Published online October 3, 2006
Correlated Response in Plant Height and Heading Date to Selection in Perennial Ryegrass Populations Laurent Hazard,* Miche`le Betin, and Nicolas Molinari adaptations did not drastically jeopardize dry matter (DM) productivity or digestibility. Repeated defoliation selects for short and prostrate ecotypes while competition for light in infrequently defoliated habitats results in tall and erect ecotypes (McNeilly, 1981; Carman and Briske, 1985; Westoby, 1989; Painter et al., 1993; Hazard et al., 2001). In addition to morphological differentiation, phenological differentiation responds to defoliation: increasing cutting frequency results in late-maturity ecotypes (Sonneveld, 1955; Charles, 1973; Clements et al., 1983; Hayward, 1985; Painter et al., 1993; Humphreys, 1995; Smith et al., 2000). A reduction in plant size in response to grazing has been interpreted as an avoidance mechanism that reduces the amount of biomass removed by herbivores (Briske, 1989). The shortest plants are not only more tolerant of defoliation (Hazard et al., 2001), but they are also less severely defoliated in a canopy than neighboring taller plants (Anderson and Briske, 1995; Hazard and Ghesquie`re, 1995). Moreover, a reduction in plant size could be associated with an increase in the number of tillers per plant (Horst et al., 1978; Hazard and Ghesquie`re, 1997), enhancing persistency under grazing for short, prostrate plants with numerous tillers (Hazard et al., 2001). A late heading date in response to grazing could also be interpreted as an adaptive response that favors tillering. During the vegetative phase, the apical meristem remains close to ground level, where it is protected against damage from grazing. Once the reproductive phase is initiated, the meristem elevates within the leaf sheaths and becomes vulnerable to decapitation when the canopy is grazed or harvested (Brown, 1982). Tillering is inhibited as self-shading increases and resources are allocated to the inflorescence formation (Warringa and Kreuzer, 1996; Bahmani et al., 2000b). Delayed flowering gives more time to the plants in early spring to build up a reserve of vegetative tillers before flowering. This increase in vegetative tiller proportion within the sward could reduce the impact of reproductive tillers dying from decapitation (Laidlaw, 2005). Delayed flowering time in grazed ecotypes could be viewed as an adaptive process that provides an opportunity for more tillers to develop under frequent cutting. This study was undertaken to establish the causes of covariation between vegetative and reproductive traits associated with grazing adaptation in a forage grass breeding project. Reproductive development has often been neglected in grassland studies, probably because reproductive pathways are irrelevant compared with
Reproduced from Agronomy Journal. Published by American Society of Agronomy. All copyrights reserved.
ABSTRACT Understanding how management practices apply selective pressures that shape adaptive traits in forage grass is essential for managing genetic resources and breeding improved cultivars. The defoliation regime, for instance, induces a genotypic differentiation in plant height and heading. To explore these changes, plant height, heading date, and their relationship were studied in a nursery experiment among populations of perennial ryegrass (Lolium perenne L.) that had experienced different selection pressures: the cultivar ‘Clerpin’ from the original seed lot and its derived population from a pasture intensively grazed, 23 commercial cultivars, and 86 families coming from a divergent and recurrent mass selection for leaf length. A controlled experiment was performed on two populations contrasted on both plant height and heading date to compare their phyllochron, leaf elongation rate, leaf elongation duration, tiller number, and biomass per plant. Dissections were performed to monitor apex development and stem elongation. Genetic differentiations in plant height and heading date were found as a result of grazing pressure, commercial breeding, and selection for leaf length. Heading date was negatively correlated to plant height. Delayed flowering and reduction in plant stature resulted from a reduction in both leaf and stem elongation rate and were associated with a delay in the production of double ridge at the apex level. Selecting for a short plant resulted in an overall decrease in both vegetative and reproductive plant growth. Genetic differentiation in heading date in response to defoliation regime could be interpreted as an indirect effect of an adaptive differentiation in plant height.
P
to a defoliation regime could lead to genetic differentiation of ecotypes (Linhart and Grant, 1996). The possibility of a genotypic differentiation via natural selection is increased in forage species and cultivars that are often cross-fertilized and consequently polygenotypic. Understanding this adaptive process and its impact on agronomic characteristics of forage grasses is essential for managing genetic resources, breeding, and matching pasture management to cultivar requirements. Identifying adaptation syndromes in grasses is of interest for collecting and pooling germplasm. If a defoliation regime induced a rapid, adaptive genetic differentiation, a diversity of defoliation practices (continuous or rotational grazing, making hay, stockpiling, etc.) would help to conserve in situ the genetic resources for forage species. Moreover, advantage could be taken from these adaptation syndromes by using specialized ecotypes in a grass breeding program if their LANT ADAPTATION
L. Hazard, INRA-SAD, UMR Agir, BP 52627, F-31326 CastanetTolosan Cedex, France; M. Betin, INRA-UGAPF, F-86600 Lusignan, France; and N. Molinari, Biostatistique, UFR Me´decine, 640 avenue du Doyen Giraud, F-34295 Montpellier Cedex 5, France. Received 20 Apr. 2005. *Corresponding author (
[email protected]). Published in Agron. J. 98:1384–1391 (2006). Forages doi:10.2134/agronj2005.0115 ª American Society of Agronomy 677 S. Segoe Rd., Madison, WI 53711 USA
Abbreviations: DM, dry matter; DR, double ridges; gdd, growing degree day; HS, half-sib; LED, leaf elongation duration; LER, leaf elongation rate.
1384
Reproduced from Agronomy Journal. Published by American Society of Agronomy. All copyrights reserved.
HAZARD ET AL.: PLANT HEIGHT AND HEADING DATE CORRELATION IN RYEGRASS
vegetative propagation in predicting grassland dynamics (Chapman, 1987). The interaction between vegetative growth and reproductive development is viewed as a “trade-off” in resource allocation between clonal propagation and reproduction (Kotanen and Bergelson, 2000). Moreover, this “trade-off” is assumed to be fixed in the grass populations on a larger time scale than that of agronomic practices and rotation times of sown pastures. The objectives of our study were to show that: (i) genetic differentiation in a pasture could occur on a time scale that is relevant with rotation times of sown pastures; and (ii) covariation between vegetative and reproductive traits in forage grasses is constrained by their genetic architecture, which prevails over a tradeoff in resource allocation. The genetic differentiation of plant stature and reproductive development was studied in perennial ryegrass after 3 yr of intensive grazing. The effect of grazing on this relationship was compared with the genetic differences induced in improved cultivars by breeding and by experimental selection for leaf length. Studying families derived from selection for a single trait and originating from the same population prevented the risk of confounding the effects of different plant traits with various selection pressures that occur in natural habitats (Coughenour, 1985). The relationship between plant size and heading was analyzed in terms of rates and duration of growth. Heading date was split into apex initiation and stem elongation rate. MATERIALS AND METHODS In one experiment (Exp. 1), heading date and plant stature were assessed on 111 populations of L. perenne grown in a nursery. In a second experiment (Exp. 2), morphogenesis and reproductive development were characterized on two populations of L. perenne chosen for their contrasting morphogenesis.
Experiment 1: Heading Date and Plant Stature Characterization in a Nursery In 1997, a nursery trial was performed with 111 populations of perennial ryegrass: the cultivar Clerpin, a population derived from Clerpin (grazed Clerpin), 86 half-sib (HS) families coming from a divergent selection for leaf length, and 23 cultivars. Both Clerpin and grazed Clerpin were selected by F. Balfourier (INRA Clermont-Ferrand, France). The seeds of grazed Clerpin were harvested from a polycross done with 120 plants of Clerpin that had survived 3 yr in a paddock intensively grazed by sheep. The 86 HS families originated from a divergent mass selection on leaf length that was performed from 1989 to 1995 at the INRA station of Lusignan (France, 468269 N, 08099 E; for more details, see Hazard et al., 1996). The source population was composed of 58 HS families obtained by polycrossing 58 plants selected from natural late-flowering populations collected in France in 1983 and 1984 (Charmet et al., 1990). During the first cycle of selection done in 1990–1991, 2 3 20 plants out of 1160 were selected based on long and short leaves. They were polycrossed in 1991 to produce 20 HS families for both the “short-” and the “long-leaved” populations. A second and a third cycle of selection were done in 1992–1993 and 1994–1995. For both cycles and both branches of the divergent selection, 20 plants were selected from 400 plants according to their extreme leaf lengths and polycrossed to obtain 20 HS families. The 86 HS families studied in this
1385
experiment were families where at least one plant was chosen to contribute to the next generation during the course of the selection work: 15 HS families from the source population; 7 and 10 HS families of “short-” and “long-leaved” populations, respectively, from the first cycle; 10 HS families each of “short-” and “long-leaved” populations issued from the second breeding cycle; and 20 HS families and 14 HS families of “short-” and “long-leaved” populations, respectively, of the third breeding cycle. The 23 cultivars were chosen to represent the different maturity groups of commercial cultivars. They ranged from early (17 May) to late (8 June) maturity. The 111 populations were sown in a glasshouse on 27 Jan. 1997. On 24 Mar. 1997, 30 plants per population were transplanted and spaced 70 cm apart in a field nursery at the INRA station of Lusignan according to a randomized complete block design. Each population was represented by a single plant per block. The populations were fully randomized within the 30 blocks. The nursery plants were cut in November 1997 and 1998. In March 1998 and 1999, the nursery was fertilized with NH4NO3 at 40 kg N ha21. Heading date and leaf length were monitored. Heading was defined as the emergence of the top spikelet through the leaf sheath of the flag leaf. Heading date was the date when a plant exhibited at least three heading tillers. The nursery plants were surveyed twice per week for heading from 15 May until the end of June in 1998 and 1999. The whole leaf length (sheath and lamina) of the longest leaf on each plant was measured at the vegetative stage on 21 Apr. 1998 and on 3 May 1999.
Experiment 2: Morphogenesis of “Short-” and “Long-Leaved” Families Vegetative and reproductive morphogenesis was characterized more precisely for one “short-” and one “long-leaved” family from the third cycle of selection. They were chosen for their contrasting heading dates recorded in 1998 in the nursery experiment: 23 May for the “long-leaved” family vs. 10 June for the “short-leaved” family. Characterizing their reproduction phases required dissections. The beginning of the destructive harvesting on these two populations was set on the basis of the apex development monitored on plants of ‘Bree’, an early-maturity cultivar (12 May), grown under the same environmental conditions. Seeds of the three populations were placed in petri dishes to germinate in September 2000. They were transplanted into pots on 19 Sept. 2000, and in six trays (1.25 by 0.85 by 0.20 m) containing a sterilized mixture of one-third topsoil, one-third sand, and one-third peat on 1 Dec 2000. The plants were arranged according to a complete randomized design, 0.10 by 0.14 m apart, and surrounded by a continuous row of extra guard plants to reduce border effects. One hundred and twenty-six plants for both “short-” and “long-leaved” populations and 20 for Bree were grown. The trays were maintained in a glasshouse until 15 Dec. 2000, and then placed outside to accomplish vernalization. The trays were watered when necessary and received a dressing of NH4NO3, 40 kg N ha21, on 30 Mar. 2001. In January 2001, three tillers per plant were labeled with a ring until dissection to be sure that the tiller to be dissected had experienced vernalization conditions. Apical dissections started for Bree on 2 February and for both “short-” and “long-leaved” families on 20 Feb. 2001. Thereafter, three plants and three marked tillers per plant were dissected per family twice a week until the heading date, i.e., for 18 wk. Apices were dissected and observed under a binocular microscope. At each date, the number of primodia on the apex was counted. A leaf was regarded as a primordium when it was too
1386
AGRONOMY JOURNAL, VOL. 98, NOVEMBER–DECEMBER 2006
small to enclose the shoot apex. Apex stage was noted according to the following phenological stages:
Reproduced from Agronomy Journal. Published by American Society of Agronomy. All copyrights reserved.
. white stria (WS) that correspond to the internode formation and the beginning of stem elongation . double ridges (DR) when the spikelet primordia appear . B stage when the glumes appear . C stage when the stamens appear
Stem length was measured. The number of tillers per plant and DM biomass per plant were assessed for each plant collected. Vegetative morphogenesis was characterized for 18 plants per family and three labeled tillers per plant during the period from 6 Mar. to 11 May 2001. Three successive leaves per tiller were monitored every day except on Saturday for the date of leaf appearance, date of ligule appearance, and lamina length. The phyllochron was calculated as the time between the appearance of two successive leaves. Leaf elongation duration (LED) was the time between leaf appearance and ligule appearance. Leaf elongation rate (LER) was calculated as the ratio between leaf length and LED. This calculation approximates sigmoidal growth by assuming linear growth (Bultynck et al., 1999). The heading date of each marked tiller was noted. Thermal time was calculated as growing degree days (gdd) using a simple sum of mean daily air temperature .08C.
Statistical Analysis An ANOVA was performed on nursery variables using the GLM procedure of SAS (SAS Institute, 2001). In Exp. 1, heading date and plant height were analyzed according to the following model: population 1 year 1 block 1 population 3 year. No significant block or population 3 year interactions were found. Means and standard deviation were calculated using the LSMEANS statement in the GLM procedure. Population means were used to calculate linear correlations between heading date and plant height for the 111 populations and for the cultivars alone. In Exp. 2, primordium number, lamina length, phyllochron, LED, and LER were averaged for the three leaves observed per plant. The family effect was tested in the GLM procedure for lamina length, LER, LED, and phyllochron. Primodium number, tiller number per plant, and plant DM, and family were analyzed according to the model: family 1 date 1 family 3 date. Means and standard deviations were calculated for morphogenetic variables by family or by family 3 date using LSMEANS. Mean comparisons were done using a Tukey’s Studentized range test in the MEANS statement of GLM. Overall significance of main effects was declared at P , 0.05. To determine the shift between slow and rapid stem elongation phases, free knot spline functions were used in the analysis of these two-dimensional data (x,y) with thermal time as x and stem length as y. Spline functions are defined as piecewise polynomials of degree d. The pieces join in the so-called knots j with continuity conditions for the function itself and the first d 2 1 derivatives (de Boor, 1978). A useful basis B(j), for the spline representation is given by Schoenberg’s B-splines, or basic splines. A spline s(.,b,j) could be written as the linear combination of the basic functions s(.,b,j) 5 B(j) b. The vector of coefficients b and the vector of knots j are considered tuning parameters. The spline function was computed based on a least-square approach (for further details, see Molinari et al., 2002). For both slow and rapid stem elongation phases, a variance–covariance analysis of stem length and thermal time was performed using the GLM procedure of SAS according to the following model: family 1 thermal time 1 family 3 thermal time.
RESULTS AND DISCUSSION Genetic Differentiation in Heading Date and Plant Stature as a Result of Sheep Grazing Experiment 1 showed that grazed Clerpin was phonologically and morphologically different from the cultivar Clerpin. Compared with Clerpin, grazed Clerpin was composed of shorter vegetative plants (279 vs. 311 mm, P , 0.05) with a later heading date (8 June vs. 31 May, P , 0.05). Such a rapid ecotypic differentiation in the field was expected in perennial ryegrass because a genotypic shift in response to frequent defoliation was demonstrated after only 4 mo in controlled environment studies (Hazard and Ghesquie`re, 1995).
Correlated Changes in Heading Date and Plant Stature The observed changes in plant stature and heading date occurred in response to both natural and recurrent mass selection. As a result of artificial selection, both experimental progenies selected for divergent leaf length and commercial cultivars studied in Exp. 1 showed a negative correlation between leaf length and heading date (r 5 20.59 within HS families, r 5 20.42 within cultivars; Fig. 1). Experimental selection for leaf length at the vegetative stage led to an indirect effect on reproductive development: the shorter the leaf length, the later the heading date. The correlated response of heading date was highly significant (P , 0.001) between the source population and both “short-” and “long-leaved” populations after three cycles of recurrent mass selection on leaf length. Even if it did not get much attention, this correlation is empirically known by forage grass breeders as they acknowledge that plants of cultivars from the early maturity group are taller than those of the late maturity group. Such a correlation has been interpreted so far as the result of a difference in phenological development: compared at the same time in spring, plants of early-flowering cultivars are taller than those of late-flowering cultivars because their stem elongation is more advanced. Our study showed that this difference in plant height between early- and late-flowering plants exists during the vegetative phase when the stem does not elongate.
Reproductive Morphogenesis in “Short-” and “Long-Leaved” Families Experiment 2 showed that the difference in heading date (8 June vs. 25 May, P , 0.05) between “short-” and “long-leaved” families was partly explained by their difference in stem elongation rate. Stem elongation shifted from a slow to a rapid stem elongation period (Fig. 2). The stem elongation rates were faster for both periods in the “long-leaved” family than in the “short-leaved” family (0.031 vs. 0.012 mm gdd21 during the slow elongation phase, P , 0.03; 0.75 vs. 0.35 mm gdd21, P , 0.001 during the fast elongation phase). The first phase corresponded to the node formation. The formation of nodes along the stem can be visualized after dissection as white
1387
HAZARD ET AL.: PLANT HEIGHT AND HEADING DATE CORRELATION IN RYEGRASS
165
Heading date (day of the year)
SL3 155
SL2 SL1
150
LL1
145
140
135
130 16
LL2 LL3
Col 27 vs Col 28 cultivar "short-leaved" population after 1 (SL1), 2 (SL2), and 3(SL3) cycles of selection for long leaves "long-leaved" population after 1 (LL1), 2 (LL2), and 3 (LL3) cycles of selection for short leaves 18
20
22
24
26
28
Vegetative plant height (cm) Fig. 1. Negative correlation between heading date and plant height in HS families (original, short-, and long-leaved populations) and cultivars of perennial ryegrass at the vegetative stage (r 5 20.59 and 20.42, respectively). Bars indicate 95% confidence interval.
stria below the apex. White stria were observed from 15 to 29 March in the “long-leaved” family (Fig. 3). In the “short-leaved” family, white stria appeared earlier, on 9 March, but they were observed until 12 April. Wilman et al. (1994) demonstrated that the stem elongation measured during this first phase was due to the accumulation of leaf primodia at the apex since the plastochron is higher than the phyllochron. In cereals, several researchers consider that the beginning of floral initiation is correlated with a change in plastochron (Jamieson et al., 1998; Miralles and Richards, 2000). This correlation does not occur in all grass species: in Lolium
temulentum L., changes in plastochron occur independently of floral initiation (Evans and Blundell, 1996). In our study, no evidence of change in the rate of leaf primordia initiation was found when tillers shifted from the vegetative to the reproductive phase; however, leaf primordia accumulated at a higher rate in the “longleaved” family than in the “short-leaved” family (0.0098 vs. 0.0047 primordia gdd21). At that stage, Gillet (1980) has shown that the primordia that accumulate do not develop into leaves but spikelets; consequently the “longleaved” family exhibited a potentially higher number of spikelets per inflorescence than the “short-leaved” family.
500
400
Stem length (mm)
Reproduced from Agronomy Journal. Published by American Society of Agronomy. All copyrights reserved.
160
slow stem elongation rate regression slopes LL: y = 0.031 x - 11.9 (r 2 = 0.37) SL: y = 0.012 x - 3.9 (r 2 = 0.33)
rapid stem elongation rate regression slopes LL: y = 0.75 x - 726 (r 2 = 0.81) SL: y = 0.35 x - 321 (r 2 = 0.77)
300
200
100
"short-leaved" family (SL) "long-leaved" family (LL)
0 200
400
600
800
1000
1200
1400
1600
Growing Degree Days (baseline = 0°C) Fig. 2. Time course of stem length in cumulative degree days since 1 January. All regressions were significant at P , 0.05.
1388
AGRONOMY JOURNAL, VOL. 98, NOVEMBER–DECEMBER 2006
C stage B stage DR WS Shift from slow to rapid stem elongation Heading
Short leaved family
Reproduced from Agronomy Journal. Published by American Society of Agronomy. All copyrights reserved.
Long leaved family
Heading
C stage B stage DR WS
200
400
600
800
1000
1200
1400
1600
1800
2000
Growing degree days (baseline = 0°C) Fig. 3. Time course of the phenological stages of the apex in the “short-” and “long-leaved” families: white stria (WS), double ridges (DR), appearance of glumes (B), appearance of stamens (C), and heading date expressed in cumulative degree days from 1 January.
The second phase of rapid stem elongation brought the inflorescence out of the cornet. The shift from slow to rapid elongation has been considered as a good indicator of the onset of the reproductive phase in cereals (Williams, 1966; McMaster, 1997); however, no difference was found between the “short-” and “long-leaved” families for the time of the shift (5 May vs. 9 May, P . 0.05; Fig. 2). During the phase of rapid elongation, stem elongation rate was slower for the “short-leaved” family than for the “long-leaved” populations, which explained the difference in heading date between “short-” and “long-leaved” families. The onset of the rapid stem elongation phase was not synchronized with the appearance of a specific stage of the apex (Fig. 3). The shift from slow to rapid stem elongation occurred at the DR stage, as described in cereals by McMaster (1997). There was some variation in the occurrence of the two events, however; rapid stem elongation started at the beginning of the DR stage for the “short-leaved” family and at the end of DR stage for the “long-leaved” family. According to the DR stage, the “short-leaved” population initiated flowers later than the “long-leaved” population. Differences in heading dates were associated with differences in the date of DR formation. Double-ridge appearance could be a good indicator of the onset of flower initiation, as already stated by Slafer and Rawson (1994).
Vegetative Morphogenesis in “Short-” and “Long-Leaved” Families Like differences in heading dates, differences in leaf length between “short-” and “long-leaved” plants (68 vs. 126 mm, P , 0.001) resulted from their difference in growth rates. These differences in leaf length were positively correlated to differences in LER (Fig. 4). The dif-
ference in LER (0.373 vs. 0.547 mm gdd21, P , 0.001) between “short-” and “long-leaved” families was greater than that for LED (183 vs. 229 gdd, P , 0.001). The phyllochron that determines the developmental rate in grasses (Skinner and Nelson, 1995) differed significantly between “short-” and “long-leaved” families (131 vs. 136 gdd, P , 0.001), but this difference was relatively low compared with that of the LER. These results are in good agreement with the previous study of these “short-” and “long-leaved” families (Hazard et al., 1996; Bahmani et al., 2000a). No significant difference in tiller number per plant between “short-” and “long-leaved” families was found at heading, but tillering dynamics differed. A flush of tiller production was observed at 900 to 950 gdd (Fig. 5), which corresponded to optimal conditions for growth in early May. At that time, the “short-leaved” family attained a higher number of tillers per plant than the “long-leaved” family, but this increase was followed by a higher rate of tiller mortality as the stem rapidly elongated. Ong (1978) showed that the developing head and elongating stem make large demands on plant resources once stem elongation starts and so young tillers that are not competitive usually die once rapid stem elongation starts. Tiller number per plant did not compensate for the difference in leaf length. As a result, DM per plant was significantly greater for the “long-leaved” family than for the “short-leaved” family (P , 0.05, Fig. 6).
Adaptive Significance Brown (1982) stated that “for survival and seed production of individual plants, the species most favored under grazing will be the ones in which the meristem spends as much time as possible near the base of the
HAZARD ET AL.: PLANT HEIGHT AND HEADING DATE CORRELATION IN RYEGRASS
1389
180
160
Leaf length (mm)
y = 302.9x - 41.86 r 2 = 0.97 P < 0.05
120
100
80
"short-leaved" family "long-leaved" family
60
40 0.2
0.3
0.4
0.5
0.6
0.7
0.8
-1
Leaf elongation rate (mm gdd ) Fig. 4. Relationship between adult leaf length and leaf elongation rate. Bars indicate 95% confidence interval.
tion rate of leaves and stems results in “short-leaved” plants flowering later than “long-leaved” plants. The potential reproductive output was reduced in the “shortleaved” population. When these experimental populations were selected before this study, “short-leaved” populations were found to produce fewer and lighter seeds than “long-leaved” populations (unpublished data, 2000). This response to grazing involved an overall decline in growth rate, which was described by Chapin (1991) as a common integrated response of plants to stress. It is definitely not a “trade-off” between vegetative and reproductive growth: being shorter and investing less in reproduction did not result in greater phyllochron or tiller bud production that could enhance
tiller and then elevates as rapidly as possible during inflorescence exertion.” Since the stem elongation rate of “short-leaved” plants was significantly lower than that of “long-leaved” plants, the shift in heading date is a “nonadaptive” consequence of the change in leaf size. Short leaves are selected under a frequent defoliation regime. The selection pressure exerted by frequent defoliation seems to impact intrinsic growth rate much more than developmental rate. Preliminary results showed that the underlying process involves cell division (Gastal, personal communication, 2003). As a consequence, grazing that selects for short leaves is likely to induce an overall decrease in growth rate that affects both vegetative and reproductive growth. Reduced elonga180 160
Number of tillers per plant
Reproduced from Agronomy Journal. Published by American Society of Agronomy. All copyrights reserved.
140
140 120 100 80 60 40
"short-leaved" family "long-leaved" family
20 0 400
500
600
700
800
900
1000 1100 1200 1300 1400 1500
Growing Degree Day (baseline = 0°C) Fig. 5. Time course of number of tillers per plant expressed in growing degree days since the 1 January. Bars indicate 95% confidence interval.
1390
AGRONOMY JOURNAL, VOL. 98, NOVEMBER–DECEMBER 2006
25
Plant dry matter (g)
Reproduced from Agronomy Journal. Published by American Society of Agronomy. All copyrights reserved.
20
15
10
5
0
400
"short-leaved" family "long-leaved" family 600
800
1000
1200
1400
1600
Growing Degree Day (baseline = 0°C) Fig. 6. Time course of plant dry matter expressed in growing degree days since 1 January. The overall difference in dry matter per plant was significant between “long-” and “short-leaved” families (P , 0.05). Bars indicate 95% confidence interval.
tillering. Being shorter and slow growing in a canopy frequently cut allows the ryegrass plant to be less severely defoliated than its neighbors and able to sustain a higher number of tillers per plant. Such a strategy probably enhances perennation under a frequent defoliation regime.
Forage Grass Breeding Implications Intensive grazing favoring short and “late-heading” plants results in ecotypic differentiation in perennial ryegrass. Such ecotypes that are very persistent under grazing could be collected from old pastures and used in breeding programs to improve defoliation tolerance of forage or turf cultivars. Such a strategy would probably be counterproductive, since defoliation tolerance of these ecotypes results from their slow growing rates that reduce herbage and seed yields. While low herbage production is a desirable trait for turf cultivars, a slow growing rate slows down sward establishment, reduces seed yield, and is prejudicial for both turf and forage cultivars. Breeding late-heading perennial ryegrass cultivars, which are increasingly used under grazing management, results in higher tolerance to grazing but lower herbage and seed yields than earlier heading cultivars. Since late heading results from an indirect response to grazing through the reduction in plant growth, promoting late cultivars should not be recommended to increase grazing tolerance but it can be used to reduce the peak of DM production occurring in spring. On the other hand, selecting and planting tall, early-heading perennial ryegrass cultivars is inappropriate under intensive grazing because populations rapidly shift to short genotypes. This ecotypic differentiation has an agronomic significance since it occurs in ,3 yr, while a perennial ryegrass meadow is sown to last at least 5 yr. Is that to say that grazing tolerance and high yield are mutually exclusive?
Probably not. The negative correlation between heading date and plant size could be broken in the course of selection in at least one way: selection of short and earlyheading plants. Because seed cost is a primary factor for success of turf cultivars, turf breeders succeeded in creating ryegrass plants that are short and early in maturity because of a rapid developmental rate (rapid turnover of leaves) and not because of a slow growing rate. To do so, they selected simultaneously for short and vigorous early-flowering types of perennial ryegrass. To increase grazing tolerance, forage grass breeders did try to improved the productivity of late-flowering types. This work was grounded on the hypothesis that late heading has an adaptive significance to grazing. This hypothesis is not supported by our study. Moreover, it does not seem possible to improve the growth rate of late-heading types without changing the heading date to be earlier in the season. Early-heading types of perennial ryegrass probably offer breeders more combinations of sward structure and growth rate that will improve both productivity and persistency under grazing. ACKNOWLEDGMENTS We thank François Balfourier and Jean-Claude Vezine for providing plant material, as well as Kimberly Cassida, Bernadette Julier-Koubaiti, and Philippe Barre for their help to improve our manuscript.
REFERENCES Anderson, V.J., and D.D. Briske. 1995. Herbivore-induced species replacement in grasslands: Is it driven by herbivory tolerance or avoidance. Ecol. Appl. 5:1014–1024. Bahmani, I., L. Hazard, C. Varlet-Grancher, M. Betin, C. Matthew, E.R. Thom, and G. Lemaire. 2000a. Differences of tillering in perennial ryegrass cultivars under a simulated shading in relation to leaf development. Crop Sci. 40:1095–1102. Bahmani, I., C. Varlet-Grancher, L. Hazard, C. Matthew, M. Betin, A. Langlais, G. Lemaire, and E.R. Thom. 2000b. Post-flowering
Reproduced from Agronomy Journal. Published by American Society of Agronomy. All copyrights reserved.
HAZARD ET AL.: PLANT HEIGHT AND HEADING DATE CORRELATION IN RYEGRASS
tillering in contrasting light environments of two New Zealand perennial ryegrass cultivars with different perennation strategies. Grass Forage Sci. 55:367–371. Briske, D.D. 1989. Developmental morphology and physiology of grasses. p. 85–89. In R.K. Heitschmidt and J.W. Stuth (ed.) Grazing management: An ecological perspective. Timber Press, Portland, OR. Brown, R.F. 1982. Tiller development as a possible factor in the survival of the two grasses, Aristida armata and Thyridolepis mitchelliana. Aust. Rangeland J. 4:34–38. Bultynck, L., C. Fiorani, and H. Lambers. 1999. Control of leaf growth and its role in determining variation in plant growth rate from an ecological perspective. Plant Biol. 1:13–18. Carman, J., and D. Briske. 1985. Morphologic and allozymic variation between long-term grazed and non-grazed populations of the bunchgrass Schizachyrium scoparium var. frequens. Oecologia 66: 332–337. Chapin, F.S., III. 1991. Integrated responses of plants to stress. A centralized system of physiological responses. BioScience 41:29–36. Chapman, D.F. 1987. Natural re-seeding and Trifolium repens demography in grazed hill pastures: II. Seedling appearance and survival. J. Appl. Ecol. 24:1037–1043. Charles, A.H. 1973. A comparison of ryegrass populations from intensively managed pastures and leys. J. Agric. Sci. 81:99–106. Charmet, G., F. Balfourier, and A. Bion. 1990. Agronomic evaluation of a collection of French perennial ryegrass populations: Multivariate classification using genotype 3 environment interactions. Agronomie 10:807–823. Clements, R.J., M.D. Hayward, and D.E. Byth. 1983. Genetic adaptation in pasture plants. p. 101–115. In J.G. McIvor and R.A. Bray (ed.) Genetic resources of forage plants. CSIRO, Melbourne. Coughenour, M.B. 1985. Graminoid responses to grazing by large herbivores: Adaptations, exaptations and interacting processes. Ann. Missouri Botanical Garden 72:852–863. de Boor, C. 1978. A practical guide to splines. Springer, New York. Evans, L.T., and C. Blundell. 1996. The acceleration of primordium initiation as a component of floral evocation in Lolium temulentum L. Austr. J. Plant Physiol. 23:569–576. Gillet, M. 1980. Les gramine´es fourrage`res. Bordas, Paris. Hayward, M.D. 1985. Adaptation, differentiation and reproductive systems in Lolium perenne. p. 83–94. In P. Jacquard et al. (ed.) Genetic differentiation and dispersal in plants. Springer Verlag, Berlin. Hazard, L., D. Barker, and S. Easton. 2001. Morphogenetic adaptation to defoliation and soil fertility in perennial ryegrass. N.Z. J. Agric. Res. 44:1–12. Hazard, L., and M. Ghesquie`re. 1995. Evidence from the use of isozyme markers of competition in swards between long-leaved and short-leaved perennial ryegrass. Grass Forage Sci. 50:241–248. Hazard, L., and M. Ghesquie`re. 1997. Productivity under contrasting cutting regimes of perennial ryegrass selected for short and long leaves. Euphytica 95:295–299. Hazard, L., M. Ghesquie`re, and C. Barraux. 1996. Genetic variability for leaf development in perennial ryegrass populations. Can. J. Plant Sci. 76:113–118. Horst, G.L., C.J. Nelson, and K.H. Asay. 1978. Relationship of leaf elongation to forage yield of tall fescue genotypes. Crop Sci. 18: 715–719.
1391
Humphreys, M.O. 1995. Multitrait response to selection in Lolium perenne L. (perennial ryegrass) populations. Heredity 74:510–517. Jamieson, P.D., I.R. Brooking, M.A. Semenov, and J.R. Porter. 1998. Making sense of wheat development: A critique of methodology. Field Crops Res. 55:117–127. Kotanen, P.M., and J. Bergelson. 2000. Effects of simulated grazing on different genotypes of Bouteloua gracilis: How important is morphology? Oecologia 123:66–74. Laidlaw, A.S. 2005. The relationship between tiller appearance in spring and contribution to dry-matter yield in perennial ryegrass (Lolium perenne L.) cultivars differing in heading date. Grass Forage Sci. 60:200–209. Linhart, Y.B., and M.C. Grant. 1996. Evolutionary significance of local genetic differentiation in plants. Annu. Rev. Ecol. Syst. 27: 237–277. McMaster, G.S. 1997. Phenology, development, and growth of the wheat (Triticum aestivum L.) shoot apex: A review. Adv. Agron. 59: 63–118. McNeilly, T. 1981. Ecotypic differentiation in Poa annua: Interpopulation differences in response to competition and cutting. New Phytol. 88:539–547. Miralles, D.J., and R.A. Richards. 2000. Responses of leaf and tiller emergence and primordium initiation in wheat and barley to interchanged photoperiod. Ann. Bot. 85:655–663. Molinari, N., M. Morena, J.P. Cristol, and J.P. Daures. 2002. Free knot splines for biochemical data. Comput. Methods Programs Biomed. 67:163–167. Ong, C.K. 1978. The physiology of tiller death in grasses: I. The influence of tiller age, size and position. J. Br. Grassl. Soc. 33:197–203. Painter, E.L., J.K. Detling, and D.A. Steingraeber. 1993. Plant morphology and grazing history—relationships between native grasses and herbivores. Vegetatio 106:37–62. SAS Institute. 2001. Statistical analysis software. Version 8. SAS Inst., Cary, NC. Skinner, R.H., and C.J. Nelson. 1995. Elongation of the grass leaf and its relationship to the phyllochron. Crop Sci. 35:4–10. Slafer, G.A., and H.M. Rawson. 1994. CO2 effects on phasic development, leaf number and rate of leaf appearance in wheat. Ann. Bot. 79:75–81. Smith, S.E., R. Mosher, and D. Fendenheim. 2000. Seed production in sideoats grama populations with different grazing histories. J. Range Manage. 53:550–555. Sonneveld, A. 1955. The response of different types of perennial ryegrass and timothy to divergent managements. Versl. Landbouwkd. Onderz. 58–66. Warringa, J.W., and A.D.H. Kreuzer. 1996. The effect of new tiller growth on carbohydrates, nitrogen and seed yield per ear in Lolium perenne L. Ann. Bot. 78:749–757. Westoby, M. 1989. Selective forces exerted by vertebrate herbivores on plants. Trends Ecol. Evol. 4:115–117. Williams, R.F. 1966. Development of the inflorescence in Gramineae. p. 74–87. In F.L. Milthorpe and J.D. Ivins (ed.) The growth of cereals and grasses. Butterworth, London. Wilman, D., Y. Gao, and P.J. Michaud. 1994. Morphology and position of the shoot apex in some temperate grasses. J. Agric. Sci. 122: 375–383.
