GRAFTED SWEET ORANGE [CITRUS SINENSIS (L.) OSBECK] INF'LU- ENCED BY VARIABILITY IN WOODAPPLE [FERONIA LIMONIA (L.) SWINGLE] ROOTSTOCK

Abs t r ac t : Sweet orange [Cllrus sinensls (L.) Osbeclz I can he grown in the lowcountry dry zone as a grafted plant on woodapple rootstock. However, a wide variation in growth and fruit characteristics was observed in these plants. This experiment was therefore conducted to study the variability of sweet orange grafted onto woodapple [Feronia lin.onia (L.) Swinglel rootstoclc with special emphasis on fruit quality characteristics. The rootstoclz seedlings were sortecl out into three groups based on length of spines. The growth rate and the intensity of pigmentation were higher in t he longer spine rootstoclc group. Magnitude of growth difference a t the bud-union, canopy size, fruit weight and cumulative fruit yieldlyear of the graftecl plants were higher when long spine rootstoclc was used. However, the same group of root,stoclrs procluced fruits with very poor skin quality, juice with low citric acid and low brix content. l1npvoved citric acid, brix contents and fruit skin quality were observed in the orange plants grafted onto rootstocl~ seedlings with short spines.


INTRODUCTION
Sweet orange [Citrz~s sinensis (L.) Osbeckl grafted onto woodapple [Feronin limolzia (L.) Swinglel rootstock has successfully adapted to the dry zone and shows a tendency to produce fruits year round.' Drought tolerance is considered advantageous to the upland rain-fed farming. However, it was found that en.ormous variability has occurred in fruit quality among the plants grafted onto woodapple even with scion taken from a single mother tree of orange.' Variability in woodapple rootstock plants may play a vital role in fruit quality and other growth characteristics in sweet orange plants grafted onto woodapple plants. Variability in latex yield was previously observed in monoclon.al, rubber grafted on rootstock seedlings cv Tjir-12,3 and it was proposed that this could be due to variability in rootstock seedlings4 The stocWscion ratio is the relative size of the rootstock and scion at the bud-union of the grafted plants. The ratio varied previously in grafts of sweet orange on woodapple and average stocW scion ratio became greater than one.' The differential growth at the bud-union was previously explained as incompatibility of the graft~.~However, no attempt has been made so far to study the variability in sweet orange caused by 'Present address: Regional Agricultural Research Centre, Dept. ofAgriculture, Djyatalawa Road, Bandarawela. woodapple rootstock. Therefore, this research was conducted to find out the effect of rootstock variability of woodapple on the variability of growth, yield and fruit quality i n sweet orange.

METHODS AND MATERIALS
Raising woodapple seedlings: Fully ripened woodapple fi-uits were collected from several trees. Seeds were extracted by removing the pulp and washed thoroughly to remove all mucilage. Thereafter, the seeds were kept i n shade for about 12 h taking precautions to avoid drying. Seeds were sown in a sand nursery. Watering was done twice a day so as to keep the sand wet until the seed.s germinated. Copper oxycllloride a t a concentration of 25 gl12 1 water was sprayed onto seedlings for control of fungal diseases. When they attained about 5 cm height they were transplanted in polyethylene bags of 15 cm width and 20 cm length filled with potting mixture of 1:l:l sand : cattle manure : top soil. The plants were maintained in a greenhouse to receive uniform environmental conditions. Fertiliser mixture [16:20:12 (N:P,O,:K,O)] was applied a t the rate of 2 g per plant once i n 4 wks. After six months of potting, woodapple seedlings were grouped as spine length of 20 P4 (lopg), 10 +2 (intermediate) and 4 H . 5 (small) mm. Bud-wood was obtained from a healthy sweet orange var. Bibile sweet tree. Inverted 'T' budding was done.
Field establishment: The experiment was performed a t the Regional Agricultural Research and Development Centre, Angunakolapelessa, where a11 average rainfall of 120 cm is received. Soil type of the area is reddish-brown earth (Rhoclustalfs), with the depth varying from 0.8 to 1.6 m. Soil pH of the site ranges from 6.5 to 7.0. Planting holes of 45 x 45 x 45 cm were filled with compost. Grafts were planted i n the field 2 m apart in March 1993. Three groups of woodapple seedlings (as mentioned above) were the treatments tested. Each. treatment consisted of a single grafted plant. Number of replicates was 24. Treatments were laid out in a completely randomized design. The crop was maintained purely under rain-fed conditions. Fertilizer mixture [16:20:12 (N : P,O, : &0)1 was applied a t the rate of 500 glplant once every 6 montl~s. The first flowers observed in February 1994 were removed since the plants were too young to bear fruits a t this stage.
Duta collection: The growth rate of the woodapple seedlings was measurccl as elongation rate during the 5th month after germination a t weekly intervals. The pigmentation of the seedlings was recorded a t the 5t11montl~ after germination by giving scores for the intensity of the pigments as follows: l= no pigments; 2= light pink colour on1.y in some parts of the seedling; 3=ligllt pink colour i n entire seedling; 4=dark pink colour in the entire seedling. Th.e height and the radius of the canopy, circumference of stoclr and scion a t the bud-union of the plants were recorded. Records were taken from the fruits harvested during the months of April to May in order to avoid variability in qua1i.t~ due to seasonal effects. The peel appearance was measured visually giving scores for peel roughness ( i.e., score 1 to 3 indicate soft peel and 4 to 5 indicate rough peel). Peel weight, fruit weight, number of fvuits/plant/year, peel thiclmess, juice volume/fruit, brix value and acidity of the juice were also recorded. The peel/ fruit ratio was computed by fvuit weight and the coi-responding peel weight. Growtl~, yield and fruit characteristics were measured in the years 1995 and 1996. A comparison was done at the significant level of 5% for each character measured in the years 1995 and 1996.

Variability i n rootstock seecllilzgs:
The weekly elongation rate of the seed lings was 1. Stock /scioa ratio: The stocldscion ratio did not vary between the year 1995 and 1996 (Table 1) Growth chal-a.cteristics: Canopy radius of the grafted plants varied significantly between three plant groups in both years (Table 1). However, canopy height of' the grafted plants varied only in 1996 between the three groups. The variation o-f canopy radius and canopy height in different plant groups occurred ill the following order: short spine < intermediate spine < long spine.
Fruit yield: The average fruit weight did not vary significantly between three groups ( Table 1). The number of fruits/plant/year varied significantly between all three groups in 1995 and 1996 in the foll.owing order: short spine< intermediate spine < long spine (Table 1).
Fruit qzsality: The peellfruit ratio did not vary significantly between the three plant groups (Table 1). However, peel thickness and peel appearance varied significantly between the three groups of plants in both years in the following order: short spine< intermediate spine< long spine (Table 1

DISCUSSION
Growth rate and anthocynin pigmentation were varied in woodapple rootstock seedlings. The growth rate and pigmentation were found to be positively related with the spine length of the woodapple seedlings. Woodapple is a cross pollinated crop6and the presence of spine is one of the xerophytic characteristics of the plants that is an adaptation to adverse climate^.^ The variability of the seedlings may be of the result from segregation of the heterozygous mother plants from which the seeds are extracted.
Stocldscion ratio of the grafted plants was related to the type ofwoodappl.e rootstock used. The ratio was higher when the spine length was higher in stock seedlings. This could be due to high growth rate of the long spine seedlings causing unequal growth rates of stock and scion tissues at the budunion.. Compatibility was explained as equalisation of radial growth rate of rootstock and scion a t the bud u n i~n .~.~ Therefore, the unequal growth at the bud-unj.on in the grafted plants is a sign of incompatibility. Despite the incompatibility indicated in the grafted plants under long spine rootstoclr group, their canopy radius, canopy height and cunlulative fiuit yield per year were greater than those in the short or intermediate spine groups. Consistently, canopy volu.me of grapefruit was influenced by the rootstock used for grafting"' and also the latex yield in grafted rubber was influenced by the vigour of the rootstoclrr seedling-s.:'.~ Therefore, the development of large canopy and the high yield in g~afted sweet orange on long spine woodapple rootstock could be due to its higher vigour.
Variability in fruit peel qualities of orange grafted onto woodapple has also been noted previously." It was revealed that significant variability exists in orange peel qualities when grafted onto different citrus rootstocks."~'" Moreover, peel quality was also determined by rootstock and seasolla1 interact i o n~.~ However, peel and fruit qualities were measured from April to May in the present study. Therefore, occurrence of the rough and thick peel qualities could. be d~l e to high vigour of the long spine rootstock not compatible with growth. of sweet orange scion. The soft and thin peel qualities of orange may be brought about by the low vigour of the short spine rootstocks which is compatible with the growth of orange scion. The juice volume of the fruit of orangc varied in. a previous study when different citrus rootstocks were used for grafting orange.I1-'" Therefore, high juice volume, high brix content and high citric acid content in orange fruits on short spine rootstocks may be due to the influence of the short spine rootstoclis and their compatibility with orange scion.
The results suggest that selection of woodapple rootstock seedlings in the nursery for grafting of sweet orange will be useful for production of good quality sweet orange fruits. The presence of short or no visible spines and the absence of anthocyanin pigmentation on the woodapple seedlings are preferable for selecting rootstocks. Areduction in the canopy volume in grafted plants by selection of such rootstock seedlings can be compensated by increasing the plant density i n the cultivation.