SciELO - Scientific Electronic Library Online

vol.26 issue4Studies of blood groups and protein polymorphisms in the Brazilian horse breeds Mangalarga Marchador and Mangalarga (Equus caballus)Genotype-environment interaction and phenotypic stability for girth growth and rubber yield of Hevea clones in São Paulo State, Brazil author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Genetics and Molecular Biology

Print version ISSN 1415-4757

Genet. Mol. Biol. vol.26 no.4 São Paulo Dec. 2003 



Contributions of weather variables for specific adaptation of rubber tree (Hevea brasiliensis Muell.- Arg) clones



P.M. Priyadarshan

Rubber Research Institute of India, Regional Station, Agartala, India





The specific adaptation of 15 rubber tree (Hevea brasiliensis) clones was assessed by analyzing yield during a normal year (1997-98) and a year (1998-99) in which the yield was exceptional. Differences in yield in response to changes in weather conditions over the years were evident with clones RRII 203, RRIM 703, PB 5/51 and PB 235 which all exhibited a negative trend with increasing wind velocity during 1997-98, these clones also exhibited a negative correlation with minimum temperature during 1998-99. The prominent yield differences across the years made selection based on both yield and stability inevitable through computing weather variables and environmental index as covariant. To determine the contribution of variable(s) to genotype-environment (GE) interactions, the GE interaction was partitioned into heterogeneity and residual GE interaction. Heterogeneity only for environmental index was highly significant (p = 0.01), meaning that stability or instability of clones was due to a linear effect of the environmental index. The non-significant values of heterogeneity for the weather variables revealed that none of these factors individually was sufficient to explain heterogeneity. A QBASIC computer program called STABLE was used to select simultaneously for yield and stability. Clones PB 235, RRII 118, RRII 203, RRIM 703 and RRIM 600 were stable over the years investigated.

Key words: rubber clones, weather variables, specific adaptation, Hevea, stability, environmental index.




The importance of using selection procedures to assess the specific adaptation of plants to moderate or marginal conditions has been discussed by Ceccarelli (1994). Estimation of specific adaptation is normally based on genotype-environment (GE) interactions, although the partitioning of genotypic effects and phenotypic plasticity effects in response to changes in the environment is debatable (Vega, 1996). Although yield is the primary attribute evaluated in agricultural GE interaction studies, it is not solely under genetic control, being modulated by genotype-genotype (GG) and genotype-environment (GE) interactions as well as physiological and biochemical processes controlled by specific genes (Blum, 1988). In rubber tree (Hevea brasiliensis), these processes are even more complex because yield is based on latex production.

Tripura state in Northeast India (22-24° N, 91-92° E) is a non-traditional environment for rubber cultivation with sub-optimal conditions (Rao et al., 1993; Priyadarshan et al., 1998a; Priyadarshan, 2003), where H. brasiliensis clones have different yields compared with traditional rubber-growing areas because of specific adaptation (Priyadarshan et al., 1998b). According to Priyadarshan et al. (2000), two yield regimes are prevalent in Tripura state, low-yielding Regime I that occurs from May to September (sub-optimal environment) and high-yielding Regime II that occurs from October to January (optimal environment). This significant difference in yield is caused by the range of latitude and longitude covered by the state, which results in areas that are sub-optimal. It is possible to control the edaphic factors at will while the climatic attributes are uncontrollable but predictable. The responses of rubber trees to this type of environment can be assessed by measuring phenotypic attributes such as the girth of the tree and/or the yield of latex and analyzing resulting GE interactions. Such ontogenetic changes occur in response to changes in the seasons and weather that prevents important stages in the development of the plant coinciding with adverse environmental conditions (Roberts et al., 1993). In Tripura state, selecting rubber clones that have specific adaptation traits is vital since the exceptional climate of Tripura also offers low- and high-yielding environments (Priyadarshan et al., 1998b). Ceccarelli (1994), states that genotypes exhibiting high yield-potential in otherwise low-yielding environments should be preferred when selecting for specific adaptation.

Although GE interactions can be assessed by analyzing phenotypic stability, only very few papers have been published on how weather variables affect GE interactions in perennial crops. Devi et al. (1998) investigated the impact of weather variables on the rubber yield of a polyclonal rubber seedling population and concluded that both temperature and evaporation influenced yield. In another study, Priyadarshan et al. (2000) found that minimum temperature, evaporation and wind velocities were negatively correlated with dry rubber yield under the conditions found in Tripura state. They also found that the GE interactions that occurred were of the cross-over type. The major factors (covariants) that contribute to heterogeneity (non-additivity) in GE interactions can be determined using a BASIC program developed by Kang (1988). Kang (1993) developed a statistic called YSi or yield-stability statistic to simultaneously select for yield and stability. Kang and Magari (1995) wrote a QBASIC program that allows the calculation of the YSi statistic. The present study evaluated the impact of weather variables (minimum temperature, wind velocity, evaporation) on the yield of rubber in Tripura during a normal year (1997-98) and an exceptionally high-yielding year (1998-99) in relation to changed macro-environmental attributes.


Material and Methods

The experiment was conducted at the Regional Research Farm of the Rubber Research Institute of India at Taranagar (Tripura State, North East India - 23°53' N; 91°15' E; 30 m above sea level (a.s.l.). For this research, yield data of dry rubber during a normal year (1997-98) and an exceptionally high-yielding year (1998-99) was used. Yielding period for rubber in Tripura is from May to January. The trial involved 15 clones from different geographic origins (Table 1), which had been planted during 1979 in a completely randomized design using a 5 x 5 m spacing between plants. Initially 40 trees per clone were planted that were multiplied by bud grafting. Latex was collected from each tree of each clone once a month, coagulated, squeezed through rollers to remove excess water, dried in a smoke house, weighed and mean dry rubber yield in grams per tree per tap (g tree-1 tap-1) calculated. Data on three weather variables (minimum temperature, wind velocity and evaporation) were collected each day using standard meteorological methods (Rao et al., 1993).



Analysis of variance and other statistical calculations were done using standard procedures (Snedecor and Cochran, 1980). Genotype-environment (GE) interaction was partitioned into si2 components that were assigned to each genotype using Kang's BASIC program (Kang, 1988). The si2 is Shukla's stability variance statistic (Shukla, 1972). The program calculates stability variance (si2) prior to and after the use of a covariant (si2) and analysis can be used to determine the major factor (covariant) that contributes to heterogeneity in the GE interactions. In this case, weather variables were used as covariants and an analysis was also made using environmental index (Xj – X.., where Xj is the mean yield of all clones in environment j and X.. is the mean of all clones across all environments) as a default covariant to remove heterogeneity (non-additivity or a linear effect of a covariant) from the GE interaction (Kang and Gorman, 1989). The program STABLE was used to simultaneously select for yield and stability (Kang, 1993; Kang and Magari, 1995), which is based on si2 of Shukla (1972).


Results and Discussion

This investigation was conducted not only to compare the yield of clones during a year (1997-98) when the yield was normal and a year (1998-99) when the yield was exceptionally high but also to gauge the stability of yield of the different clones. Since the edaphic factors (soil type, texture, drainage etc.) remained the same during the two years, it is probable that the exceptional yield during 1989-99 was due to a favorable macro-environment (e.g., climatic conditions). An earlier study on the yielding trends of these clones established that minimum temperature, wind velocity and evaporation were the key attributes contributing to fluctuations in rubber yield in Tripura (Priyadarshan et al., 2000). In any year, yields are lower during May to September (Regime I) and higher during October to January (Regime II), high yields being mainly due to the lower temperatures (23-26 °C) during the cooler winter months of October to January because lower temperatures stimulate prolonged latex flow, it having been established by Shuochang and Yagang (1990) that temperatures in the range of 18-24 °C are conducive to higher latex flow.

Effect of climatic variables on yield

The monthly mean values of minimum temperature, wind velocity and evaporation are given in Table 2. A variety of relationships was evident from the calculated correlations between yield and weather variables. For example, during 1997-98 none of the clones exhibited a correlation between yield and evaporation but almost all clones showed a significant negative correlation between yield and wind velocity (i.e., the higher the wind velocity, the lower the yield), with clones RRII 203, RRIM 703, PB 5/51 and PB 235 giving highly significant correlations (Table 3). When we examined the 1998-99 data, RRII 105, RRII 118, RRIM 605, RRIC 52, RRIC 105 and Gl 1 exhibited significant negative correlations between yield and minimum temperature, i.e., the lower the temperature the higher the yield. Although the relationship between yield and wind velocity was not as important as in 1997-98 (Table 4).







These differences in yield should be seen in the light of the effects that climatic variables have on the micro-environment, the lower yields during 1997-98 probably being caused by the higher minimum temperature and stronger winds that reduced atmospheric vapor pressure and hence the higher turgour pressure, being directly proportional to the osmotic pressure of the laticifers that control latex flow (Jacob et al., 1998). On the other hand, lower wind velocities during 1998-99 assisted in enhancing yield.

Simultaneous selection for yield and stability

The differences in yield over months and seasons mean that it is important to select cultivars based on both yield and stability (Figure 1), in other words stability of yield should be taken into consideration as well as high yield. In our trials, the year 1998-99 was exceptional in regard to yield, such high yields not having been recorded during the preceding 10 years.



The weather variables and environmental index were used as covariants when analyzing the yield data to determine the variable that contributed to heterogeneity in the GE interactions. The heterogeneity caused by environmental index was highly significant (Tables 5 and 6), wind velocity and other climate variables causing only non-significant heterogeneity in GE interaction.





The yield stability statistic (YSi) can be used to select clones that can adapt to normal and exceptional years (Tables 7 and 8). While eight clones (RRII 118, RRII 203, RRIM 600, RRIM 703, PB 235, RRIC 52 and RRIC 105) were stable during the normal-yielding year, only seven clones (RRII 105, RRII 118, RRII 203, RRIM 600, RRIM 703, PB 235 and GT 1) were stable during the exceptional-yielding year. When the cultivars were ranked in order of yield (highest-yielding clones first) the rank-order for 1998-99 was PB 235, RRII 203, RRIM 703, RRIM 600 and RRII 118, while the rank order (again starting with the highest-yielding clones) for 1997-98 was PB 235, RRII 203, RRII 118, RRIM 703 and RRIM 600, although these all gave lower yields than the highest-yielding clones for 1998-99. Calculation of yield-stability using the YSi statistic showed that clones PB 235, RRII 118, RRII 203, RRIM 703 and RRIM 600 were stable in yield during both the years (these cultivars can therefore be considered as showing adaptation under specific environments) while the same statistic showed that clones RRII 118 needed an exceptional climate to produce a high yield.






The author thanks Dr. M.S. Kang (Louisiana State University, Baton Rouge, USA) for supplying the STABLE software and for valuable comments. Dr. D.H. Wallace (Emeritus Professor of Plant Breeding and Biometry, Cornell University Ithaca, USA) also provided useful criticism and comments. The facilities provided by Dr. N.M. Mathew (Director, Rubber Research Institute of India) are gratefully acknowledged.



Blum, A (1988) Plant breeding for stress environments. CRC Press Inc, Florida.        [ Links ]

Ceccarelli, S (1994) Specific adaptation and breeding for marginal conditions. Euphytica 77:205-219.        [ Links ]

Devi TS, Nair RB, Kothandaraman R and Sethuraj MR (1998) Impact of weather parameters on seasonal and inter-year variations in yield of rubber. In: Mathew, NM and Jacob, KC (eds) Developments in Plantation Crops Research, New Delhi: Allied Publishers, pp 170-174.        [ Links ]

Jacob JL, Prevot JC, Lacote R, Gohet E, Clement A, Gallois R, Joet T, Pujade-Renaud V and D'Auzac J (1998) The biological mechanisms controlling Hevea brasiliensis rubber yield. Plantations, Recherche, Developpement. Jan-Feb, pp 14-16.        [ Links ]

Kang MS (1988) Interactive BASIC program for calculating stability-variance parameters. Agron J 80:153.        [ Links ]

Kang MS (1993) Simultaneous selection for yield and stability in crop performance trials: Consequences for growers. Agron J 85:754-757.        [ Links ]

Kang MS and Gorman DP (1989) Genotype x environment interaction in maize. Agron J 81:662-664.        [ Links ]

Kang MS and Magari R (1995) STABLE: A basic program for calculating stability and yield-stability statistic. Agron J 87:276-277.        [ Links ]

Priyadarshan PM (2003) Breeding Hevea brasiliensis for environmental constraints. Advances in Agronomy 79:351-400.        [ Links ]

Priyadarshan PM, Vinod KK, Rajeswari MR, Pothen J, Sudhdaowmyalatha MK, Sasikumar S, Raj S and Sethuraj MR (1998a) Breeding Hevea brasiliensis in Tripura (NE India): I. Performance of a few stress tolerant clones in the early phase. In: Mathew NM and Jacob KC(eds) Developments in Plantation Crops Research. New Delhi: Allied publishers, pp 63-68.        [ Links ]

Priyadarshan PM, Sowmyalatha MKS, Sasikumar S, Varghese YA and Dey SK (1998b) Relative performance of six Hevea brasiliensis clones during two yielding regimes in Tripura. Indian Journal of Natural Rubber Research 11(1&2):67-72.        [ Links ]

Priyadarshan PM, Sowmyalatha MKS, Sasikumar S, Varghese YA and Dey SK (2000) Evaluation of Hevea brasiliensis clones for yielding trends in Tripura. Indian Journal of Natural Rubber Research 13:56-63.        [ Links ]

Rao GG, Rao PS, Rajagopal R, Devakumar AS, Vijayakumar KR and Sethuraj MR (1990) Influence of soil, plant and meteorological factors on water relations and yield in Hevea brasiliensis. International Journal of Biometeorology, 34:175-180.        [ Links ]

Rao PS, Jayaratnam K and Sethuraj MR (1993) An index to assess areas hydrothermally suitable for rubber cultivation. Indian Journal of Natural Rubber Research 6:80-91.        [ Links ]

Roberts E, Summerfield R, Ellis R, and Qi A (1993) Adaptation of flowering in crops to climate. Outlook Agri 22:105-110.        [ Links ]

Shuogang A and Yagang G (1990) Extrapolation of the high yield physiological regulation of Hevea brasiliensis in Xishuangbanna. Proc. IRRDB Symp. on Physiology and Exploitation of Hevea brasiliensis, pp 83-92. Kunming, China, 1990.        [ Links ]

Snedecor, CW and Cochran WG (1980) Statistical methods. The Iowa State University Press.        [ Links ]

Shukla GK (1972) Some statistical aspects of partitioning genotype-environmental components of variability. Heredity 29:237-245.        [ Links ]

Vega PM (1996) Plant genetic adaptedness to climatic and edaphic environment. Euphytica 92:27-38.        [ Links ]


Editor: Natal Antonio Vello



Correspondence to
P.M. Priyadarshan
Rubber Research Institute of India, Regional Station
799 006, Agartala, India

Received: January 20, 2003; Accepted: August 25, 2003

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License