Desmodium intortum (Mill.) Urb.

Desmodium aparines (Link.) DC.

Greenleaf desmodium (Australia), pega-pega (Philippines, Costa Rica, Venezuela), kuru vine (Zimbabwe), amor seco (Colombia).

Large trailing and climbing perennial; roots at the nodes and has a deep taproot; long, pubescent stems branch freely and are often reddish brown. Has shorter internodes than D. uncinatum and is leafier. Leaves usually have reddish-brown to purple flecking on the upper surface. Leaflets, 2 to 7 cm long and 1.5 to 5.5 cm broad, with a length-width ratio of 1.4 to 1, are shorter and more rounded than in D. uncinatum. Terminal raceme compact, flower deep lilac to deep pink. Seed pod narrow, bears 8 to 12 seeds, recurves to the main rachis; seed adheres to animals and to clothing, but not as tenaciously as that of D. uncinatum (Barnard, 1967).

Common in northern parts of South America, on the eastern slopes of the Andes and in a restricted area of Brazil between 18 and 25°S. Native in Panama, Colombia, Guatemala, Venezuela, Ecuador and Peru. Has spread throughout the tropics and is being widely tested for fodder value.

A summer-growing perennial, with main growth from mid-October to mid-May (Queensland).

Favours a well-distributed rainfall in excess of 875 mm a year. Kelly (1964) found it outstanding in a rainfall up to 3 475 mm at Topaz on the Atherton Tableland, Queensland, but in similar high rainfall in Panama and Guatemala it was badly affected by disease and leaf-eating insects.

Grows on a wider range of soils than D. uncinatum and does not do quite as well on sandy soils as S. guianensis (Stobbs, 1969f). Will grow in a range of soils from light to clay loams. Requires a soil with a pH in excess of 5.0 (Andrew and Bryan, 1958; Moomaw and Takahashi, 1962) . Has no tolerance to salinity, and is depressed by high chloride levels (Andrew and Robins, 1969b).

Requires the specific "Desmodium" culture (Date, 1969) . The current Australian recommendation is CB 627 (1970) . Boultwood (1964) found it unnecessary to inoculate it in Zimbabwe. Whiteman (1970) showed that peak nodule formation occurred three months before flowering in D. intortum. Does not spread well from natural seed sources; better by natural vegetative means with its stoloniferous habit.

Because of its small seed, Desmodium intortum requires a well-prepared seed bed (Younge, Plucknett and Rotar, 1964). Will establish from broadcasting into ashes from the air.

Can be sown by drilling, broadcasting from ground machines or from the air. Has been established by cuttings in Zaire (Risopoulos, 1966) and on steep slopes in Guatemala on contour ridges (Johnston, personal communication). Risopoulos found that planting by cuttings gave only 30 to 40 percent strike but, if the cuttings were rooted in banana fibre baskets under light shade and later transplanted, results were good. Boultwood (1964) established it by transplanted cuttings spaced at 1 x 1 m and also by undersowing in maize early in the season. Calma, Valera and Santos (1959) obtained best yields with 20 x 25 cm spacings. It generally does not establish when oversown into existing pastures because of low seedling vigour. Extensive pasture renovation would be required to give any success. Should be sown at no greater depth than 1 cm (Suttie and Ogada, 1967) and rolled or very lightly harrowed.

Can be sown from spring to midsummer or later in frost-free environments at a rate of 1 to 2 kg./ha. Middleton (1970) found no seedling competition at rates of 1. 1,3.3 and 9.9 kg./ha, and under high rainfall conditions seedling density was proportional to sowing rate.

594 000. Percentage of hard seed is fairly high. Risopoulos (1966) found that only 25 percent of fresh seed germinated; this figure did not improve up to 18 months after collection.

To break dormancy: if seed is harvested mechanically, there is no need to treat. Boultwood (1964) advised treatment in strong sulphuric acid for five minutes, then thoroughly washing in water and drying. Inoculation is necessary. Pelleting with lime depresses nodulation (Norris, 1967). Luck (personal communication) obtained nil, 35 and 77 percent nodulation, however, from (a) uninoculated, (b) inoculated and band-sown, (c) inoculated, pelleted and band-sown seed respectively. If pelleting is required, Norris (1967) recommends rock phosphate.

D. intortum usually requires adequate levels of phosphorus, sulphur, potash and molybdenum for growth. Heavy dressings of potassium chloride, however, can cause chlorine toxicity (Andrew and Robins, 1969b). Younge, Plucknett and Rotar (1964) proved that when D. intortum is adequately fertilized with P, K, Mo, and Zn, it is able to compete with Kikuyu and pangola grass. On acid ferruginous and aluminous latosols in Hawaii, Younge and Plucknett (1966b) found that heavy P fertilization was needed. Treatment with 1 650 kg. P/ha allowed grazing of 6 beasts/ha compared with 3 beasts/ha pasture treated with 275 kg. P/ha.

• Calcium: 

In Hawaii, Younge, Plucknett and Rotar (1964) found that D. intortum grew satisfactorily on a soil of pH 5.5 and showed little response to application of lime.

• Nitrogen: 

Nitrogen fertilization at 14 kg./ha reduced the D. intortum component in mixtures with pangola grass and Kikuyu in Hawaii to less than 10 percent in the ten-week cutting treatment, and to less than 1 percent in the five-week cutting treatment (Whitney, 1970).

• Phosphorus: 

The critical level for P in the dry matter of the leaves at the immediately preflowering stage is 0.23 percent. On a Samford gley soil fertilized with the equivalent of 60 kg. superphosphate per ha, D. intortum yielded 43 percent of its maximum yield at 1 230 kg./ha (Andrew and Robins, 1969b). Plucknett and Fox (1966) found that the phosphorus content of unfertilized D. intortum remained at 0.11 percent for two years and regarded this as the minimum value for survival of this species.

• Potash: 

In Swaziland, l'Ons (1968) obtained a linear response to K at rates up to 55 kg./ha, the increase being about 11 kg./ha K. Andrew and Robins (1969c) found that D. intortum contained relatively high concentrations of potash. They found that deficiency occurs when the dry matter in the tops at the immediate preflowering period is less than 0.80 percent. Andrew and Robins (1969d) found no deficiency symptoms in plants which had 1.39 g of K per 100 in the dry matter of the tops, but deficiency symptoms were apparent in a plant with 0.47 g/100 g. The first sign of potassium deficiency in D. intortum was interveinal necrotic spotting on the midpositioned leaves of the plants. The necrotic spotting was not preceded by any chlorotic effects; however, in severe cases of deficiency some interveinal chlorosis occurred, usually in those areas with intense necrosis. The necrosis and chlorosis occurred in a symmetrical pattern on the tips and margins of the leaflets. Necrotic spots were irregular in shape, equally visible on both leaf surfaces, and mid-brown in colour. In severe form, the small spots coalesced to give large areas of necrotic tissue which usually encompassed the leaf margins. Associated with this there was an inward curling of the leaf margins. In extreme deficiency, even the younger, fully expanded leaves became pale and chlorotic, with the margins of the leaflets curled or rolled inward. Very little leaf abscission occurred in this species (Andrew and Pieters, 1970a) .

Grows well with Setaria spp., Paspalum commersonii, Panicum maximum, Pennisetum purpureum, Melinis minutiflora and, if adequately fertilized, with Kikuyu and pangola (Younge, Plucknett and Rotar, 1964). Also grows well with siratro and glycine. Middleton (1970) found siratro more competitive with Setaria anceps than with D. intortum. Bryan (1966) recorded that D. intortum had invaded swards of at least 12 species of Paspalum.

Exhibits extremely useful tolerance to 2,4-D . At 3 /2 weeks of age, 1.65 kg. of acid equivalent per hectare can be used, thus there is no need to use 2,4-DB. Diquat should be used only at 140 g of cation per hectare from six weeks of age. Diquat at 280 g of cation per hectare is safe on established greenleaf desmodium (Bailey, 1970).

Directly related to yield. D. intortum fixed over 300 kg. N/ha/year in Hawaii (Whitney, 1970). Row width had an effect. Whitney and Green (1969b) found that it fixed 213 kg./ha/year in 90-cm rows, and 264 kg./ ha/year in 45-cm rows. Whitney, Kanehiro and Sherman (1967) found that it fixed 375 kg. N/ha, of which it transferred only 5 percent to the associated grass. Leaf fall could add an additional 1.3 kg. N/ha/week.

Jones (1967) imposed three cutting heights and three cutting frequencies on a sward of D. intortum. It disappeared when cut at a height of 3.75 cm every four weeks, but survived cutting at 3.75 cm at frequencies of 8 and 12 weeks and cutting at 7.5 or 15 cm at 4, 8 and 12 weeks respectively. Risopoulos (1966) found that cutting at 12-week intervals gave the highest yield of cut fodder. Whitney (1970) found a cutting height of 13 cm was better than at 5 cm, N fertilization reduced the legume component to less than 10 percent in the ten-week cutting treatment and to less than 1 percent in the five-week cutting treatment. Under heavy stocking (1 beast to 0.4 ha) the legume decreased; under light stocking (1 beast to 0.8 ha) it increased. Riveros and Wilson (1970) found no significant difference in yield with cutting heights of 7.5 and 15 cm at the first season but a cutting interval of three weeks gave the highest yield; in the second season a cutting height of 15 cm at five-week intervals gave the highest yield. Whiteman (1970) found that the severity of initial defoliation affects the proportion of the original nodule population on the plant and subsequent defoliations affect the number of new nodules developed and their rate of increase in size.

Bryan (1966) illustrates the way cattle graze Desmodium pastures: stock normally remove the last part of the shoot and then browse the leaves, leaving large numbers of axillary buds which ensure rapid regrowth. If grazing is intermittent and intense, a greater proportion of the stem and buds may be removed or damaged, with consequent reduction in bud sites and residual leaf material. Recovery from grazing would then be much slower and would ultimately affect persistence. Grazing management must, therefore, first allow the legume to become established and then adjust grazing pressure to allow for retention of bud sites and leaf material. This also involves the companion grass and a compromise must be established to protect the sward.

It is not advisable to allow fire to pass through a sward of D. intortum, but it will not be entirely destroyed by it. In an established stand the top is killed but the taproot will shoot again (Boultwood, 1964).

Generally self-fertile, but pollination can be improved by tripping the flowers. Chromosome number 2n = 22 (Rotar and Ukio, 1967).
Crosses have been made between D. intortum and D. sandwicense (Hutton and Gray, 1967; McWhirter, 1969).

Boultwood (1964) recorded seasonal yields of 19 tonnes/ha of green material with a crude protein percentage of 18.8 percent. Younge and Plucknett (1966b) recorded a five-year average of 19 000 kg./ha/ year from a pangola/D. intortum pasture fertilized with 1 320 kg. P/ha in Hawaii. Whitney, Kanehiro and Sherman (1967) recorded a yield of 19 000 kg. dry matter/ha/year. Roe and Jones (1966) recorded a dry-matter yield of 12 500 kg./ha at Gympie, Queensland. Riveros (1969) recorded over 17 000 kg. DM/ha during a eight-month growing season at Redland Bay, Queensland. Calma, Valera and Santos (1958) obtained 5 875 kg. DM/ha in four cuttings spaced at 60, 59, 86 and 101 days respectively.

Risopoulos (1966) made good hay at Mulungu, Zaire. It has also been made successfully in Brazil, Queensland and Guatemala (Calma, Valera and Santos, 1958). In Guatemala the hay is ground into meal for stock-feeding. Without the addition of molasses it made reasonable silage with 12.2 percent dry-matter loss and 0.03 percent N loss in Brazil. The pH was 5.0 and was better with the addition of 8 percent molasses on a green-weight basis. Boultwood (1964) made good silage by flail harvesting, adding 2 percent molasses by green weight and compacting well, because the material is light and fluffy. Catchpoole (1970) made stable lactic acid silage from D. intortum to which molasses had been added up to 8 percent of the green weight of the material.

Its late flowering allows it to provide good standover feed in frost-free areas.

The meal is an excellent source of protein, riboflavin and vitamin A for chickens (Squibb et al., 1950, 1953; Huang, 1967). In Puerto Rico, Warmke and Freyre (1952) found high intake and good palatability when grazed by cattle.

None recorded. Boultwood (1964) found no bloat even when comprised the whole ration. Rotar (1965) found 3.2 to X.8 percent tannin in the leaves and 1.5 to 3.7 percent in the stems. Hutton and Coote (1966) found 7 percent tannin in the dry matter of leaves. Bindon and Lamond (1966) found no toxicity in mice fed with leaves and seed.

A difficult cultivar to harvest. In Queensland the crop is mown when 40 to 50 percent of the seed is ripe, as seed matures unevenly and progressively from the bottom. Unripe seed matures in the swath, and the ripe seed should not fall in the mowing process. Crops can be either windrowed or left in a swath. Heavy crops are best left in the swath. Leave the crop to dry for 1 0 to 14 days after cutting, then use a pick-up threshing harvester with the highest possible drum speed and close concave settings. All the material must be threshed in its first passage through the drum or the seed will ball up and be lost over the sieves. Use either adjustable sieves on a fairly close setting or a 25-mm punched-hole bottom sieve. Suttie and Ogada (1967) used this method in Kenya.

In areas where early frosts occur, seed production may be severely curtailed. Greenleaf desmodium yields about 80 to 100 kg./ha clean seed.

There is only one Australian cultivar­cv. Greenleaf. It is derived from three introductions, CPI 17916, CPI 18009 and CPI 23189 from El Salvador, Hawaii (ex Guatemala) and the Philippines respectively, each with similar characteristics. In Tanzania there is a cultivar named Tengeru. Rotar (1970) has drawn attention to the great amount of variability within the species and suggests that there is opportunity to develop a fairly large gene pool for Desmodium utilization in many different environments in the tropics and subtropics.

D. intortum is more resistant to legume little-leaf than D. uncinatum (Hutton and Grylls, 1956), but under very wet conditions it, too, is likely to be infected.

A well-grazed legume with high yield potential in frost-free areas of good rainfall. Has a long growing season and makes vigorous growth in association with grasses; gives early spring growth and is a good fertility builder.

Low seedling vigour, poor drought tolerance, poor salt tolerance and relatively low digestibility. It is grazed out unless heavily fertilized. Leaves, flowers and roots subject to attack by various pests.

Stobbs (1969c) at Serere, Uganda, obtained 500 kg./ha/year live-weight gain from a Panicum maximum/D. intortum sward. Moomaw and Takahashi (1962) obtained a beef yield of 660 kg./ha per year over a period of 2/2 years from a fertilized pangola grass/D. intortum pasture in Hawaii­a tenfold increase over normal ranch performance. Villareal (1967), at Beerwah, Queensland, obtained a live-weight gain of more than 500 kg./ha from a pangola grass/D. intortum pasture containing some Lotononis bainesii.

Bryan (1969); Compère (1961); Younge, Plucknett and Rotar (1964).

Requires a long, warm growing season. Whiteman (1968) assessed the optimum temperature for growth at 30/25°C + 3°C. Withstands hot weather better than D. uncinatum.l

Grows well into the autumn, but is susceptible to heavy frosts; retains leaf fairly well after frosting.

In the early stages poor, but when well established will suppress weeds (Boultwood, 1964).

Meloid beetles eat the flowers (Horrell, 1958). Amnemus weevil has attacked plants at Gympie, Queensland (Roe and Jones, 1966).

Is susceptible to extended dry spells, but persists well where soils are fertile. Wilts less readily than D. uncinatum (Ostrowski, 1966). Stobbs (1969e) showed it to be less drought-resistant than Stylosanthes guianensis at Serere, Uganda. It carries little foliage in the dry season, when most of the leaves drop and form a mulch (Horrell, 1958). Will survive temporary flooding and some waterlogging (Boultwood, 1964) but is susceptible to extended waterlogging. Performs better on slopes.

Extends to 25°S (Bryan, 1969). Is known to grow at 2 400 m near the equator in Kenya. In tropical latitudes in Latin America it occurs from 600 to 2 500 m. In the subtropics it occurs at a lower elevation but is rarely found in warm temperate zones, even at sea level.

Once established, D. intortum makes vigorous growth in the wet season; commences growth later in the spring than D. uncinatum.

Younge and Plucknett (1966b) recorded the chemical composition of a pangola grass/D. intortum pasture (mainly D. intortum) as 13.1 percent crude protein and 0.34 percent P, 1.44 percent Ca and 0.34 percent Mg in the dry matter. Dry-matter yields for D. intortum leaves and stems obtained by Compère (1961) are presented in Table 14.3.
Digestibility of the leaf protein was 54.08 percent, of the stem protein 61.88 percent, of the organic matter, 72.44 and 48.65 percent respectively. Risopoulos (1966) recorded 5.5 percent crude protein of 69 percent digestibility for the green material of D. intortum in Zaire.

Seventy percent germination with 94.5 percent pure seed (Queensland).