Comm. Appl. Biol. Sci, Ghent University, 72/1, 2007 1
ROOT ARCHI TECTURE OF THE PROMI SI NG BI O-DI ESEL
PLANT JATROPHA

W.M.J. ACHTEN1, B. REUBENS1, W. MAES1, E. MATHI JS2, L. VERCHOT3, V.P.
SI NGH4, J. POESEN5 & B. MUYS1
1Division Forest, Nature and Landscape, K.U.Leuven, Celestijnenlaan 200 E box 2411, BE-3001
Leuven, Belgium
2Division Agricultutral and Food Economics, K.U.Leuven, de Croylaan 42, BE-3001 Leuven,
Belgium
3World Agroforestry Centre, Head Quarters, United Nations Avenue P.O. Box 30677-001000
Nairobi, Kenya
4World Agroforestry Centre, Regional office South Asia, Pusa Campus, New Delhi – 110 012,
India
5Physical and Regional geography Reasearch Group, K.U.Leuven, Celestijenlaan 200 E box
2409, BE-3001 Leuven, Belgium

I NTRODUCTI ON

Ja tropha cu rcas L. (Eu ph orb ia cea e) receives a lot of a t t en t ion a s a sou rce of
ren ewab le en ergy. Th e d rou gh t res is ta n t , oil bea r in g sm a ll t ree is well
a da p ted to t rop ica l s em i-a rid region s a n d m a rgin a l s ites . As a p ion eer s pe-
cies , or igin a t in g from Cen t ra l Am erica , bu t n ow growin g p a n trop ic, J . cu rcas
is ea s ily p ropa ga ted a n d ca n es tab lish qu ick ly in a wide va r iety of soils , a l-
th ou gh th e p la n t s u ffers im m ed ia tely from fros t a n d wa ter loggin g (Heller ,
1996 ). Th e J . cu rcas s eeds con ta in 30 -35% in ed ib le oil wh ich ca n be ea s ily
con verted to b io-d ies el th a t m eets th e Am erica n a n d Eu ropea n s t an da rds
(Aza m et a l., 20 05 ). Th e b io-d ies el p rodu ct ion ch a in a lso res u lt s in som e
va lu a b le by-p rod u cts (e.g. s eedca ke, fru it h u s ks , glycer in ) (fig. 1 ). Next to th e
b io-d iesel th e p la n t is a ls o u sefu l a s livin g fen ce. S in ce J . cu rcas is n ot
b rows ed , th e fen ces ca n be u s ed to p rotect food crop p rodu ct ion (Gu b it z et
a l., 1999 ). Fu r th er th e fen ces a re u sed for ecologica l res tora t ion of degra ded
a rea s (Za h awi, 2005 ). J . cu rcas cu lt iva t ion is a ls o rep orted to p reven t an d
con trol eros ion (Gu b itz e t a l. , 1999 ), a lth ou gh n ot ba s ed on s cien t ific da ta .
Th e p rotect ive role of vegeta t ion h a s been p roven in m a n y s tu d ies a n d es pe-
cia lly p la n t roots cou ld p lay a n im porta n t role in s lope s t a b iliza t ion a n d ero-
s ion con trol (Abe & Ziem er , 1999 ; Cou tts , 1983 ). Alth ou gh s t ill a h ypoth es is ,
J . cu rcas prob ab ly cou ld be u sed s im u lta n eou s ly for b io-d iesel p rodu ct ion a s
well a s eros ion con t rol in t rop ica l a rea s . In th is con text th ere is a n eed for
s cien t ific resea rch to bet ter u n ders ta n d th e effect s of th is s pecies a n d it s
roots on geom orph ologic p roces s es . In gen era l, a s is th e ca s e with m an y
p la n t s pecies , belowgrou n d s t ru ctu ra l a n d fu n ct ion a l developm en t of J . cu r-
cas h a s received very lit t le a t t en t ion so fa r , m a in ly a s a con sequ en ce of
m eth odologica l d ifficu lt ies (Gys sels & Poesen , 2003 ; Wa is el et a l. , 2002 ).
Never th eles s , in order to su s t a in ab ly gu ide th e boom in g u s e of th is p rom is -
in g s pecies , it is im p or ta n t to a lso fu lly u n d ers ta n d it s b elowgrou n d a s m u ch
a s it s a bovegrou n d fu n ct ion a l an d s t ru ctu ra l developm en t . Bes id es th e s ta -
b ilis in g role m en t ion ed b efore, a n oth er p r im a ry role of roots is off cou rs e th e
u p ta ke an d s tora ge of wa ter a n d n u t r ien ts (Gregory, 200 6 ). Su m m a ris ed , th e
m a in rea son s for s tu dyin g roots a re ecologica l s ign ifica n ce, resou rce ca p tu re
a n d a lloca t ion , ca rbon flow, p la n t in tera ct ion s , effect on soil s t ru ctu re, an -
ch ora ge, u s e of root p rodu cts , a s well a s s im p ly ba s ic b iologica l in form a t ion

2
on a p a rt of th e p lan t wh ich is of ph ys iologica l a n d d evelopm en ta l in teres t in
it s own r igh t (Sm it e t a l. , 2 000 ).
In th is com m u n ica t ion a n on goin g s tu dy is h igh ligh ted in wh ich a t ten t ion is
p r im a r ily pa id to th e root s t ru ctu ra l developm en t of J . cu rcas s eed lin gs , a n d
th e lin k of a rch itectu ra l ch a ra cter is t ics with eros ion con t rol poten t ia l. Du r-
in g th e la s t d eca d es , t h is rela t ion sh ip received m ore a tten t ion (Reu ben s et
a l., 2007 ). At th e m om en t of wr it in g on ly p relim in a ry res u lts a re a va ila b le.
Th e m a in a im of th e resea rch is to a s ses s th e eros ion con trol poten t ia l of th is
p la n t , a lth ou gh th e resu lts m igh t a ls o be u s efu l for oth er a pp lica t ion s , s u ch
a s determ in in g th e op t im a l a grofores t ry a n d p la n ta t ion la you t .

MATERI AL AND METHODS

Plant material and growth conditions

Th e resea rch is orga n ised in th e fa cu lty green h ou s e com plex of th e fa cu lty of
b io-s cien ce en gin eer in g, K.U.Leu ven , ca m p u s Aren berg. Da y t em pera tu re is
kep t b etween 22°C a n d 25°C. At n igh t th e tem pera tu re is a llowed to d rop t ill
15°C. Th e rela t ive h u m id ity is kep t on 70%. Th e CO 2 con cen tra t ion in th is
a r t ificia l en viron m en t va r ies a rou n d 60 0 ppm . In order to m ea s u re root a r -
ch itectu re a t d ifferen t a ges (i.c. 10 da ys a ft er germ in a t ion , 1 m on th a fter
germ in a t ion a n d 2 m on th s a fter germ in a t ion ), pot s with d ifferen t d im en s ion s
a re u s ed . Th e p lan t wh ich a re a llowed to grow 10 da ys a re sown in pots with
a d ia m eter of 24cm an d dep th of 21 cm , wh ile th e oth er p la n ts a re s own in
pot s with a d ia m eter of 41cm an d depth of 28cm . Th e p ots were filled with a

 

  fiff
– pea t m ixtu re. E th iop ia n J . cu rcas s eeds , ob ta in ed
from FACT-fu els , E in dh oven (th e Neth er la n ds ), were sown a fter n ick in g a n d
cold wa ter t rea tm en t . Soil wa s kept m ois t a n d two s eeds were sown in ea ch
pot . Aft er germ in a t ion sola r ra d ia t ion depen d in g d r ip irr iga t ion wa s in s ta lled .
Fer t iga t ion wa s a pp lied a fter th e s p rou t in g of th e s econ d lea f.

Measurement and analysis

Before des t ru ct ive m ea s u rem en t of th e root sys t em s t ru ctu re, a lim ited s et of
a bovegrou n d ch a ra cter is t ics is m ea s u red . Th es e in clu d e m ea s u rem en t of
sh oot len gth u s in g a m ea su r in g t a pe (a ccu ra cy 1m m ), d ia m eter a t th e ba se
with a d igit a l ca lliper (a ccu ra cy 0 .01m m ), n u m ber of lea ves a n d in d ivid u a l
lea f d im en s ion s , u s in g a d igita l ca lliper .
After rem ova l of th e a er ia l pa r t , root s a re exposed b y ca refu lly rem ovin g th e
soil a rou n d th em with a spa tu la , ta k in g ca re to m a in ta in or igin a l root pos i-
t ion . Root a rch itectu ra l m ea s u rem en ts a re perform ed by record in g th e d i-
a m eter with a d igita l ca lliper (a ccu ra cy 0 .01m m ) a n d th e XYZ coord in a tes
(a ccu ra cy 0 .5m m ) of every in d ividu a l root s egm en t with a fra m e con s is t in g of
m ovea b le ru lers in X, Y a n d Z d irect ion . By s h ift in g th es e ru lers to th e r igh t
pos it ion , coord in a tes of ea ch root s egm en t ca n be determ in ed Hen derson , et
a l., 19 83 ; Kh u der et a l., 200 6 ; Da n jon & Reu ben s , u n pu b lis h ed ). A root s eg-
m en t en ds wh en a root b ra n ch es or ch an ges or ien ta t ion or d iam eter . All root
s egm en ts u p to a lim it of 0 .5m m a t th e begin n in g of a n ew root axis a re
m ea s u red . Add it ion a l in form a t ion , s u ch a s th e p resen ce of fin e roots , growth
a lon g th e pot border or b rea ka ge of a root , is ca refu lly recorded .

Comm. Appl. Biol. Sci, Ghent University, 72/1, 2007 3
Fin a lly both a bove- a n d belowgrou n d p la n t m a ter ia l is weigh ed fresh a s well
a s a ft er 24 h ou rs of oven -dryin g a t 70°C.

Th e root sys tem s t ru ctu re is u lt im a tely rep resen ted a s a m u lt i-s ca le t ree
gra ph (MTG) (God in e t a l., 1999 ) a n d a n a lys ed in th e AMAPm od freewa re
(God in e t a l. , 19 99 ) a n d R open s ta t is t ica l pa cka ge (Ih a ka a n d Gen t lem a n ,
1996 ). AMAPm od in clu des a n u t ility to con trol th e s t ru ctu re of th e in pu t file,
n u m erou s bu ilt -in fu n ct ion s wh ich can be a s s em b led in u s er -m a de fu n c-
t ion s to com pu te pa ra m eters , a p rogra m to d is p la y 3D gra ph ics of va r iou s
elem en ts of th e root s ys tem a n d fu n ct ion s requ ired for a d eta iled a rch itec-
tu ra l exp lora t ion , an a lys is or m odellin g of root sys tem s (Da n jon et a l.,
1999 a ; Da n jon et a l. , 199 9b ; Da n jon & Reu ben s , u n pu b lis h ed ).

RESULTS AND DI SCUSSI ON

As th is res ea rch p roject h a s on ly very recen t ly been s et u p , res u lts rep re-
sen ted in th is com m u n ica t ion a re s t ill p relim in a ry, a n d ba sed on a very lim -
it ed da ta set . In th e n ext m on th s , deta iled m ea s u rem en ts will con t in u e a n d
fu ll a n a lys is of th e exten ded d a ta set will ta ke p la ce.

Never th eles s , som e in teres t in g fir s t obs erva t ion s h a ve been m ade. A firs t
s t ep in an a lys is is th e vis u a liza t ion of th e root s ys t em s t ru ctu re, a s rep re-
sen ted in figu re 1 . Mos t rem a rka b le obs erva t ion is t h e developm en t of a p re-
d icta b le root s t ru ctu re con s is t in g of on e cen tra l tap root a n d fou r la tera ls
(figu re 1 a n d figu re 2 ). Heller (1996 ) a s s u m ed th is con s is ten t root form a t ion
from s eed lin gs , b u t th e root form a t ion an d s t ru ctu re h a s n ever s ys t em a t i-
ca lly been des cr ibed . All of th es e la tera ls or igin a te a t on e an d th e s a m e pos i-
t ion of th e ta p root , a n d grow in fou r perpen d icu la r d irect ion s . A h igh n u m -
ber of fin e roots is a lso obs erved , a n d root b iom a s s in crea ses a t h igh s peed .

A la rge s et of qu a lita t ive root sys tem ch a ra cter is t ics ca n be ob ta in ed th rou gh
fu r th er a n a lys is in AMAPm od a n d R. A su b set of im porta n t a bove- a n d
belowgrou n d ch a ra cter is t ics is rep resen ted below in Ta b le 1 . Aga in , s om e
rem a rka b le fir s t con clu s ion s ca n be ta ken , s u ch a s e.g. th e low Root Pa r t i-
t ion in g Coefficien t (RPC) an d Root-Sh oot ra t io (R:S). An oth er in teres t in g ob -
serva t ion is th e very h igh m ois tu re con ten t both in root a n d sh oot , a s can be
der ived from th e d ifferen ce in fres h a n d d ry weigh t .

Alth ou gh m ore res ea rch s t ill h a s to b e u n dertak en , in clu d in g d ifferen t s ta ges
of develop m en t , va r iet ies a n d growin g s u bs t ra t es , th e obs erved 3D s t ru ctu re
revea ls a poten t ia l for con t rol of d ifferen t types of eros ion a n d s lop e s ta b ilit y.
It h a s been h ypoth es ized th a t in gen era l woody sp ecies with a la rge, deep
a n d s t ron g cen tra l pa rt of th e root sys tem h a vin g som e r igid ver t ica l roots
pen et ra t in g deep ly in to th e s oil a n d an ch or in g in to firm s t ra ta , a s well a s a
la rge n u m ber of fin er roots n u m erou s ly b ra n ch in g from th e m a in la tera l
roots , wou ld be m os t effect ive to in crea se sh a llow s lope s tab ility (Reu ben s et
a l., 2 007 ).

4
Table 1 : Set of a bove a n d belowgrou n d ch a ra cter is t ics of 10-da ys -old J. curcas s eed-
lin gs .
Variable (u n it ) Ave rage ± s .d.
Heigh t (cm ) 22 .1 8 ± 4 .70
N° leaves 2 .00 ± 0 .00
Dba s e 1 (m m ) 6 .32 ± 0 .35
Fresh weigh t S tem (g) 6 .93 ± 1 .13
Fresh weigh t Root (g) 1 .33 ± 0 .20
Dry weigh t S tem (g) 0 .62 ± 0 .09
Dry weigh t Root (g) 0 .12 ± 0 .01
Stem den s ity (g/ cm ³) 0 .16 ± 0 .03
Root den s ity (g/ cm ³) 0 .08 ± 0 .00
R:S 2 0 .20 ± 0 .02
CSD_2 n d 3 (cm ) 0 .08 ± 0 .00
RPC4 16 .4 8 ± 1 .40
Tota l root len gth (cm ) 104 .97 ± 10 .34
In dividu a l root len gth (cm ) 21 .0 3 ± 2 .05
Tota l root volu m e (cm 3) 1 .46 ± 0 .11
Tota l s tem volu m e (cm 3) 3 .98 ± 1 .10
In dividu a l root volu m e (cm 3) 0 .29 ± 0 .02
In dividu a l root volu m e (n o t a p 5) (cm 3) O.10 ± 0 .02
SRL6 (cm cm -3) 71 .1 3 ± 5 .09
1
Dba s e = dia m eter a t th e ba se of th e s tem
2
R:S = (Mroot / Msh oot) with M = m a s s
3
CSD_2 n d = Cr os s -Sect ion a l Dia m eter of 2 n d or der roots
4
RPC = (Vroot / (Vroot + Vs tu m p + Vstem)) with V = volu m e (Dan jon e t al. , 2 005)
5
n o ta p=with ou t ta kin g t a p root in to a ccou n t
6
SRL= S pecific Root Len gth = (tot a l len gth / tota l volu m e).

For soil eros ion p roces s es by wa ter opera t in g in th e topsoil, s pecies colon iz-
in g th e u pper soil h or izon s with a fa s t -growin g a n d h igh ly b ra n ch in g root in g
sys tem wou ld h a ve th e h igh es t ca pa city to h old soil pa r t icles togeth er (Reu -
ben s e t a l. , 200 7 ). At fir s t s igh t , J . cu rcas s eem s to be p rom is in g for both
s itu a t ion s . It s fa s t growth is a n oth er im por ta n t fa vorab le fa ct . However ,
som e oth er ch a ra cter is t ics , su ch a s th e low den s ity, R:S ra t io an d RPC, s eem
to h a ve a ra th er a dvers e effect on eros ion con trol poten t ia l. At lea s t a t s eed -
lin g s ta ge J .curcas root s a re very fra gile. Th is rem in ds u s of th e fa ct th a t n ot
on ly s t ru ctu ra l b u t a ls o m ech a n ica l a spects sh ou ld be ta ken in to a ccou n t .

Figure 1 : Visu a lisa t ion of root s t ru ctu re of fou r 10 -da ys -old J . curcas s eed lin gs
grown in green h ou se con d it ion s . Th e r oot con figu ra t ion of a s in gle ta p root (b la ck ) a n d
fou r perpen dicu la r la tera ls (b leu ) is con s is ten t between differen t s eedlin gs .

Comm. Appl. Biol. Sci, Ghent University, 72/1, 2007 5

Figure 2 : Root form a t ion du r in g germ in a t ion of J . curcas .

CONCLUSI ON

Alth ou gh it is too ea r ly to d ra w con clu s ion s , th e root s t ru ctu re of th e you n g
J . cu rcas s h ows to be very in teres t in g. Th e root s ys t em , con s is t in g of on e
ta p root a n d fou r perpen d icu la r or ien ted la tera ls , is rem a rka b ly p red ict ab le
a n d is t h ou gh t to h a ve p rom is in g eros ion con t rol poten t ia l. Oth er ch a ra cter -
is t ics h owever s eem to be les s fa vou ra b le. In ord er to t a ke th e r igh t con clu -
s ion s m ore resea rch is n eces s a ry. On -field in ves t iga t ion of th e root s t ru ctu re
of m a tu re in d ividu a ls a s well a s m ea su r in g root ten s ile or ben d in g s t ren gth
is n eeded . Deta iled in form a t ion on th e root in g s ys t em will ela bora te th e op -
t im iza t ion of a grofores t ry a n d p la n ta t ion s ys tem s .

ACKNOWLEDGMENTS

Th is res ea rch is fu n ded by th e Flem is h In teru n ivers ity Cou n cil – Un ivers ity
Developm en t Co-opera t ion (VLIR-UDC) a n d th e K.U.Leu ven Res ea rch Fu n d,
a n d is a colla bora t ion between K.U.Leu ven a n d th e World Agrofores t ry Cen -
t re (ICRAF). Ywe J a n Fran ken of FACT-fu els is k in d ly a ckn owledged for th e
p rovis ion of th e s eed m a ter ia l. We a lso wa n t to th a n k Frédér ic Dan jon (INRA-
EPh ys E, Bordea u x) for th e tech n ica l s u pport rela t ed to th e AMAPm od an d R
softwa re.

REFERENCES

ABE K. & ZIEMER R.R. (1991). E ffect of t ree roots on a sh ea r Zon e - m odelin g rein forced
s h ea r -s t res s . Ca n a dia n J ou rn a l of Fores t Resea rch , 2 1 :1012-1019 .
AZAM M.M., WARIS A. & NAHAR N.M. (2005). Pros pects a n d poten t ia l of fa t ty a cid m eth yl
es ters of s om e n on -tra d it ion a l s eed oils for u se a s b iod iesel in In d ia . Biom a ss &
Bioen er gy, 2 9 :293-302 .
COUTTS M.P. (1983). Root a rch itectu re a n d t ree s ta bility. Pla n t a n d S oil, 7 1 :171-188 .
DANJ ON F., BERT D., GODIN C. & TRICHET P. (1 999a ). S tr u ctu ra l root a rch itectu re of 5-
Yea r -Old Pinus Pinas ter m easu red by 3d d igit is in g a n d a n a lysed with Am a pm od .
Pla n t a n d S oil, 2 1 7 :49-63 .
DANJ ON F., S INOQUET H., GODIN C., COLIN F. & DREXHAGE M. (1999 b). Ch ar a cter isa t ion
of s t ru ctu r a l t ree root a rch itectu re u s in g 3d d igit is in g a n d Am a pm od s oftwa re.
Pla n t a n d S oil, 2 1 1 :241-258 .
GODIN C., COSTES E. & S INOQUET H. (1999). A m eth od for descr ib in g p la n t a rch itectu re
wh ich in tegra tes topology a n d geom etry. An n a ls of Bota n y, 8 4 :343-357.
GREGORY P. (2006). Pla n t r oot s : growth , a ct ivity a n d in tera ct ion with s oils . Oxford , UK,
Bla ckwell Pu b lish in g, pp. 31 8.
GUBITZ G.M., MITTELBACH M. & TRABI M. (1999). Exp loita t ion of th e t rop ica l oil s eed
pla n t J atroph a curcas L. Bioresou rce Tech n ology, 6 7 :73 -82.

6
GYSSELS G. & POESEN J . (2003). Th e im por ta n ce of p la n t root ch a ra cter is t ics in con tr ol-
lin g con cen t ra ted flow eros ion ra tes . Ea rth Su r fa ce Proces ses a n d Lan dform s , 2 8 :
371-38 4.
HELLER J . (1996). Ph ys ic n u t . J atroph a curcas L. Prom ot in g th e con serva t ion an d u se
of u n deru t ilized an d n eglected crops . 1 . Ph D d is ser ta t ion , In s titu te of Pla n t Gen et ic
a n d Crop Pla n t Resea rch , Ga ters leben , Germ a n y & In tern a t ion a l Pla n t Gen et ic Re-
s ou rce In s t itu te, Rom e, Ita ly, pp . 66 .
h t tp :/ / www.ipgr i.cgia r .org/ Pu b lica t ion s / pdf/ 161 .pdf
HENDERSON R., FORD E.D., RENSHAW E. & DEANS J .D. (1983). Morph ology of th e s t ru c-
tu r a l r oot -s ys tem of S itka S pru ce - 1 . An a lys is a n d qu an t ita t ive des cr ipt ion . For -
es t ry, 5 6 :121-135.
KHUDER H., DANJ ON F., STOKES S. & FOURCAUD T. (2006). Growth res pon se a n d r oot
a r ch itectu re of Black locu s t s eed lin gs growin g on s lopes a n d su b jected to m e-
ch a n ica l per tu r ba t ion . 5 th Pla n t Biom ech a n ics Con feren ce.
REUBENS B., POESEN J ., DANJ ON F., GEUDENS G. & MUYS B. (2007). Th e role of fin e a n d
coa rse roots in sh a llow s lope s t a b ilit y a n d s oil eros ion con trol with a focu s on r oot
s ys tem ar ch itectu re: a review. Tr ees , DOI 10 .107/ s 00 468-00 7-0132-4
SMIT A L, BENGHOUGH A G, ENGELS C, VAN NOORDWIJ K M & PELLERIN S. (2000). Root
m eth ods : a h a n dbook. Ber lin Heidelberg New York , Spr in ger , pp . 587 .
WAISEL Y, ESHEL A & KAFKAFI U. (2002) Pla n t root s : th e h idden h a lf. New Yor k - Basel,
Ma rcel Dekker In c., pp. 1120 .
ZAHAWI R.A. (2005). Es ta b lis h m en t an d growth of livin g fen ce species : a n over looked
tool for th e res tora t ion of degra ded a rea s in th e t rop ics . Res tora t ion Ecology,
1 3 :92-102.