Our long-term goal is to improve persistence and yield of alfalfa (Medicago sativa L.) and other forage legumes by identifying and manipulating genes that affect these traits. Future improvements by genetic manipulation depend, however, on new insights into basic physiological and biochemical plant processes. Currently we lack knowledge of discrete traits controlling agronomic performance that can serve as targets for manipulation using modern genetic techniques. Our work, and recent work of others, has failed to show a positive association between root total nonstructural carbohydrate (TNC) levels and genetic variation in regrowth and winterhardiness of forage legumes. We are exploring alternatives to the conventional thinking that root TNC reserves control alfalfa regrowth and persistence. Recent results indicate that root N declines during herbage regrowth after defoliation, and again in spring when shoot growth resumes. Labeling studies have proven that much of the N found in shoots during early regrowth is derived from root N pools. In alfalfa, certain root N pools, especially root vegetative storage proteins (VSPs) are preferentially used as N reserves during the early stages of shoot regrowth. The VSPs represent 25% of the root protein pool. They are unique to alfalfa roots, and their synthesis is developmentally regulated. Work is underway to isolate and characterize the cDNAs for the VSPs to learn more about regulation of VSP synthesis and degradation in alfalfa roots.

Recommendations regarding P and K fertilization and critical soil and tissue concentrations for alfalfa (Medicago sativa L.) are inconsistent. Our objective was to use cluster analysis and binary logistic regression to determine if soil and herbage P and K concentrations can be used as diagnostics of P‐ and K‐limited alfalfa yields across soils with variable P and K soil test values. Samples were obtained from 28 consecutive harvests (four per year) of plots fertilized with four rates of P and five rates of K. Moderate to high rates of both nutrients were required to maintain high yields. Cluster analysis using May 2004 data for mass shoot−1 and shoots m−2 identified six clusters that also differed in yields. Averaged over the seven May harvests the two lowest yielding clusters had low herbage K (10.3–15.9 g kg−1), while intermediate yields were associated with low herbage P (1.6 g kg−1). Herbage P and K of the highest yielding cluster were 2.5 and 21.5 g kg−1, respectively. Plant populations in 2004 were the greatest with intermediate yields (61 m−2) and lowest for the poorest yielding plots (6 m−2). High taproot N reserves were associated with high yields. Prediction of acceptable (high‐yielding clusters) vs. unacceptable (low yielding clusters) performance was best achieved using the combination of herbage K and P concentrations and their ratio as independent variables. Results illustrate a varying, interdependent nature for alfalfa critical P and K tissue concentrations and suggest the need to account for both nutrients simultaneously.

In the summer dry environment of cool temperate Tasmania, summer irrigation is used to maximise forage production. For lucerne (Medicago sativa L.) this irrigation is likely to interact with winter-dormancy genotypes to influence seasonal changes in taproot reserves and thus, the process of cold acclimation. To test this hypothesis four lucerne cultivars with contrasting levels of winter dormancy (DuPuits: winter-dormant; Grasslands Kaituna: semi winter-dormant; SARDI 7: winter-active: SARDI 10, highly winter-active) were grown in small plots at Elliott, Tasmania, under irrigated or dryland conditions. At each defoliation taproots were sampled and assayed for the concentration of soluble sugars, starch, amino acids, soluble protein, the abundance of vegetative storage proteins (VSP), and the abundance of mRNA transcripts associated with cold acclimation and VSP. Taproot-soluble protein concentrations in DuPuits significantly increased from summer to autumn when plants were grown under dryland conditions. When grown under irrigated conditions, taproot-soluble protein concentrations decreased over summer and increased in autumn for all cultivars. The abundance of VSP increased in summer in all cultivars grown under dryland conditions. Taproot-soluble sugar concentrations increased and starch decreased in autumn for all cultivars grown under both water regimes. Plants grown under dryland conditions showed little change in RNA transcript abundance of cold acclimation genes across all cultivars and sampling dates, while in those plants grown under irrigated conditions, transcript abundance was influenced by sampling date, and for some genes, by cultivar. There was a clear carry-over effect from the exposure of summer drought on the winter-dormancy response. The expression of winter dormancy at an agronomic and molecular level was greater under dryland conditions.

Phosphorus (P) and potassium (K) increase alfalfa (Medicago sativa L.) yield, but little information is available on how forage nutritive value is affected by P and K fertilization. The objective of this study was to investigate the effect of long-term P and K fertilization on alfalfa yield-forage nutritive value relationships. A factorial experiment with four P and five K treatments was replicated four times. Beginning in 1998 and continuing through 2004, herbage samples were collected in May, June, July, and September. Samples were analyzed for crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), in vitro true dry matter disappearance (IVTDMD), and acid detergent lignin (ADL) using near-infrared reflectance spectroscopy. Soils were also analyzed for P and K concentrations. Concentrations of NDF, ADF, and ADL were reduced by low P and K fertility. Addition of P fertilizer increased CP concentrations, while addition of K fertilizer reduced CP concentrations. Concentrations of NDF, ADF, and ADL and dry matter yield decreased with stand age and harvest within year. The low forage dry matter yield of unfertilized alfalfa was associated with higher IVTDMD. Yield of digestible nutrients per hectare (yield x IVTDMD) was highly correlated with yield but not IVTDMD concentration. Fertilizing for high yield, despite the slight reduction in forage nutritive value, remains the most viable strategy for maximizing digestible nutrient production per hectare in alfalfa.

Fertilization with K and P impacts alfalfa (Medicago sativa L.) yield, but how these nutrients influence taproot reserves and gene expression is unknown. Our objectives were to determine how P and K impact (i) alfalfa yield and yield components, (ii) accumulation and use of taproot carbon (C) and nitrogen (N) pools, and (iii) transcript levels for β‐amylase, sucrose synthase, and the high molecular weight vegetative storage protein in alfalfa taproots. Yield and yield components were determined at 30‐d intervals beginning in late May. Roots were sampled after the late June harvest (Day 0) and 1, 3, 6, 10, 14, 21, and 30 d thereafter. Addition of P and K increased forage yield by enhancing mass per shoot. High P resulted in rapid starch use, while taproots of plants fertilized with K had low sugar concentrations. Transcripts decline by Day 6 and by Day 10 were below detection limits. Transcripts for β‐amylase and sucrose synthase accumulated rapidly after Day 10 in taproots of plants fertilized with both P and K. Balanced P and K nutrition resulted in the accumulation and effective utilization of C and N reserves and in improved alfalfa adaptation to defoliation.

Addition of P and K fertilizer can increase alfalfa (Medicago sativa L.) yield and stand persistence, but the yield components associated with P‐ and K‐induced variation in agronomic performance are not clear. Our objectives were: (i) to determine the impact of P and K nutrition on productivity of a relatively old alfalfa stand; and (ii) determine which yield components are associated with changes in alfalfa forage yield. Treatments were a factorial combination of four P and five K rates replicated four times. Forage harvests occurred four times annually. Plant populations were determined in early December and late May each year. When compared to unfertilized plots, addition of P and K increased forage yield each year. Fertilization with P decreased plants m−2 at all K application rates, but especially in plots fertilized with P, but not K. By comparison, plots fertilized with K, but not fertilized with P, had the higher plant population densities. Although regression analysis eventually revealed a positive association between forage yield and shoots m−2 in 2003 and 2004, the greatest forage yields were not obtained in plots with the greatest plant population densities, shoots plant−1 or shoots m−2 Regression and path analysis revealed that improved forage yield in P‐ and K‐fertilized plots was consistently associated with greater mass shoot−1

Phosphorus deficiency reduces forage yield and stand persistence of alfalfa (Medicago sativa L.). Our objectives were to isolate and characterize a high-affinity phosphate-transporter (P-transporter) from alfalfa roots (Medicago sativa L.); determine how phosphorus (P) nutrition impacts P-uptake, growth, and carbohydrate and protein metabolism of alfalfa cells; and learn how expression of the P-transporter is influenced by P nutrition. An 1087-base pair (bp) sequence was isolated using RT-PCR that possessed high nucleotide and amino acid sequence similarity to high-affinity P-transporters. Cultured cells were sampled at 3-day intervals for 9 days while growing in media containing P concentrations ranging from 0 to 10 mM. Media P concentrations declined rapidly in all P treatments by day 6. Low media P concentrations (0, 0.1 and 0.5 mM) reduced cell growth rates compared to higher media P levels (2.5, 5 and 10 mM). Suspension cell cultures supplied 0.5, 2.5, 5, and 10 mM P had lower starch concentrations by day 3 compared to cells cultured in media containing 0 and 0.1 mM P. Steady-state transcript levels for the high-affinity P- transporter were high in P-deprived cells, but declined within 1 day when cells were provided 10 mM P.

Phosphorus and K fertilization increases alfalfa (Medicago sativa L.) yield and stand persistence, but the changes in yield components as affected by P and K fertility level are not known. Our hypothesis is that P and (or) K fertilization will increase one or more alfalfa yield components, and those component responses may change with stand age. The objectives of this field study were to determine the impact of P and K fertilization on alfalfa forage yield and yield components during the initial 3 yr after establishment. Treatments were the factorial combinations of four P rates (0, 25, 50, and 75 kg P ha(-1)) and five K rates (0, 100, 200, 300, and 400 kg K ha(-1)) arranged in a randomized complete block design with four replications. Forage harvests occurred four times annually, and yield, mass shoot(-1), and shoots area(-1) were determined. Plant populations were determined in early December and late May each year. Incremental additions of P and K increased alfalfa yield in each year. Potassium fertilization did not influence plant population, while robust P-responsive alfalfa plants apparently crowded out smaller, less vigorous plants thus decreasing plants m(-2). Stand assessments based on shoot counts, or aboveground plant counts may not accurately indicate alfalfa yield potential. Shoots plant(-1) was not affected by application of either nutrient, while shoots m(-2) generally declined with increased P and K fertilization. Improved forage yield of P- and K-fertilized plots was consistently associated with greater mass shoot(-1). Because fertilizer-responsiveness is closely associated with greater mass shoot(-1), cultivars possessing this trait may be relatively more productive under well-fertilized conditions.

This study presents the effects of methyl jasmonate (MeJA) on growth, N uptake, N partitioning, and N storage in taproots of non-nodulated alfalfa (cv. Lodi). When compared to untreated plants, addition of 100 micro M MeJA to the nutrient solution for 14 days reduced total growth and modified biomass partitioning between shoots and roots in favour of taproots and lateral roots. MeJA decreased N uptake (after 7 days) and increased N partitioning towards roots after 14 days. This preferential N partitioning to roots was accompanied by increased N storage in taproots as soluble proteins. Compared to total soluble proteins, VSP accumulation occurred earlier (7 days), and was greater (2-fold increase) in plants treated with 100 micro M MeJA. Steady-state transcript levels for two VSPs (32 and 57 kDa) also increased markedly (about 4-fold) in roots of plants treated with 100 micro M MeJA. This suggests that MeJA could act directly (transcriptional regulation) or indirectly (via the changes of N partitioning among alfalfa organs) on N storage as soluble proteins and in particular, VSPs. Because the deduced amino acid sequence of the 32 kDa VSP clone reveals high homology with Class III chitinases, we propose that the 32 kDa VSP may have a role in pathogen defense, in addition to its function as a storage protein.

In perennial forage legumes such as alfalfa (Medicago sativa L.) and white clover (Trifolium repens L.), vegetative storage proteins are extensively mobilized to meet the nitrogen requirements of new shoot growth in spring or after cutting in summer. The 32-kDa alfalfa storage protein possesses high homology with class III chitinases, belonging to a group of pathogenesis-related proteins that possess antifreeze protein properties in some species and exhibit chitinolytic activity in vitro. This protein and the corresponding mRNA accumulate in taproots of cold-hardy culti vars during acclimation for winter, and in response to short-day conditions in controlled environments. The 17.3-kDa storage protein of white clover possesses high homology with pathogenesis-related proteins and abscisic- acid-responsive proteins from several legume species and has characteristics common to stress-responsive proteins. Low temperature enhances accumulation of this 17.3-kDa protein and its corresponding transcript. Exogenous abscisic acid stimulates the accumulation of vegetative storage proteins and their transcripts in both legume species. These observations suggest that vegetative storage proteins do not exclusively serve as nitrogen reserves during specific phases of legume development, but may play important adaptive roles in plant protection against abiotic (low temperature) and biotic (pathogen attack) stresses.Key words: nitrogen reserves, vegetative storage proteins, regulation, cold tolerance, chitinase, pathogenesis-related proteins.

Fall dormancy results in decumbent, slow shoot growth of alfalfa (Medicago sativa L.) in autumn and reduced shoot regrowth rates after herbage removal in summer. Although fall dormancy is used to predict alfalfa adaptation, we possess a poor understanding of the biological mechanisms underlying fall dormancy. Our objective was to examine growth and carbohydrate metabolism of suspension cell cultures derived from contrasting alfalfa cultivars that genetically differed in fall dormancy. Suspension cells were grown in B5h media containing 2% sucrose. Cells derived from fall non-dormant plants accumulated sugars more rapidly after transfer to fresh media and to higher concentrations than did cells derived from fall dormant alfalfa cultivars. Dark respiration rates of cells derived from non-dormant plants were similar to those derived from fall dormant plants when growth was limited at low cell sugar concentrations. However, both cell growth and dark respiration rates increased in cells derived from non-dormant cultivars in response to greater cell sugar concentrations. High growth rates of cells derived from rapid growing, fall non-dormant alfalfa cultivars were associated with rapid sugar uptake and higher cell respiration rates when compared to cells derived from dormant alfalfa cultivars.

Our objective was to study the effect of short-day photoperiod for 28, 42 and 56 d on growth, N uptake and N partitioning, particularly vegetative storage protein (VSP) accumulation in taproots of two alfalfa (Medicagosativa L.) cultivars (Lodi and Europe). For both varieties, the reduction of daylength from 16 h (long day,LD) to 8 h (short day, SD) for 28 d reduced total plant growth by decreasing shoot growth. Nitrogen uptake and N distribution within the plant was determined by 15N labeling. N uptake decreased with SD treatment duration, and was 2- and 3-fold lower for Europe and Lodi, respectively, for 56 d in SD conditions when compared with LD plants. The SD treatment resulted in preferential partitioning of N to taproots in comparison with LD conditions (19vs 9% for Lodi and 12 vs 5% for Europe after 28 d). For both cultivars, the SD-induced changes in N allocation to taproots did not significantly affect taproot soluble protein concentrations during 42 d of daylength treatment. In contrast, VSP accumulation occurred after only 28 d for plants grown in SD conditions (6.2 vs 4.8 mg g-1 DW for Lodi and 5.1 vs 1.4 mg g-1 DW for Europe). SD exposure also increased vsp 57 and vsp 32 mRNA transcript levels in Lodi and Europe (up to 2-fold higher) taproots in SD for 28 d compared with LD conditions. Overall results indicate that photoperiod modulates taproot N accumulation in alfalfa by enhancing both β-amylase (vsp 57) and vsp 32 gene expression and accumulation. The enhanced VSP accumulation by short-day photoperiod may result from altered VSP gene expression / transcript stability or occur indirectly through altered N source-sink relationships. Additionally, when SD treatment included a night break with 15 min illumination with sodium high pressure light or red light, our results suggest that the induction of vsp 57 and vsp 32 gene expressions by SD signal is mediated by the phytochrome system.

before a killing freeze was imperative for winter survival and plant persistence. Cutting during cold hardening Harvesting alfalfa (Medicago sativa L.) after mid-September in lowered both plant persistence and root total nonstructhe North-Central USA often reduces plant winter survival, but the physiological mechanisms associated with poor winter survival are not tural carbohydrate (TNC) concentrations. However, understood. Our objective was to determine how autumn harvesting other research has shown that cutting during this critical affects alfalfa root physiology, gene expression, and plant winter surfall rest period may not impact persistence or subsevival. In Exp. 1, seven fall harvest dates were used to identify 1 to quent yield (Tesar and Yager, 1985; Sheaffer et al., 1986; 15 October as a critical interval where significant changes in alfalfa Edmisten et al., 1988; Bélanger et al., 1992). In Virginia, winter survival and root physiology occur in Indiana. In Exp. 2, rows harvesting in late rather than early fall generally inof six alfalfa cultivars possessing contrasting fall dormancy (FD) were creased shoot growth the following spring (Edmisten et established in May. Plants in one-half of each row were defoliated al., 1988). Bélanger et al. (1992) used growing degree in mid-October, and roots were sampled at this defoliation and again days instead of calendar dates to show that winter injury in December. Winter injury was determined in mid-April. Shoot rewas reduced if the interval between the final harvest in moval in mid-October increased winter injury and reduced plant vigor in spring. As expected, the October defoliation reduced root protein late autumn and the previous harvest was 500 growingand starch concentrations in December, but unexpectedly increased degree-days or greater. Improved plant persistence obroot sugar concentrations. In addition, defoliation did not reduce the tained by delaying the final fall harvest is generally besteady state transcript levels of several cold-acclimation responsive lieved to result from increased root TNC levels (Tesar (car) genes that are associated with genetic variation in winter survival. and Yager, 1985; Sheaffer et al., 1986), but this hypotheAlthough positively associated with genetic differences in winter harsis has not been rigorously tested. Edmisten and Wolf diness, factors other than root sugar accumulation and expression of (1988) speculated that an extended period of slow these car genes regulate defoliation-induced changes in winter survival growth, low dark respiration rates, and high rates of of these alfalfa cultivars. photosynthesis during autumn months stimulates TNC accumulation and improves winter hardiness. Recently, Dhont et al. (2002) reported that mass of TNC in roots P in the North-Central USA sow fall dorin autumn had a stronger correlation with shoot growth mant alfalfa cultivars because of the positive associain spring than does the root TNC concentration. tion of FD with winter survival. One negative conseThe physiological mechanisms for reduced persisquence of growing fall dormant cultivars is their reduced tence and vigor in spring resulting from autumn defoliashoot elongation and leaf area expansion rates after tion are not clearly understood. In addition, it is not forage harvest in summer (Volenec, 1985). This limits known how autumn defoliation alters the expression of the number of annual cuttings, and subsequently respecific cold hardiness genes whose expression has been duces seasonal forage yield potential. When compared consistently associated with genetic differences in alfalfa with dormant cultivars, those with intermediate FD [5 winter hardiness (Cunningham et al., 2001). The goal to 7, a relative measure as described by Teuber et al. of this research was to determine the effects of autumn (1998)] may provide enough forage for additional hardefoliation on root physiology and winter survival of vests each year. If less fall dormant cultivars were made alfalfa. This objective was achieved through two experiavailable to northern production areas, benefits could ments: (i) assessment of several autumn cutting dates include increased yield potential. However, a potential on the accumulation of reserves in alfalfa roots; (ii) risk of raising cultivars possessing intermediate FD is evaluating the effects of autumn defoliation on root their lower winter hardiness, a problem exacerbated by physiology and winter survival of six alfalfa cultivars improper fall harvest management. possessing contrasting FD and winter hardiness. We Past research indicates that autumn defoliation may expected that autumn defoliation would prevent accudecrease winter survival and stand persistence (Silkett mulation of sugars, starch, and protein in roots, and et al., 1937; Grandfield, 1943). Smith (1972) reported be accompanied by reduced winter survival. We also that a period of 4 to 6 weeks of uninterrupted growth expected untimely autumn defoliation to reduce the expression of specific car genes whose expression is Dep. of Agronomy, Purdue Univ., West Lafayette, IN 47907-1150 USA. Contribution from the Purdue Univ. Agric. Exp. Stn., Journal Abbreviations: car, cold-acclimation response gene; FD, fall dorSeries No. 16820. Received 4 July 2002. *Corresponding author mancy; HPLC, high performance liquid chromatography; RFO, raffi(jvolenec@purdue.edu). nose family oligosaccharides; TNC, total nonstructural carbohydrate; VSP, vegetative storage protein. Published in Crop Sci. 43:1340–1348 (2003).

Large differences in winter hardiness exist among alfalfa (Medicago sativa L.) cultivars, but the physiological and molecular bases for these differences are not understood. Our objective was to determine how raffinose family oligosaccharide (RFO) accumulation and steady state mRNA levels for galactinol synthase (GaS) in roots relate to genetic variation in alfalfa winter survival. A GaS cDNA was isolated that possesses over 70% identity with GaS clones from other plant species. Induction of GaS transcripts in crowns of winter hardy alfalfa cultivars occurred within 8 h of exposure to 2°C, and was intensified by exposing plants to −2°C for 2 wk. Galactinol synthase transcripts increased in November in crown and root tissues of winter hardy alfalfa plants. This increase was accompanied by large increases in root RFO concentrations between October and December. A close positive association between RFO accumulation in roots in December and genetic differences in winter survival was observed in these alfalfa populations. Although roots and crowns of nondormant alfalfa cultivars accumulated both GaS transcripts and RFO, accumulation was delayed until December and these cultivars did not survive winter. Understanding the mechanisms regulating GaS gene expression and subsequent RFO accumulation in roots and crowns provides opportunity to genetically improve alfalfa winter hardiness.

The physiological mechanisms causing fall dormancy (FD)-induced differences in alfalfa (Medicago sativa L.) shoot growth in autumn and winter hardiness are not understood. The objective of this research was to examine root physiology of experimental germplasms selected for decreased FD that also were selected simultaneously for high winter hardiness. Dormant and semi-dormant cultivars and germplasms had high root sugar concentrations that were positively associated with winter hardiness. Root amino N and protein levels in December were greater for germplasms selected for decreased FD and increased winter hardiness than for cultivars with comparable levels of winter hardiness. Among the five most fall dormant cultivars Vernal incurred the greatest amount of winter injury and it had lower root amino-N concentrations when compared with three of the other four dormant cultivars and germplasms. Germplasm 98-132 with an intermediate FD, incurred relatively low winter injury, similar to that of fall dormant Vernal, when compared with other intermediate dormancy cultivars and germplasms. This germplasm had root sugar concentrations that were similar to plants with FD ratings of 1 to 3. Creation of even less FD germplasms that possess high winter hardiness would facilitate our understanding of the physiological and molecular mechanisms controlling these two very important agronomic traits of alfalfa.

Harvesting alfalfa (Medicago sativa L.) after mid‐September in the North‐Central USA often reduces plant winter survival, but the physiological mechanisms associated with poor winter survival are not understood. Our objective was to determine how autumn harvesting affects alfalfa root physiology, gene expression, and plant winter survival. In Exp. 1, seven fall harvest dates were used to identify 1 to 15 October as a critical interval where significant changes in alfalfa winter survival and root physiology occur in Indiana. In Exp. 2, rows of six alfalfa cultivars possessing contrasting fall dormancy (FD) were established in May. Plants in one‐half of each row were defoliated in mid‐October, and roots were sampled at this defoliation and again in December. Winter injury was determined in mid‐April. Shoot removal in mid‐October increased winter injury and reduced plant vigor in spring. As expected, the October defoliation reduced root protein and starch concentrations in December, but unexpectedly increased root sugar concentrations. In addition, defoliation did not reduce the steady state transcript levels of several cold‐acclimation responsive (car) genes that are associated with genetic variation in winter survival. Although positively associated with genetic differences in winter hardiness, factors other than root sugar accumulation and expression of these car genes regulate defoliation‐induced changes in winter survival of these alfalfa cultivars.

The objective of this paper is to assess the effectiveness of alfalfa (Medicago sativa L.) improvement efforts over the last century, and with the advent of molecular biology, identify challenges for alfalfa improvement in the future. Yield trials conducted between 1986 and 1998 from around the US were used to compare yield and persistence of older alfalfa cultivars to those released in the 1990s. First and second harvest forage yield of recently released alfalfa cultivars were not improved over those of older cultivars. New cultivars had higher forage yield at fourth harvest, in early September, possibly due to a reduction in fall dormancy. Efforts to improve alfalfa persistence by breeding for improved disease resistance and greater winter hardiness also have not been effective at most locations. Use of molecular biology for alfalfa improvement depends upon identifying genes that control important agronomic traits that translate into greater yield, improved persistence, and enhanced forage quality. Few such genes have been identified in alfalfa, and their use might be complicated by the polyploid nature of this outcrossing species. The Medicago truncatula genome project is providing large amounts of sequence information, but little is known about the regulation of these genes and the function of their protein products in planta. Uncertainty exists regarding the effectiveness of transferring these genes to alfalfa to obtain a desired phenotype. Much remains to be done to identify key genes that determine agronomic performance of crop plants, including alfalfa, and to clarify mechanisms that regulate the expression of genes and the function(s) of their protein products under field conditions. Future efforts to improve agronomic performance of alfalfa will be enhanced by partnerships between public and private scientists because companies now dominate commercial release of new alfalfa cultivars.

Addition of phosphorus (P) and potassium (K) fertilizer is vital to maintain alfalfa productivity. The objective of this study was to examine how P and K fertilizer application increases alfalfa yield, nutrient use, and plant persistence. Replicate plots of P (0, 50, 100, 150 lbs P2O5/acre) and K (0, 100, 200, 300, 400 lbs K2O/acre) treatments were arranged in a factorial design. Forage harvests occur-red four times annually for 3 years and yield, mass per shoot, shoots per area, and herbage nutrient concentrations were determined. Roots were dug in May and December of each year to ascertain plant populations and determine whether plants were dying during the summer (May to December) or during the winter (December to May). Total annual yield increased with application of P and K, but K application did not increase first harvest yield. Yield increases were due to greater mass per shoot. Addition of P decreased plant populations, whereas K did not alter stand populations. Removal rates of K exceeded K application each year even at the highest (400 lbs K2O/acre/yr) rate.