fao.gif (5805 bytes)

Astragalus cicer L.

Photo1.jpg (1333 bytes)

Author: John Frame


Cicer milkvetch, chickpea milkvetch.


Perennial, prostrate to decumbent, with succulent hollow stems, height 0.6-1.0 m. Leaves imparipinnate with 8-17 pairs of lanceolate leaflets plus the terminal leaflet. Undersides of leaflets slightly pubescent. Stipules triangular-oblong. Plants have branched tap roots and spread by rhizomes. Inflorescences, arising from leaf axils, are compact racemes with up to 60 white to pale yellow florets. Inflated black seed pods contain 9-15 oval, yellow seeds. High proportion of hard seeds. Number of seeds kg-1: 250 000-285 000. Average 1000-seed weight: 3.8 g.


It is indigenous to central and eastern Europe, and introduced to southern Europe and to North and South America.


Growth It is adapted to a wide range of soil and climatic conditions, from drylands with more than 400 mm rainfall annually to irrigated soils (Miller and Hoveland, 1995). It prefers calcareous soils but grows on moderately acid and alkaline soils, although intolerant of very acid soils, possibly because of Al toxicity and P deficiency (Davis, 1981) and so responds to liming to raise soil pH (Vickers, Zak and Oduruke, 1977; Hill, Mulcahy and Rapp, 1996). It has good drought tolerance, though less so than alfalfa (Peterson, Sheaffer and Hall, 1992). It is winter hardy. It has slow germination, seedling emergence and sward development (Smoliak, Johnston and Hanna, 1972; Townsend and McGinnies, 1972b), and is slow to start growth in spring at low to moderate temperatures (7-18C) compared with alfalfa and sainfoin (Smoliak, Johnston and Hanna, 1972). It is a long-day species that requires a period of vernalization for maximum flowering (Townsend, 1981). It is intolerant of shade. Slow to moderately vigorous regrowth emanates from axillary growth buds, the plant crown and rhizomes, and is more dependent on residual leaf area after defoliation, as is birdsfoot trefoil, than on reserves of total non-structural carbohydrates in the roots, as is alfalfa (Gabrielsen, Smith and Townsend, 1985). It is suited to infrequent defoliation.

Ability to spread naturally It spreads by rhizomatous growth, though mainly after the initial year of establishment. Being a prolific seed producer, ripened shed seed are a potential source of new plants.

Compatibility in mixture It is compatible with a number of cool-season grasses, e.g. smooth bromegrass (Bromus inermis), creeping foxtail (Alopecurus arundinaceus) and wheatgrasses (Thinopyrum spp.), though not with some others, e.g. timothy (Phleum pratense) smooth-stalked meadow-grass (Poa pratensis) and switchgrass (Panicum virgatum) (Townsend, 1993). Cicer milkvetch was not competitive with birdsfoot trefoil, red clover or sainfoin when stands were cut for hay (Cooper, 1979). In a short-term greenhouse trial, leaf, root and total seedling weights were higher than those from birdsfoot trefoil, while in mixed culture cicer milkvetch was more aggressive than birdsfoot trefoil but less aggressive than sainfoin (Smoliak and Hanna, 1977).

Nitrogen fixation Seed inoculation with an effective and competitive strain of Rhizobium leguminosarum is necessary when sowing on to land without a previous history of growing cicer milkvetch. Astragalus cicer is reputed to have good N2 fixation ability (Gervais, 2000).


Diploid with 2n = 2x = 64 chromosomes. Cross-pollinated mainly by bumble bees (Bombus spp.), but also honey (Apis mellifera) and leaf-cutter (Megachile rotunda) bees. Little breeding effort reported in recent years. Improvements are required in seedling vigour, development of establishment phase, including rhizome spread, and length of vegetation growing period.

Cultivars Examples are Lutana, Monarch and Windsor, bred in the United States of America, and Oxley, a Canadian cultivar. More recently, a synthetic cultivar, AC Oxley II, bred for improved seedling vigour and forage yield, has been released in Canada (Acharya, 2001).

Seed production Seed crops are usually cut, windrowed and combine harvested. Since cicer milkvetch has indeterminate flowering, some immature seed pods are normally present. Row spacings up to 90 cm are used under irrigation, but even wider under unirrigated dryland conditions (Townsend, 1993). Typically, commercial seed yields are circa 500 kg ha-1 (Miller and Hoveland, 1995; Acharya, 2001).

In seed storage experiments, hard seed content fell significantly with age of seed lot and with freeze-thaw treatment, thus significantly increasing seed germination and seedling emergence (Acharya, Kokko and Fraser, 1993).


Establishment A well cultivated, uniform and firm seed bed is required for good results. Seeds require scarification, but preferably immediately before sowing, since viability rapidly declines following scarification (Townsend and McGinnies, 1972b). Seeds are best drilled without a cover crop at 15-20 mm depth, and the seed bed then consolidated. Sown at 9 kg ha-1 in monocultures and 6 kg ha-1 in grass/cicer milkvetch mixtures. Legume content was adversely affected in the establishment year by competition from weeds (Moyer, 1989); when sown with cool-season grasses, legume contents were less when the components were sown mixed in the same rows or when cross-sown than when sown in alternate rows (Kenno, Brick and Townsend, 1987). By the third year, legume contents - at 80-90 percent - were similar for the three sowing patterns (Townsend, Kenno and Brick, 1990). Cicer milkvetch has also been direct drilled (sod seeded) successfully in autumn or spring following suppression of the existing sward by glyphosate application (Malik and Waddington, 1990). Cicer milkvetch has proved useful for erosion control, and for reclamation of land disturbed by mining (Holecheck et al., 1982).

Nutrient requirements Adequate soil fertility is needed to ensure satisfactory yields.

Weeds Ability to compete with weeds is poor in the early establishment phase because of lack of vigour, and therefore problem weeds should be mechanically, chemically or culturally controlled before sowing. Established swards, with their dense cover, are competitive towards weed ingress. A number of "cicer milkvetch-safe" herbicides have been identified for establishing and established swards (Moyer, 1989; Townsend, 1993), e.g. trifluralin (pre-sowing); 2,4-DB (post-emergence); trifluralin or 2,4-DB + bromoxynil (established swards).

Pests A number of pests have been observed on cicer milkvetch, but their effects do not seem to be severe (Townsend, 1993). Within seven species of forage legumes, cicer milkvetch was one of the least preferred as a food plant by rangeland grasshoppers (Melanoplus spp.) (Hewitt, Wilton and Lokenz, 1982). It is less susceptible than alfalfa to damage from insects such as potato leafhopper (Empoasca fabae), pea aphid (Acyrthosiphon pisum) and alfalfa weevil (Hypera postica) (Kephart et al., 1990).

Diseases Information is sparse on the susceptibility or tolerance of the legume to specific diseases and though several root, crown and stem rot diseases have been observed, its rhizomatous growth habit has enabled it to resist any adverse effects (Townsend, 1993).

Forage production In a review of DM yields relative to other forage legumes in the same trials at a range of North American sites, irrigated and non-irrigated cicer milkvetch yields were generally lower than alfalfa yields, sometimes considerably so, though also higher at a few sites and somewhat similar to birdsfoot trefoil yields, but superior to those of Caucasian clover, crownvetch , red and alsike clovers (Townsend, 1993); actual DM yields of cicer milkvetch varied from 1.2 to 15.8 t ha-1 because of different growing conditions, but were often in the 8-11 t ha-1 range. Under irrigation and differing cutting frequencies (2-7 cuts annually), DM yields ranged from 9.0 to 10.7 t ha-1, with the highest yields associated with infrequent cutting (Townsend, Christensen and Dotzenko, 1978); the percentage contribution to annual yields of the primary cut ranged from 62 percent (2-cut system) to 36 percent (7-cut system). Over a 6-year trial period, cicer milkvetch maintained an annual DM production similar to alfalfa, with lower yields than alfalfa in early season and higher yields in late season (Loeppky et al., 1996). In cultivar trials in western Canada, DM yields of cv. AC Oxley II ranged between 1.0 and 11.7 t ha-1 (mean 4.8), over 26 site years in dryland conditions, and between 8.1 and 9.4 t ha-1 (mean 8.8) over 8 site years with irrigation; these yields were considerably superior to those of the old control cultivar, Oxley (Acharya, 2001).


Cicer milkvetch has excellent nutritive value that compares favourably with several other forage legumes such as alfalfa, birdsfoot trefoil and sainfoin (McGraw and Marten, 1986; Kephart et al., 1990). Cicer milkvetch has a higher leaf : stem ratio than alfalfa, especially under drought conditions (Peterson, Sheaffer and Hall, 1992) and it retains its leaves longer in late season (White and Wight, 1981), as a result of which it has a higher in vitro dry matter digestibility (DDM) (Loeppky et al., 1996). Compositional data from trials in Quebec, Canada, are shown in Tables 1 and 2. The forage was less acceptable to grazing heifers than alfalfa, birdsfoot trefoil or sainfoin (Marten, Ehle and Ristau, 1987) and less acceptable to grazing sheep than alfalfa, birdsfoot trefoil and red clover (Marten, Jordan and Ristau, 1990). Its non-bloat-inducing characteristic is probably related to the physical make-up of the foliage - thicker cuticles, epidermal and mesophyll cell walls than bloat-causing legumes such as white and red clovers, and hence slower disintegration and digestion by rumen bacteria (Lees et al., 1981; Lees, Howarth and Goplen, 1982).

Table 1  Nutritive value of cicer milkvetch (g kg-1 DM) at three stages of primary growth (Means of 2 cultivars over 3 years)

Stage of growth


Cell contents




Flower bud






10 percent flowering






Full flowering






Key: CP = crude protein; NDF= neutral detergent fibre; ADF = acid detergent fibre.
Source: Based on Gervais, 2000.

Table 2  Mineral composition of cicer milkvetch at three stages of primary growth (Means of 2 cultivars over 3 years).


Stage of growth

Flower bud

10 percent flowering

Full flowering

Ca g kg-1 DM




P g kg-1 DM




K g kg-1 DM




Mg g kg-1 DM




Na g kg-1 DM




S g kg-1 DM




Mn mg kg-1 DM




Zn mg kg-1 DM




Cu mg kg-1 DM




Fe mg kg-1 DM




Source: Based on Gervais, 2000.

Anti-quality factors In trials at a location in north-central United States of America, dairy heifers and sheep grazing pure-sown cicer milkvetch developed a photosensitization response, possibly a secondary (hepatogenous) photosensitization associated with high levels of phyllocrythrin, a normal breakdown product of chlorophyll in the blood, (Marten, Ehle and Ristau, 1987; Marten, Jordan and Ristau, 1990). However, this response has not occurred elsewhere.

Grazing management
It is suited to rotational grazing systems rather than continuous stocking.

Conservation management It is suited to conservation as hay, although its dense foliage and high moisture content make the cut herbage more difficult to dry than other legumes such as alfalfa. During haymaking, care in handling is necessary to avoid loss of the nutritious leaf fraction.

Satisfactory liveweight gains have been obtained by sheep and cattle from pure-sown cicer milkvetch and grass/cicer milkvetch mixtures, the gains being generally less than from N-fertilized grass swards but nevertheless in approximate proportion to the forage yields obtained (Townsend, 1993). In a 3-year study, lambs grazing cicer milkvetch had greater production (826 kg ha-1) than lambs grazing alfalfa (784 kg ha-1), red clover (739 kg ha-1) or birdsfoot trefoil (744 kg ha-1), even though cicer milkvetch caused a photosensitization response in 33 to 50 percent of the lambs (Marten, Jordan and Ristau, 1990); average daily liveweight gains of lambs were highest for birdsfoot trefoil (228 g) and cicer milkvetch (224 g) and lowest for alfalfa (215 g) and red clover (217 g). However, in a grazing trial with heifers, seasonal liveweight gains ha-1 were 434 kg for birdsfoot trefoil, 392 kg for alfalfa, 366 kg for sainfoin and 282 kg for cicer milkvetch (Marten, Ehle and Ristau, 1987). Corresponding average daily liveweight gains of heifers were 0.81, 0.67, 0.81 and 0.42 kg, the poor performance of cicer milkvetch reflecting low acceptability and intake, and possibly the effects of photosensitization that occurred in this trial.

It has high forage yields of good nutritive value, is persistent, with good N2-fixing ability and is non-bloat-inducing. It is suitable for erosion control and reclamation of disturbed land.

It has a slow establishment phase and has only moderate acceptability.


ACHARYA, S.N. (2001) AC Oxley II cicer milkvetch. Canadian Journal of Plant Science, 81, 749-751.
ACHARYA, S.N., KOKKO, E.G. and FRASER, J. (1993) Storage duration and freeze-thaw effects on germination and emergence of cicer milkvetch (Astragalus cicer) seeds. Journal of Seed Technology, 17, 1-13.
COOPER, C.S. (1979) Evaluation of legume mixtures for hay plantings. Agronomy Journal, 71, 81-83
DAVIS, M.R. (1981) Growth and nutrition of legumes on a high country yellow-brown earth subsoil. L. Phosphate response of Lotus, Trifolium, Lupinus, Astragalus and Coronilla species and cultivars. New Zealand Journal of Agricultural Research, 24, 321-332
GABRIELSEN, B.C., SMITH, D.H. and TOWNSEND, C.E. (1985) Cicer milkvetch and alfalfa as influenced by two cutting schedules. Agronomy Journal, 77, 416-422.
GERVAIS, P. (2000) L’astragale pois chiche, la coronille bigarrée et le sainfoin (cicer milkvetch, crown vetch and sainfoin). Université Laval, Québec, 190pp
HEWITT, G.B., WILTON, A.C. and LOKENZ, R.J. (1982) The suitability of legumes for rangeland interseeding and as grasshopper food plants. Journal of Range Management, 35, 653-656.
HILL, M.J., MULCAHY, C. and RAPP, G.G. (1996) Perennial legumes for the high rainfall zone of eastern Australia. 1. Evaluaiton in single rows and selection of Caucasian clover (Trifolium ambiguum M. Bieb). Australian Journal of Experimental Agriculture, 36, 151-163.
HOLECHECK, J.L., DEPUIT, E.J., COENENBERG, J.G. and VALDEZ, R. (1982) Legume establishment on strip mined lands in southeastern Montana. Journal of Range Management, 35, 298-300.
KENNO, H., BRICK, M.A. and TOWNSEND, C.E. (1987) Establishment of cicer milkvetch with four cool-season grasses. Crop Science, 27, 810-812.
KEPHART, K.D., HIGLEY, L.G., BUXTON, D.R. and PEDIGO, L.P. (1990) Cicer milkvetch forage yield, quality and acceptability to insects. Agronomy Journal, 82, 477-483.
LEES, G.L., HOWARTH, R.E., GOPLEN, B.P. and FESSER, A.C. (1981) Mechanical disruption of leaf tissues and cells in some bloat-causing and bloat-safe forage legumes. Crop Science, 21, 444-448.
LEES, G.L., HOWARTH, R.E. and GOPLEN, B.P. (1982) Morphological characteristics of leaves from some legume forages: Relation to digestion and mechanical strength. Canadian Journal of Botany, 60, 2126-2132.
LOEPPKY, H.A., BITTMAN, S., HILTZ, M.R. and FRICK, B. (1996) Seasonal changes in yield and nutritional quality of cicer milkvetch and alfalfa in northeastern Saskatchewan. Canadian Journal of Plant Science, 76, 441-446.
MALIK, N. and WADDINGTON, J. (1990) No-till pasture renovation after sward suppression by herbicides. Canadian Journal of Plant Science, 70, 261-267.
MARTEN, G.C., EHLE, F.R. and RISTAU, E.A. (l987) Performance and photosensitization of cattle related to forage quality of four legumes. Crop Science, 27, l38-l45.
MARTEN, G.C., JORDAN, R.M. and RISTAU, E.A. (1990) Performance and adverse response of sheep during grazing of four legumes. Crop Science, 30, 860-866.
McGRAW, R.L. and MARTEN, G.C. (1986) Analysis of primary spring growth of four pasture legume species. Agronomy Journal, 78, 704-710.
MILLER, D.A. and HOVELAND, C.S. (l995) Other temperate legumes. In: Barnes, R.F., Miller, D.A. and Nelson, C.J. (eds) Forages, 5th edn, Vol.1, An Introduction to Grassland Agriculture. pp. 273-281. Iowa State University Press, Ames, Iowa.
MOYER, J.R. (1989) Weed control during cicer milkvetch establishment and yields in subsequent years. Canadian Journal of Plant Science, 69, 213-222.
PETERSON, P.R., SHEAFFER, C.C. and HALL, M.W. (l992) Drought effects of perennial forage legume yield and quality. Agronomy Journal, 84, 774-779.
SMOLIAK, S. and HANNA, M.R. (1977) Seedling competition of some forage legumes in mono- and mixed culture under greenhouse conditions. Canadian Journal of Plant Science, 57, 897-903.
SMOLIAK, S., JOHNSTON, A. and HANNA, M.R. (1972) Germination and seedling growth of alfalfa, sainfoin and cicer milkvetch. Canadian Journal of Plant Science, 52, 757-762.
TOWNSEND, S.E. and McGINNIES, W.J. (1972b) Establishment of nine forage legumes in the central Great Plains. Agronomy Journal, 64, 699-702.
TOWNSEND, C.E., CHRISTENSEN, D.K. and DOTZENKO, A.D. (1978) Yield, quality and persistence of cicer milkvetch as influenced by cutting frequency. Agronomy Journal, 70, 109-113.
TOWNSEND, C.E. (1981) Vernalization and photoperiod requirements for maximum flowering of cicer milkvetch. Crop Science, 21, 917-921.
TOWNSEND, C.E., KENNO, H. and BRICK, M.A. (1990) Compatibility of cicer milkvetch in mixtures with cool-season grasses. Agronomy Journal, 82, 262-266.
TOWNSEND, C.E. (1993) Breeding, physiology, culture and utilization of cicer milkvetch (Astragalus cicer L.) Advances in Agronomy, 49, 253-308.
VICKERS, J.C., ZAK, J.M. and ODURUKE, S.O. (1977) Effects of pH and Al on the growth and chemical composition of cicer milkvetch. Agronomy Jurnal, 69, 511-513.
WHITE, L.M. and WIGHT, J.R. (1981) Seasonal dry matter yield and digestibility of seven grass species, alfalfa, and cicer milkvetch. Agronomy Journal, 73, 457-462.