Performance of 'Oneco' mandarin on six rootstocks in South Brazil

The objective of this work was to assess the effect of six rootstocks on yield, fruit quality, and growth of 'Oneco' mandarin during the first seven harvesting seasons, in Butiá, Rio Grande do Sul State, Brazil. The rootstocks evaluated were: 'Swingle' citrumelo (Citrus paradisi × Poncirus trifoliata), 'Caipira' orange (C. sinensis), 'Troyer' citrange (C. sinensis × P. trifoliata), 'Rangpur' lime (C. limonia), 'Volkamer' lemon (C. volkameriana), and 'Flying Dragon' trifoliata orange (P. trifoliata var. monstrosa). Plants budded onto 'Flying Dragon' had the lowest vegetative development, which indicates the dwarfing characteristics of this rootstock, and had the highest mean production efficiency, despite low yield. Plants grafted on 'Volkamer' lemon and 'Rangpur' lime had the highest alternate bearing. Under the experimental conditions evaluated, the most adequate rootstocks for mandarin 'Oneco' are 'Swingle' citrumelo and 'Troyer' citrange, regarding fruit yield and quality.

'Oneco' was identified as a potential alternative cultivar to 'Ponkan', due to its high productivity and longer harvesting period evidenced by a higher capacity of fruit storage (Sartori et al., 1998).Furthermore, 'Ponkan' has a very concentrated harvesting period that contributes to the rise of seasonal prices (Amaro & Caser, 2003).However, there is a lack of information on the effects of different rootstocks on 'Oneco' mandarin fruit quality, tree development and productivity.
Rootstocks affect canopy growth and development, production, and fruit quality in citrus, as well as tolerance to various stresses (Castle, 1987).'Rangpur' lime (Citrus limonia Osbeck) is the main rootstock in Brazil, and is used in 80% of Brazilian citrus orchards because of its drought tolerance (Pompeu Júnior, 2005).
High disease incidence, to which rootstock resistance is crucial, has prompted the need to investigate rootstock assortment.Diseases, such as citrus tristeza virus, gummosis (Phytophthora sp.), and citrus sudden death, reinforce the call for research on diverse citrus species and hybrids as potential rootstock material.
The objective of this work was to assess the effect of six rootstocks on yield, fruit quality, and growth of 'Oneco' mandarin during the first seven harvesting seasons, in Butiá, Rio Grande do Sul state, Brazil.The experiment was carried out in a private grove (30°07'5"S, 51°51'22"W) close to the city of Butiá, about 70 km west of Porto Alegre, Rio Grande do Sul State, Brazil.The soil is classified as Paleudult (United States Department of Agriculture, 1999), and the climate as Cfa, according to Köppen.The average annual temperature is 19.2°C, and the average temperatures of the coldest (June) and warmest (January) months are 13.5 and 24.6°C, respectively.Annual rainfall is approximately 1.440 mm (Bergamaschi et al., 2003).

Materials and Methods
Trees were planted in August 1997, in a 6.0x2.5 m spacing, with 667 trees per hectare.The experiment was managed according to the usual cultural practices (Koller, 2009), with plants being hand-thinned every year between February and March, and pruned in the winter of 2004 to eliminate dead wood and crooked branches.
Individual measurements of tree height (H), perpendicular (D r ) and parallel (D l ) canopy diameters, and trunk circumference at 5 cm above the bud union were taken from 2003 to 2007, after the summer growth flush.Canopy volume (V) was calculated using the formula: (Zekri, 2000).
Production efficiency (kg m -3 ) was determined by the ratio of total fruit mass to canopy volume.The canopy projection area (CPA) was calculated using the formula CPA = π[(D l + D r )/4]² (Schäfer et al., 2001b).
An alternate bearing index (ABI) was calculated as described by Stenzel et al. (2003): ABI = [(1/(n -1)] × {[│(a 2 -a 1 )│/(a 2 + a 1 )] + [│(a 3 -a 2 )│/(a 3 + a 2 )] +...+ [│(a n -a n-1 )│/(a n + a n-1 )]}, in which n is the number of years, and a 1 , a 2 , ..., a n-1 , a n is the yield per tree in the corresponding year.The ABI ranges from 0 to 1, and the closest it is to 0, the less variation there is in yield between years.This trait was evaluated for the following periods : 2001/2003, 2002/2004, 2003/2005, 2004/2006, 2005/2007, 2001/2004, 2004/2007, and 2001/2007.Fruit harvest occurred from 2001 to 2007.Total annual yield per tree and accumulated yield were calculated for this period.Starting in 2002, 20 fruits harvested from each replicate of each treatment were randomly selected to determine fruit quality.The measurements taken were: juice content (JC, %), calculated by the difference of mass between juice and fruits; total soluble solids (TSS), with a table refractometer; and titratable acidity (TA, % of citric acid), determined by titration of 10 mL juice diluted in 90 mL of distilled water up to pH 8.1.Mass (M), length (L), and diameter (D) of individual fruits were also determined, in which L and D were used to estimate fruit shape through the ratio L:D.
The experiment was set up in a completely randomized block design with four replicates and four trees as an experimental unit, for each rootstock.Data for total yield, vegetative growth, and fruit quality were analyzed using analysis of variance, and means were compared with Duncan's multiple range test, at 5% probability.

Results and Discussion
Annual yield (Table 1) was affected by rootstocks, except in 2002 and 2006, when production was low and variable due to alternate bearing.In general, trees on 'Swingle' consistently had greater yields than those on the other rootstocks.Trees budded onto 'Rangpur' and 'Swingle' were the most precocious, having the highest yields in 2001.The highest cumulative yields Pesq.agropec.bras., Brasília, v.46, n.4, p.406-411, abr. 2011 were found for trees grafted on 'Swingle', whereas no significance was found for yields of trees on 'Caipira', 'Troyer', or 'Rangpur'.The lowest cumulative yields were obtained for plants budded onto 'Flying Dragon', while plants on 'Volkamer' had intermediate yields.Stenzel et al. (2003), who compared seven rootstocks for 'Ponkan' mandarin grown in Paraná state, Brazil, also found equivalent yields for plants grafted on 'Rangpur' lime and 'Volkamer' lemon, while plants on P. trifoliata had the lowest yields.Mourão Filho et al. ( 2007) observed that cumulative yield for plants budded on 'Rangpur' lime was higher than for plants on 'Swingle'.Georgiou (2000), while evaluating 'Nova' mandarin, reported that plants grafted on 'Volkamer' lemon were superior to those on 'Swingle' and 'Troyer'.
The ABIs were affected by rootstocks especially during the first years of the orchard's production.No effect from the rootstocks was observed in 2004/2007, which indicates that this effect can be associated with initial plant yield instability (Smith et al., 2004).Production efficiency was influenced by rootstocks, except for the last two years of evaluation.Highest cumulative production efficiency was observed, from 2003 to 2007, for trees grafted onto 'Flying Dragon', despite low yields.Therefore, this rootstock has potential for high-density plantings of citrus cultivars with characteristics similar to 'Oneco', in spite of previous reports with disappointing results regarding this rootstock (Wheaton et al., 1991).Cantuarias-Avilés et al. ( 2010) also found high production efficiency when working with the cultivar Okitsu satsuma.
Vegetative growth, which was measured in January 2008, was influenced by rootstocks (Table 3).Trunk perimeter and canopy height were highest in trees budded onto 'Caipira' orange, while canopy volume and CPA were highest on 'Caipira' orange and 'Swingle' citrumelo trees.'Flying Dragon' trifoliate trees had the lowest vegetative growth, with a 37% decrease in tree height, a 75% decrease in canopy volume, a 62% decrease in the projected canopy area, and a 59% decrease in trunk perimeter in comparison to the most vigorous rootstocks.Similar vegetative responses of trees grafted on 'Rangpur' lime and 'Volkamer' lemon were found in previous studies (Parente et al., 1993;Georgiou, 2000;Stenzel et al., 2003).However, Parente et al. (1993) did not find differences in tree height or trunk circumferences for 'Ponkan' mandarin trees budded onto 'Caipira' orange, 'Rangpur' lime, 'Volkamer' lemon, 'Swingle' citrumelo or 'Troyer' citrange.
Plants grafted on 'Swingle' had greater vegetative growth than those on 'Rangpur' lime, probably due to the edaphoclimatic conditions in which the present study took place, with colder winters, since 'Rangpur' lime is sensitive to cold and 'Swingle' citrumelo shows some cold resistance (Castle, 1987).Parente et al. (1993), Georgiou (2000), Espinoza-Núñez et al. (2007), andMourão Filho et al. (2007) did not find any differences among these rootstocks.
'Oneco' trees budded onto citrange 'Troyer' had a lower CPA than trees on 'Swingle'.Schäfer et al. (2001b), working with eight-year-old 'Montenegrina' trees, under the same environmental conditions, did not find differences in CPA between these rootstocks.
The reduction in vegetative growth observed in trees grafted on 'Flying Dragon' indicates the dwarfing characteristics of this rootstock, which is reported to maintain plant height under 2.5 m (Cheng & Roose, 1995).These plants showed similar vegetative characteristics to eight-year-old 'Monte Parnaso' orange (Reis et al., 2008) and ten-year-old 'Michal' mandarin (Brugnara et al., 2009) trees in the same orchard.Canopy height and volume reduction of plants budded onto 'Flying Dragon', in comparison to 'Troyer', are similar to the reduction observed by Roose (1986) for 'Valencia' orange.
Mandarins harvested from plants on 'Flying Dragon' had the smallest fruit among the tested rootstocks, while 'Volkamer' lemon produced the largest and heaviest ones (Table 4).Fruits from trees grown on 'Flying Dragon' were more spherical than those on the other tested rootstocks, which were more oblong.
The highest average fruit weight for plants budded on 'Volkamer' lemon is similar to that found for 'Nova' tangerine (Georgiou, 2000), although, in the present work, there were no differences in fruit diameter for 'Rangpur', 'Troyer', 'Swingle', and 'Volkamer' rootstocks.The effect of rootstocks on the L:D ratio was also observed by Koller et al. (2000) for 'Hamlin' orange, under the same climatic conditions.

Conclusions
1.The most adequate rootstocks for mandarin 'Oneco' in the tested region are 'Swingle' citrumelo and 'Troyer' citrange, regarding fruit yield and quality.
2. Plants grafted onto 'Flying Dragon' rootstock have higher production efficiency, in spite of lower canopy volume.

Table 1 .
Annual and cumulative yields, and production efficiency of 'Oneco' mandarin grafted onto six different rootstocks grown in Butiá, Rio Grande do Sul state, Brazil(1).Means followed by equal letters, in the columns, do not differ significantly by Duncan's multiple range test, at 5% probability.

Table 2 .
Alternate bearing index of 'Oneco' mandarin grafted onto six different rootstocks grown in Butiá, Rio Grande do Sul state, Brazil(1).Means followed by equal letters, in the columns, do not differ significantly by Duncan's multiple range test, at 5% probability.

Table 3 .
Tree height, volume, canopy projection area (CPA) and trunk perimeter of 'Oneco' mandarin grafted onto six different rootstocks grown in Butiá, Rio Grande do Sul state, Brazil(1).