| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
(American Journal of Botany. 2008;95:1063-1071.) doi: 10.3732/ajb.0800042 © 2008 Botanical Society of America, Inc. |
What's this? |
Ecology |
Department of Biology, College of Staten Island, City University of New York, Staten Island, New York 10314 USA
Received for publication 1 February 2008. Accepted for publication 11 June 2008.
ABSTRACT
Leaves of many cool-season grasses are infected by endophytic fungi that can impact their populations. A common garden experiment with Lolium perenne was established in a lawn in New Jersey, USA, to investigate the impact of endophyte infection and host genotype on tiller and spike production over three years. Infected (E+) and uninfected (E–) plants of each genotype were monitored every 2–3 mo. Infection intensity within plants varied significantly among genotypes and years, but there was no evidence of directional change over time. Tiller production varied significantly among genotypes and was affected by endophytes: E+ plants of several genotypes produced more tillers than E– plants during the third year. E+ plants had greater aboveground biomass, but host genotype explained a far greater proportion of variation in tiller production, number, and biomass than infection. Plant survival, percentage flowering, flowering date, number of spikes, and mean tiller mass were unaffected by endophytes. However, the last three variables showed significant variation among host genotypes. Although studies have demonstrated a positive growth effect of endophytes on several grass hosts, in this experiment host genotype accounted for far more of the variation in tiller and spike production and in biomass of Lolium perenne than endophyte infection.
Key Words: common garden • endophytic fungi • flowering • infection intensity • Neotyphodium lolii • perennial ryegrass • population ecology • tiller production
Interactions among animal, plant, and microbial species are especially common in natural communities and are the focus of many investigations in population ecology and evolution. There is little doubt that coevolutionary processes have shaped many aspects of the population ecology of interacting species (Thompson, 2005
). Associations among species in nature can be especially close and symbiotic. For example, symbioses between microbial organisms and their host plants can greatly influence survival, growth, and reproduction within plant populations (Burdon, 1987; Clay, 1990b, 1998; Faeth and Hamilton, 2006; Olejniczak and Lembicz, 2007), with each symbiotic partner acting as an agent of natural selection on the other. Biotic interactions have recently been the focus of several collections of papers in plant biology in which the scales of investigations range from molecular to ecological (Osbourn and He, 2006; Glazebrook and Ton, 2007). From an ecological perspective, improved understanding of the microevolution of genetically diverse plant populations depends on delimiting the potentially important effects of clandestine, symbiotic microbes on survival, growth, and reproduction of the host genotypes they infect.
Fungal endophytes, which are symbiotic within the aboveground tissues of many widely distributed, cool-season grasses, are reported to affect a variety of organismal characteristics of their hosts (e.g., tiller and biomass production) under experimentally controlled conditions (review and references in Cheplick and Faeth, in press). For some host species such as Lolium arundinaceum (Schreb.) S. J. Darbyshire [tall fescue, formerly Festuca arundinacea (Schreb.)], endophyte-mediated effects on plant growth are mostly positive (Clay, 1987, 1990b; Belesky and Malinowski, 2000). For other hosts such as Lolium perenne L. (perennial ryegrass), effects of endophytes are much more variable and inconsistent (Cheplick and Cho, 2003; Cheplick, 2004, 2007; Hesse et al., 2004; Lewis 2004). Across all systems investigated to date, the grass–endophyte symbiosis clearly spans a continuum from antagonistic to mutualistic (Saikkonen et al., 1998, 2006; Faeth and Fagan, 2002; Müller and Krauss, 2005; Schardl and Leuchtmann, 2005).
As with most coevolutionary interactions (Thompson, 2005), the grass–endophyte symbiosis has pronounced geographic variation, and the nature of the interaction (i.e., antagonistic, commensalistic, mutualistic) depends greatly on environmental conditions and host–endophyte genotypic combinations (Cheplick and Faeth, in press). Endophytes sometimes benefit their host under abiotically stressful conditions (West, 1994; Belesky and Malinowski, 2000; Malinowski et al., 2005), but growth responses to endophyte infection often vary among host genotypes (Faeth and Fagan, 2002; Cheplick and Cho, 2003; Cheplick, 2004, 2007). Microevolution within a grass population of a species known to harbor endophytic fungi could therefore depend not only on whether an individual is infected, but also on particular host genotype–endophyte combinations that differentially affect phenotypic traits.
Despite the potential importance of endophytes to the population biology of their hosts, only a few studies have provided demographic data for host grasses in the field that span several years (Clay, 1990a; Hill et al., 1998; Clay and Holah, 1999; Faeth and Hamilton, 2006; Olejniczak and Lembicz, 2007). For the majority of endophyte-infected, perennial grass species that have been monitored for a least a few years, survivorship has been found to be very high and mostly unaffected by endophytes (Cheplick and Faeth, in press). However, tiller production and long-term persistence in the field has been shown to be improved in a well-studied, agronomic cultivar (Kentucky 31) of tall fescue infected by the endophyte Neotyphodium coenophialum (Morgan-Jones & Gams) Glenn, Bacon & Hanlin (Clay, 1990a; Hill et al., 1998; Clay and Holah, 1999; Clay et al., 2005). The high infection frequency found in natural populations of several cool-season grasses supports the notion that the grass–endophyte symbiosis often functions as a mutualism.
Less is known about the dynamics of perennial ryegrass and its endophyte N. lolii (Latch, Christensen & Samuels) Glenn, Bacon & Hanlin under field conditions. Growth enhancement due to endophyte infection was first reported in controlled-environment studies in the 1980s by some workers (Latch et al., 1985; Clay, 1987), but not by others (Keogh and Lawrence, 1987). In addition, some evidence suggests that endophyte-mediated enhancement of growth and persistence is possible in the field (Ravel et al., 1995; Hesse et al., 2003, 2005; Jensen and Roulund, 2004). However, in a field experiment in New Zealand, Eerens et al. (1998) did not detect any effect of endophyte infection in L. perenne on tiller densities or pasture production.
Unfortunately, in most of these studies individual genets, both infected and uninfected by endophytes, were not followed. However, Hesse et al. (2004) tracked 13 L. perenne genotypes, each replicated as infected and uninfected individuals, in a field trial at Halle, Germany over three years. They showed how host genotype interacted with infection status to ultimately determine host biomass and inflorescence production. In that field study (and in additional studies conducted indoors in controlled environments, Cheplick and Cho, 2003; Cheplick, 2004, 2007), the effect of N. lolii on L. perenne has not, on average, been consistently positive, tending instead to vary greatly from one host genotype to another.
In the current study, a subset of the phenotypically variable L. perenne genotypes used for prior greenhouse experiments (Cheplick et al., 2000; Cheplick and Cho, 2003; Cheplick, 2004) were planted into a common garden established within a lawn in central New Jersey, USA, and closely monitored for almost three years. Tiller and spike production of both endophyte-infected and uninfected replicate plants of each genotype were followed. The overall goal was to determine whether endophyte infection matters to the population ecology of L. perenne plants growing in a common environment under noncompetitive conditions. The common garden is a classic ecological technique used to distinguish the environmental and genetic components of phenotypic variation by "growing plants in a uniform environment" (Silvertown and Charlesworth, 2001, p. 22). Common garden experiments can reveal genetic differences among locally adapted populations or among individual genotypes within a population (Gibson, 2002). As a first step in examining genotype vs. endophyte effects on L. perenne in a temperate region, all plants were grown under relatively uniform field conditions without the complications in interpretation that would arise from variation in competitive (or other biotic) interactions. Specific questions addressed were (1) Does the intensity of endophyte infection vary among host genotypes and over time? (2) Are the effects of endophyte infection on tiller and spike production more important than the effects of host genotype? (3) Does long-term endophyte infection improve growth over time and, thus, eventually increase the number of tillers and aboveground biomass of host genotypes?
MATERIALS AND METHODS
Perennial ryegrass and its endophyte
Perennial ryegrass (Lolium perenne L.) is a widespread, globally important caespitose grass that has been introduced into most parts of the temperate world (Beddows, 1967; Jung et al., 1996). Because of its prevalence in many pastures and successional fields and its economic importance as a forage crop, L. perenne has a long history of investigation of its tiller dynamics and growth in relation to environmental conditions (e.g., Luxmoore and Millington, 1971; Kays and Harper, 1974; Davies and Thomas, 1983; Gautier et al., 1999). In nature the species varies phenotypically and genotypically (Beddows, 1967; McNeilly and Roose, 1984; Hayward, 1985), perhaps due to a breeding system that requires cross-pollination (Jung et al., 1996). In addition, L. perenne has pronounced genetic variation in morphological and molecular features within and between its many cultivars (Casler, 1995; Huff, 1997; Kubik et al., 2001).
The endophyte of perennial ryegrass, Neotyphodium lolii, is exclusively asexual and can only be vertically transmitted within the seeds of its host (Schardl and Leuchtmann, 2005). When infected, hosts are asymptomatic, but microscopic examination reveals intercellular hyphae within the leaf sheaths (Wilson et al., 1991). Due to asexual reproduction, Neotyphodium spp. endophytes often have low genetic diversity relative to other sexual endophytes (Cheplick and Faeth, in press). Molecular markers revealed limited polymorphism in N. lolii (van Zijll de Jong et al., 2003), and six isolates of N. lolii from one cultivar of L. perenne had identical isozyme profiles (Leuchtmann and Clay, 1990). Because a single cultivar (Yorktown III) of L. perenne was used in the current study, it is assumed that different host genotypes contain a consistent endophyte genotype. Genotypes of this cultivar vary significantly in tiller and biomass production in the greenhouse (Cheplick and Cho, 2003; Cheplick, 2004).
Endophyte extermination and assessment
Stock cultures of 10 of the 13 L. perenne genotypes used in the experiments of Cheplick et al. (2000) were available as both endophyte-infected (E+) and uninfected (E–) individuals. The E– individuals were obtained from the original infected genotypes in 1998 after treatment of E+ ramets with a systemic fungicide (Benlate; Dupont, Kansas City, Kansas, USA); additional details are available in Cheplick et al. (2000).
In 1998, infection intensity, defined as the relative level of endophyte infection within each individual host plant (Cheplick and Faeth, in press) was assessed by counting the number of hyphae within leaf sheath tissue across a microscopic field of view at 400x. Each leaf sheath sample was 5–10 mm long and collected from a single, young vegetative tiller per plant. Standard protocols were employed to directly observe endophytic hyphae: clearing of sampled leaf tissue in 70% ethanol followed by staining with aniline blue dissolved in lactic acid (Hignight et al., 1993; Bacon and White, 1994). Although only one tiller leaf sheath was examined per plant, 8–10 replicate plants per genotype were sampled. Thus, each host genotype is represented by 8–10 estimates of infection intensity. The entire leaf sheath sample was microscopically examined, and hyphae were counted in a least three separate fields of view (and averaged) to estimate infection intensity for each plant. To examine possible variation in endophytic infection under field conditions, I again assessed infection intensity for all surviving E+ individuals within the common garden in late June, one and two years after the experiment began in 2001 (see Common garden experiment). E– individuals were also sampled to determine whether they remained endophyte-free.
Prior to planting in the common garden, E+ and E– ramets of each genotype were individually grown in a polystyrene planter within inverted, pyramid-shaped cells. Each cell was a 5 x 5 cm square that tapered to 7 cm depth. The soil was a 1:1:1 mixture of topsoil, peat moss, and medium-grade vermiculite. Ten E+ and 10 E– ramets per host genotype were planted on 3 October 2001. These were permitted to establish in the greenhouse (dead ramets were replaced as necessary) until 24 October when they were placed into an incubator on a daily cycle of 15°C, 12-h light and 10°C, 12-h dark. These conditions were used to precondition the plants to the cool autumn temperatures expected upon planting into the field in November.
To ensure transplants had no major size differences at the start of the experiment, I recorded the number of tillers >1cm (including the original ramet planted) and the summed length of all tillers a few days before planting outdoors. Initial size, assessed as total tiller length, did not differ between E+ and E– plants (F1,186 = 0.15, P = 0.70; E+ = 25.9 ± 1.0 cm, E– = 26.4 ± 1.0 cm). In addition, there was no significant interaction of genotype with infection status (F8,186 = 0.71, P = 0.68). Mean (± SE) number of tillers for E+ plants was 1.14 ± 0.05, slightly less than that for E– plants (1.26 ± 0.05; F1,186 = 4.42, P = 0.04).
Common garden experiment
A 3 x 4 m plot was selected for use as a common garden within a low-maintenance lawn in Millstone Township, Monmouth County in central New Jersey, USA (40.25°N, 74.47°W). This part of New Jersey (inner coastal plain) has a temperate climate, with monthly temperature varying from –1.6°C in January to 23.3°C in July (30-yr annual mean = 11.1°C; all data for Hightstown, NJ, USA, 4.5 km from the site, http://www.worldclimate.com). The length of the frost-free period ranges from 170 to 180 d (Robichaud and Buell, 1983). Monthly precipitation is greatest in July (124 mm) and August (114 mm) but is relatively evenly distributed throughout the year. Soils are fine-textured, mesic, and relatively fertile (Robichaud and Buell, 1983).
From 15 to 17 November 2001, each randomly chosen individual (typically 1–3 tillers) was removed from the planting tray and placed into an 8-cm diameter hole cut into the plot with a bulb planter. Individuals were planted along a grid of 15 columns and 15 rows, spaced at 20-cm intervals, with a 60-cm wide access path through the middle of the plot. Because L. perenne is a caespitose (bunch) grass without rhizomes or stolons and the root systems do not spread extensively as plants add tillers, this degree of spacing was maintained. Thus, I reasoned that plants did not interact belowground during the experiment. Although 10 genotypes were thought to be available as E+ and E– plants, I later determined that all 20 individuals of one genotype (B) planted into the plot were infected; thus, nine genotypes were represented by both E+ and E– plants (N = 10–12 plants per infection status per genotype). Although 20 E+ individuals of genotype B were planted, these plants were only used to estimate infection intensity and to add to the survival and flowering data of the infected plants—statistical comparisons of the performance of E+ (N = 91) and E– plants (N = 94) do not include this genotype.
At the time of planting, all other vegetation was removed by hand, and the plot was weeded of other plants that recruited within the plot at least three times per year. Most of these recruits were small annual weeds (e.g., Poa annua L.) that were easily extracted by pulling, with minimal soil disturbance. Thus, for the duration of the experiment, all perennial ryegrass plants were free of competition. They received ambient rainfall and were never fertilized. At no times were plants clipped, and senescent leaf material, culms, and spikes were not removed.
Depending on season, plants received 5–8 h•d–1 of direct sunlight. Light readings recorded with a quantum photometer (Li-Cor Model LI-189, Lincoln, Nebraska, USA) at 10 positions within the plot on a sunny day (21 June 2002) averaged 1674 (SD = 40) µmol•m–2•s–1. Following planting, the autumn was mild during the first week (mean daily temperatures from 7 to 14°C, with a maximum high of 21°C), cooler during the following week (mean = 7°C), and moderate throughout December 2001 (mean = 5°C, +3.8°C above the long-term average for December). These conditions were evidently favorable to establishment, and all transplants survived through the end of 2001 and the first winter, which also had above average temperatures (monthly mean = 3°C in both January and February 2002, +4.8°C and +3.6°C above the long-term average, respectively).
Beginning at 6 wk (27 December 2001), the number of tillers was recorded for all plants at 2–3 mo intervals for the duration of the experiment. The number of spikes (inflorescences, one per flowering tiller) and the date on which the first spike was at least 50% emerged above the flag leaf was recorded each spring when flowering occurred. Some rabbit herbivory was noted in June 2002, but did not affect data on tiller or spike production. Mole damage resulted in the death of five plants (3 E+ and 2 E– individuals of 4 genotypes). Other recorded mortality was presumably due to abiotic factors or natural die-off following a reproductive episode (following inflorescence production, flowering tillers die in this species).
For the three years of the experiment, precipitation was relatively close to the 58-yr, long-term average for the region. In the first growing season (2002), rainfall was below average in June and July, but precipitation for the year (94.9 cm) was 83.2% of the long-term average (114.1 cm). Rainfall for the second growing season (2003) was mostly above average and precipitation for the year (131.7 cm) was 115.5% of the yearly average. For the third growing season (2004), rainfall was below average in May and June; precipitation for the first half of the year (the final 6 mo of the experiment)(44.8 cm) was 83.5% of the long-term average (53.6 cm).
In late June 2002 and 2003, leaf sheath samples were collected for microscopic examination and assessment of infection intensity (details in Endophyte extermination and assessment). Plants were clipped at the soil surface, and all aboveground tissues were collected after flowering in the third growing season (16 July 2004). These tissues included mostly living tillers with a small quantity of senescent tissues (e.g., flowering tillers, which normally senesce after spikes mature). Tissues were dried to constant mass at 60°C, separated into vegetative and flowering tillers and weighed.
Data analysis
Endophyte infection intensity was statistically compared between the first (2002) and second (2003) growing seasons with analysis of variance (ANOVA), where genotype, year, and their interaction were the sources of variation. Data on infection intensity collected in 1998 on the original source plants are presented for comparative purposes only (Fig. 1) and were not included in these analyses. To determine whether or not infection affected survival or the proportion of plants that flowered, row-by-column G-tests of independence (Sokal and Rohlf, 1981) were performed.
|
). Between-subject effects were genotype, infection status, and their interaction. Within-subject effects were time, and time by genotype, time by infection, and time by genotype by infection interactions. To comply with ANOVA assumptions, both tiller and spike numbers were log-transformed. Flowering date, expressed numerically as the log-transformed day of the year, was also analyzed with repeated-measures ANOVA. The final mass of aboveground tissues was analyzed separately for vegetative and reproductive tillers (log-transformed) with two-way ANOVA. Genotype, infection status, and their interaction were the sources of variation. Analyses were conducted with the Statistical Analysis System, version 9.1 (SAS Institute, Cary, North Carolina, USA).
When both host genotype and endophyte infection had significant effects in ANOVA, the relative magnitude of these effects were quantified using mean square estimates as outlined in Underwood (1997, p. 348). These estimates provided a simple means to compare the relative importance of genotype and endophytes in explaining phenotypic variation in tiller production, final tiller number, and biomass.
RESULTS
Endophyte infection intensity
In 1998, 3 years before the field experiment began, mean (± SE) infection intensity (number of hyphae visible per 400x field) for the 10 host genotypes ranged from 2.67 ± 0.50 (genotype L) to 8.30 ± 0.42 (genotype H; Fig. 1). Across all genotypes infection intensity was 6.14 ± 0.39.
The first growing season (summer 2002) after planting in the field, mean infection intensity was 6.68 ± 0.39. The following year, infection intensity declined to 5.85 ± 0.41 (Fyear, 1,173 = 9.51, P < 0.01). Variation among genotypes was highly significant (F9,173 = 6.28, P < 0.0001), and there was a significant genotype by year interaction (F9,173 = 2.59, P < 0.01). Despite the variation in infection intensity, some genotypes (e.g., H and K) had a consistently high intensity of endophyte infection (Fig. 1), while others were typically relatively low throughout (e.g., G and L).
Although endophyte infection appeared to decline slightly in some genotypes over the 5-yr between 1998 when the plants were initially examined (3 yr before the common garden was established) and 2003 (e.g., G and H in Fig. 1), in others (e.g., B, K and T) infection intensity was remarkably constant. There was no evidence of complete endophyte loss from any infected individual. In addition, in both 2002 and 2003, all individuals originally designated in 1998 as E– were still endophyte-free.
Survival
Plant survival during the field experiment was uniformly high, ranging from 17/20 in three genotypes to 20/20 in one genotype. There was no relation of survival to infection status: by the third year, 87.4% of the E+ plants and 90.4% of the E– plants were alive (G = 0.03, not significant).
Tiller production
The nine perennial ryegrass genotypes had highly significant differences in tiller production (Table 1, Fig. 2). In addition, tiller numbers generally increased over time, but the changes in tiller production varied among genotypes as evidenced by the highly significant time by genotype term in the repeated-measures ANOVA (Table 1).
|
|
With mean square estimates from the repeated-measures ANOVA output for tiller numbers over time, the relative magnitude of the time effect was 95.9%. The time by genotype effect was 3.8%, 19 times greater than the time by infection effect (0.2%). For the final number of tillers, the effect of genotype was 81.7%, which was 10 times greater than the effect of endophyte infection (8.2%).
The temporal patterns in tiller production over the three years of growth in the common garden reflected seasonal patterns in growth and reproduction. The first peak at 26 wk (Fig. 2) reflects rapid tiller production during the spring after set up of the experimental population when temperatures at the field site are optimal for growth of perennial ryegrass (18–20°C; Jung et al., 1996). After spike production in late spring, tiller numbers decline because flowering tillers are monocarpic (Kays and Harper, 1974). Another growth increase at 79 wk is evident during spring of the second year, followed again by a postflowering decline (Fig. 2). The final increase in tiller numbers at 127 and 138 wk represents new tiller production in spring and early summer of the third growing season.
Flowering
The proportion of plants that flowered (i.e., produced at least one spike) was greatest in the first year, but did not differ between E+ (0.982) and E– plants (0.979; G < 0.01, P >> 0.05). Flowering was lower the second and third years, but again did not differ between E+ and E– plants (2nd yr: E+ = 0.782, E– = 0.841, G = 0.11, P >> 0.05; 3rd yr: E+ = 0.877, E– = 0.835, G = 0.05, P >> 0.05).
Flowering date was significantly affected by genotype (F9,111 = 7.14, P < 0.0001), but not by endophyte infection (F1,111 = 0.28, P = 0.60). Flowering date varied among years (F2,222 = 238.11, P < 0.0001), and there was a significant year by genotype interaction (F18,222 = 3.02, P < 0.0001). However, there was no year by endophyte interaction (F2,222 = 1.46, P = 0.23). Mean (±SE) calendar days to flowering in years 1, 2, and 3 was 163.0 ± 5.0, 163.2 ± 7.2, and 151.4 ± 6.0, respectively, for E+ plants. Comparable values for E– plants were 162.6 ± 5.5, 164.2 ± 7.7, and 152.6 ± 7.2, respectively.
The number of spikes produced varied significantly with genotype and year (Table 1). Year and genotype interacted significantly. Many, but not all, genotypes produced the most spikes the first year (Fig. 3). During this episode of flowering, from 17.1% to 51.9% of the total number of tillers made by a genotype produced spikes. However, endophyte infection had no significant effect on spike production, and there was no interaction between year and infection status (Table 1, Fig. 3).
|
|
Mean tiller mass varied greatly among genotypes (F9,163 = 4.96, P < 0.0001): morphologically, genotypes ranged from those with many, small tillers (e.g., genotypes T and R) to those with fewer, but larger tillers (e.g., genotypes K and E; Fig. 4). Endophyte infection had no effect on mean tiller mass (F1,163 = 0.25, P = 0.62). Across all genotypes (including E+ and E– groups), mean tiller mass was negatively correlated with the number of tillers (r2 = 0.36, P < 0.01; tiller mass = 38.9 – 0.23[tiller number]; Fig. 4).
|
The intensity of infection by the endophyte Neotyphodium lolii was high in the Lolium perenne plants used for the common garden experiment, but varied significantly among host genotypes. Similar to infection intensities reported in the current study (Fig. 1), Cheplick (1997) also reported variability in the intensity of endophyte infection for six genotypes of L. perenne, ranging from 3.3 to 7.0 hyphae per 400x field of view. Despite some variation between years, there was no evidence of a consistent decline in infection intensity across all host genotypes or complete loss of endophytic hyphae over the first 2 years of the common garden experiment (Fig. 1). Furthermore, plants of the same infected host genotypes treated with systemic fungicide never regained infection, presumably because Neotyphodium spp. endophytes are completely asexual and do not show horizontal transmission (Schardl and Leuchtmann, 2005). Thus, any effect that the endophyte might have on survival, tiller production, biomass, and/or sexual reproduction of its host should be readily detected by comparison of infected (E+) to uninfected (E–) groups of the same genotypic composition. This type of comparison made over more than a single season on genotypically diverse plants growing in a field environment is necessary if plant ecologists are ever to determine whether endophytic fungi really matter to the population dynamics of their hosts (Cheplick and Faeth, in press).
In this L. perenne population monitored in a common garden in a temperate climate for almost three years, the effect of host genotype on tiller and spike production was far greater than that of endophyte infection status (E+ vs. E–). Although endophyte infection had increased total tiller production (Fig. 2) and dry mass (Table 2) by the end of the experiment, the influence of genotype on all recorded variables was much more pronounced. Infection status was not related to the proportion of plants that survived or flowered, number of days to flowering, number of spikes produced (Fig. 3), or mean tiller mass. Although survival, tiller production, and (sometimes) flowering have been reported to be greater for E+ plants in some grass and sedge species in nature (Clay, 1990a, 1998), for other perennial grasses asexual endophytes do not enhance long-term survival or reproduction (Faeth and Hamilton, 2006; Olejniczak and Lembicz, 2007). In a multiyear study of tiller dynamics of E+ and E– L. perenne, in a field experiment in New Zealand, Eerens et al. (1998) did not detect any effect of endophytes on tiller densities or biomass production over a 4-yr period. The proportion of E+ tillers within the study pasture remained relatively stable over that time (Eerens et al., 1998), suggesting that there was no particular growth or survival advantage to infected hosts.
Despite the lack of solid evidence in the present study and in Eerens et al. (1998) that endophytes impact L. perenne population dynamics, field conditions in both studies were relatively benign, mostly presenting cool, moist conditions in temperate climates that are favorable to growth of perennial ryegrass (Jung et al., 1996). In the 3 years of this experiment, annual precipitation was from 85 to 115% of the long-term average for the region, and there were no prolonged drought periods. Mutualistic benefits of endosymbiosis can be more or less pronounced under abiotically stressful conditions (Cheplick and Faeth, in press), and some have suggested that environmental stresses can favor endophyte-infected L. perenne (Ravel et al., 1995; Lewis et al., 1997; Hesse et al., 2003; Jensen and Roulund, 2004). Plants in the present experiment were also well spaced and grown in the absence of competitors, which is conducive to rapid leaf and tiller production (Kays and Harper, 1974; Davies and Thomas, 1983). Putative host benefits to endophyte infection may take a long time to become manifest, as illustrated by the increase in tiller production by E+ plants of some genotypes, which was not apparent until the third growing season (Fig. 2).
Many have noted that endophyte-mediated effects on grasses are highly contingent on environmental conditions and host (and endophyte) genotype (Saikkonen et al., 1998; Cheplick and Cho, 2003; Cheplick, 2004; Hesse et al., 2004). However, even though genotypes varied greatly in tiller, spike, and biomass production in the experimental L. perenne population (Figs. 2, 3; Table 2), there were never any genotype by infection interactions (Table 1). This lack of interaction implies that natural selection could differentiate among host genotypes, but that the selection process will be independent of whether individuals are infected by N. lolii.
The ability to produce tillers is an important aspect of the population ecology of caespitose grasses (Briske and Derner, 1998); for L. perenne, tiller number is closely tied to aboveground mass and leaf area (Luxmoore and Millington, 1971; Gautier et al., 1999; Cheplick, 2004). The caespitose growth form that results from a closely packed cluster of tillers may allow genets with high tillering ability to consolidate and monopolize resources within its immediate environment (Briske and Derner, 1998). Size, whether assessed as tiller numbers or dry mass, can also be correlated with sexual reproduction. For example, in this L. perenne population, spike production in the third year was positively correlated with the number of vegetative tillers in both E+ and E– groups (E+: F = 14.98, P = 0.0002; E–: F = 13.99, P = 0.0003). Increased reproductive tiller production in L. perenne can result in a greater seed yield (Hesse et al., 2004). In addition, vegetative tillers and developing seeds do not appear to compete for carbohydrates in L. perenne (Warringa and Kreuzer, 1996). Hence, selection might be expected to favor genotypes with a well-developed tillering ability (e.g., genotypes R and T in Fig. 2) because this should improve competitiveness and reproductive fitness. However, tillers in these genotypes tend to be smaller; there is a trade-off between the number of tillers and mean tiller mass (Fig. 4). It is currently not known what the ecological consequences are of genotypic variation in tiller size and number within a L. perenne population. However, tiller survival, the probability of new tiller recruitment, and inflorescence production are positively related to tiller size in other perennial grasses (e.g., Pedro et al., 1997).
In conclusion, this 3-year common garden experiment has revealed genotypic variation for morphological and life history traits in a Lolium perenne population containing a mixture of endophyte-infected and uninfected individuals. Such genotypic diversity is not unexpected in perennial grasses with a predominantly outcrossing breeding system (McNeilly and Roose, 1984; Hayward, 1985; Casler, 1995; Fair et al., 1999). Although endophyte-mediated enhancement of host growth has been documented for some grasses, especially under adverse conditions (Belesky and Malinowski, 2000; Malinowski et al., 2005; Cheplick and Faeth, in press), in the benign conditions of the common garden, endophyte effects were minor and limited to a slight enhancement of tiller (and biomass) production over the 3 years. Thus, the evidence is weak that the fungal endosymbiont functioned as an agent of natural selection among these host genotypes under common garden conditions. Without experimental evidence, plant ecologists should never assume that biotic interactions are invariably significant to the microevolution of particular plant populations. Future work should be directed at interpreting the variation in "ecological outcomes" of symbiotic interactions within the framework of coevolutionary theory (Thompson, 2005): that is, the grass–endophyte association should be explored under a range of natural environmental conditions that incorporates variation in both abiotic and biotic components.
FOOTNOTES
1 The author thanks K. H. Kane for help with data collection and two anonymous reviewers and the associate editor for helpful comments on the manuscript. 
2 e-mail: cheplick{at}mail.csi.cuny.edu
LITERATURE CITED
Bacon, C. W., AND J. F. White Jr. 1994. Stains, media, and procedures for analyzing endophytes. In C. W. Bacon, and J. F. White [eds.], Biotechnology of endophytic fungi of grasses, 47–56. CRC Press, Boca Raton, Florida, USA.
Beddows, A. R. 1967. Biological flora of the British Isles: Lolium perenne L. Journal of Ecology 55: 567–587.[CrossRef][Web of Science]
Belesky, D. P., AND D. P. Malinowski. 2000. Abiotic stresses and morphological plasticity and chemical adaptations of Neotyphodium-infected tall fescue plants. In C. W. Bacon, and J. F. White [eds.], Microbial endophytes, 455–485. Marcel Dekker, New York, New York, USA.
Briske, D. D., AND J. D. Derner. 1998. Clonal biology of caespitose grasses. In G. P. Cheplick [ed.], Population biology of grasses, 106–135. Cambridge University Press, Cambridge, UK.
Burdon, J. J. 1987. Diseases and plant population biology. Cambridge University Press, Cambridge, UK.
Casler, M. D. 1995. Patterns of variation in a collection of perennial ryegrass accessions. Crop Science 35: 1169–1177.
Cheplick, G. P. 1997. Effects of endophytic fungi on the phenotypic plasticity of Loliumperenne (Poaceae). American Journal of Botany 84: 34–40.[Abstract]
Cheplick, G. P. 2004. Recovery from drought stress in Lolium perenne (Poaceae): Are fungal endophytes detrimental? American Journal of Botany 91: 1960–1968.
Cheplick, G. P. 2007. Costs of fungal endophyte infection in Lolium perenne genotypes from Eurasia and North Africa under extreme resource limitation. Environmental and Experimental Botany 60: 202–210.[CrossRef][Web of Science]
Cheplick, G. P., AND R. Cho. 2003. Interactive effects of fungal endophyte infection and host genotype on growth and storage in Lolium perenne. New Phytologist 158: 183–191.[Web of Science]
Cheplick, G. P., AND S. H. Faeth. In press. Ecology and evolution of the grass–endophyte symbiosis. Oxford University Press, New York, New York, USA.
Cheplick, G. P., A. Perera, AND K. Koulouris. 2000. Effect of drought on the growth of Lolium perenne genotypes with and without fungal endophytes. Functional Ecology 14: 657–667.[CrossRef]
Clay, K. 1987. Effects of fungal endophytes on the seed and seedling biology of Loliumperenne and Festuca arundinacea. Oecologia 73: 358–362.[CrossRef][Web of Science]
Clay, K. 1990a. Comparative demography of three graminoids infected by systemic, clavicipitaceous fungi. Ecology 71: 558–570.[CrossRef][Web of Science]
Clay, K. 1990b. The impact of parasitic and mutualistic fungi on competitive interactions among plants. In J. B. Grace, and D. Tilman [eds.], Perspectives on plant competition, 391–412. Academic Press, San Diego, California, USA.
Clay, K. 1998. Fungal endophyte infection and the population dynamics of grasses. In G. P. Cheplick [ed.], Population biology of grasses, 255–285. Cambridge University Press, Cambridge, UK.
Clay, K., AND J. Holah. 1999. Fungal endophyte symbiosis and plant diversity in successional fields. Science 285: 1742–1744.
Clay, K., J. Holah, AND J. A. Rudgers. 2005. Herbivores cause a rapid increase in hereditary symbiosis and alter plant community composition. Proceedings of the National Academy of Sciences, USA 102: 12465–12470.
Davies, A., AND H. Thomas. 1983. Rates of leaf and tiller production in young spaced perennial ryegrass plants in relation to soil temperature and solar radiation. Annals of Botany 57: 591–597.
Eerens, J. P. J., R. J. Lucas, H. S. Easton, AND J. G. H. White. 1998. Influence of the ryegrass endophyte (Neotyphodium lolii) in a cool moist environment. I. Pasture production. New Zealand Journal of Agricultural Research 41: 39–48.[Web of Science]
Faeth, S. H., AND W. F. Fagan. 2002. Fungal endophytes: Common host plant symbionts but uncommon mutualists. Integrative and Comparative Biology 42: 360–368.
Faeth, S. H., AND C. E. Hamilton. 2006. Does an asexual endophyte symbiont alter life stage and long-term survival in a perennial host grass? Microbial Ecology 52: 748–755.[CrossRef][Medline]
Fair, J., W. K. Lauenroth, AND D. P. Coffin. 1999. Demography of Bouteloua gracilis in a mixed prairie: Analysis of genets and individuals. Journal of Ecology 87: 233–243.[CrossRef][Web of Science]
Gautier, H., C. Varlet-Grancher, AND L. Hazard. 1999. Tillering responses to the light environment and to defoliation in populations of perennial ryegrass (Lolium perenne L.) selected for contrasting leaf length. Annals of Botany 83: 423–429.
Gibson, D. J. 2002. Methods in comparative plant population ecology. Oxford University Press, New York, New York, USA.
Glazebrook, J., AND J. Ton. 2007. Biotic interactions: Recurring themes and expanding scales. Current Opinion in Plant Biology 10: 331–334.[CrossRef][Web of Science]
Hayward, M. D. 1985. Adaptation, differentiation and reproductive systems in Lolium perenne. In P. Jacquard, G. Heim, and J. Antonovics [eds.], Genetic differentiation and dispersal in plants, 83–93. Springer-Verlag, Berlin, Germany.
Hesse, U., H. Hahn, K. Andreeva, K. Förster, K. Warnstorff, W. Schöberlein, AND W. Diepenbrock. 2004. Investigations on the influence of Neotyphodium endophytes on plant growth and seed yield of Lolium perenne genotypes. Crop Science 44: 1689–1695.
Hesse, U., W. Schöberlein, L. Wittenmayer, K. Förster, K. Warnstorff, W. Diepenbrock, AND W. Merbach. 2003. Effects of Neotyphodium endophytes on growth, reproduction and drought-stress tolerance of three Lolium perenne L. genotypes. Grass and Forage Science 58: 407–415.[CrossRef][Web of Science]
Hesse, U., W. Schöberlein, L. Wittenmayer, K. Förster, K. Warnstorff, W. Diepenbrock, AND W. Merbach. 2005. Influence of water supply and endophyte infection (Neotyphodium spp.) on vegetative and reproductive growth of two Lolium perenne L. genotypes. European Journal of Agronomy 22: 45–54.[CrossRef][Web of Science]
Hignight, K. W., G. A. Muilenburg, AND A. J. P. van Wijk. 1993. A clearing technique for detecting the fungal endophyte Acremonium sp. in grasses. Biotechnic & Histochemistry 68: 87–90.[CrossRef][Web of Science][Medline]
Hill, N. S., D. P. Belesky, AND W. C. Stringer. 1998. Encroachment of endophyte-infected on endophyte-free tall fescue. Annals of Botany 81: 483–488.
Huff, D. R. 1997. RAPD characterization of heterogeneous perennial ryegrass cultivars. Crop Science 37: 557–564.
Jensen, A. M. D., AND N. Roulund. 2004. Occurrence of Neotyphodium endophytes in permanent grassland with perennial ryegrass (Lolium perenne) in Denmark. Agriculture Ecosystems & Environment 104: 419–427.[CrossRef][Web of Science]
Jung, G. A., A. J. P. van Wijk, W. F. Hunt, AND C. E. Watson. 1996. Ryegrasses. In L. E. Moser, D. R. Buxton, and M. D. Casler [eds.], Cool-season forage grasses, 605–641. American Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, Wisconsin, USA.
Kays, S., AND J. L. Harper. 1974. The regulation of plant and tiller density in a grass sward. Journal of Ecology 62: 97–105.[CrossRef][Web of Science]
Keogh, R. G., AND T. Lawrence. 1987. Influence of Acremonium lolii presence on emergence and growth of ryegrass seedlings. New Zealand Journal of Agricultural Research 30: 507–510.[Web of Science]
Kubik, C., M. Sawkins, W. A. Meyer, AND B. S. Gaut. 2001. Genetic diversity in seven perennial ryegrass (Lolium perenne L.) cultivars based on SSR markers. Crop Science 41: 1565–1572.
Latch, G. C. M., W. F. Hunt, AND D. R. Musgrave. 1985. Endophytic fungi affect growth of perennial ryegrass. New Zealand Journal of Agricultural Research 28: 165–168.[Web of Science]
Leuchtmann, A., AND K. Clay. 1990. Isozyme variation in the Acremonium/Epichloë fungal endophyte complex. Phytopathology 80: 1133–1139.[CrossRef][Web of Science]
Lewis, G. C. 2004. Effects of biotic and abiotic stress on the growth of three genotypes of Lolium perenne with and without infection by the fungal endophyte Neotyphodium lolii. Annals of Applied Biology 144: 53–63.[CrossRef][Web of Science]
Lewis, G. C., C. Ravel, W. Naffaa, C. Astier, AND G. Charmet. 1997. Occurrence of Acremonium endophytes in wild populations of Lolium spp. in European countries and a relationship between level of infection and climate in France. Annals of Applied Biology 130: 227–238.[CrossRef][Web of Science]
Luxmoore, R. J., AND R. J. Millington. 1971. Growth of perennial ryegrass (Lolium perenne L.) in relation to water, nitrogen, and light intensity. Plant and Soil 34: 269–281.[CrossRef][Web of Science]
Malinowski, D. P., D. P. Belesky, AND G. C. Lewis. 2005. Abiotic stresses in endophytic grasses. In C. A. Roberts, C. P. West, and D. E. Spiers [eds.], Neotyphodium in cool-season grasses, 187–199. Blackwell, Ames, Iowa, USA.
McNeilly, T., AND M. L. Roose. 1984. The distribution of perennial ryegrass genotypes in swards. New Phytologist 98: 503–513.[CrossRef][Web of Science]
Müller, C. B., AND J. Krauss. 2005. Symbiosis between grasses and asexual fungal endophytes. Current Opinion in Plant Biology 8: 450–456.[CrossRef][Web of Science][Medline]
Olejniczak, P., AND M. Lembicz. 2007. Age-specific response of the grass Puccinellia distans to the presence of a fungal endophyte. Oecologia 152: 485–494.[CrossRef][Web of Science][Medline]
Osbourn, A., AND S. Y. He. 2006. Biotic interactions: Towards a unifying and balanced view. Current Opinion in Plant Biology 9: 347–350.[CrossRef][Web of Science][Medline]
Pedro, L., V. A. Deregibus, AND N. O. Maceira. 1997. Demographic variability in tiller populations of two perennial pampa grasses. Journal of Vegetation Science 8: 369–376.[CrossRef][Web of Science]
Ravel, C., G. Charmet, AND F. Balfourier. 1995. Influence of the fungal endophyte Acremonium lolii on agronomic traits of perennial ryegrass in France. Grass and Forage Science 50: 75–80.[CrossRef][Web of Science]
Robichaud, B., AND M. F. Buell. 1983. Vegetation of New Jersey. Rutgers University Press, New Brunswick, New Jersey, USA.
Saikkonen, K., S. H. Faeth, M. Helander, AND T. J. Sullivan. 1998. Fungal endophytes: A continuum of interactions with host plants. Annual Review of Ecology and Systematics 29: 319–343.[CrossRef][Web of Science]
Saikkonen, K., P. Lehtonen, M. Helander, J. Koricheva, AND S. H. Faeth. 2006. Model systems in ecology: Dissecting the endophyte–grass literature. Trends in Plant Science 11: 428–433.[CrossRef][Web of Science][Medline]
Schardl, C. L., AND A. Leuchtmann. 2005. The Epichloë endophytes of grasses and the symbiotic continuum. In J. Dighton, J. F. White, and P. Oudemans [eds.], The fungal community: Its organization and role in the ecosystem, 3rd ed. 475–503. CRC Press, New York, New York, USA.
Silvertown, J., AND D. Charlesworth. 2001. Introduction to plant population biology, 4th ed. Blackwell Science, Oxford, UK.
Sokal, R. R., AND F. J. Rohlf. 1981. Biometry, 2nd ed. Freeman, New York, New York, USA.
Thompson, J. N. 2005. The geographic mosaic of coevolution. University of Chicago Press, Chicago, Illinois, USA.
Underwood, A. J. 1997. Experiments in ecology: Their logical design and interpretation using analysis of variance. Cambridge University Press, Cambridge, UK.
van Zijll de Jong, E., K. M. Guthridge, G. C. Spangenberg, AND J. W. Foster. 2003. Development and characterization of EST-derived simple sequence repeat (SSR) markers for pasture grass endophytes. Genome 46: 277–290.[Medline]
von Ende, C. N. 2001. Repeated-measures analysis: growth and other time-dependent measures. In S. M. Scheiner, and J. Gurevitch [eds.], Design and analysis of ecological experiments, 134–157. Oxford University Press, New York, New York, USA.
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. Annals of Botany 78: 749–757.
West, C. P. 1994. Physiology and drought tolerance of endophyte-infected grasses. In C. W. Bacon, and J. F. White [eds.], Biotechnology of endophytic fungi of grasses, 87–99. CRC Press, Boca Raton, Florida, USA.
Wilson, A. D., S. L. Clement, AND W. J. Kaiser. 1991. Survey and detection of endophytic fungi in Lolium germplasm by direct staining and aphid assays. Plant Disease 75: 169–173.[Web of Science]
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
