Animal feed resources information system

Karanja (Millettia pinnata)


Click on the "Nutritional aspects" tab for recommendations for ruminants, pigs, poultry, rabbits, horses, fish and crustaceans
Common names 

Karanja, karanj, pongam, Indian beech, Pongamia tree, Indian pongamia, pongamia [English]; arbre de pongolote [French]; malapari, mempari [Indonesian]; bangkong [Javanese]; Igiti cy’Ubuhinde [Kinyarwanda]; kranji [Madurese]; pokok mempari, mempari, kacang kayu laut, biansu [Malaysian]; bani [Philippines]; ki pahang laut [Sundanese]; pongami, honge [Swahili]; Đậu dầu, cây sồi Ấn Độ, cây Pongam, cây Honge [Vienamese]; করচ [Bengali]; 水黄皮 [Chinese]; करंज [Hindi; Marathi]; ഉങ്ങ് [Mayalayam]; புங்கை [Tamil]; కానుగ [Telugu]; หยีน้ำ [Thai]

Products: karanja oil cake, karanja oil meal, karanja press cake, karanja oilcake, karanja oilmeal, karanja presscake, karanj oil cake, karanj oil meal, karanj press cake, karanj oilcake, karanj oilmeal, karanj presscake, pongam oil cake, pongam oil meal, pongam press cake, pongam oilcake, pongam oilmeal, pongam presscake


Cytisus pinnatus L.; Derris indica (Lam.) Bennet; Galedupa indica Lam. Galedupa pinnata (L.) Taub.; Pongamia glabra Vent.; Pongamia mitis Kurz; Pongamia pinnata (L.) Pierre; Millettia novo-guineensis Kanehira & Hatusima

Though the taxon Millettia pinnata was suggested as early as 1988, karanja is still often referred to as Derris indica or Pongamia glabra in the literature (Orwa et al., 2009; Daniel, 1997).

Related feed(s) 

Karanja (Millettia pinnata (L.) Panigrahi) is a fast-growing, multipurpose tree of the humid tropic. It is one of the few N-fixing trees that produce oilseeds (Sangwan et al., 2010Orwa et al., 2009). With the increasing production of oil for biofuel, large amounts of oil cake are available for livestock feeding. "Karanja" is the popular name of Millettia pinnata in Hindi while it is called "pongam" in tamil (Punitha et al., 2006).


Karanja is a medium-sized evergreen or briefly deciduous tree, usually about 8 m high but that can grow to 15-25 m (Sangwan et al., 2010; Orwa et al., 2009). The trunk is straight or crooked, 50 cm in diameter, covered with grey to greyish brown bark, smooth or vertically fissured. Karanja has a deep and thick taproot with several secondary lateral roots (Daniel, 1997). The branches are spreading or dropping, and form a broad hemispherical crown of dark green leaves. Branchlets are hairless with pale stipule scars. Leaves are alternate, pinnately compound, pinkish-red when young becoming glossy dark green at maturity (Orwa et al., 2009). The leaves consist of 5 to 7 glabrous leaflets borne in pairs (2 or 3) on slender stalks, and a single end leaf. The leaflets are ovate-elliptical, about 5-10 cm long and 4-6 cm wide, pointed at the tip. The inflorescence is a 6-27 cm long raceme of typically papilionaceous, very fragrant flowers. The flowers are lavender, pink white in colour, finely pubescent, and 15-18 mm long. Fruits are numerous, elliptical, 3-6 cm long x 2-3 cm broad, hard and woody indehiscent pods. Pods contain 1-2 seeds (Orwa et al., 2009Daniel, 1997). Seeds are bean-like, 1.5-2.5 cm long x 1.2-2.0 cm wide, dark brown in colour. They are oily, containing about 30-40% oil (Daniel, 1997). The tree sheds its pods which produce long lasting ground litter. The seeds germinate after the pod husks decay, several months after dropping (Morton, 1990).


Karanja is increasingly used for oil production due to its use in biodiesel. The oil was formerly used for lighting, as a raw material for soaps, varnishes and paints, to repel insects in storage instalments, and as a mosquito repellent (George et al., 2005;Wood et al., 2001). Karanja wood can be used for fuel, and the resulting ashes as a dyeing agent. Roots yield pinnatin, a dyeing pigment. The bark is fibrous and can be turned into rope. The leaves are potential sources of fodder. The fragrant flowers are a source of pollen and nectar from which bees produce dark honey. The karanja tree hosts lac insects and is valued as an ornamental plant. Many parts of the tree are used in ethnomedicine (Orwa et al., 2009Daniel, 1997). The karanja tree is an important species for afforestation. Leaf shedding produces large amounts of organic litter (Morton, 1990).

Oil extraction yields a press cake that can be used as a fertilizer or as animal feed for ruminants and poultry (Sreedevi et al., 2009Scott et al., 2008). However, its feeding value is disputed because of its bitterness and antinutritional factors (see Potential constraints on the "Nutritional aspects" tab) (Scott et al., 2008). Three main types of karanja oil cakes are available, namely rotary pressed, expeller pressed and solvent-extracted, the composition of which depending on the degree of decortication and the method of oil extraction (Dutta et al., 2012).


Karanja is a fast growing species occurring in the humid tropics from sea level up to 1200 m (not above 600 m in the Himalayan foothills). It requires an annual rainfall ranging from 500 to 2500 mm with a 2-6 months dry period, and temperatures between 1°C and 38°C. Karanja is highly tolerant of salinity and alkalinity, and can grow on seashores. However, saline soils are thought to reduce tree nodulation and growth (Murphy et al., 2012). Karanja does well on most soils (sandy, stony to clayey) at a pH ranging from 6 to 9. The karanja tree prefers well-drained soils with assured moisture for optimal growth, particularly at the early stages of growth (Murphy et al., 2012Sangwan et al., 2010Orwa et al., 2009).

Forage management 

Karanja can be sown or propagated by branch cuttings or root suckers (Daniel, 1997). Its growth is fairly slow in its early stages of development, and annual weed control is necessary during the first 3 years after planting (Orwa et al., 2009). It should be planted in blocks with 2 x 2 m or 5 x 5 m spacing. It produces profuse root suckers and is not suitable as an agroforestry species (Orwa et al., 2009). The karanja tree starts fruiting 4 to 7 years after planting and full production is achieved within 10 years (Murphy et al., 2012). It can produce 0-30 kg seeds per year. Planting at approximately 5 x 5 m intervals allows 350 tree/ha. At 20 kg/tree average yield, an annual production of 7 t of seeds can be expected (Murphy et al., 2012).

Environmental impact 

Afforestation and soil reclamation

Karanja is promoted as being able to produce oil on poor, degraded or saline soils (Murphy et al., 2012). Karanja trees have been used for soil reclamation around coal mines and revegetation in India (Maiti, 2012).

Soil binder and improver

In India, karanja has been traditionally used by villagers on slopy uplands to bind the soil (Kumar, 2004).

Nutritional aspects
Nutritional attributes 

Karanja oil cake has a low fibre content, the crude fibre level varying from 4 to 7% of DM. The protein content of expeller cake ranges from 24 to 30% of DM, and from 27 to 33% of DM for solvent-extracted cake. Karanja oil cakes obtained by rotary and expeller extraction have a higher content of residual oil (6-15% of DM), while solvent-extracted ones contain little oil (under 2% of DM). 

Potential constraints 

Karanjin and pongamol

Karanja oil cake, and particularly its lipid fraction, contains two furanoflavones, karanjin and pongamol, that are toxic and make the cake unpalatable. The oil also contains polyphenolic compounds. Many treatments have been tested since the 1970s to remove these undesirable substances, including physical treatments (autoclaving, pressure cooking, water soaking, water washing, solvent extraction and partial deoiling), chemical treatments (sodium hydroxide, lime, addition of binder) and microbiological treatments (S. cerevisiae and Aspergillus oryzae) (Scott et al., 2008; Nagalakshmi et al., 2011; Dutta et al., 2012). However, while some treatments such as deoiling or pressure cooking are more effective than others in reducing karanjin, none of them appeared to make the processed cake completely suitable, feasible and wholesome for animal feeding. Also, the cost of detoxification discourages the use of karanja oil cake (Nagalakshmi et al., 2011Dutta et al., 2012).

Protease inhibitors

Karanja oil cake contains protease inhibitors. Different methods have been assessed to reduce their activity, as summarized in the following table.

Table 1: effects of treatments on trypsin and chymotrypsin inhibitors of karanja (Rattansi et al., 1997):

Technological treatment Trypsin inhibitor reduction Chymotrypsin inhibitor reduction
Solvent-extraction 34% 15%
Autoclaving 83% 15%
Fermentation 33% 15%
Gamma radiation (1, 5, 10 and 50 kGy) 4-83% 3.7-78%
2.4% HCl 100% 100%

A more recent method consists in autoclaving the cake with lime, then refluxing with 2% HCl and finally neutralizing with NaOH. This treatment improves the protein digestibility of the cake (Scott et al., 2008).


Karanja oil cake

Most studies on the use of karanja for ruminants deal with oil cakes obtained after various oil extraction processes aiming at reducing or eliminating antinutritional factors. Based on results obtained on cattle, sheep and goats, it is recommended to limit karanja cake at 15-20% of the diet (DM basis as sole protein source, or partly replacing another oil meal). Beyond this level, karanja cake tends to decrease growth and milk production. The variability between results could be due to the oil extraction process undergone by the seed.


With crossbred dairy cows fed solvent-extracted karanja meal, partially (50%) replacing groundnut meal, there was no effect on DM intake, daily weight gain, milk production and milk composition. Higher inclusion rates decreased voluntary feed intake (Konwar et al., 1987b). Similar results were observed in bulls when solvent-extracted karanja oil meal replaced more than 50% groundnut meal in their straw-based diet (Konwar et al., 1987d). However, in bull calves, expeller karanja oil cake included at 4% in a concentrate diet depressed feed intake, caused histopathological changes in vital organs, and produced toxicity symptoms (Gupta et al., 1981 cited by Dutta et al., 2012).


In vivo DM and OM digestibilities of karanja cake were reported to be high (85% and 80% respectively) in adult ewes fed Brachiaria hay (Chandrasekaran et al., 1989).

All the experiments on feeding karanja oil cake to sheep have been conducted in India using male lambs at different ages and over long periods. Raw, expeller, solvent-extracted or detoxified karanja oil cake have been tested in lamb diets at levels ranging from 5 to 16% of DM. The main objective was the total or partial replacement of commercial oil meals (soybean, sunflower, groundnut or mustard). The main results are presented and summarized in the table below.

Table 2: Effects of the inclusion of karanja oil cake in lamb diets

Breed Experiment Substitution rate Dietary level Main results Reference
Native male lambs (13 kg, 6 m) Detoxified karanja cake replacing soybean meal (SBM) in a concentrate fed with finger millet straw at 3% BW 0, 50 or 75% of SBM 9-15% After 140 days, lambs fed with the 75% level weighted less than control lambs (18 vs. 24 kg). DMI of each component (concentrate and straw) were not different. Dineshkumar et al., 2013
Male lambs (13.7 kg, 6.5 m) Detoxified karanja cake replacing SBM in a concentrate fed with finger millet straw at 3% BW (concentrate:straw ratio 50:50) 0, 25, 50 or 75% of SBM 5-15% Over 140 days, total DMI decreased (from 685 to 511 g/d) with increasing karanja cake levels. OMD was lower with the 2 higher levels of karanja cake. Rumen parameters were not modified. Final BW was lower with the highest karanja cake level compared to 0 and 25% (17.8 vs. 24.4 kg). Rao et al., 2015
Native male lambs (10 kg, 6-7 m) Expeller karanja cake (24%) or solvent extracted karanja cake (20%) replacing 50% GNC in a concentrate fed (230 g/d) with oat hay 0 or 50% of GNC 9-10% Over 98 days, DMI of concentrate or hay was not different. DMD (53.5 vs. 59 %) and DWG were lower (49 vs. 60 g/d) with expeller treatment than with control or solvent treatment. Ravi et al., 2000
Native male lambs (10,5 kg, 6-7 m) Expeller karanja cake (24%) or solvent extracted karanja cake (20%) replacing GNC in a concentrate (275 g/d) fed with straw 0 or 50% of GNC 14-16% Over 255 days, DMI of concentrate or straw was not different. No effect on DMD but CP digestibility was lower with expeller karanja cake. Up to 5 week experiment had no effect on DWG then it is null in both expeller and solvent karanja cake. Singh et al., 2006
Native male lambs (13 kg, 3.5 m) Solvent extracted karanja cake treated with toxin binder (B-karanja cake), lime (L-karanja cake) or washed (W-karanja cake) replacing SBM in a concentrate fed with oat straw 0 or 50% of SBM 11-14% After 90 days feeding, BW was lower with B-karanja cake and L-karanja cake (17.3 vs. 23 kg). Concentrate DMI was also lower (184 vs. 278 g/d). Straw DMI was lower with the 3 solvent extracted karanja cakes than with the control (155 vs. 240 g/d). Digestibility was not modified. Soren et al., 2009a,Soren et al., 2009b
Nellore male lambs (28 kg, 9-10 m) Karanja cake compared to sunflower and GNC in a total mixed ration based on sorghum forage (40%) and concentrate (60%) 100% 12% Over 155 days, DMI (77 vs. 101 g/d) and DWG (28.3 vs. 79 g/d) were lower with karanja cake than with sunflower and GNC (75.3-82.7 g/d). OM digestibility was higher (70.4 vs. 60,5%) but CP digestibility was lower (53 vs. 59%). Nagalakshmi et al., 2011

CP: crude protein; DMD: dry matter digestibility; DMI: dry matter intake; DWG: daily weight gain; GNC: groundnut cake; OMD: organic matter digestibility; SBM: soybean meal

Karanja oil cake generally had no effect on diet intake, except in one study (Nagalakshmi et al., 2011). The DM or OM digestibility of the diet was either unchanged (Singh et al., 2006) or decreased with higher levels of karanja cake (Ravi et al., 2000Rao et al., 2015). In comparison with other oil meals, diet digestibility was higher with karanja cake (Nagalakshmi et al., 2011). Protein digestibility decreased with expeller karanja cake and was not improved with solvent-extracted karanja meal (Ravi et al., 2000; Singh et al., 2006; Nagalakshmi et al., 2011).

The lower DM intake and daily weight gain observed with karanja oil cake could be due to either tannin and/or trypsin inhibitors, which reduced nitrogen utilization but had no effects on rumen parameters. Daily weight gain was almost always lower with the highest levels of karanja cake whatever the treatment, except in some cases when the cake was solvent-extracted or washed (Ravi et al., 2000; Singh et al., 2006). Expeller or solvent-extracted karanja cake had a negative effect on carcass characteristics with heavier liver and kidneys, and lighter testis (Singh et al., 2006). Detoxified karanja cake had no effect on carcass characteristics or organ weight (Nagalakshmi et al., 2011; Rao et al., 2015), except for the liver that was lighter (271 vs. 325-371 g) at 15% karanja cake level (Rao et al., 2015). Blood parameters were not modified, except for urea content that was higher with expeller karanja cake (Ravi et al., 2001). Karanja cake depressed immunocompetence and had deleterious effects on the testicular function of ram lambs, probably due to the presence of the toxic compound karanjin (Dineshkumar et al., 2013Nagalakshmi et al., 2011).

Expeller karanja cake (Ravi et al., 2000Singh et al., 2006), and to some extent solvent-extracted karanja cake (Singh et al., 2006), are not recommended for young animals for long term feeding because they are more sensitive to the toxic compounds in the cake (Ravi et al., 2000). Washing solvent-extracted karanja cake seems to be an efficient way to remove residual karanjin. It is recommended to limit the amount of detoxified karanja cake in a concentrate at less than 18% of DM (about 10% of the total diet) when fed to young male lambs, in order to avoid reproduction and health problems (see Potential constraints above).


In growing kids (about 6 kg and 5-6 months), solvent-extracted karanja meal was fed for a long growing period (266 days) at up to 30% (replacing 45% of groundnut cake) of a concentrate fed at 200 g/d with maize forage, without marked effect on DM intake, DM digestibility and daily weight gain. At 40% of the concentrate, daily weight gain was lower (22 vs. 33-38 g/d) and DM and protein digestibilities also tended to be lower (Srivastava et al., 1990).

Karanja leaves

Karanja leaves are not liked by livestock. In Nigeria, karanja leaves were reported to be unpalatable (Wood et al., 2001). In steers, the DM digestibility of green fresh leaves and litter leaves of karanja were 37% and 34%, respectively, and were the lowest among 27 other tropical forage tree species (Lowry, 1995).



Karanja is toxic to broilers. Karanja seeds included at 5% in broiler diets led to a 50% reduction in the growth rate (Natanam et al., 1989c). This mainly comes from the lipid fraction, since 1% or 2% oil reduced performance as much as 10% to 20% expeller cake. Toxicity resulted in increased liver and pancreas weights (Natanam et al., 1989b). With raw expeller karanja cake, when fed at 10 or 20% of the diet, there was a marked reduction in broiler growth (Natanam et al., 1989b). 40% karanja cake led to a 100% mortality. Solvent extracted karanja cake also lowered growth performance and feed efficiency (Panda et al., 2008a; Panda et al., 2008b). The effect was observed with an inclusion level of 6%, and increased at higher levels of inclusion.

Several treatments have been tested for karanja detoxification. Simple treatments such as autoclaving or water soaking were inefficient (Natanam et al., 1989bNatanam et al., 1989c). Treatment with NaOH (1.5 to 3%) decreased the toxicity of karanja (Panda et al., 2008aPanda et al., 2008b; Panda et al., 2005a). However, even at low inclusion levels (about 6%), some reduction in performance was recorded over the whole growth period (Panda et al., 2005aPanda et al., 2008a). It is noticeable that expeller cake is still more harmful than solvent-extracted cake, even after alkali treatment (Panda et al., 2008a). Methionine addition could help alleviate the residual effects of detoxified karanja meal (Panda et al., 2008b).

It is concluded that it is not advisable to use karanja meal in broiler diets, even after detoxification.


The performance of layers markedly decreased with the addition of 10% karanja cake in the diet (Natanam et al., 1989a). The effect was higher with expeller cake than with solvent-extracted meal. However, in another study, the use of 10% detoxified solvent-extracted meal maintained laying performance and egg quality (Babu et al., 2015).


The addition of 6% solvent-extracted karanja meal in the diet of quails decreased growth performance (Krishna et al., 2013). The effect was attenuated by detoxification (with isopropyl alcohol) but there was still a reduction in growth.


Literature on the use of Millettia pinnata to feed rabbits does not seem to exist (October 2015). As karanja oil cake can be detoxified for limited use in ruminants and layers (see Ruminants and Poultry sections), karanja cake could be a potential feed resource for rabbit feeding. However, given the toxicity and unpalatability of the leaves and cake in other species, in vivo experiments are necessary before karanja products can be fed to rabbits.

Nutritional tables
Tables of chemical composition and nutritional value 

Avg: average or predicted value; SD: standard deviation; Min: minimum value; Max: maximum value; Nb: number of values (samples) used

Main analysis Unit Avg SD Min Max Nb  
Dry matter % as fed 86.6       1  
Crude protein % DM 26.6 2.5 24.3 29.7 5  
Crude fibre % DM 5.6 1.5 3.9 6.8 3  
NDF % DM 30.4 10.3 18.2 39.9 4  
ADF % DM 11.8 2.3 10.3 15.2 4  
Lignin % DM 2.9 0.8 2.0 3.5 3  
Ether extract % DM 11.0 4.0 6.4 14.5 5  
Ash % DM 4.9 0.9 3.9 6.1 6  
Gross energy MJ/kg DM 20.7         *
Minerals Unit Avg SD Min Max Nb  
Calcium g/kg DM 7.0 0.7 6.0 7.6 4  
Phosphorus g/kg DM 6.2 1.9 4.8 8.9 4  
Potassium g/kg DM 2.3       1  
Magnesium g/kg DM 2.4   2.0 2.7 2  
Manganese mg/kg DM 76       1  
Zinc mg/kg DM 199       1  
Copper mg/kg DM 12   2 22 2  
Iron mg/kg DM 23   14 31 2  
Amino acids Unit Avg SD Min Max Nb  
Alanine % protein 3.7   2.3 5.1 2  
Arginine % protein 4.5   4.3 4.6 2  
Aspartic acid % protein 8.6   6.2 10.9 2  
Cystine % protein 3.6   1.2 6.1 2  
Glutamic acid % protein 15.5   13.5 17.6 2  
Glycine % protein 3.6   3.4 3.9 2  
Histidine % protein 3.6   2.1 5.1 2  
Isoleucine % protein 4.8   3.6 6.0 2  
Leucine % protein 7.8   7.7 7.9 2  
Lysine % protein 4.5   3.3 5.6 2  
Methionine % protein 1.2   1.0 1.4 2  
Phenylalanine % protein 4.4   3.6 5.2 2  
Proline % protein 4.0   2.9 5.0 2  
Serine % protein 4.3   4.0 4.5 2  
Threonine % protein 3.4   2.6 4.1 2  
Tyrosine % protein 2.5   1.3 3.7 2  
Valine % protein 5.9   5.3 6.4 2  
Secondary metabolites Unit Avg SD Min Max Nb  
Tannins (eq. tannic acid) g/kg DM 26.4       1  
Ruminant nutritive values Unit Avg SD Min Max Nb  
Nitrogen digestibility, ruminants % 85.9       1  

The asterisk * indicates that the average value was obtained by an equation.


Chandrasekaran et al., 1989; Gowda et al., 2004; Kumar et al., 2007; Mandal et al., 1975; Nagalakshmi et al., 2011; Natanam et al., 1989; Ravi et al., 2000

Last updated on 15/09/2016 00:35:20

Main analysis Unit Avg SD Min Max Nb  
Crude protein % DM 30.8 2.9 26.9 33.2 4  
Crude fibre % DM 5.5       1  
NDF % DM 30.1 4.6 26.8 35.4 3  
ADF % DM 17.2 4.0 12.6 20.0 3  
Lignin % DM 3.9   3.6 4.2 2  
Ether extract % DM 0.9 0.6 0.2 1.7 4  
Ash % DM 6.0 0.6 5.4 6.9 4  
Gross energy MJ/kg DM 18.5         *
Minerals Unit Avg SD Min Max Nb  
Calcium g/kg DM 7.1 1.5 5.7 8.7 3  
Phosphorus g/kg DM 5.9 0.4 5.5 6.3 3  
Potassium g/kg DM 4.9       1  
Magnesium g/kg DM 2.0       1  
Manganese mg/kg DM 71       1  
Copper mg/kg DM 2       1  
Iron mg/kg DM 18       1  
Secondary metabolites Unit Avg SD Min Max Nb  
Tannins (eq. tannic acid) g/kg DM 34.1       1  

The asterisk * indicates that the average value was obtained by an equation.


Kumar et al., 2007; Natanam et al., 1989; Ravi et al., 2000; Soren et al., 2009

Last updated on 15/09/2016 00:36:18

Datasheet citation 

Heuzé V., Tran G., Delagarde R., Hassoun P., Bastianelli D., Lebas F., 2017. Karanja (Millettia pinnata). Feedipedia, a programme by INRA, CIRAD, AFZ and FAO. http://www.feedipedia.org/node/636 Last updated on May 10, 2017, 11:09

English correction by Tim Smith (Animal Science consultant) and Hélène Thiollet (AFZ)