Skip to main content
NCBI home page
BMC Genomics logoLink to BMC Genomics
. 2025 Dec 12;27:53. doi: 10.1186/s12864-025-12331-0

The relationships among alkaloid concentration, endophyte mycelial concentration and host genetics in the tall fescue Epichloë symbiosis

Darrian R Talamantes 1,✉, Courtney Phillips 2, Carolyn Young 3, Jason G Wallace 1,4,5
PMCID: PMC12817529  PMID: 41388512

Abstract

Tall fescue (Lolium arundinaceum) is an important forage and turf grass that covers 35 million acres (140,000 square kilometers) in the transition zone of the southeastern United States. Most tall fescue in the US is infected with a symbiotic fungus, Epichloë coenophiala, which confers biotic and abiotic stress tolerance for the plant but also produces toxic alkaloids that harm livestock. Although there has been prior evidence that the grass host can influence alkaloid production levels, these have never been precisely quantified. Here, we report on testing alkaloid concentration and relative fungal biomass on > 1000 genetically distinct tall fescue plants. We find that these two traits are weakly correlated, and that both show evidence of moderate to high influence by the host genome. Genome-wide associations find only a single marginally significant hit, however, implying that any genetic control by the host is spread among a large number of genes. These results indicate that the host plant exerts moderate influence on these endophytic traits, that the two are largely independent of each other, and that the host’s influence is likely due to a large number of genes of small effect. These results have relevance for breeding tall fescue for forage and turf production, and especially for optimizing the endophyte relationship for tall fescue management.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-025-12331-0.

Keywords: Tall fescue, Epichloë, Ergot alkaloids, Heritability, Genomics, Symbiosis, RNA sequencing, Heritability, SNP

Background

Tall Fescue (Lolium arundinaceum) is an important forage and turf grass that covers 35 million acres (140,000 km2) in the transition zone of the southeastern United States [1]. Tall fescue is valued because of its high heat tolerance, adaptability to many soil types, high forage production, and ability to survive heavy grazing from cattle [2–6].

The most widely planted cultivar of tall fescue in the United States is Kentucky 31, which was released In 1942 and is valued for its widespread adaptation and hardiness [1, 7]. One of the reasons that Kentucky 31 is so hardy is that it contains a fungal endophyte, Epichloë coenophiala, which improves drought tolerance, deters insect herbivory, and promotes overall plant vigor. E. coenophiala is an asexually reproducing fungus that propagates by growing within its tall fescue host and infecting developing seeds, ensuring it is vertically transmitted to the next generation [8–10].

Fungi in the genus Epichloë are thought to help grasses grow in times of drought and environmental stress [11–14]. E. coenophiala defends its host from herbivores such as insects, nematodes, and mammals [15–17] due to producing various alkaloids [18] that are only produced when associated with the plant host [19]. Endophyte-free tall fescue lacks persistence in the field and shows a clear decrease in stress tolerance when compared to its endophyte-infected counterpart [20].

E. coenophiala produces several classes of alkaloids, including indole-diterpenes, pyrrolopyrazines (e.g. peramine), pyrrolizidines (e.g. N-formyllolines), and ergot alkaloids [21]. Peramine and the lolines are most well known for their anti-insect properties [22, 23]. Indole-diterpenes can be insecticidal but are best known for the mammalian mycotoxin, lolitrem B that causes ryegrass staggers [24]. However, of these alkaloids, the most relevant for detrimental effects on grazing mammals are ergot alkaloids, which cause vasoconstriction within cattle that can lead to the loss of weight, gangrenous lesions in extremities, spontaneous abortion, and other aspects of “fescue toxicosis” [25]. Ergot alkaloid production is encoded by the EAS gene cluster [26, 27]. Isolates that lack key genes in this cluster have been identified and introduced into tall fescue to produce “non-toxic” varieties [28, 29].These “novel” non-toxic isolates do not produce ergot alkaloids and have been shown to eliminate fescue toxicosis. However, due to decades of toxic fescue use, there can be a high cost to pasture renovation in both time and money (e.g. cost of seed), so much of the tall fescue in the U.S. remains the toxic variety [1, 30].

Studying the mechanisms that control alkaloid concentration made within Epichloë has been difficult. In part this is because the Epichloë gene clusters that encode for alkaloids can create a wide range of alkaloid diversity profiles, even within the same species [31, 32]. E. coenophiala can also only reliably produce alkaloids when living inside a host [19]. Nonetheless, some factors have been identified that appear to influence alkaloids in planta. For example, alkaloid levels increase when nitrogen fertilizer is applied, when seed heads have sprouted, during the months of April-May and August-September when there is active plant growth [13, 33–35]. There is some evidence that the plant influences the amount of alkaloids that Epichloë produces [36–39].

We report here the analysis of the interaction between tall fescue and its fungal endophyte, Epichloë coenophiala. Our specific goal was to determine the degree to which the plant host can affect fungal growth and alkaloid concentration in planta, using a pair of biparental mapping populations sharing a common parent and a larger open cross among 17 parents. We report on tall fescue’s estimated heritability on these Epichloë-related phenotypes and a lack of clear genomic regions in tall fescue influencing these traits. With 1083 genetically distinct individuals tested for both alkaloids and fungal biomass, this represents the largest direct comparison of alkaloid concentration in tall fescue to date.

Methods

Population design

The plant materials used in this study were generated from an open-pollinated cross among 17 tall fescue genotypes similar to the cultivar KY-31. All genotypes harbored a strain of Epichloë coenophiala with a known ability to produce ergot alkaloids, peramine, and lolines. Production of ergovaline was confirmed in all 17 parental genotypes using the Phytoscreen tiller ergot alkaloid kit (Agrinostics, Watkinsville, GA) [40]. Once the 17 genotypes were established, they were allowed to open pollinate among themselves (Supplementary Figure S1). Seeds were collected from each mother plant In 2015, 2016, and 2017 In 2019 these seeds were planted in a greenhouse at the University of Georgia. Since the paternal plant was unknown, approximately 176 plants per maternal parent were planted. All plants were tested for endophyte with Agronostics immunoblot kit [41], and plants without endophyte were discarded. All plants were grown out to the vegetative state and trimmed at least once a week to ensure reproductive tillers did not mature.

After identifying paternal parentage (see Sequencing and Genotyping, below), a second set of genotyping and phenotyping was performed on a subset of the original cross. This follow-up sampling focused on the two largest sets of progeny from the same original parents (314 × 310 and 314 × 312); since these groups shared the parent 314, these are functionally a pair of half-sib families, with a total of 175 progeny (Supplementary Figure S1).

Sampling

To characterize plant genotypes, DNA was extracted from three freeze-dried leaf blades. To determine the ergot alkaloid concentrations, 9 whole tillers were sampled and leaf blades were discarded, as leaf blades contain little to no endophyte [42]. For relative biomass ratios, the grass was sampled from the bottom 1.5–2 inches of 3 to 4 tillers where the fungus is most abundant (Fig. 1). Both the ergot alkaloid samples and the relative biomass samples were determined once in February of 2023 for the entire population and again in February 2024 for only the two half-sib families.

Fig. 1.

Fig. 1

Open in a new tab

Tall Fescue sampling. Leaf blades (blue box) were sampled for genotyping, whereas the pseudostem (orange box) was taken for alkaloid analysis and fungal biomass since the fungus is concentrated here (Christensen et al., 2008). Figure courtesy of Chloe Mootz

Fungal preliminary data

Differences among original parental lines were tested by selecting 7 parent plants (301, 305, 312, 314, 315, 319, 320) from the 17 founders and creating 5 subclones for each of them. Clones were grown in the greenhouse and each clone was sampled once for both alkaloid testing and biomass testing, with one random plant sampled twice to evaluate technical variation. Parents 315 and 320 had plants die, so additional technical replicates from these lines were taken from other clones. ANOVA was used to determine if any differences existed among the genotypes, followed by Tukey’s Honest Significant Difference test to determine the differences among them. Similar ANOVA and HSD tests were used later to determine differences among the two half-sib families and the entire population.

Sequencing and genotyping

Since a reference genome for tall fescue was not available at the time, we performed skim sequencing on the 17 parental lines at a depth of 5x using the Illumina platform. These sequences were used to create a rough assembly for later mapping. Specifically, the reads were trimmed using TrimGalore [43], and then assigned taxonomy using Kraken2 (Wood et al., 2019). Any reads flagged as non-plant species were removed. De-novo assemblies of the parental lines were created using SPAdes [44]. From these parental assemblies, we chose the most complete assembly to use as a reference for the next steps. Using BWA [45], BCF tools [46], and VCF tools [47] a VCF file was created and filtered for SNPs with a minor allele frequency of 5%, a mean depth between 20 and 1000, a minimum quality score of 50, and present in 95% of the parents. From the parental SNPs identified this way, a FlexSeq panel (Rapid Genomics, Gainsville, FL) was developed and 3000 progeny were tested with this panel.

Because only the maternal parent was known for each progeny, paternal parents were identified based on shared k-mers. K-mers were tabulated with KMC3 [48] and filtered for those unique to a single parent and that appeared in at least six progeny. From this reduced list, the number of unique k-mers that were shared between each possible parent-progeny pair were calculated. Then, for each progeny, the counts of shared k-mers with each possible parent were converted into z-scores based on the distribution across all possible parents (so that they could be compared across progeny despite different read depths). Z-scores were then clustered using k-means clustering (k = 4 clusters) and resampled 100 times. Parentage was assigned when (1) the potential parent-progeny pair appeared in the top cluster (with the highest number of shared SNPs/k-mers) at least 65 out of 100 times; (2) exactly 2 parents passed criterion #1; and (3) one of those two parents was the known maternal parent. Using these cutoffs, we were able to infer parentage for 2407 of the 3000 genotyped progeny.

Later, the two largest crosses, 314 × 310 and 314 × 312, and all 17 parents were re-genotyped to greater depth using tunable genotype by sequencing (tGBS) [49] outsourced to Freedom Markers. SNPs were called using a draft scaffold-level tall fescue genome (Bushman et al. personal communication; In Press, G3). tGBS SNPs were called by quality trimming the reads and aligning to the tall fescue draft genome using GSNAP [50]; only reads that aligned uniquely were used to identify polymorphic markers. This resulted in a VCF file with 896,240 SNPs. VCFtools [47] and BCFtools [47] were then used to filter SNPs to a minimum allele frequency of 0.05, minimum site depth of 8, maximum site depth of 30, and maximum 50% missing across samples. This resulted in a final genotype set of 12,103 SNPs with an average read depth of 15.9, compared to the average depth of 4.3 for the original Flex-Seq data.

To find genetic outliers, we conducted a PCA on the filtered SNP dataset. Most individuals from each cross grouped with members of their own cross; however, some individuals from 314 × 312 separated from the main grouping, likely due to misassignment of parentage. Those individuals were filtered out, as well as any individuals that were separate from the main population cluster.

Phenotyping

Total ergot alkaloids were measured via quantitative ELISA as a fee-for-service via the company Agronostics (Watkinsville, GA). While high performance liquid chromatography (HPLC) is the gold standard for ergovaline measurements, the large number of samples in this analysis made HPLC impractical from both a time and resources perspective, hence relying on ELISA as a faster and less expensive protocol that has been shown to get comparable results [51]. Fungal biomass was estimated based on relative DNA amounts for the fungus and tall fescue hosts, measured by quantitative polymerase chain reaction (qPCR). Tillers were harvested as described in the sampling section and DNA extracted with ZYMO fungal/bacterial miniprep kit (Zymo #D3024) according to the manufacturer’s instructions. Extracted DNA was diluted to 10 ng/µl.

All qPCR reactions used 10 µl of SYBR Green master mix (ROCHE # 04707516001), 1 µl of forward and reverse primers at 10 nm, 3 µl of nuclease free water, and 5 µl of the diluted sample. Using primer sets G3P4 targeting tall fescue glyceraldehyde-3-phosphate dehydrogenase gene and DMAW4 targeting Epichloë dimethylallyl-tryptophan synthase gene [52], all reactions were run in triplicate on a Roche 480 II machine with the following settings: pre-incubation at 95 °C for 5 min, followed by 45 cycles of 95 °C for 10 s, 50 °C for 15 s, and 72 °C for 16 s. The crossing threshold (CT) values were measured for Epichloë and tall fescue to use as a proxy for the biomass of each organism. Any CT values that were 3 standard deviations from the mean for their respective qPCR plate or were measured as 0 were removed before averaging.

Every run had a standard curve made of tenfold serial dilutions (5 concentrations) of purified PCR product. G3P4 product initial concentration was 0.001 ng/µl, and DMAW4 was 0.01 ng/µl. Each run had two sets of serially diluted products in case one standard curve failed. These standard curves allowed for the calculation of qPCR efficiency and r2 of the standard curve. Any runs with standard curves with r2 values below 0.95 or primer efficiency below 85% were redone. The averaged CT values were normalized and adjusted for primer amplification efficiency according to the following equation:

graphic file with name d33e541.gif
graphic file with name d33e544.gif

To get the efficiency-adjusted CT ratio, the efficiency-adjusted CT value for Epichloë was divided by the efficiency-adjusted CT value for tall fescue.

Phenotyping was done for 1083 of the 3000 total individuals, excluding individuals where we could not identify parentage or that had died between the genotyping In 2019 and phenotyping In 2023. Phenotyping was performed one more time for the two half-sib families In 2024. After filtering, the two half-sib families totaled 133 individuals.

Residuals for the data were then taken by running the lm() function in R and removing known environmental factors (sampling date for alkaloids, and sampling date and assay plate for qPCR). Using residuals like this was meant to remove potential batch effects in the data, since the size of the population required sampling to be spread across multiple days (and thus slightly different environmental conditions, etc.). For the half-sib families, the phenotypes were averaged across the two years they were taken, and then the residuals were found as above.

Heritability calculation and GWAS

Heritability was estimated for the full population using the Flex Seq genotypes and the phenotypes taken In 2023. For the half-sib population, Heritability was calculated using the Freedom Markers genotypes, and a separate calculation was made for each set of phenotypes: 2023, 2024, and the average of both years. First, TASSEL [53] was used to convert the filtered VCF file into a numerical genotype file. The numerical genotype file was then used in rrBLUP [54] to create an additive kinship matrix. The kinship matrix and the phenotypes were then used to do genotypic value prediction based on kinship. The equation below was used to estimate narrow-sense heritability.

graphic file with name d33e566.gif
graphic file with name d33e569.gif

GWAS was conducted on the half-sib families with TASSEL. First, both principal components and a kinship matrix were created from the filtered SNPs (described in the sequencing and genotyping section). The genotypes, first 5 PCs, and residual phenotype files were joined and combined with the kinship matrix to run a mixed linear model (MLM). The MLM was also run with the separated 2023 and 2024 residual phenotypes. P-value cut-offs for GWAS were calculated using both the false discovery rate and the Bonferroni correction.

GWAS and heritability calculations were also run on the larger population in a similar manner using the 1083 individuals that had phenotypes and genotypes.

Results

Phenotypic relationship between alkaloids and fungal biomass

Seven of the 17 parental lines were selected for an initial test to determine how fungal biomass and alkaloid concentrations varied by genotype. Significant differences were detected among these lines for total ergot alkaloid concentration, but not for relative fungal biomass (Fig. 2A). These phenotypic traits were weakly but significantly correlated (r2 = 0.132; p = 0.02) among the parental lines.

Fig. 2.

Fig. 2

Open in a new tab

The relationship between alkaloid concentration and fungal biomass. Analyses are shown for a subset of parents (A), the entire population (B), and the two half-sib families for which additional data was collected (C). (Boxplots in B are separated based on maternal parent since the endophyte is only inherited maternally.) For each population group, the leftmost plot shows the amount of ergot alkaloids present as measured by ELISA. (Actual values in A, and residual values in B & C after fitting a statistical model to correct for batch effects.) The center column shows relative fungal biomass, as measured by delta CT of qPCR for either the plant or the fungus. (B and C use residual instead of raw values to correct for batch effects.) The correlations between alkaloids and fungal biomass are shown at right

In 2023, all 17 parents and 1083 progeny derived from an open cross among them (see Methods section) were phenotyped for total ergot alkaloid concentration and relative fungal biomass. To estimate fungal biomass, the difference in qPCR crossing threshold (“delta CT”) ratio of fungus/plant PCR product was used. This data was filtered for outliers, normalized, and corrected for potential batch effects (see Methods), then grouped by maternal parent. This approach was chosen because the maternal parentage is known, while paternal parentage was inferred. Additionally, grouping by maternal lineage ensures consistency in Epichloë presence across groups, as the endophyte is maternally inherited. We find significant differences in alkaloid concentration among the maternal groups but no differences in the relative fungal biomass (Fig. 2B). When grouped by the inferred paternal plant (Supplementary Figure S2), we again find significant differences in alkaloid concentrations along with minor differences in fungal biomass. The correlation between these phenotypes is numerically small (r2 = 0.03) but highly significant (p = 8.89 × 10− 9), likely due to the large sample size.

As a focused test, we selected out the two largest crosses (based on known maternal and genotype-inferred paternal parents) for deeper genotyping and a second round of phenotyping. These families share a common parent (314) and so are half-sib families. We did not detect any significant differences in phenotype for these families (Fig. 2), which again shows a statistically significant but numerically small relationship between fungal biomass and alkaloid concentration (r2 = 0.122, p = 4.39 × 10− 5) (Fig. 2C). This pattern holds when individual timepoints are analyzed separately (Fig. 3). When comparing across timepoints, correlations for both traits were relatively low (r2 = 0.063; p = 3.06 × 10− 3 for alkaloids and r2 = 0.133; p = 9.78 × 10− 6 for relative biomass).

Fig. 3.

Fig. 3

Open in a new tab

Correlation between the phenotypic traits of alkaloid concentration and fungal biomass in the half-sib population. The efficiency-adjusted CT ratio (X axis) is a proxy for relative fungal biomass (mycelial concentration), plotted against alkaloids (Y axis) to identify any relationship between them. Data are shown for 2023 (left) and 2024 (right) sample points. The Spearman correlation (R-squared) and significance (p-value) are displayed at top for each dataset

Heritability estimates of alkaloid concentration and fungal biomass

“Heritability” refers to the amount of variability in a trait that is explainable by genetics (specifically the genetics of the host plant in this study). Narrow-sense heritability for the overall population was calculated using genetic and trait data for 1083 individuals, finding an estimated h2 of 0.352 for ergot alkaloids and 0.672 for relative fungal biomass.

Narrow-sense heritability was similarly estimated for the values of the half-sib families. When calculated for both years, heritability was 0.603 for alkaloid concentration and 0.702 for relative fungal biomass. In 2023, heritability was lowest for both traits (0.245 for alkaloids and 0.171 for relative fungal biomass), while In 2024, heritability increased to 0.465 for alkaloids and 0.730 for relative fungal biomass. Genetic mapping of these data sets found no sites that passed a 5% false discovery rate threshold (data not shown).

Genome-wide association was conducted on these 1083 individuals, which resulted in a single significant SNP (FaChr7G1: 214311125) for relative biomass (Fig. 4). The 25 kb region up- and downstream of this SNP were BLASTed against the Lolium Perenne [55] and rice [56] genomes to identify potential candidate genes, but only uncharacterized proteins were found (Table S1).

Fig. 4.

Fig. 4

Open in a new tab

Manhattan plot of genome-wide association for relative fungal biomass. A mixed linear model was fit in TASSEL (Bradbury et al. 2007) for relative biomass (delta CT) across 1083 individuals. Only a single SNP (circled, at FaChr7G1: 214311125) passes the 5% False Discovery Rate cutoff (blue), though not the Bonferroni-corrected cutoff (red)

Discussion

In this study, we compared the relative fungal biomass and alkaloid concentration of over 1000 genetically unique tall fescue plants grown in a common greenhouse environment. Our principal aims were to quantify the relationship between these traits and the degree to which both are influenced by the genome of their host. Our results support the following conclusions:

Fungal biomass is weakly correlated to alkaloid concentration

Across all population groups, the relative biomass of Epichloë was weakly but significantly correlated with fungal alkaloid concentration. We found that in the half-sib families, the correlation varied ~ 3.5x in two consecutive years (2023 and 2024; 0.07 vs. 0.25), implying that environmental or other factors may change the relationship between the traits. Alkaloid levels tend to respond to environmental variability [57]; if fungal biomass is more constant, then it could explain the variable relationship between these traits. Our data indicates that fungal biomass is marginally more stable (r2 of 0.133 vs. 0.063), though the correlation is still weak. This low stability of fungal biomass may be due to the host continually producing new tillers, which would allow for different rates of fungal infection over time.

Other studies that have investigated the correlation between Epichloë biomass and alkaloid concentrations consistently have higher R2 values [58, 59], though they are also looking at related species (Lolium perenne and Epichloë festucae) and not tall fescue/Epichloë coenophiala specifically. Differences may also be due to methodology. Liquid chromatography/mass spectrometry on > 1000 individuals proved to be logistically unfeasible, hence why we chose ELISA quantification and qPCR [52], the latter of which is similar to how Epichloë festucae was measured within Lolium perenne [60].

Both biomass and alkaloid concentration show evidence of plant genetic influence

Although the two traits examined here (alkaloid concentration and fungal biomass) show only weak relationship to each other, they both show a moderate to significant relationship to the genetics of their host plant. Evidence of host genetic control has been shown before [36] and to our knowledge, our results are the largest and most thorough quantification of this effect to date. Both ergot alkaloid concentration and relative fungal biomass were found to be moderately heritable traits within tall fescue, with values ranging from 0.245 to 0.603 for alkaloid concentration and from 0.171 to 0.730 for relative fungal biomass. With one exception (half-sib families In 2023, with the lowest heritability overall), fungal biomass was consistently more heritable than alkaloid concentration.

Previous studies that have measured heritability have relied on phenotypic data alone and also found alkaloid concentration and fungal biomass to be relatively heritable traits (e.g., heritabilities of 0.49–0.72) [37, 61]. Our study found similar heritability estimates for both traits, supporting the idea that tall fescue exerts influence over fungal phenotypes. Combined with the low correlations between traits, this suggests that alkaloid levels and fungal biomass are influenced by plant genetics but likely governed by different genes or genetic networks.

Investigation into possible genes that governed these traits found only one weakly associated SNP, at FaChr7G1: 214,311,125. However, no protein encoding genes with known functions were identified within 25 kb of this SNP. Combined with the moderate heritability results of 0.35 (alkaloids) and 0.67 (fungal biomass), these results imply that the host genetic influence on these traits is highly quantitative and diffuse, meaning that many genes across the genome are each likely contributing a small effect. These results contrast with a QTL analysis on the same traits of alkaloid amounts and fungal biomass with perennial ryegrass and Epichloë festucae, which yielded 11 genomic regions affecting these traits [38]. Given the close relationship between tall fescue and perennial ryegrass, it is curious that we find such different results; one possibility is that since perennial ryegrass is a diploid species whereas tall fescue is hexaploid, the latter’s genetic effects are simply spread among its 3 genome copies and thus more difficult to uniquely identify.

Limitations and future directions

A limitation of our study is that we do not know how similar the genotypes of Epichloë are in our population past their ability to produce certain alkaloids. Although a previous study reported that ergovaline concentrations were more strongly determined by the host genotype than by Epichloë genotype [36], we note that variation in fungal genotype may also affect fungal phenotypes [62] thus we cannot completely rule out a contribution of endophyte genetic variation to the phenotypes we measured. In our experiment, plants were also grown in 3× 3 × 5 inch pots under greenhouse conditions, which may not fully represent field environments. Heritability estimates, including narrow-sense heritability, can vary depending on both the genetic composition of the population and the environment in which it is measured. It is possible that plants grown to maturity under field conditions would show different expression of Epichloë phenotypic traits [63]. Since alkaloid concentrations vary throughout the plant [64], sampling a different part of the plant e.g., leaf blade) could also conceivably give different results. Our sampling of the pseudostem where the fungus and alkaloids are most present was chosen specifically to get the most direct measurement of production.

Conclusion

Tall fescue covers 140,000 km2 of the southern US, and ergot alkaloid toxicity is estimated to cost US growers between $600 million to $1 billion per year [65]. Although newer cultivars now exist that eliminate livestock toxicity, adoption has been slow [28]. Most pastures still consist primarily of toxic cultivars because remediation is costly and non-toxic cultivars require more management, as cattle tend to overgraze them [1]; efforts to accelerate this transition through public-private partnerships and education are underway and showing significant promise [66]. Understanding the mechanisms that control both fungal presence and alkaloid concentration will provide insight into this intimate and highly beneficial symbiosis, but also potentially identify ways to more economically remediate these pastures and/or mitigate the amount of alkaloids produced and ingested by cattle.

In this study we demonstrate that while fungal biomass and alkaloid concentration in tall fescue are weakly correlated, both demonstrate moderate heritability based on host plant genetics. When conducting GWAS only one SNP was found to be significant, suggesting that the host genetic control may be spread amongst many genes that all contribute small effects. Such diffuse genetic architecture is still amenable to both traditional and modern breeding techniques (e.g., genomic selection). Ultimately, these results aim to improve the lives of both livestock and ranchers dependent on tall fescue and provide insight for superior management of pasture resources.

Supplementary Information

Supplementary Material 1. (307.2KB, docx)

Acknowledgements

The authors would like to thank members of the Wallace lab for assistance with sampling fescue plants, especially Naomi Rodman,, and Michael Trammell (Oklahoma State University) for the production of tall fescue seeds at the Noble Research Institute, Chloe Mootz for providing Fig.1, and Drs. Shaun Bushman and Matthew Robbins for early access to the tall fescue genome assembly.

Abbreviations

CT

Crossing Threshold

GWAS

Genome Wide Association Study

MLM

Mixed Linear Model

SNP

Single Nucleotide Polymorphism

Authors’ contributions

DRT: Methodology, Investigation, Data Curation, Formal Analysis, Visualization. Writing-Original Draft. CY: Resources, Writing – Review and Editing. CP: Plant care and sampling. JGW: Conceptualization, Writing – Review and Editing, Supervision, Project Administration, Funding acquisition.

Funding

This work was supported by NSF grant # 1764127.

Data availability

Raw RNA sequence data generated by this study is available at Sequence Read Archive (SRA) Bio Project accession number PRJNA648970. (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA648970)Scripts for the original assembly creation and parentage determination can be found here at git hub (https://github.com/wallacelab/paper-talamantes-fescue-parentage-2025)Scripts for phenotype and genotype analysis can be found here at git hub (https://github.com/wallacelab/paper-talamantes-fescue-mapping-2025).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Hmielowski T. The fascinating tale of tall fescue. CSA News. 2016;61(12):4–9. [Google Scholar]
  • 2.Breuillin-Sessoms F, Watkins E. Performance of multiple turfgrass species during prolonged heat stress and recovery in a controlled environment. Crop Sci. 2020;60(6):3344–61. [Google Scholar]
  • 3.Jiang Y, Huang B. Physiological Responses to Heat Stress Alone or in Combination with Drought: A Comparison between Tall Fescue and Perennial Ryegrass. 2001. Available from: https://journals.ashs.org/hortsci/view/journals/hortsci/36/4/article-p682.xml. Cited 23 Dec 2024.
  • 4.Morris K. 2006 National Tall Fescue Test. Beltsville, Maryland: USDA-ARS; 2010. (National Turfgrass Evaluation Program).
  • 5.Novello A, Pornaro C, Fidanza M, Macolino S. Adaptability and Character Traits of Turf-type Tall Fescue Cultivars Grown under Limited Irrigation in Northern Italy. 2025. Available from: https://journals.ashs.org/horttech/view/journals/horttech/35/1/article-p5.xml. Cited 23 Dec 2024.
  • 6.Richard AM, Gervais R, Tremblay GF, Bélanger G, Charbonneau É. Tall fescue as an alternative to timothy fed with or without alfalfa to dairy cows. J Dairy Sci. 2020;103(9):8062–73. [DOI] [PubMed] [Google Scholar]
  • 7.Fergus EN, Buckner RC. Registration of Kentucky 31 Tall Fescue (Reg. No. 7). Crop Science. 1972;12(5):cropsci1972.0011183X001200050061x.
  • 8.Christensen MJ, Bennett RJ, Ansari HA, Koga H, Johnson RD, Bryan GT, et al. Epichloë endophytes grow by intercalary hyphal extension in elongating grass leaves. Fungal Genet Biol. 2008;45(2):84–93. [DOI] [PubMed] [Google Scholar]
  • 9.Clay K, Schardl C. Evolutionary origins and ecological consequences of endophyte symbiosis with grasses. Am Nat. 2002;160(Suppl 4):S99–127. [DOI] [PubMed] [Google Scholar]
  • 10.Zhang W, Card SD, Mace WJ, Christensen MJ, McGill CR, Matthew C. Defining the pathways of symbiotic Epichloë colonization in grass embryos with confocal microscopy. Mycologia. 2017;109(1):153–61. [DOI] [PubMed] [Google Scholar]
  • 11.Bacon CW. Abiotic stress tolerances (moisture, nutrients) and photosynthesis in endophyte-infected tall fescue. Agric Ecosyst Environ. 1993;44(1):123–41. [Google Scholar]
  • 12.Islam MS, Krom N, Kwon T, Li G, Saha MC. Transcriptome of Endophyte-Positive and Endophyte-Free Tall Fescue Under Field Stresses. Front Plant Sci. 2022;13. Available from: https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2022.803400/full. Cited 9 Dec 2024. [DOI] [PMC free article] [PubMed]
  • 13.McCulley RL, Bush LP, Carlisle AE, Ji H, Nelson JA. Warming reduces tall fescue abundance but stimulates toxic alkaloid concentrations in transition zone pastures of the U.S. Front Chem. 2014;2:88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Xu L, Li X, Han L, Li D, Song G. Epichloe endophyte infection improved drought and heat tolerance of tall fescue through altered antioxidant enzyme activity. Eur J Hortic Sci. 2017;82:90–7. [Google Scholar]
  • 15.Bacetty AA, Snook ME, Glenn AE, Noe JP, Hill N, Culbreath A, et al. Toxicity of endophyte-infected tall fescue alkaloids and grass metabolites on Pratylenchus scribneri. Phytopathology. 2009;99(12):1336–45. [DOI] [PubMed] [Google Scholar]
  • 16.Clay K. Endophytes as antagonists of plant pests. In: Andrews JH, Hirano SS, editors. Microbial ecology of leaves. New York, NY: Springer; 1991. pp. 331–57. (Brock/Springer Series in Contemporary Bioscience). [Google Scholar]
  • 17.Liebe DM, White RR. Meta-analysis of endophyte-infected tall fescue effects on cattle growth rates. J Anim Sci. 2018;96(4):1350–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Schardl CL, Panaccione DG, Tudzynski P. Ergot alkaloids–biology and molecular biology. Alkaloids Chem Biol. 2006;63:45–86. [DOI] [PubMed] [Google Scholar]
  • 19.Young CA, Bryant MK, Christensen MJ, Tapper BA, Bryan GT, Scott B. Molecular cloning and genetic analysis of a symbiosis-expressed gene cluster for lolitrem biosynthesis from a mutualistic endophyte of perennial ryegrass. Mol Genet Genomics. 2005;274(1):13–29. [DOI] [PubMed] [Google Scholar]
  • 20.Bouton JH, Gates RN, Belesky DP, Owsley M. Yield and persistence of tall fescue in the southeastern coastal plain after removal of its endophyte. Agron J. 1993;85(1):52–5. [Google Scholar]
  • 21.Takach JE, Young CA. Alkaloid genotype diversity of tall fescue endophytes. Crop Sci. 2014;54(2):667–78. [Google Scholar]
  • 22.Nelli MR, Scheerer JR. Synthesis of peramine, an anti-insect defensive alkaloid produced by endophytic fungi of cool season grasses. J Nat Prod. 2016;79(4):1189–92. [DOI] [PubMed] [Google Scholar]
  • 23.Schardl CL, Grossman RB, Nagabhyru P, Faulkner JR, Mallik UP. Loline alkaloids: currencies of mutualism. Phytochemistry. 2007;68(7):980–96. [DOI] [PubMed] [Google Scholar]
  • 24.Niu J, Qi J, Wang P, Liu C, Gao J. ming. The chemical structures and biological activities of indole diterpenoids. Natural Products and Bioprospecting. 2023;13(1):3. [DOI] [PMC free article] [PubMed]
  • 25.Gunter SA, Beck PA. Novel endophyte-infected tall fescue for growing beef cattle. J Anim Sci. 2004;82(suppl13):E75–82. [DOI] [PubMed] [Google Scholar]
  • 26.Fleetwood DJ, Scott B, Lane GA, Tanaka A, Johnson RD. A complex ergovaline gene cluster in Epichloë endophytes of grasses. Appl Environ Microbiol. 2007;73(8):2571–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Florea S, Panaccione DG, Schardl CL. Ergot alkaloids of the family Clavicipitaceae. Phytopathology®. 2017;107(5):504–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Bouton J, Latch G, Hill N, Hoveland C, McCann M, Watson R et al. Reinfection of Tall Fescue Cultivars with Non-Ergot Alkaloid–Producing Endophytes. Agronomy Journal - AGRON J. 2002;94:567–74.
  • 29.Hopkins AA, Young CA, Panaccione DG, Simpson WR, Mittal S, Bouton JH. Agronomic performance and lamb health among several tall fescue novel endophyte combinations in the South-Central USA. Crop Sci. 2010;50(4):1552–61. [Google Scholar]
  • 30.Ball DM, Lacefield GD, Hoveland CS. The Wonder Grass: The Story of Tall Fescue in the United States. 2019. Available from: https://aurora.auburn.edu/handle/11200/49449. Cited 25 Oct 2022.
  • 31.Bony S, Pichon N, Ravel C, Durix A, Balfourier F, Guillaumin JJ. The relationship between mycotoxin synthesis and isolate morphology in fungal endophytes of lolium perenne. New Phytol. 2001;152(1):125–37. [DOI] [PubMed] [Google Scholar]
  • 32.Young CA, Schardl CL, Panaccione DG, Florea S, Takach JE, Charlton ND, et al. Genetics, genomics and evolution of ergot alkaloid diversity. Toxins (Basel). 2015;7(4):1273–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Hemken RW, Boling JA, Bull LS, Hatton RH, Buckner RC, Bush LP. Interaction of environmental temperature and anti-quality factors on the severity of summer fescue toxicosis. J Anim Sci. 1981;52(4):710–4. [DOI] [PubMed] [Google Scholar]
  • 34.Lea KLM, Smith SR. Using on-farm monitoring of ergovaline and tall fescue composition for horse pasture management. Toxins. 2021;13(10):683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Repussard C, Zbib N, Tardieu D, Guerre P, ACS Publications. American Chemical Society. ; 2014. Endophyte Infection of Tall Fescue and the Impact of Climatic Factors on Ergovaline Concentrations in Field Crops Cultivated in Southern France. Available from: https://pubs.acs.org/doi/full/10.1021/jf503015m. Cited 2 Jan 2025. [DOI] [PubMed]
  • 36.Agee C, Hill N. Ergovaline variability in Acremonium-Infected tall fescue due to environment and plant genotype. Crop Science - CROP SCI; 1994;34:221–6.
  • 37.Easton HS, Latch GCM, Tapper BA, Ball OJP. Ryegrass host genetic control of concentrations of endophyte-derived alkaloids. Crop Sci. 2002;42(1):51–7. [DOI] [PubMed] [Google Scholar]
  • 38.Faville MJ, Briggs L, Cao M, Koulman A, Jahufer MZZ, Koolaard J, et al. A QTL analysis of host plant effects on fungal endophyte biomass and alkaloid expression in perennial ryegrass. Mol Breed. 2015;35(8):161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.West CP. In. Plant influences on endophyte expression. 2007;117–21.
  • 40.Hill NS, Agee CS. Detection of Ergoline alkaloids in Endophyte-Infected tall fescue by immunoassay. Crop Sci. 1994;34(2):cropsci19940011183X003400020041x.
  • 41.Hiatt EE III, S Hill N, Bouton JH, Stuedemann JA. Tall fescue endophyte detection: commercial Immunoblot test kit compared with microscopic analysis. Crop Sci. 1999;39(3):cropsci19990011183X003900030030x. [Google Scholar]
  • 42.Siegel MR. A fungal endophyte in tall fescue: incidence and dissemination. Phytopathology. 1984;74(8):932. [Google Scholar]
  • 43.Krueger F, James F, Ewels P, Afyounian E, Schuster-Boeckler B. FelixKrueger/TrimGalore: v0.6.7 - DOI via Zenodo. Zenodo; 2021. Available from: https://zenodo.org/records/5127899. Cited 17 Feb 2025.
  • 44.Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, et al. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol. 2012;19(5):455–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinf 2009;25(14):1754–60. [DOI] [PMC free article] [PubMed]
  • 46.Li H. A statistical framework for SNP calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data. Bioinformatics. 2011;27(21):2987–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Danecek P, Auton A, Abecasis G, Albers CA, Banks E, DePristo MA, et al. The variant call format and vcftools. Bioinformatics. 2011;27(15):2156–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Kokot M, Długosz M, Deorowicz S. KMC 3: counting and manipulating k-mer statistics. Bioinf 2017;33(17):2759–61. [DOI] [PubMed]
  • 49.Ott A, Liu S, Schnable JC, Yeh CT, ‘Eddy,’ Wang KS, Schnable PS. tGBS® genotyping-by-sequencing enables reliable genotyping of heterozygous loci. Nucleic Acids Res. 2017;45(21):e178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Wu TD, Nacu S. Fast and SNP-tolerant detection of complex variants and splicing in short reads. Bioinformatics. 2010;26(7):873–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Schnitzius JM, Hill NS, Thompson CS, Craig AM. Semiquantitative determination of ergot alkaloids in seed, straw, and digesta samples using a competitive enzyme-linked immunosorbent assay. J Vet Diagn Invest. 2001;13(3):230–7. [DOI] [PubMed] [Google Scholar]
  • 52.Talamantes D, Kirkpatrick C, Wallace J. Developing robust quantitative PCR primers for comparative biomass analysis of Tall Fescue (Festuca arundinacea) and its Epichloë endophyte. microPublication Biology. 2024. Available from: https://www.micropublication.org/journals/biology/micropub-biology-001275. Cited 23 Jan 025. [DOI] [PMC free article] [PubMed]
  • 53.Bradbury PJ, Zhang Z, Kroon DE, Casstevens TM, Ramdoss Y, Buckler ES. TASSEL: software for association mapping of complex traits in diverse samples. Bioinformatics. 2007;23(19):2633–5. [DOI] [PubMed] [Google Scholar]
  • 54.Endelman JB. Ridge Regression and Other Kernels for Genomic Selection with R Package rrBLUP. The Plant Genome. 2011;4(3). Available from:https://onlinelibrary.wiley.com/doi/abs/10.3835/plantgenome2011.08.0024. Cited 2 May 2025.
  • 55.Frei D, Veekman E, Grogg D, Stoffel-Studer I, Morishima A, Shimizu-Inatsugi R, et al. Ultralong Oxford nanopore reads enable the development of a Reference-Grade perennial ryegrass genome assembly. Genome Biol Evol. 2021;13(8):evab159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Shang L, He W, Wang T, Yang Y, Xu Q, Zhao X, et al. A complete assembly of the rice Nipponbare reference genome. Mol Plant. 2023;16(8):1232–6. [DOI] [PubMed] [Google Scholar]
  • 57.Dinkins RD, Coe BL, Phillips TD, Ji H. Accumulation of alkaloids in different tall fescue KY31 clones harboring the common toxic Epichloë coenophiala endophyte under field conditions. Agronomy. 2023;13(2):356. [Google Scholar]
  • 58.Cagnano G, Lenk I, Roulund N, Jensen CS, Cox MP, Asp T. Mycelial biomass and concentration of Loline alkaloids driven by complex population structure in Epichloë uncinata and meadow fescue (Schedonorus pratensis). Mycologia. 2020;112(3):474–90. [DOI] [PubMed] [Google Scholar]
  • 59.Fuchs B, Krischke M, Mueller MJ, Krauss J. Plant age and seasonal timing determine endophyte growth and alkaloid biosynthesis. Fungal Ecol. 2017;29:52–8. [Google Scholar]
  • 60.Fuchs B, Krischke M, Mueller MJ, Krauss J. Herbivore-specific induction of defence metabolites in a grass–endophyte association. Funct Ecol. 2017;31(2):318–24. [Google Scholar]
  • 61.Adcock R, Hill N, Bouton J, Boerma H, Ware G. Symbiont regulation and reducing ergot alkaloid concentration by breeding Endophyte-Infected tall fescue. J Chem Ecol - J CHEM ECOL. 1997;23:691–704. [Google Scholar]
  • 62.Roylance JT, Hill NS, Agee CS. Ergovaline and peramine production in endophyte-infected tall fescue: independent regulation and effects of plant and endophyte genotype. J Chem Ecol. 1994;20(9):2171–83. [DOI] [PubMed] [Google Scholar]
  • 63.Visscher PM, Hill WG, Wray NR. Heritability in the genomics era — concepts and misconceptions. Nat Rev Genet. 2008;9(4):255–66. [DOI] [PubMed] [Google Scholar]
  • 64.Rottinghaus GE, Garner GB, Cornell CN, Ellis JL. HPLC method for quantitating ergovaline in endophyte-infested tall fescue: seasonal variation of ergovaline levels in stems with leaf sheaths, leaf blades, and seed heads. J Agric Food Chem. 1991;39(1):112–5. [Google Scholar]
  • 65.Hancock DW, Andrae J. Novel Endophyte-Infected Tall Fescue. University of Georgia Extension; 2017. Available from: https://secure.caes.uga.edu/extension/publications/files/pdf/C%20861_4.PDF
  • 66.Roberts CA, Andrae JG, Smith SR, Poore MH, Young CA, Hancock DW et al. The Alliance for Grassland Renewal: A Model for Teaching Endophyte Technology. World Academy of Science, Engineering and Technology International Journal of Animal and Veterinary Sciences [Internet]. 2020;14. Available from: https://publications.waset.org/10011192/the-alliance-for-grassland-renewal-a-model-for-teaching-endophyte-technology. Cited 30 May 2025.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1. (307.2KB, docx)

Data Availability Statement

Raw RNA sequence data generated by this study is available at Sequence Read Archive (SRA) Bio Project accession number PRJNA648970. (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA648970)Scripts for the original assembly creation and parentage determination can be found here at git hub (https://github.com/wallacelab/paper-talamantes-fescue-parentage-2025)Scripts for phenotype and genotype analysis can be found here at git hub (https://github.com/wallacelab/paper-talamantes-fescue-mapping-2025).

Articles from BMC Genomics are provided here courtesy of BMC

RESOURCES