Cool‐season grasses infected with Neotyphodium spp. endophytes have an extraordinary impact on the ecology and economy of pasture and turf. A range of adaptations of endophyte‐infected grasses to biotic and abiotic stresses has been identified but mechanisms of these adaptations are not clearly understood. In this review, we present recent research progress on endophyte‐related mechanisms affecting abiotic (drought, mineral) and selected aspects of biotic stress tolerance in cool‐season grasses. Endophytes induce mechanisms of drought avoidance (morphological adaptations), drought tolerance (physiological and biochemical adaptations), and drought recovery in infected grasses. Mineral nutrition (nitrogen, phosphorus, calcium) affects production of ergot alkaloids, thus understanding mechanisms involved in mineral economy of endophyte‐infected grasses will help in developing management practices to reduce forage toxicity to livestock. Previous research resolved the role of endophyte in nitrogen (N) economy of tall fescue. We identified two endophyte‐related mechanisms in tall fescue operating in response to phosphorus (P) deficiency. The mechanisms are altered root morphology (reduced root diameters and longer root hairs) and chemical modification of the rhizosphere resulting from exudation of phenolic‐like compounds. These mechanisms were shown to benefit endophyte‐infected plants grown under P deficiency. We also report a mechanism of aluminum (Al) sequestration on root surfaces in endophyte‐infected tall fescue, which appears to be related to exudation of phenolic‐like compounds with Al‐chelating activity. Understanding mechanisms of abiotic stress tolerance in endophyte‐infected grasses is essential for continued improvement and persistence of grasses for a range of applications, e.g., forage for semi‐arid areas or cover plants for soil renovation.

The coastal plain of the southeastern USA lacks a dependable per· ennial cool season forage crop, but tall fescue (Festuca arundinacea Schreb.) gennplasms have been developed which show greater per· sistence and yield than currently marketed cultivars in this region. Since these gennplasms were found to be infected with the tall fescue endophyte (Acremonium coenophialum Morgan-Jones and Gams), the contribution of the endophyte to their perfonnance was unknown. The objective of this study was to compare yield and stand survival of endophyte-infected (EI) and endophyte-free (EF) (achieved by re· mo~·aJ of its endophyte) versions of four persistent gennplasms (GA· 5, GA-Jesup, GA·Jesup Improved, and GA-Ecotype KY31) in clipped small plots at three locations in Georgia (Americus and Tifton in the coastal plain and Watkinsville located in the fescue groWing area of the southern piedmont region) for a 3-yr period. The El version of each gennplasm showed greater stand survival and yield than its EF version at Americus and Tifton, but no differences were recorded for infection status for the same parameters at Watkinsville. These find· ings indicate that endophyte removal greatly reduces the ecological fitness of tall fescue by possibly allowing less tolerance to summer drought. Presently, only El tall fescue can be dependably recom· mended for perennial pasture in the southeastern coastal plain.

Endophyte (Acremonium coenophialum Morgan-Jones et Gams) -infected tall fescue (Festuca arundinacea Scbreb.) produces ergopeptine alkaloids (EP) that may be associated with fescue toxicosis symptoms in cattle (Bos taurus). A study was conducted during 1983 and 1984 at Watkinsville, GA, to determine the effects of endophyte infection, season, and N fertilization on EP concentration and steer average daily gain (ADG) in grazed 0.7-ha tall fescue paddocks. Treatments included low (LF) and high (HF) endophyte infection frequency 'Kentucky-31' (KY-31) tall fescue fertilized with 134 or 336 kg N ha-• (2x2 factorial) replicated three times. The cultivars AU Triumph and Johnstone were also evaluated. Animals were managed using put-and-take grazing system. Ergopeptine alkaloid concentration (ergovaline was 80% of total) was greater in HF KY-31 than LF KY-31 in both years. Peak ergovaline concentrations occurred in spring and again in autumn (1984) when concentrations in HF KY-31 approached 1 mk kg-•. Mid-summer concentrations were about one-third the autumn maximum for HF and LF KY-31. Nitrogen fertilization bad minimal effect on ergovaline concentration in the spring-summer of 1983 but increased concentrations in LF KY-31 in 1984. Ergopeptine alkaloids were not detected in endophyte-free AU Triumph, and were detected in low amounts ( < 0.1 mg kg-1 ) in a pre-varietal release of Johnstone containing 7 to 10% endophyte. Increased ergovaline was associated with decreased ADG in the spring-summer of 1983 and 1984; however, there was no significant association between ergovaline and ADG when the entire (April-December) 1984 growing season was considered, suggesting that ergopeptine alkaloids may operate in conjunction with other environmental and nutritional factors to elicit fescue toxicosis symptoms.

Neotyphodium spp. fungal endophytes form symbiotic associations with agronomic grasses with expression ranging from mutualistic to parasitic. In general, endophyte infection frequencies seem more variable in natural compared to pastoral situations, suggesting that expression and benefits attributable to endophyte infection depend on the resource environment in which the host–endophyte complex was formed and the conditions under which it is being grown. Emerging evidence suggests that expression is not always mutualistic or beneficial to the host in terms of productivity and persistence in resource‐limited environments. Expression of host–endophyte associations will vary depending on resource availability and environmental conditions that influence host physiology and growth, and herbivore behavior. New research suggest alkaloids and secondary metabolites produced by host–endophyte associations can have physiological functions in addition to acting as herbivore deterrents. Novel host–endophyte associations are being created that maintain insect and disease resistance, while having minimal detrimental impact on mammalian herbivores. The influence of host–endophyte associations on sward composition and soil food webs is only now coming to light. While novel endophyte associations and their attendant mechanisms can be considered beneficial in some respect, the associations may not be as adaptable to stressful or marginal resource environments and could have long‐term ecological impacts measured in terms of persistence and total productivity of the sward. We review the extensive published work on host–endophyte interactions and illustrate the complexity of host–endophyte associations and their interactions with environment, and the range of responses that occur.

Neotyphodium coenophialum, (Morgan-Jones & Gams) Glenn, Bacon & Hanlin, infected tall fescue (Festuca arundinacea Schreb.) plants perform better than non-infected isolines on phosphorus (P)-deficient soils. Our objective was to characterize growth and P uptake dynamics of tall fescue in response to endophyte infection and P source at low P availability in soil. Two tall fescue genotypes (DN2 and DN4) infected with their naturally occurring N. coenophialum strains (E+), and in noninfected (E-) forms were grown in Lily soil (fine loamy siliceous, mesic Typic Hapludult) in a greenhouse for 20 weeks. Three soil P treatments were imposed: no P supplied (control) and P supplied as commercial fertilizer (PF) or as phosphate rock (PR) at the level of 25 mg P kg -1 soil. Interaction of tall fescue genotype and endophyte status had a significant influence on mineral element uptake suggesting high 1 Corresponding author. 835 836 MALINOWSKI AND BELESKY specificity of endophyte-tall fescue associations. Endophyte infection did not affect root dry matter (DM) when no P was supplied but shoot DM was reduced by 20%. More biomass was produced and greater P uptake rate occurred in PR than PF treatment. Root DM was greater in E+ DN4 than E-DN4 when supplied with either PF or PR. In contrast, endophyte infection did not affect root DM of DN2, regardless of P source. Relative growth rate (RGR) of E+ plants grown with PR was 16% greater than that of E-plants. Endophyte infection did not improve growth or P uptake in PF treatment. When PR was supplied, P uptake rate was 24% greater in E+ DN2 than E-DN2, but endophyte infection did not benefit DN4. Phosphorus-use efficiency was 6% less in E+ DN2 but 16% greater in E+ DN4 compared to E-plants, regardless of P source. Root exudates of E+ DN2, but not E+ DN4 solubilized more P from PR than those of E-plants. The correlation between root RGR and P uptake rate was relatively high for E-plants (r=0.76), but low for E+ plants (r=0.27) grown with PR. Results suggest that P uptake by E+ tall fescue might rely on mechanisms other than an increase in root biomass (surface area). Endophyte infection modified tall fescue responses to P source. This phenomenon was associated with modes of P acquisition which included enhanced activity of root exudates in releasing P from PR in E+ plants (DN2), and increased root biomass (DN4). The dominant means of P acquisition may be determined by a specific association of endophyte and tall fescue genomes. Endophyte-tall fescue association plasticity contributes to widespread success of symbiotic in marginal resource conditions.

Tall fescue (Festuca arundinacea Schreb.) is a common pasture grass adapted to a wide range of environmental conditions. An endophyte, Acremoninm coenophialum Morgan‐Jones & Gams, growing in association with tall fescue is known to produce ergopeptine alkaloids that are thought to be detrimental to animal health. Endophyteinfection changes how tall fescue plants grow and may be involved in the adaptability of tall fescue to environmental extremes. The objective of this study was to determine whether plant‐endophyte associations that had previously tested as high and low ergopeptine alkaloid producers varied in competitive ability when their endophytes were removed. Two identical field studies were conducted on a Cecil sandy clay loam and a Pacolet sandy clay loam soil (both clayey, kaolinitic, thermic Typic Hapluduits) in 1987 and 1988, respectively. In each study, a modified Nelder's design was used to test the competitiveness between infected and noninfected forms of two tall fescue genotypes. Endophyte‐infected plants were larger and, generally, more competitive in mixtures with noninfected plants. When infected with endophyte, the high ergopeptine alkaloid‐ producing genotype of tall fescue had similar or reduced competitiveness than the low‐alkaloid genotype when compared with their noninfected forms. We conclude that if ergopeptine alkaloids can be genetically removed from tall fescue‐endophyte associations, the competitiveness of the plant would be unaffected in nongrazed conditions. Reduction or elimination of ergopeptine alkaloids would help improve animal health when consuming endophyte‐infected tall fescue forage.

Pure stands of 'Grasslands Puna' chicory (Cichorium intybus L.) are productive and responsive to N fertilization in the eastern USA.We conducted a field experiment for 3 yr to investigate productivity and nutritive value of swards including chicory, orchard grass (Dactylis g/omerata L.), and birdsfoot trefoil (Lotus comicu/atus L.) as a func· tion of clipping frequency (3-and 6-wk intervals). Each treatment was replicated three times on an upland site of Dekalb series soil (loamy-skeletal, mixed, subactive, mesic Typic Dystrochrept). Modest rates of N, P, and K were applied annually. Herbage mass, botanical composition, in vitro organic matter disappearance (IVOMD), and crude protein (CP) were determined. Cumulative herbage yield was not influenced by initial sward composition with chicory, chicoryorchardgrass, and chicory-orchardgrass-trefoil averaging 6.8 Mg ha-• during a growing season. Clipping frequency influenced yield with canopies clipped at 6-wk intervals producing 26% more herbage than those clipped at 3-wk intervals. Growth rates of canopies clipped at 6-wk intervals averaged 51 kg ha-• d-1 early in the growing season, suggesting that swards including chicory would be responsive to nutrient inputs and where rapid herbage regrowth and nutrient use by the growing crop are needed. Chicory declined in swards with time, regardless of initial sward composition or clipping frequency. Fluctuations in IVOMD and CP were related to changes in sward composition arising from the interaction of time and clipping frequency. Including orchardgrass and trefoil retarded invasion of less desirable species and so provided a means to control overall sward productivity and herbage composition.

Infection of tall fescue (Festuca arundinacea Schreb.) with its endemic Neotyphodium coenophialum-endophyte (Morgan-Jones and Gams) Glenn, Bacon and Hanlin appears to reduce copper (Cu) concentrations in forage and serum of grazing animals, contributing to a range of immune-related disorders. A greenhouse experiment was conducted to identify effects of novel endophyte strains on Cu acquisition by tall fescue (Festuca arundinacea Schreb.) varieties Grasslands Flecha and Jesup infected with a novel, non ergot producing endophyte strain AR542, and two perennial ryegrass (Lolium perenne L.) varieties Aries and Quartet infected with a novel, non lolitrem B producing strain AR1, and their noninfected (E−) forms. Individual endophyte/grass associations were cultivated in nutrient solutions at 1.0 (P+) and 0.0 mM (P−) phosphorus concentrations. The Cu 2+ -binding activity of extracellular root exudates, and concentrations of Cu and other heavy metals in roots and shoots were measured. Extracellular root exudates of AR542-infected vs. E− tall fescue had higher Cu 2+ -binding activity only in P− nutrient solution as shown by lower concentration of free Cu 2+ (0.096 vs. 0.188 mmol Cu 2+ g −1 root DM, respectively). The Cu 2+ -binding activity by root exudates of perennial ryegrass was not affected by endophyte infection, but was higher (i.e., lower concentration of free Cu 2+ ) in P− vs. P+ nutrient solution (0.068 vs. 0.114 mmol Cu 2+ g −1 root DM). In this hydroponic experiment, Cu concentrations in shoots of both grasses were not a function of Cu 2+ -binding activity and endophyte effects on heavy metal concentrations in shoots and roots were specific for each variety. The Cu 2+ -binding activity of extracellular root exudates may affect Cu accumulation by field-grown, endophyte-infected tall fescue under P-limiting growth conditions and warrants verification by more specific methods.