Crop Science Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 24 June 2005
Published in Crop Sci 45:1456-1463 (2005)
© 2005 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in Crop Science
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Piano, E.
Right arrow Articles by Pecetti, L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Piano, E.
Right arrow Articles by Pecetti, L.
Agricola
Right arrow Articles by Piano, E.
Right arrow Articles by Pecetti, L.
Related Collections
Right arrow Other Fiber Crops
Right arrow Plant Genetic Resources

CROP BREEDING, GENETICS & CYTOLOGY

Specificity of Host-Endophyte Association in Tall Fescue Populations from Sardinia, Italy

E. Pianoa, F. B. Bertolia,*, M. Romania, A. Tavaa, L. Riccionib, M. Valvassorib, A. M. Carronic and L. Pecettia

a Istituto Sperimentale per le Colture Foraggere, viale Piacenza 29, 26900 Lodi, Italy
b Istituto Sperimentale per la Patologia Vegetale, via C.G. Bertero 22, 00156 Rome, Italy
c Istituto Sperimentale per le Colture Foraggere, via Crespellani 4, 09121 Cagliari, Italy

* Corresponding author (bred{at}iscf.it)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Tall fescue [Festuca arundinacea Schreb. var. arundinacea Schreb. (2n = 6x = 42)] breeding objectives are to exploit the natural variation of the associated endophytic fungi and to select specific plant–fungus combinations that optimize the host fitness but do not cause detrimental effects on grazing animals. This study investigated the presence of endophytes in 60 tall fescue natural populations from Sardinia, Italy; identified the endophytes and assessed some important metabolites produced by their interaction with the host plant; and characterized the tall fescue populations for morphological traits, relating the variation among populations to possible differences in the associated endophytes. The high frequency of infected populations (58 out of 60), and high levels of infection (on average, 80% infected seed), suggested an adaptive advantage of E+ plants under harsh Mediterranean conditions. Morphological identification of fungal isolates, in comparison with Neotyphodium coenophialum (Morgan-Jones et Gams) Glenn, Bacon et Hanlin check isolates, made it possible to separate two groups of populations. One, infrequent, associated with a long-conidia endophyte (attributable to N. coenophialum), and another associated with a short-conidia form, likely belonging to the Festuca arundinacea Taxonomic Grouping-2 (FaTG-2), which was previously isolated in a few Mediterranean accessions. Chemical analysis revealed the presence of loline alkaloids only in populations associated to the long-conidia endophyte, thus corroborating that form's attribution to the loline-inducing N. coenophialum. A difference was also observed in ergovaline concentration between long-conidia (= N. coenophialum) and short-conidia endophyte variants, the latter producing only about 25% of the ergovaline produced by the former. A coevolutionary specificity between the native Sardinian fescue germplasm and its associated endophyte was suggested by the agreement between morphology of the host plant (distinct from germplasm originating in temperate environments) and morphological and biochemical characteristics of the harbored fungus.

Abbreviations: E+, endophyte-infected germplasm • E–, endophyte-free germplasm


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
TALL FESCUE is often infected by endophytic fungi of the genus Neotyphodium, which grow asymptomatically within plant tissues (Neill, 1941). These endophytes, which evolved from sexual Epichloë species, do not produce fruiting structures and are transmitted only through infected plant seed (Schardl, 1996). The relationship between plant and fungus is regarded as mutualistic: the host provides the endophyte with shelter, nutrients, and an easy means of propagation; the fungus improves its host survival through enhanced growth and fertility, better drought tolerance, increased resistance to pests and diseases, and a more efficient utilization of soil nitrogen and phosphorus (Wilkinson and Schardl, 1997). These features make the endophyte-infected (E+) tall fescue potentially very important when breeding new turf cultivars, particularly for low-input management. On the other hand, ingestion of E+ plants has been associated to serious toxicoses in grazing animals (Bacon et al., 1977; Hoveland et al., 1983), thus hampering the exploitation of E+ cultivar benefits for pastures. Evidences indicate that a range of secondary metabolites, including different toxic alkaloids, produced by the plant–fungus association, mediate the effects of the endophyte on both host performance and animal health [for a review see Siegel and Bush (1997)].

Tall fescue is generally reported to be associated with the endophytic species N. coenophialum (Morgan-Jones et Gams) Glenn, Bacon et Hanlin (Glenn et al., 1996). Investigations by Christensen et al. (1993) performed on endophyte isolates from tall fescue populations of diverse geographic origin, and based on isozyme differentiation, alkaloid pattern production, and conidia and colony morphology, revealed remarkable morphological and biochemical variation among isolates. They described two new taxonomic groupings (FaTG = Festuca arundinacea Taxonomic Grouping) distinct from N. coenophialum (FaTG-1) as originally described (Morgan-Jones and Gams, 1982). These new groups, defined as FaTG-2 and FaTG-3, are characterized by shorter conidia than N. coenophialum. In addition, FaTG-2 does not induce production of loline alkaloids in the host, different from FaTG-1 and FaTG-3. Ergot alkaloids have not been reported to be as resolutive as the loline alkaloids in the taxonomic distinction of endophytes. However, they represent a group of compounds of primary agronomic and economic importance, being the main ones responsible for the tall fescue toxicity on animals (Siegel and Bush, 1997).

Taxonomy of the hexaploid (6x) tall fescue is still debated [for a review see Clayton and Renvoize (1986) and Craven et al. (2003)]. It has recently been proposed to expand the genus Lolium L., which is also associated to specific Neotyphodium endophytes, to include tall fescue [Lolium arundinaceum (Schreb.) S.J. Darbyshire = F. arundinacea]. Borrill (1972) and Chandrasekharan and Thomas (1971) supported the origin of 6x tall fescues from hybridization between diploid (2x) F. pratensis Huds. and tetraploid (4x) F. arundinacea subsp. fenas (Lag.) Arcang. However, the heterogeneity among populations of 6x tall fescue may indicate multiple origins with different progenitor species. In particular, it has been proposed that 6x tall fescue evolved separately in the northern and southern sides of the Mediterranean Sea (Borrill et al., 1971; Sleper, 1985). According to ecological, morphological, and genetic considerations, at least two "geographic races" are recognized in tall fescue, viz., the "European" and the "Mediterranean" races. Hybrids between races are reported as frequently being sterile, causing the two races to be reproductively isolated (Jadas-Hécart and Gillet, 1973; Hunt and Sleper, 1981, Ghesquière and Jadas-Hécart, 1995). The European race is widespread worldwide and includes the commonly cultivated varieties both in the USA and in Europe, while the Mediterranean race has been described on the basis of accessions from south Spain, north Africa, and west Asia (Ghesquière and Jadas-Hécart, 1995).

The distribution of the different taxonomic groups of Neotyphodium endophytes, originated through different hybridization processes (Schardl, 1996), seems to support and reflect the distribution of taxonomic groups in 6x tall fescue. This could be the result of the mutualistic nature of the plant–fungus association in tall fescue, the unique maternal transmission of the fungus by host seed, and the physical and genetic isolation of the fungus within its host. Interestingly, both FaTG-2 and FaTG-3 were isolated only from tall fescues collected in south Spain and Algeria (Christensen et al., 1993). Thus, a coevolution pattern is postulated (Schardl, 1996), but a comprehensive investigation attempting to relate variation in the endophyte to the variation in the host plant is still lacking.

The Sardinian populations have proved different for morphological traits from germplasm originating from central and northern Europe (Piano and Pusceddu, 1982), which has always been associated with FaTG-1 (Christensen et al., 1993). Previous investigations, based on a limited number of samples, indicated the presence of short-conidia endophytes in Sardinian tall fescues, likely belonging to the FaTG-2 and/or FaTG-3 (Riccioni and Piano, 1994; Clement et al., 2001). The current investigation was performed to increase the knowledge of the relationships between endophytes and tall fescue populations originating from the Mediterranean basin and relied on a large collection of native germplasm from Sardinia.

To verify the presence of a possible specificity of association between endophyte and Sardinian tall fescues, the objectives of this investigation were (i) to assess the presence and level of Neotyphodium infection in natural tall fescue populations from Sardinia, (ii) to characterize and identify the fungi by means of morphological and biochemical (alkaloid production) diagnostic traits, and (iii) to characterize the relation between possible taxonomic variation in the harbored endophyte and possible morphologic and taxonomic variation in the host plant.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Seed of tall fescue natural populations was collected at 60 different locations of Sardinia, Italy, in June 1998, by harvesting at least 50 plants on areas of 500 to 2000 m2 at any collection site (Fig. 1). Locations had altitude ranging from 0 to 700 m above sea-level, average annual rainfall from 400 to 1100 mm, and soil pH from 6 to 8.5. Physical soil characteristics were also variable. The collected seed samples were stored in controlled environment at 4°C and 45% relative humidity. These conditions are known to generally maintain endophyte viability (Williams et al., 1984).



View larger version (71K):
[in this window]
[in a new window]
  		Fig. 1. Sites of collection of 60 tall fescue populations in Sardinia, Italy, superimposed on an altitude map.

 
Presence and infection rate of Neotyphodium spp. in the 60 populations were assessed within a few weeks from collection by microscopic analysis of 100-kernel samples as described by Shelby and Dalrymple (1987). Seeds were soaked in 5% (w/v) NaOH at room temperature for 16 h, rinsed in running tap water for 3 min, dehulled, stained with aniline blue-lactic acid for 8 h (1 g of aniline blue + 100 mL of water + 200 mL of 85% lactic acid), and then squashed with a cover slip and examined under a microscope for the presence of the endophyte mycelium.

Endophytes of all infected populations were isolated, as described by Latch et al. (1984) with partial modifications. Leaf tissue was harvested from nine seedlings per population previously checked for the presence of endophyte infection. The whole seedling was chopped into pieces of approximately 1-cm length, including both sheaths and blades, and at least five pieces per seedling were examined. These pieces were surface sterilized with 2% (v/v) NaClO for 10 min, placed on potato dextrose agar in Petri dishes and incubated at 23°C for 3 to 4 wk in the dark. To assess the homogeneity of endophyte isolates from each host population, the developed mycelium was identified according to the literature descriptions (Morgan-Jones and Gams, 1982; Latch et al., 1984; White and Morgan-Jones, 1987; Christensen and Latch, 1991) on the basis of macroscopic and microscopic observations made on individual colonies (i.e., size, shape, color, margin, amount of aerial mycelium, growth rate, and conidia size). Having observed uniformity of endophyte isolates for each population, only one isolate per host population was kept to carry out further morphometric measurements and assess the level of variation among isolates. The length of conidia is an easily measurable trait in vitro, the variation of which proved to be closely related to the distinctness among isozyme phenotypes and the subsequent definition of Taxonomic Groupings by Christensen et al. (1993). In the present study, a mycelial plug from each isolate was removed, squashed with a cover slip, and then examined under a microscope to measure the conidia length. The same procedure was applied to two N. coenophialum control isolates, previously obtained from the commercial tall fescue cultivars Jesup and Kentucky-31. Twenty conidia were measured in each of the population isolates and the control N. coenophialum isolates. The variation among isolates for length was tested by a one-way analysis of variance (ANOVA), with the variation among the 20 individual measurements within isolates pooled and used as the error term. A second ANOVA was performed to test the variation between the two groups of populations that were shown clearly by the first ANOVA to have "short" and "long" conidia. For the second ANOVA, the error term was the variation among isolates within the short and long groups.

For a further characterization of each population–fungus association, a quantitative determination of loline alkaloid concentration was performed by capillary gas-chromatographic (GC) analysis of seed samples, as these alkaloids also proved useful indicators of the taxonomic groupings of most isolates in Christensen et al. (1993) and, therefore, may contribute to differentiate tall fescue Neotyphodium endophytes. Since no commercial standards of lolines were available for the GC analysis, lolines were extracted and purified from certified E+ Kentucky-31 seed (International Seeds Inc., Halsey, OR), and N-acetylloline and N-formylloline were synthesized from loline as reported in Petroski et al. (1989) with partial modification. One kilogram of seed was defatted with hexane (3 L), air dried, and ground in a Cyclotec mill (Foss-Tecator, Högnäs, Sweden) to pass through a 2-mm mesh. Flour was extracted with CH3OH (3x 3 L) by shaking overnight at room temperature, and the methanol solution was concentrated in vacuum (10x) at 35-40°C with a rotating evaporator, then diluted with 1% aqueous citric acid (5 volumes), and washed with CHCl3 (3x 1 L). The remaining aqueous solution was adjusted to pH 11 with NaOH and loline-type alkaloids were extracted with CHCl3 (5x 300 mL). The organic phase was concentrated to 500 mL with a rotating evaporator and extracted with 0.2 M aqueous HCl (5x 200 mL), and the aqueous HCl phase was basified with NaOH to pH 11, extracted with CHCl3 (5x 300 mL), concentrated in vacuum, dried over Na2SO4, and filtered. Loline-type alkaloids were precipitated by bubbling anhydrous HCl gas through the CHCl3 solution. Nearly 2 g of mixed loline-alkaloid dihydrochloride salts were extracted (0.2% yield); pure loline was obtained from the pyrrolizidine-alkaloid mix after 3 h 1 M HCl hydrolysis at 80°C. This compound was characterized by means of nuclear magnetic resonance (NMR) and gas-chromatography/mass-spectrometry (GC/MS) analyses (data not reported); N-acetylnorloline and norloline were also found as impurities in the mixture. Calibration curves for each loline-pyrrolizidine alkaloid were assessed by chromatographing mixtures of the pure individual alkaloids with the internal standard (4-phenilmorpholine, Aldrich, Milwaukee, WI). A linear response of lolines (concentration/internal standard concentration) vs. peak area (area/internal standard area) was obtained between 10 and 1800 ng per injection. The GC conditions used in this determination allowed a practical measurable sensitivity of 10 ng µL–1.

A seed sample of each natural population and of N. coenophialum-infected and uninfected (E–) Georgia 5, Jesup, and Kentucky-31 was divided into two subsamples, which were frozen in liquid N2, finely ground, and extracted as described by Yates et al. (1990) with partial modification. Two hundred milligrams of flour sample were extracted with 4 mL of a CH2Cl2:MeOH:NH4OH = 75:25:0.5 solution to which 200 µg of the internal standard were added. Mixture was shaken overnight at room temperature, filtered, and concentrated under vacuum. Three hundred microliters of CHCl3 were added, the mixture was dried over anhydrous Na2SO4 and injected in "splitless" mode in the GC (1.5 µL). GC was a PerkinElmer 8500 (Foster City, CA) equipped with flame ionization detectors and a DB-5 capillary column (30 m long, 0.32-mm diameter, 0.23-µm film thickness). The oven temperature was held at 70°C for 2 min, then programmed to 300°C at 4°C/min and maintained at this level for 10 min; the injector and the detector temperatures were 300 and 320°C, respectively. The helium head pressure was 103.5 Pa (15 psi). To compare the loline concentration of tall fescue accessions with different levels of infection, absolute concentrations (Ca) were weighed (Cw) on the percentages of infection:

For each accession, the whole pattern of loline alkaloids was considered, summing the concentrations of individual, quantified alkaloids. Data were submitted to ANOVA to test variation among accessions, with the two seed subsamples, analyzed separately, as replicates of the factor "accession." Mean values of accession groups, as defined on the basis of the isolation of the endophytes they harbored, were tested by another ANOVA, using the variation among accessions within groups as the error term. Variances within groups were not homogeneous, some groups having near-zero or zero values and others showing values in the thousands. Therefore, the data were subjected to a logarithmic transformation before the statistical analysis (Snedecor and Cochran, 1972).

Ergot alkaloid concentration was also determined on a seed sample of each natural population and check cultivar finely ground after treatment with liquid nitrogen. As the chemical determination was not immediately performed, 500-mg flour samples were stored at –80°C after addition of two drops of chloroform. The alkaloid extraction was performed according to Shelby et al. (1997) with partial modification. Two hundred milligrams of seed flour, to which 2 µg of ergotamine was added as internal standard (Sigma Chemicals Co., St. Louis, MO), were treated with 4 mL of CH3OH/H2O (30:70) pH 8.5 with NH4OH, and gently shaken for 8 h. After centrifugation, the solution was concentrated with a rotary evaporator (pH exactly adjusted to 8.5 with NH4OH) and extracted by partitioning with 3x 3 mL of CHCl3 in a centrifuge tube. The organic solution was evaporated at room temperature and the residue was dissolved in 1 mL of 70% (w/v) alkaline CH3OH. Precipitates were removed by centrifugation and filtration through a 0.2-µm filter and the solution obtained was directly injected (1 µL) into the HPLC apparatus. The alkaloid concentration was determined by the HPLC method described by Shelby and Flieger (1997) with partial modification. Primary standard solutions of ergotamine (internal standard) and ergovaline (0.5–1 mg mL–1) in CH3OH were prepared. These solutions were stored at –20°C for several days without apparent loss of detectable alkaloids, other than the expected epimerization. Working standard solutions (50–500 ng mL–1) in methanol/water (2:1) were prepared daily. The analyses were performed on a PerkinElmer chromatograph equipped with LC250 binary pump and a fluorimeter (excitation at 310 nm and emission at 415 nm; Waters Corporation, Milford, MA) using an XTerra (Waters) column (RP18, 5 µm, 4.6 x 250 mm). The mobile phase was CH3OH/H2O (40:60) plus 0.03% NH4OH (A), and CH3OH/H2O (80:20) plus 0.03% NH4OH (B). The flow rate was 1 mL/min in a linear gradient from 100% A to 100% B in 45 min, holding 100% B for 10 min. The quantitative analysis was performed by an integration dedicated software (TotalChrom, PerkinElmer). The areas of integrated peaks were compared by linear regression to standard curves previously generated from chromatographic analyses of ergotamine tartrate and ergovaline analytical standards. The ergot alkaloid concentration was expressed as µg kg–1 of dry matter, on the basis of the recovery of the ergotamine internal standard. The same ANOVA previously described to test loline mean values of accession groups was also applied to the ergot alkaloid concentration (both absolute and weighed on infection percentage) after logarithmic transformation of original values.

The 60 natural tall fescue populations collected in Sardinia and four control cultivars were grown at Sanluri, south Sardinia (39°30' N, 8°50' E; 68 m above sea-level; 450 mm long-term average annual rainfall). The control cultivars were Barcel (E–, from the Netherlands), Magno (E–, from northern Italy), Sopline (E–, from France), and Tanit (E+, obtained from Sardinian germplasm) (Riccioni et al., 1994). Seeds were germinated in Petri dishes in September 1998; seedlings were moved to plugtrays and grown in the greenhouse until they were transplanted in the field on 27 January. Each population was represented by three replications of a row of six spaced plants (75 cm apart) in a randomized complete block design. The following traits were recorded on three central plants per plot: heading time (days from 1 January), foliage index (a visual appraisal of leafiness with score from 1 = min to 9 = max) and number of panicles per plant; on the main stem of each of the three plants, flag-leaf length and width, panicle length, length of the first panicle-internode, and number of primary and total branches per panicle were recorded. To verify whether natural populations harboring different endophyte types were also distinguishable on a morphological basis, a canonical discriminant analysis was performed holding the conidia type as the classification variable and the recorded plant characters as original variables. Mean values of each population were computed, and for all the observed E+ populations they were submitted to the analysis according to the DISCRIM procedure of SAS Software (SAS Institute Inc., 1989). The allocation of each individual population to its closest conidia group was checked by a "jackknife classification" using the Crossvalidate option of the DISCRIM procedure in SAS. The computed discriminant function, separating the two conidia groups, was also used to assign a posteriori to either group the four control cultivars and the populations resulted E–. Finally, to assess in a univariate sense the distinctness of the two groups of E+ populations thus formed, an ANOVA was performed on the average values per plot of each recorded character, testing the factor "group" on the error term represented by the group x block interaction.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
Microscopic analysis revealed that 58 of the 60 tall fescue natural populations from Sardinia were endophyte-infected. Moreover, the mean infection rate per population was 80%, ranging from 25 to 100%. More than two-thirds of the populations had an infection level higher than 70% (Fig. 2). These results agree with those reported by Riccioni and Piano (1994) and Clement et al. (2001), where 75 and 100%, respectively, of the tested Sardinian populations were infected. The presence of the endophyte in most populations, and the high endophyte infection rates provide indirect evidence of the adaptive advantage conferred by the endophyte to the host survival in the stressful environments of Sardinia, and contrast with the lower levels of infection reported in grass populations from cool environments in Europe (Lewis, 1994; Ravel et al., 1994; Oldenburg, 1997). In our research, the only two E– populations (coded as FA98/059 and FA98/060) were collected at mountain sites, where drought and heat stresses are generally lower than in the rest of Sardinia (Arrigoni, 1968).



View larger version (9K):
[in this window]
[in a new window]
  		Fig. 2. Relative frequency of endophyte infection percentage assessed in seed samples of tall fescue natural populations from Sardinia, Italy.

 
As in Christensen and Latch (1991), a wide range for the length of individual conidia (3.6–16.8 µm) was found in the present study. Nonetheless, the one-way ANOVA (data not reported) among Sardinian population isolates indicated significant (P = 0.05) differences among mean conidia lengths of populations and resulted in two population groups with contrasting length. A short-conidia group, which included 50 of the collected populations, had mean lengths per isolate shorter (5.5–6.3 µm) than previously reported for N. coenophialum (Latch et al., 1984; Morgan-Jones and Gams, 1982). Most of the individual conidia in these isolates were shorter than 6.5 µm, which is the lower limit for N. coenophialum given in the taxonomic key for the species (White and Morgan-Jones, 1987). The long-conidia group comprised the remaining eight E+ populations, with mean lengths (11.5–11.7 µm) included in the range (6.5–13 µm) given in White and Morgan-Jones (1987) for N. coenophialum, and much longer than the lower limit for the species (8 µm) indicated by Christensen et al. (1993). Hereafter, we shall refer to the tall fescue populations in the first group as "short-conidia" populations, and to those in the second group as "long-conidia" populations. These findings are consistent with those reported in Clement et al. (2001), who found both short- (<6.5 µm) and long-conidia (>6.5 µm) endophytes in 10 Sardinian accessions. However, unlike in Clement et al. (2001), where 60% of populations hosted the long-conidia endophyte, our data, based on a more comprehensive representation of Sardinian germplasm, suggest that the majority of populations in Sardinia harbor short-conidia endophytes.

Both long-conidia endophytes and N. coenophialum differed significantly (P = 0.05) for conidia length from the short-conidia variant. No significant difference for this trait was found between long-conidia endophytes and N. coenophialum isolates from commercial cultivars (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Mean and range values of conidia length of endophytic variants isolated from 58 tall fescue natural populations from Sardinia, Italy, and of two test-isolates of Neotyphodium coenophialum from Jesup and Kentucky-31 tall fescue cultivars.

 
The GC analysis of loline alkaloids on seed samples allowed the identification and quantification of N-formylloline, N-acetylloline, N-metylloline, and loline. Variation among accessions in total loline concentration agreed with the grouping resulting from the morphological characterization of the isolates. Lolines were present in all the long-conidia populations and in the N. coenophialum controls. Loline content was not significantly different from zero (P = 0.05) (represented by the E– control cultivars) for the two E– populations, and for all the short-conidia populations but one. Loline mean content of short-conidia populations was significantly lower than the mean content of both long-conidia populations and N. coenophialum-infected controls (Table 2).


View this table:
[in this window]
[in a new window]
  		Table 2. Mean total loline concentration (absolute and weighed on infection percentage) in seed of 58 endophyte-infected (grouped by conidia variants) and two uninfected tall fescue natural populations from Sardinia, Italy, and in Neotyphodium coenophialum-infected and uninfected seed of the control cultivars Georgia 5, Jesup, and Kentucky-31.

 
By the HPLC analysis, different ergot alkaloids were identified by comparison with chromatograms published in Shelby et al. (1997). However, the alkaloids ergosine, ergonine, ergine, and their epimers were detected in low amounts, while ergovaline and its epimer ergovalinine were the major components (about 80% of total) of the ergopeptine fraction, in agreement with Siegel and Bush (1997). Because of the quantitative relevance of ergovaline, and since ergovaline and ergovalinine alone have been used as indicators of ergot alkaloid synthesis in Neotyphodium-infected grasses (Siegel and Bush, 1997), the concentration of (ergovaline + ergovalinine) was statistically analyzed and is reported in Table 3. Just as with lolines, long-conidia populations and N. coenophialum-infected cultivars showed similar ergot alkaloid concentration. The concentration in short-conidia populations was markedly lower (on average, only about 25%) than, and significantly (P = 0.05) different from those in long-conidia populations and infected cultivars. Uninfected populations and cultivars showed no synthesis of ergot alkaloids (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Mean concentration of ergovaline + ergovalinine (absolute and weighed on infection percentage) in seed of 58 endophyte-infected (grouped by conidia variants) and two uninfected tall fescue natural populations from Sardinia, Italy, and in Neotyphodium coenophialum-infected and uninfected seed of the control cultivars Georgia 5, Jesup, and Kentucky-31.

 
On the basis of both morphological and biochemical characterization, it appears that long-conidia endophytes, all of which gave rise to the synthesis of loline alkaloids, may be ascribed to N. coenophialum (FaTG-1). The fact that all the short-conidia isolates (except one with intermediate loline content) did not result in lolines suggests their belonging to the Taxonomic Grouping-2 (FaTG-2) of Neotyphodium. This group was previously described by Christensen et al. (1993) in a few tall fescue accessions from south Spain and North Africa, and by Riccioni and Piano (1994) and Clement et al. (2001) in some accessions from Sardinia. The short-conidia endophyte isolate showing intermediate loline production could possibly be ascribed to the Taxonomic Grouping-3 (FaTG-3), described by Christensen et al. (1993) in two tall fescue accessions from the Mediterranean basin. A sampling effect cannot be excluded in the rare occurrence of this taxonomic grouping within the collected germplasm. The presence of the FaTG-3 variant in the tall fescue Sardinian germplasm could also be inferred from Clement et al. (2001), who found a short-conidia endophyte in three populations which were resistant to the bird cherry-oat aphid (Rhopalosiphum padi L.), and it is known that loline alkaloids are likely responsible for aphid deterrence (Johnson et al., 1985; Siegel and Bush, 1997). In addition to toxic effects against insects, Malinowski and Belesky (2000) have also suggested a role of loline alkaloids in the enhanced drought tolerance of E+ tall fescues. Hence, it remains to be explained what is the adaptive meaning of the lack of lolines in the majority of the tall fescues collected in Sardinia.

For the first time to our knowledge, in the Sardinian germplasm there appeared to be a difference in ergovaline concentration between long-conidia (= N. coenophialum) and short-conidia endophyte variants. The lower concentration in tall fescue populations harboring the short-conidia endophyte still does not warrant their safety toward grazing animals. However, having analyzed seed bulks deriving from many different genotypes, it might be of some interest to assess whether the low mean concentration of certain short-conidia populations implies the occurrence of nonergovaline-producing plant–endophyte combinations in the population, which could be exploited by breeding.

The canonical discriminant analysis performed on plant morphological traits resulted in a discriminant function clearly separating the tall fescue population groups harboring endophytes with different conidia length (Fig. 3). An impression of the relative effect of each original variable on the discriminant function was given by the standardized canonical coefficients reported in Table 4. The size of the flag-leaf and panicle, as well as the foliage index seemed to have a larger effect on the discriminant function than other characters. The analysis correctly allocated 100% of natural populations to the a priori grouping based on the harbored endophyte. This result highlighted a perfect agreement between the taxonomic distinction of the endophytes [both morphological (length of conidia) and biochemical (presence/absence of loline alkaloids)] and the morphological features of their host accessions. The a posteriori classification of control cultivars and the two E– populations allocated the cultivar Tanit, developed from Sardinian germplasm, to the short-conidia group, confirming its attribution to the FaTG-2 variant (Riccioni and Piano, 1994), and the remaining E– cultivars, originating from continental Europe, as well as the two uninfected populations, to the long-conidia group (Fig. 3). It was possible, therefore, to distinguish accessions harboring different conidia variants on the basis of the morphology of the host plant. In particular, the ANOVA showed that short-conidia populations were slightly earlier in heading, had lower leafiness, longer and narrower leaves, and laxer panicles than the long-conidia populations (Table 4).



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. Frequency distribution of the canonical variable values calculated by discriminant analysis on nine morphological traits of 60 tall fescue populations collected in Sardinia, Italy, and of four control cultivars (Barcel, Magno, Sopline, and Tanit).

 

View this table:
[in this window]
[in a new window]
  		Table 4. Standardized canonical coefficients of nine morphological characters in the canonical discriminant analysis; F test significance from the analysis of variance, and mean values of the nine characters in 58 tall fescue populations from Sardinia grouped by conidia variants.

 
Our results corroborate and extend the findings, on the basis of a limited number of cases, by Christensen et al. (1993), Riccioni and Piano (1994), and Clement et al. (2001) on a possible host-endophyte specificity and coevolution in the Mediterranean basin. The grouping of accessions and cultivars observed in the present study suggests that long-conidia and E– populations might be of continental origin, as were the control cultivars Barcel, Magno, and Sopline, whereas the short-conidia populations probably belong to native Mediterranean germplasm.

This investigation showed that the morphology of the host plant and the identification of the harbored endophyte could be combined to provide useful information for the settlement of the phylogeny and taxonomy of tall fescue.


   ACKNOWLEDGMENTS
 
We wish to thank Prof. M. Flieger, Institute of Microbiology, Czech Academy of Science, Prague, for kindly providing the ergovaline standard used in the HPLC analysis.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Research supported by the Project ‘Turfgrasses and technical cover crops’, funded by the Italian Ministry of Agriculture and Forestry Policy; research paper no. 60.

Received for publication May 11, 2004.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 


Related articles in Crop Science:

THIS ISSUE IN CROP SCIENCE

Crop Science 2005 45: vii. [Full Text]  




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in Crop Science
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Piano, E.
Right arrow Articles by Pecetti, L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Piano, E.
Right arrow Articles by Pecetti, L.
Agricola
Right arrow Articles by Piano, E.
Right arrow Articles by Pecetti, L.
Related Collections
Right arrow Other Fiber Crops
Right arrow Plant Genetic Resources

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Plant Registrations Soil Science Society of America Journal
Journal of Natural Resources
and Life Sciences Education		 Journal of
Environmental Quality