known to regulate activity of buds (McSteen and Leyser 2005). It is thus possible that heterogeneous radiation conditions may result in asymmetric branching, which seems to be a very powerful way for exploring radiation heterogeneity. It has been repeatedly found in trees: more daughter ramets were placed in high-light patches (Zhang et al. 2006), more and larger buds developed on branches in sunny patches (Jones and Harper 1987; Sprugel 2002), and more lateral branches occurred on forest edges or gaps (Harper 1985; Young and Hubell 1991).

In herbaceous plants I am aware of only two unsuccessful experiments testing selective branching under heterogeneous radiation conditions, i.e. Novoplansky (1996) and Leeflang (1999). In the first experiment, asymmetric branching occurred only in plants exposed to damage stress. The result of the second experiment is disputable because most of the species used did not show branching response to changes in overall radiation. In order to increase the chance of a positive result, it seems reasonable to demonstrate that the investigated plants are capable of the branching response to changes in radiation conditions because absence of this response is not rare (Murphy and Briske 1994, Stuefer and Huber 1998).

Most research dealing with response of herbaceous plants to radiation or canopy heterogeneity has been conducted using dicots as a model. Although grasses have repeatedly demonstrated to respond to the decreased R/FR ratio (Casal et al. 1985; Deregibus et al. 1985; Evers et al. 2006; Monaco and Briske 2001), intensive investigation of the shade avoidance networks has only recently been explored (Kebrom and Brutnell 2007). Grasses share some regulatory pathways with dicots, but others are unique to them (Doust 2007). Response of grasses to radiation heterogeneity might thus differ.

Grasses produce a large number of vegetative branches which are of two types: tillers are developed from young buds close to the primary shoot apex within the leaf sheaths, or from dormant buds at the tussock base. The latter are usually placed in a certain distance from the mother tussocks and connected to them with rhizomes. The two tiller groups may be an example of developmentally programmed division of labour in clonal plants (Stuefer 1998), with the second group being aimed at foraging (Hutchings and Slade 1988). During the development, grass tillers produce adventive roots, and thus achieve at least partial independence (Doust 2007). Tussocks of caespitose grasses consist of a successively produced series of connected generations (ramet hierarchies), which function autonomously, for instance in resource sharing (Briske and Derner 1998). Independency of individual tillers or tiller groups may be strengthened due to the natural break-off points within tussocks (Wilhalm 1995). Tillers or ramet hierarchies may thus plastically respond to environmental heterogeneity independently from the whole plant as it occurs in ramets of clonal plants (Preston and Ackerly 2004; de Kroon et al. 2005). Different degree of independence are expected in the two tiller types.

For this study, Festuca rubra L. was used as a model plant for testing tillering response of grass to spatial radiation heterogeneity. The species grows in communities with horizontal radiation heterogeneity on a scale comparable to the tussock size and expansion rate (Skálová et al. 1999; Wildová et al. 2007). In these communities the plants perceive different radiation signals from different directions, and selective placing of tillers seems thus reasonable. The species shows pronounced tillering response, shifts in proportion of both tiller types and rhizome system architecture due to R/FR changes (Skálová et al. 1997; Skálová 2005). Because of the occurrence of a large number of clones with different tussock structure and response to R/FR, vegetativelly multiplied material was used. Single initial tillers were planted and exposed to three radiation regimes in two green-house experiments: full sunlight, full shading simulating canopy and heterogeneous radiation. In the first experiment I used young initial tillers and took into account their orientation relative to their mother tillers due to possible morphologic constraints (de Kroon et al. 1994). Older initial tillers, which unlike the young ones were able to get rise to both tiller types, were used in the second experiment.

Methods

Study species

The perennial grass Festuca rubra L. is a common species in temperate grasslands. Similar to Lolium perenne L. or Agrostis capillaris L. it is characterised by two types of clonal growth depending on the way of tillers’ offspring (Herben et al. 1994b): new tillers can either occur ‘intravaginally’ or ‘extravaginally’. Intravaginal tillers develop from axillary buds within the leaf sheaths of the mother tiller after about two plastochrons. If new tillers are not developed during this stage buds become dormant. During this stage, new, very short module primordia are formed in the buds. Tip direction is changed from vertical to horizontal (Rytova 1969). Bud activation results in rhizome formation with daughter tussocks placed at a certain distance from the mother tussocks. These tillers are further referred as extravaginal tillers.

Clones of F. rubra differ in morphological parameters both in the field and under garden cultivation, and during their growth response to changes in R/FR in the incident radiation in growth chamber experiments (Skálová et al. 1997; Skálová 2005). Therefore, vegetativelly multiplied material was used. The clones investigated come from a species-poor mountain meadow in the Krkonoše Mountains, the Czech Republic (the Severka settlement––latitude 50o41′42′′N, longitude 15o42′25′′E, altitude approx. 1100 masl). They have been grown in an experimental garden in Průhonice near Prague for more than 15 years.

Experiment 1

Two clones, Nr. 5 and 9, investigated by Skálová et al. (1997) and Pecháčková (1999), were used to study the interactions of directional response of tussocks and morphologic constraint, i.e. orientation relative to their mother tillers. Well-developed tillers were cut from the tussocks and cultivated in pots filled with common garden soil. Two or three-leaved daughter tillers that emerged during the two-month cultivation were used in the experiment. The tillers were individually transplanted into the centre of plastic pots (12 cm in diameter, 9 cm deep) and filled with a mixture of common garden soil and perlite (1:1). The soil surface was divided into quarters with coloured wires (Fig. 1). All tillers were planted in the same position, with the ventral side facing the same wire colour, enabling the positions of the first leaf of the initial tiller and its daughter tillers to be recorded.

Fig. 1
figure 1

Daughter tillers at the base of the initial tiller (a), and arrangement in an experimental pot with an initial tiller shift into the ventral pot half (b); D dorsal, V ventral

Full size image

The plants were exposed to three basic treatments: full shading, full sunlight and heterogeneous radiation. Full shading was achieved by placing a belt of double-layered green foil along the entire pot margin (Fig. 1, for details on the foil see Skálová and Krahulec 1992). The foil simulated canopy shading in the natural communities. Plants in the full sunlight treatment were surrounded by transparent foil of the same shape. R/FR in the pot centre at the soil level (vertical radiation) was 0.25 in the fully shaded treatment and 1.1 in the pots with transparent foil. In plants exposed to heterogeneous radiation, the green foil was placed only along a half of the pot (two quarters). When the sensor faced the green foil to capture lateral radiation, R/FR was 0.25 and when it faced the transparent foil, R/FR was 1.1. The light quality was measured with a SPh 100 photometer from Optické dílny Turnov, CR.

In order to investigate the possible interaction of morphological parameters of the initial tillers and radiation conditions, the green foil was placed in different positions relative to the initial tiller. Four groups were identified within the treatment: dorsal, ventral, left and right lateral side of the initial tiller shaded. There were 11 replicates for each clone in each treatment at the beginning of the experiment (i.e. 132 pots in total). However, only 100 plants (7–11 per clone and treatment) survived until the end of the experiment and were thus involved in the analyses. The pots were situated in east–west rows, with the shaded part of 50% of the heterogeneous radiation pots facing south and 50% facing north. The experiment was carried out in a greenhouse without any temperature and humidity regulation, and without additional shading; the plants were regularly watered.

Tiller number in the individual quarters and tussock height were recorded monthly from the beginning of the experiment in May until its end in September (i.e. five censuses). If the newly born tillers pushed the initial tiller out of the pot centre, the record was made again, with the centre of the recording system imaginarily shifted to the new position of the initial tiller. Aboveground biomass in the individual quarters were cut at the end of the experiment, dried at 80°C for 10 h and weighed.

Experiment 2

Older tillers of clone Nr. 13, investigated by Skálová and Krahulec (1992), were used in order to study response of extravaginal tillers, which were not produced by young initial tillers used in Experiment 1. It was not possible to trace orientation of the older tillers relative to the mother tillers due to the complicated structure of the old large tussocks, from which the tillers came. The tillers were cut from tussocks, rooted in water and individually transplanted into the centres of plastic pots as in the previous experiment. A similar shading system was also used, but with three quadrant types within the heterogeneous radiation treatment: no shading, full shading and intermediate shading (green foil along half of the quadrant margin). 32 plants were used in the experiment (i.e. 10 plants each for the no shading and full shading treatments and 12 for the heterogeneous radiation treatment). The cultivation conditions and recording system were similar, with censuses every two weeks from April until June (i.e. 6 censuses in total), and intravaginal and extravaginal tillers being recorded separately. Tillers or groups of tillers situated at least 0.5 cm from the main tussock and connected to it with at least 0.5 cm rhizome (final decision made after the plant excavation at the end of the experiment) were considered extravaginal.

Data analysis

Data were analysed using ANOVA with a split-plot design (S-plus 2000, Mathsoft Inc 2000). Tiller numbers were square-root transformed to achieve normal distribution.

The influence of the three basic treatments (full shading, full sunlight and heterogeneous radiation, i.e. intermediate shading), clone (if there were two, i.e. in Experiment 1), time and their interactions on total tiller number and tussock height were tested in both experiments. The number of intravaginal tillers in tussocks was used as another independent variable when investigating the number of extravaginal tillers to avoid the effect of tussock size.

In Experiment 1, morphological constraints in tussock development and the influence of heterogeneous radiation were investigated. In testing morphological constraints, plants exposed to the homogeneous treatment (i.e. full shading and full sunlight) were involved. The influence of clone, time, presence of the first leaf of the initial tiller, total tiller number in tussocks (to avoid the effect of tussock size) and interaction between identity of the tussock half and time on tiller number in individual tussock halves (both with regard to the initial tiller and pot centre) were analysed. In order to investigate the influence of heterogeneous radiation, the effect of treatment––full shading, full sunlight and intermediate shading (dorsal, right, ventral or left half of the tussock shaded), time, clone, total tiller number in the tussock and interaction between clone, treatment and time on tiller number in the dorsal, respectively right half was analysed. Because data obtained from the same plants were used in more analyses, Bonferroni correction was applied.

In Experiment 2, tussock asymmetry was analysed within the heterogeneous radiation treatment only. The influence of the number of intravaginal tillers in the tussock, treatment (level of shading in the quadrant––full shading, no shading, intermediate shading), time and interaction between treatment and time on the number of intravaginal and extravaginal tillers in the individual quadrant types were tested.

In all analyses each independent variable was tested at the level at which it was measured, i.e. at the level of halves or quadrants (time) and tussocks (tiller number, treatment, clone, presence of the first leaf).

Results

Response to the degree of shading

In both experiments, tussocks exposed to full sunlight were lowest and had the highest tiller number. Tussocks exposed to full shade, however, were the tallest and had the lowest tiller number. Tussocks exposed to intermediate, i.e. heterogeneous radiation were between the two previously mentioned treatments (Fig. 2, Table 1). Tussocks developed from young initial tillers (Experiment 1) contained only intravaginal tillers while tussocks that developed from older initial tillers (Experiment 2) produced both tiller types, with extravaginal tillering occurring in the later stages of the experiment. The number of intravaginal tillers and treatment influenced the number of extravaginal tillers, with larger tussocks producing more extravaginal tillers and fully shaded tussocks having the highest share of extravaginal tillers (Fig. 3, Table 1).

Fig. 2
figure 2

Tiller number per tussock in plants of clone 5 exposed to full shading (black symbols), full sunlight (white symbols) and intermediate shading, i.e. heterogeneous radiation (grey symbols) in Experiment 1 (means and standard deviations). For the test of significance, see Table 1

Full size image
Table 1 Effect of treatment (full shading, full sunlight and intermediate shading, i.e. heterogeneous radiation), time and clone on growth of Festuca rubra tussocks in Experiments 1 and 2 tested by ANOVA; significant values in bold
Full size table
Fig. 3
figure 3

Proportion of extravaginal tillers (ratio of number of extravaginal tillers to number of intravaginal tillers) in tussocks exposed to full shade (black symbols), full sunlight (white symbols) and intermediate shading, i.e. heterogeneous radiation (grey symbols) in Experiment 2 (means and standard deviations). For the test of significance, see Table 1

Full size image

Morphological constraints in tussock development

Pronounced dorso-ventral asymmetry in tussocks developed from young tillers (Experiment 1) was found in both clones. Most daughter tillers were born at the dorsal sides of the initial tillers. There were considerable shifts in tiller position. As new tillers were born, the older tillers including the initial ones were moved in the opposite direction––towards the ventral half of the pots (Fig. 1). Thus no significant tussock asymmetry was revealed in space, i.e. with regard to the pot centre (Fig. 4––plants exposed to full sunlight and to full shading). No remarkable lateral shifts in initial tiller position nor significant lateral asymmetry was revealed in the tussocks (Fig. 5––plants exposed to full sunlight and to full shading). The position of the first leaf (i.e. its presence in the quadrant) did not affect the direction of tussock growth (Table 2).

Fig. 4
figure 4

Proportion of daughter tillers occurring dorsally from the initial tiller (a, c) and pot centre (b, d) in Experiment 1 (means and standard deviations) in plants exposed to full shading (black circles), full sunlight (white circles), plants with dorsal side shaded (grey squares) and with ventral side shaded (grey triangles). For the test of significance, see Table 2. Note that 0.5 indicates symmetric tussocks

Full size image
Fig. 5
figure 5

Proportion of daughter tillers occurring to the right of the initial tiller (a, c) and pot centre (b, d) in Experiment 1 (means and standard deviations) in plants exposed to full shading (black circles), full sunlight (white circles), plants with left side shaded (grey squares) and with right side shaded (grey triangles). For the test of significance, see Table 2. Note that 0.5 indicates symmetric tussocks

Full size image
Table 2 Morphological constraints in tussock development: effect of total tiller number, clone, time and presence of the first leaf on tiller number in individual halves with regard to pot centre (i.e. in space) and initial tiller in tussocks exposed to homogeneous conditions (full sunlight and full shading) in Experiment 1 tested by ANOVA; significant values in bold (Bonferroni correction used)
Full size table

Growth under heterogeneous radiation

Despite the significant effect of shading on tiller number (Fig. 2, Table 1), heterogeneous radiation did not influence the direction of the development of intravaginal tillers in the central part of tussock. Proportion of intravaginal tillers in the dorsal, respectively the right half in Experiment 1 (Figs. 4, 5 and Table 3) and the proportion of intravaginal tillers in individual quadrant types in Experiment 2 (Table 4) were not influenced by treatment. Except for a small increase in the proportion of tillers in the right half when shading the left side in clone Nr. 9, no clear tendency in the direction of tussock development was apparent. A tendency to develop extravaginal tillers towards the shaded quadrants (Fig. 6, Table 4) did however exist.

Table 3 Effect of treatment (full sunlight, full shading and heterogeneous radiation, i.e. shading foil along one half of pot), clone, time and presence of the first leaf on proportion of tillers in dorsal or right half with regard to pot centre (i.e. in space) and initial tiller (Experiment 1) tested by ANOVA; significant values in bold (Bonferroni correction used)
Full size table
Table 4 Effect of treatment (full sunlight, full shading and heterogeneous radiation, i.e. shading foil along half of the quadrant margin), time and tussock size (i.e. total number of intravaginal tillers in tussock) on number of intravaginal and extravaginal tillers in quadrants from the pot centres in tussocks exposed to heterogeneous radiation in Experiment 2 tested by ANOVA; significant values in bold
Full size table
Fig. 6
figure 6

Number of extravaginal tillers in individual quadrant types in plants exposed to heterogeneous radiation in Experiment 2 (means and standard deviations): quadrants with full shading (black symbols), intermediate shading, i.e. heterogeneous radiation (grey symbols) and no shading (white symbols). For the test of significance, see Table 4

Full size image

Discussion

The experiment confirmed the ability of the model plant, F. rubra, to respond to simulated canopy shade. The shade reduced intravaginal tillering, but activated dormant buds at the tussock base and stimulated extravaginal tillering. The same results were shown in some clones in previous experiments (Skálová et al. 1997; Skálová 2005). These results correspond with field observation: spatial shifts due to the rhizome growth of F. rubra decreased after the removal of other grass dominants from the grassland community (Herben et al. 1994a). There may be division of labour between intravaginal and extravaginal tillers, the task of intravaginal tillers is to consolidate patches while extravaginal tillers forage for more distant suitable sites if conditions worsen (Ikegami et al. 2007). In addition, switching to extravaginal tillering may also minimise mutual shading of tillers within a tussock, which is mainly important under worsen radiation conditions.

Increased formation of extravaginal tillers under shaded conditions is in contradiction to the traditional hypothesis of increased branch production as a common growth response to favourable conditions (Evans 1992; Huber and Stuefer 1997; Thompson and Harper 1988) and to correlative inhibition of shaded ramets (Novoplansky et al. 1989; He and Dong 2003). The results indicate the possibility of independent regulation of extravaginal and intravaginal tillering in grasses. Phenotypic plasticity seems to be expressed at the organisation level lower than whole plant (de Kroon et al. 2005).

The tiller pattern in daughter tussocks formed at the rhizome tips is probably similar to the pattern in tussocks formed from initial tillers in the experiments. I found a strong morphological constraint with more intravaginal tillers born at the dorsal side of the initial tiller. Despite the asymmetric tiller production, shifts in secondary tillers prevented the tussocks from deviating from the initial position, i.e. pot’s centre in the experiment. The reason for the tiller shifts is not clear, but it was not due to the tiller size: tiller biomass in individual pot halves did not differ.

Despite the impact on the total number of intravaginal and extravaginal tillers in tussocks, heterogeneous radiation did not influence the growth direction of the intravaginal tillers. On the other hand, there was marginally significant tendency to direct extravaginal tillers into more shaded quadrants. This may indicate that different degrees of integration occur within the two tiller groups. There is probably a close vascular connection among the tillers, which enables transport of signalling products mainly within the central part of the tussock (Stuefer et al. 2004).

Different responses of the two tiller groups may be explained by mechanisms known from dicots. Shift in auxin-to-cytokinin balance along the vertical plant axis may result in differential activation of the buds (McSteen and Leyser 2005). Tendency to the asymmetric response of older buds to the radiation heterogeneity may be explained by transport of signal substances along the vertical axis, which regulates outgrowth of buds at the same side as organs perceiving radiation conditions (Hay et al. 2001). On the other hand, the asymmetric response may be slow down by within-tussock integration (Stuefer et al. 2004), which is here probably not as strong as in the intravaginally formed central part of the tussock. It is also possible that the control of grass tillering involves specific systems, such as grass specific gene tb1 whose activity is connected with the response to changing light conditions (Doust 2007). Genes, specific for individual response types, may be involved as well (Maddonni et al. 2002).

Absence of asymmetry in intravaginal tillering under heterogeneous radiation conditions (i.e. at the margins of canopy gaps) seems to be of ecological significance. Possible directional intravaginal tillering may be rather risky due to the shifts in tiller positions: increased tillering in the part exposed to favourable radiation conditions might result in pushing more tillers in the opposite direction into spots with worse radiation conditions. Rather weak directionality of the response in extravaginal tillers may reflect the low predictability of radiation conditions in the horizontal direction (Stuefer 1996). Despite horizontal radiation heterogeneity in the original communities of F. rubra is on a scale comparable to the tussock size and expansion rate (Skálová et al. 1999; Wildová et al. 2007), there may be fluctuations due to the changes in species abundance on a fine spatio-temporal scale.

The performance of F. rubra supports the hypothesis of the spatio-temporal extent of module communication (de Kroon et al. 2005) and the association of foraging with a high degree of ramet independence (de Kroon and Schieving 1990). The modular responses to canopy crowding may be among the reasons for the successful performance of grasses in communities, described for instance by Edelkraut and Gusewell (2006). Testing the response of other genera with similar tussock architecture, such as Agrostis, Calamagrostis, Lolium, Holcus, Paspalum or Schizachyrium would be desirable to prove the hypothesis.