COLLECTION OF HEVEA MATERIALS FROM RONDONIA TERRITORY
...• •. 1
IN BRAZIL: A PRELIMINARY STUDY
2 PAULO DE SOUZA GONÇALVES
Seven characteristics were studies in 27 selected wild mother trees at Ouro Preto, Rondônia Territory namely, yield, circumference, bark tickness, total number of latex vessel, density of latex vessels per 5 mi 1imeter per ring, and average distance between consecutive latex vessel rings. AlI the trees were also evaluated for the degree of incidence of South American Leaf Bl ight (SALB). Great variability was observed among the individual selected rubber trees growing under similar envirommental conditions. Linear correlations
among the quantitative characteristics were estimated. Positive co r re 1a tion s were found between yield and circumfer~nce, bar~ thickness, total number of latex vessel rings and density of latex vessels per 5mm per ring. No relationship was found between yield and diameter of the latex vessels and between yield and average distance between consecutive latex vessel rings. Regression analysis of yield on other characteristics showed the regression coefficient to be highly significant to circumference, bark thickness, total number of latex vessel rings and density of latex vessel per 5mm per ring.
coefficient, genetic variation, selecti,
Index terms - Hevea spp., rubber tree, linear correlation, regression Collection of Hevea
o
FL-FOL10~1
1111111111111111111111 11111 1111111111111111111111111111111111111111 11111111 CPAA-11074-1
lThis research has been supported by SUDHEVEA/EMBRAPA agreement. .~ ••••••••••
2A .
gronomlst M.Sc., in Tree Improvement, Centro Nacional de Pesquisa da
FOL 1051
Se r in quei ra (CNPSe) - EMBRAPA, P.O.Box 319, CEP 69.000 - Manaus - Amazonas - Brazil.
RESUMO - COLEÇÃO DE MATERIAL DE SERINGUEIRA SELECIONADO
..
NO TERRIT6RIO DE '-'RONDÔNIA - ESTUDO PRELIMINAR.
RESUMO - Foram estudadas sete características de 27 matrizes selecionadas em Ouro
Preto, no Território Federal de Rondônia. Produção, circunferência, espessura de
casca, nGmero total de an~is de vasos laticíferos em Smm do anel, e distância
média entre os an~is consecutivos de vasos laticíferos foram determinados nas se
leções. Todas as árvores foram aval iadas para o grau de incidência do mal sul-ame
ricano das folhas. Grande variabil idade foi observada entre os indivíduos selecio
nados, vegetando sob as mesmas condições ambientais. Correlações lineares ent re
as características quantitativas foram estimadas. Correlações positivas foram en
- -";,..- ..c. -- .. ,:- -•.'
contradas entre produçao e circunferencia do tronco, espessura de casca, "umer
-
o total de anéis de vasos laticíferos e densidade dos vasos laticíferos em Smm doanel. Baixa correlação foi encontrada entre a produção com diâmetro dos vasos la
ticíferos e produção com distância m~dia entre os anéis consecutivos de vasos la
ticíferos. A anál ise de regressão de produção e os valores dos cara~teres quantl
tativos estudados mostram coeficientes de regressao altamente sig~ificativos p~
ra circunferência, numero total de an~is de vasos laticíferos e densidade de va
50S laticíferos em Smm do anel.
Tenmos para indexação - Seringueira,Hevea sp, correlação 1inear, coeficiente de
.
.
.
'
.
INTRODUCTION
".J
Conventional breeding takes a long time to provide a few clones with acceptable
all-round characteristics for large scale commercial planting. This is because the perennial nature of the erop impose long testing eyeles. In addition, multiple
selection for yield and seeondary eharacters whieh are often unfavourably assoei
ated wi th yield, reduces eonsiderably the chances of producing good all-round elo
nes (Ho 1972).
The first botanical expedition with the purpose of collecting botanical material
of native rubber trees
(H
ev~
spp) was held in 1945 in the Rondônia Territory. Thecollected seeds were planted in 84 hectares but these were destroyed by fire in
1950. Later in 1962, a new expedition was designated to recol1ect materiaIs from
the same treesin the same area (Moraes 1963). Ten years later in 1972, a further
expedition was carried out and a
Hev~
breeding programme based on the wild mothertrees in jungle was initiated. From the results of previous expeditions, the Acre
and Rondônia areas have shown the presence of high yielding vigorous germplasm (Pe
reira 1972, Gonçalves
e
t
alo 1973, Viegas & Gonçalves 1974, Gonçalves 1978, Gonça~ ves 1979).The main objective of wild mother tree selection is to supplement the breeding
programme by provinding a wider range of good all-round clones in a short time.
Data from the most recent eolleetion have been subjected to analysis (mainly the
various parameters eonnected with yield and vig~r) and the preliminary findings are
reported in this paper.
MATERIALS ANO METHODS
Selection of high yielding wild mother trees
Simmonds 1969 emphasised that multiplicity of characters selected always either
demanded large populations or caused a weakened selection for individual charac
ters. The leveI of operation to cope with the former alternative is not pratical in a conventional breeding programme. However, by resorting to commercial and other advanced generation seedl ings, it is feasible to screen large populations by mass selection.
A particularly conducive situation favouring such an approach in Brazi1 is the
ready availabi1 ity of large areas of hi~h yielding wiíd trees.
In the pre-selection of mother-trees in the fie1d, the basic objective was to
identify the best yie1ding phenotypes for further yield recording and observation .
The fLndings from the pre1iminary investigations by Gonçalves 1978, have been used
in the processo For the pre-selection of the wi1d mother-trees, the tappers (Figure
1) were contacted for location of the high yie1ding trees giving more than 1000 cc
of 1atex per tapping. This information was verified and confirmed. Freedom frorn
!vU..CJtoc.tjc.1.w.,u.i.e.i and Phtj;toph:thoJta s p.1eaf fa11 was cheked vi sua11y (Figure 2). Bud
woods from orthotropic branches were then co11ected, the cut ends waxed and stored
moist saw-dust and sent to National Rubber Research Centre at Manaus and
grafting.
Yie1d, number and heigh of panel, circumference at 1 .5m height and degree of
disease incidence were recorded for each se1ected tree. Bark samples were taken for
recording the tickness, total number of 1atex vesse1 rings, diameter of 1atex ves
se1s, density of 1atex vessels per 5 mil imeter per ring and average distance be
tween consecutive 1atex vesse1 rings based on a11 rings. The sample were taken at
precise1y 1.5m above the soi 1 surf~ce (Tab1e
3).
B~dwood from 27 trees was then co11ected.
Bark Characteristics Measurements
The fo11owing bark characteristics were measured in the bark sample:
Bark tlíickness;
Total number of latex vesse1 rings was determined by examining the
longitudinal sections of the bark;
radial
Diameter of 1atex vesse1s was observed in the tranverse section of the
bark;
Density of latex vesse1s per 5mm per ring determined by the average density
·,
.
u)Average distance between consecutive latex ves?el rings was determined based
on a l l rings. 'J •
Estimates of correlation between the seven characters were computed from data of
selected individuaIs.
RESULTS AND DISCUSSION
Variation of the Sampled Trees
•
•
Great variability was observed arno n q the individual selected rubber trees of
differentf ages growing under the same environmental conditions. (Table 3).
Ule (1905) and Baldwin (1974) have mentioned that very high yielding trees are
found in the Ouro Preto area of Rondonia Territory of Brazi 1 (Figure 3).
The morphological characters of the trees studied followed the description of
~~ •J ••• •••.••• O"
Heve.a.
bJtaf.,
,
éLéelU
,
as gu
,
iven by Seibert (1974). It was noted that the majority oftrees in this area, including those collected, had purple cortical tissue (Figure 4),
but the latex of alI trees was white. A rough estimate of the age of the trees based
on the girth showed a variation from 20 to 300 years. Total number of latex vessel
rings per tree ranged from 19 to 7l. Bark thickness ranged from 0.61 to 2.42cm in
the trees (Table 1). Part of the observed differences in the latter characteristics
was associated with the difference in tree circumference. Number of panels was also
associated with the magnitude of circunference (Table 2).
Disease Incidence
The
Hevea
tree defol iates and renews its foI iage annually and this phenomenon iscommonly known as wintering. UnI ike the trees in temperate countries where
defoliation occurs during the winter period of the year, in
Hevea
b4aD~elU
,
u,
whichgrows in the tropics this takes place during the drier months of the year .
...=.=---=
-,
.'." ,
In addition, the incidence and severity of infection of the disease is strongly
influenced by the prevail ing clirnatic c.,pnditions duriaq the wintering time.
Regions having a high and well distributed rain falI with no pronounced dry season
are observed to have severe disease incidence while areas receiving moderate rainfall but without a long dry season may have disease incidence intermediate
in severity. Therefore, the prospected area which showed a extended drought· for
about months, receiving ponderate intensities of rainfall, experienced only low
incidence of MiCJ1.oc.yc1.u...6 u1.eJ.. and Phy:toph-t.hoJUl. sp. in the fo l iage.
~EMBRAPA
CG)
C N P 5 o~ BIBLIOTECA
Yield-Bark Characteristics Relationship
Correlation studies for circumference, bark thickness, total number of latex
ve~sel rings and diameter of the latex vessels in the collected material were in
good agreement with the findings of various workers (Wycherley 1969, Narayanan
e
t
alo 1973; Ashp1ant 1928) with clonal rubber tree selections. Linear corre1ations
between yield and bark characteristics such as thickness (r=0.7290**) , total num
ber of latex vessel rings (r=0.5l39**) and density of latex vessels per 5 mil li
meters(r=0.5397**) , were significant at the 0.01 leveI (Table
4).
Thus 53 percent of the total variation was associated with bark thickness. The magnitude of thecorre1ation (r=0.7290) for bark thickness is in close agreement with the r=0.8089 reported by Gonçailves
e
t
al. (1980), who used one year old clone.. A1though Brookson (1959) showed that was positive correlation between the yields of individual mature seedl ings and c10nes derived frQm them, Wycher1ey (1969) ,pointed out that this evidence was inadequate.A similar situation may also exist for thecollected material in this study. In view of the above, it is still necessary to
repeat the test of the budgrafts made from these collections for further selec
tion. It would be interesting to see whether col1ections from different geographic
areas of Brazil show different behaviour.
Yie1ds of alI selected rubber trees were regressed on their bark characteristi~.
Table 5 shows the simp1e 1 inear regression equations of yie1d (y) on the six
variables. The analysis indicated that the slope of the regression I ine was
significant at the 0.01 leve l (F=28.36**) for bark thickness, indicating that
bark thickness is closely related to yie1d, independent of the age of the tree.
I•• ,
A similar situation possibly existed for total number of latex vessel rings and
density of latex vessels per 5mm p~r r-jng. For both cnaracteristics the slope of
regression lines (F=8.97**) and (F=IO.27**) respectively (Figure 5) indicated that
were closely related to yield of the trees.
CONCLUSIONS
Hevea breeders recognise the difficulty of producing good all-round clones
quickly. Fresh approaches have therefore been sought. One of these is by m
other-tree selection from populations of wild rubber trees. The main altractions of
such an approach are as follows:
No contraI led crossing is necessary;
No capital outlay for land is required as these populations are in the
native environment.
The early results from the wild--mother-tree selection are promising.
Positive and significative correlations were foundbetween yield and
circumference, yield and bark thickness, yield and total number of latex
vessel rings, and yield and density of latex vessels per 5mm per ring.
No relationship was found between yield and diameter of-the latex vessels
and between yield and average distance between consecutive latex vessel
rings.
_ Regression analysis of yield on other characters showed the regression
coefficient to be highly significant for circumference, bark tickness
total number of latex vessel rings and the density of latex vessel per 5mm
per ring. The regression coefficient of diameter of the latex vessels and
average distance between consecutive latex vessels rings was nor signii
icanto
There is need to repeat the tests of budgrafts made from this collection and
it would be of interest to see whether latex collections in different geographic
areas of Brazil differ in behavior.
-ACKNOWLWDGEMENTS
The autor is grateful to Mr. Afonso C.C. Valois, the Head of the Scientific
Staff of CNPSe for his valuable suggestions. The assistence of Mr. Antonio Pe~ soa Rebell0 in the determination of bark characters is grateful ly acknowledged.
Acknowledgements are also due to Misses Maria E.C. Vasconcellos and Maria A.C.
Neves for statistics work and designs.
REFERENCES
ASHPLANT, H. Investigations into He.ve.a anatomy. Bu11 Rubb.Grow., 10(7): 484. 1928.
BALDWIN, Jr. J.I. He.ve.a: A fi rst interpretation. J.Heredit., lQ(2): 54-64. 1947 .
.
BROOKSON, C.W. Seed1ings and c lones of Heve.a bJta6..ilienó.i6. In: WORLD CONGRESS OF
AGRICULTURAL RESEARCH, 1., Rome, 1959. Proceedings. Rome, s.ed., 1959. p.699.
1959.
GONÇALVES, P.de S. Seleção e coleta de seringueiras nativas
à
margem dos rios Mamo• ré, Guaporé e são Migue1 - Território Federal de Rondônia (Report). Manaus
EMBRAPA-CNPSe, 1978. 43p.
GONÇALVES, P. de S. Seleção e coleta de seringueiras nativas da região de Ouro Pre
ro - Território Federal de Rondônia (Report), Manaus, EMBRAPA-CNPSe, 1979. 44p.
GONÇALVES, P. de S.; MATOS, A.P.; MULLER, M.W. & VIEGAS, I.J.M. II coleta de mate
ria1 nativo de alta produção em seringais do Estado do Acre e Território Federal
de Rondônia (Report). Be1ém, IPEAN, 1973. 24p .
••• • J •
GONÇALVES, P. de S.; VASCONCELLOS, M.E.C.; VALOIS, A.C.C. & BARCELOS, E.S.
Herdabi-1idade, correlações genéticas e fenotípicas de algumas características de c10nes
de seringueira (He.ve.a spp). Pesq.Agropec.Bras., Brasí1ia, 15(2): 129-36, abro
1980.
HO, C.Y. Investigation on shortening the geverative cyc1e for yie1d improvement in
He.ve.a bJta6~e.nó~. New York, Corne11 University, 1972. Tese Mestrado.
MORAES, V.H.F. Seleção em seringais nativos Circ.IPEAN, Belern (])~ 1-32, 1963.
NARAYANAN,-R=;_~OMEZ, J.B. & CHEN, L.T. Some structura1 factors affecting the pr~
ductivity of He.ve.a bJta6~e.nó.i6. II Corre1ation studies between structura1 fac
tors and yie1d. J.Rubb.Res.lnst.Ma1aya, Q(4): 285, 1973.
PEREIRA,_J. da P. Coleta de material silvestre de alta produção em seringais nati
vos do Acre e Estação de Porto Velho (Re po rt }. Belém, IPEAN, 1972. 9p.
SEIBERT, R.J. A study of He.ve.a (with it's economic as pe cts ) in the Repub1ic of
SI MMONDS, N.VI.
1, 1969.
Genetica1 bases of p1ant breeding. J.Rubb.Res.lnst.Ma1aya, 21(1):
•
II
ULE, E. Die Kautschukpf1anzen der Amazonas Expedition und ihre Bedeutung fur die
Pflanzengeogr phie. l n: Eng1er's Botan. Jahr bliche r,
19
05.
VIEGAS, I. de J.M. & GONÇALVES, P. de S. 111 coleta de material nativo de alta
produção em seringais do Estado do Acre e Território Federal de Rondônia (R~
po rt }. Be1ém, IPEAN, 1974. 46p.
WYCHERLEY, P.R. Breeding in He.vea... J.Rubb.Res.lnst.Ma1aya, 21(1): 38-55,1969.
-
~
.".-, I
"
Il
d
l
l
l:
!,
'l
'i
Ilj)!:'
:'
i
i
'
I
'
,.
•
TABLE 1 - Overal 1 means, standard deviation and coefficients of variation of 27 mother trees colected from Preto, Federal Territory of Rondônia. (Brazi 1). Manaus, 1980.
I
,
"
11;
Characteri st ic Mean Range Units Standard
deviation
Coefficient of variation (%)
Yield/tapping/tree C ircumferen ce Bark thickness
Total number of latex vessel rings Diameter of latex vessel
Density of latex vessels per 5mm per ri ng
Average distance between consecutive latex vessel rings 2.85 0.50 - 6.00 4.74 0.61 - 4.13 1.49 : 0.61 - 2.42 36.40 19.00 -71.00 44.18 18.30 -24.40 375.38 208.84-521 .32 1itre 1.33 46.7 2.55 53.8 0.38 25.5 10.34 28. Lf 23.40 52.9 98.20 26.2 m mn mm 289.38
i06.
10-482. 80 mm 68.71 23.7·TABLE 2 - Number of trees and their yield, circumfe,rence range and number of panels intherrother tr~ék collected in Ouro Preto, Federal Territory of Rondônia. (Br azil) Manaus 1980.
I
i
l
III
N~ of trees Yie1d (1 itre) Ci rcumference Range '(rretr e s ) N~ of panels in exploration
0.5 0,61 2 1.0 0.85-1.71 1, 3
3
1.5 1.50-2. 16 1, 4 e 6 (3
2.0 1.88-2. 18 3, 3 e 33
2.5 1.46-2.85 2, 3 e 5 8 3.0 I 1.78-3·70 3, 3, 3, 4, 4, 5, 5 e 6 3·5 3.80 7•
2 4.0 2.29-2.60 5 e 5 3 5.0 2.65-4. 13 4, 6 e 7 6.0 3.47 8.
'
llhe col'ected rrothertrees were cloned at National Rubber Research Centre as CNS-RO 7901 to CNS-RO 7927
,.
TABLE 4 - Linear corr1e1ation coefficients between yie1d and the varlous structural parameters of the
27 mother trees co11ected in Ouro Preto, Territory of Rondônia. (Brazi 1) Manaus 1980.
I!:Ij '1111 I
I~
I.
h
"'\11,"1 , Ili
I!
,I
I' • I Parameteres 51mb,q1 Y C B T.V.R. D.L.V._
I
.
D.V.R. A.D.e.V.I
Yie1d/t/t Yi
1 0.7304** 0.7290** 0.5139** 0.3464 n.s. Bark thickness B 0.6765** 0.4302** 0.5397** 0.4536** 0.6585** 0.3355 n.s. C! rcumference C 0.7894** 0.4380* 0.4274** 0.4508** 0.5002** Total number of 1atexves sel rings T. V. R. 0.2721 n.s. 0.9623** O •1155 n. s .
Diameter of 1atex vessels D. L. V.
o .
3691 n. s . ",O. 1 155 n. s .Density of latex vessels
per 5 (mm) per ring O.V. R. 0.0534 n.s.
Average distance between consecutive 1atex vessels per ring
I,
A. D.e. V.
'.
(*) Diference significant at 5% leve1 of probabi1ity li.
(**) Diference significant at 1% 1evel of probabi 1ity
I
II "li"
1
1
1
1
1\
TABLE 5 - Linear regression equatio
n
s
bet
w
een y
i
eld and various
structura
1
!
I
F
haracters of the 27
m
o
t
h
e
r
tr
ee
s
col
1
ec
te
d
from Ouro Preto,
Fed
era
l
i
0
e
r
ritory of Rondônia
(B
r
azi
1
) - Manaus, AM. 1980
Regression of yield
(y) on
yRegression Coef
f
i
c
i
ent
I
" (e) (B)(T.V.R.)
(O.V.L.)
1.059 +
o.
46 1 (e)
0.461**
-0
.
940 + 2.540
(B)
2.540
*
*
o .
576 +0. 063 (T
.
V
.
R.)
0
:
063**
(3.724 + 0.297 (O.L.V.)
0
.
297 n. s.
-0.040 + 0
.
007 (O.V.R.)
0.007**
;.0.08
1 +
0.006 (A.o.e
.
v
.
)
0.006 n.s.
•
,
(
c
.
v.
R.)(A.o.e.v.)
',I
(
**
)
Diff
erence
s
igni
f
icant
at 1
%
.
l
e
ve1
,
of probabi lity
.
.
I••
.'
Figure 1. Natíve tapper carrying out' tapping
amecí wi1th a rifle and a kerosene head 1ight. REVISTA ,PA!5 l~AB.
Nº
_.1!
i
.
c
;
(fr
~G:C._M~~.:q
-.
J!
L
...
.
.
·
.
I. )r
.
..•
.
' ~EMBRAPA ~C N P S o BIBLIOTECA , I, , "_ ..~
-
--Figure 2
.
Víew of
"
the
tree Leaves wiít.fuJlight
incídenc~
aí M
.
ulei.
,:
~
I
I
í-'"". •
-
-
----_
.
_
-
-
-
_
...,...---,---, .' •.•.••• 0_. GROSSO STATEplrnRE 3
-Localizat:ÍlolilJ of OUro Prêt0\ where the tTees- were col.Iec ted :iilJll the Territary of Rondôni.a .'c,
- ,~
,,,,.
Fig:w:re 4 - Hev<}-a. b.1/:J::U·ilie}'f;~-.iA· wi th purp] e
ba
rk
est.irnatedto
be 60years
old frorn Ouro Pret
o
.
-0040+ 0.007x
-
3.724 +0.297x y y 6 r 0.5163 6 0.3457 5•
•
•
5•
•
•
4•
•
-4•
•
"O J 4-"O 0 3..
...
.
•
•
~3 '\. Cl> li) U1 • r l -r 2 ~ Il.O
• • 2.
•
.
•J
..
•,
'. Zc
n
1 •• 1 -:•
•
-
\o
ãi
•
•
-3)
I
01,
II
'
200 300 400 500 600 700· O 18 20 ?2 ?4 26 28Densrt y of lote). vessels per 5 rrrn ú,oMeler of ve s set rIngs (~m)
-
0.0808 + O 0:16 x-
O 576 - 0063)( y y 6 0.3355 6 04945•
5·
.
5 ••
4 • • 4•
--- • "O ~ - 3.
.
.
•
.•. a, Q' -' • >- >-- • 2•
• • 2 •.
.
•••
••
•.
, ••
•
•
•-
---
-
---ti , , , O IZOO 250 3=>8 35J 400 450 500 O 10 20 30 4C 50 60 70ú,51onct lJel we er. cO':s t-:ú.Ilie 101ex ves s e! r"Jôl/Le: o~ 1(,; ( , "t"d',é: rHI;;~
rmg5 (tJm) t. )
C~:
-
-t 04('/(' 6 O 70~4•
( L 5•
• -'-
-
.
4...
•
" "O.
.
3 11> -L r r 2 2•
•
•
I O2" 3
4Circu +erence Irn)
L ,
-
-
.
LU ••
• • ••
....
.
.
.
.
r·'r • I:..;rA P Ae
·1i(Au. N~·.l~(j..T'v(-/.i·tuLWt
t-<EC. EM.:::'.,]
-
1
:
:
f
t
• ••
•
•
•
•••
• •
•
O 05 1,0 1,5 2.0 2,5 3,0 Bork 1h;ckne55 [rnrnlFigure 5- Relatianship of yi eId with circumference, bark t.hirckness dcnsuty of latex
vessels per Srrnn)cliameteTaf ves se-ü r ings, dist.ance between consecutive Iatex ves se'I rings and number of vessel r íngs of the trees collected of Ouro Preto. Territory of Rondônia.