Comparative tree growth, phenology and fruit yield of several Japanese plum cultivars in two newly established orchards, organic and conventionally managed

The growth, phenology and fruit yield of 14 Japanese plum cultivars ( Prunus salicina Lindl) were studied in two newly established experimental orchards under organic and conventional management. The experiment was conducted during 2005-2011 in the province of Seville (SW Spain), an important region of Japanese plum culture. Trunk cross-section areas (TCSA), flowering, yield and tree defoliation before winter dormancy were analysed over several years. After one year, TCSA were larger in the organically managed orchard (OMO) for most of the cultivars, in the next two years they were equal, and from the fourth year, several cultivars showed significantly larger TCSA in the conventionally managed orchard (CMO). Flowering in the conventional orchard started from 2 to 6 days before and lasted for 3 to 5 days more than in the OMO. Several cultivars produced significantly more fruit in the CMO, being the average fruit yield in the organic orchard about 72% of the conventionally managed orchard. Autumn defoliation was significantly advanced in the organic orchard, especially in cultivars highly susceptible to rust ( Tranzschelia pruni spinosae ), a disease not adequately controlled in the organic orchard. tree


Introduction
The growth of fruit trees is affected by many different factors: plant genetics, soil type, rootstock (Webster, 1995), climatic conditions (sunlight, temperature, rainfall, humidity) (Legave et al., 2006) and there are also agricultural activities having a clear effect on fitness, vigour, and fruit quality of plants as the orchard system, irrigation, and fertilization (Hester & Cacho, 2003;Rufat et al., 2011;Kükükyumuk et al., 2012).
The different phenological stages of fruit trees, and their seasonal timing and duration, vary depending on local climatic conditions and fluctuate from year to year (Montagnon, 2007). Among the processes that directly influence the flowering of fruit trees is the accumulation of chill units during the winter (Albuquerque et al., 2007). Once the minimum number of required chill units is achieved, appropriate temperatures are needed to raise the swelling and germination of buds (Melgarejo, 1996). Also, phenology is influenced by variables such as temperature (Weinberger, 1956;Browning & Miller, 1992;Rodrigo & Herrero, 2002), wind (Dennis, 1979), frost (Rodrigo, 2000), rain and relative humidity (Gradziel & Weinbaum, 1999). In addition to the genetic characteristics of different cultivars, there are other factors that affect the phenology of flowering such as the type of fertilization (Williams, 1965), the age of the trees or branches (Robbie & Atkison, 1994) and even the orientation of the branches (Robbie et al., 1993). All these factors can also affect fruit yield and quality.
In this work, a comparative study of organic and conventionally managed orchards was conducted examining differences in tree growth, flowering, fruit yield and autumn defoliation in two similar experimental orchards of 14 Japanese plum cultivars.

Description of the experimental plots
The study was conducted during 2005-2011 in two similar experimental orchards (5,500 m 2 each) located at the IFAPA Centro "Las Torres-Tomejil" in the province of Seville in the Guadalquivir River Valley (SW Spain) (37°30' 48'' N; 5°57' 46''W). The soil was classified as a Xerofluvent (Soil Survey Staff, 1999). Both plots were selected on the basis of a similar physicochemical soil composition (Table 1) and were 200 m apart from each other to avoid interference between the different pesticide treatments. The region has a Type C Mediterranean climate, according to the Köppen classification. Relevant agroclimatic conditions registered in the zone during the time of the study are shown in Table 2.
In both plots, 14 Japanese plum cultivars (Prunus salicina Lindl), which matured from late May to mid-September, were planted in January 2005. Characteristics of the cultivars are described by García-Galavís et al. (2009). Each experimental plot was subject to a different type of management, one in organic and the other in conventional agriculture. In both orchards, the experiment was set up in a randomised block design with three replications, each containing 6 trees of each cultivar.

Major field work conducted in both orchards
Fertilization in the organic plot consisted of the application of animal manure (3-4 kg m -2 yr -1 ) and sowing of several cover greens containing legumes (Table 3). A soybean (Glycine max L.) cover green was also sown in the summer of 2004 before planting the trees. The conventional plot received annual applications of mineral manures, including complex formulations (11-11-11), ammonium nitrate and potassium sulphate, to 155 units of nitrogen, 55 of phosphorus and 150 of potassium.
Both plots were irrigated during the dry season by gravity along two rows parallel to the line of trees with identical volumes and frequency. Each year, depending on climatic conditions, from 6 to 9 irrigations were applied (350,000 L ha -1 each).
Land management consisted of reduced tillage. When necessary, the grass in the line of trees in the conventional orchard was removed through the use of authorized herbicides and in the organic one by a pass from bleachers with an orchard tractor. The plum trees had a similar pruning training vessel, taking care that pruned trees bear a similar width and height in both orchards.

Tree growth and fruit yield
Growth and vigour of the trees were evaluated yearly in November at the end of the growing season by calculating the cross-section area of the trunk (TCSA) 20 cm above the graft (Layne, 1994;Lepsis & Blanke, 2006).
The fruit yield of each variety (kg tree -1 ) was calculated taking into account the total fruit collected in each orchard, tree replicates with 6 trees each. The cumulative production was obtained by joining the yields obtained in previous years.

Phenological studies
For phenological studies, data capture was carried out on nine trees per cultivar (3 per block) and treatment. According to the BBCH scale (Hack et al., 1992), the following phenological periods were analysed: opening of the first flowers, full flowering (> 50% flowers open) and end of flowering (all petals fallen), and the duration of flowering was also determined. The kinetics of falling leaves during senescence was evaluated. The fall of the leaves was recorded weekly from October to December. Defoliation was evaluated using a 0-5 scale, where 0 is 0% fallen leaf; 1, 20% fallen leaf; 2, 40% fallen leaf; 3, 60% fallen leaf; 4, 80% fallen leaf and 5, total defoliation. Evaluation of rust disease incidence was as described by García-Galavís et al. (2009).

Statistical analysis
Statistical analyses were performed using Statistix software (version 9.0, NH Analytical Software, USA).
Analysis of variance (ANOVA) was used to analyze TCSA and fruit yield data, and defoliation and rust incidence data were analyzed with the Kruskal-Wallis nonparametric test at the p < 0.05 level of significance.

Tree growth, soil and foliar analyses and fruit yield
The TCSA from 2005 to 2011 are shown in Table 4. In 2005, eight cultivars showed significantly higher TCSA in the OMO and the other six cultivars showed non-significant differences between the two management systems. In 2006 and 2007, most of the cultivars displayed similar values in both orchards, except for 'Angeleno' in 2006 (higher in the organic) and 'Red Beaut' in 2007 (higher in the conventional). In 2008, 2009, 2010 and 2011, a total of six, seven, nine and eight cultivars, respectively, showed significantly higher TCSA values in the CMO, showing mean increases of between from 15 to 20%. In these four years, the other cultivars also showed larger TCSA in the CMO, but without significant differences. According to the TCSA values, 'Souvenir', 'Red Beaut', 'Angeleno' and 'Golden Japan' were the most vigorous and 'Friar' and 'Sapphire' the least, in both types of management.
Physico-chemical soil composition of both orchards in 2011 is shown in Table 1. Except for organic matter, significantly higher in the organic plot, no relevant differences were observed. Foliar composition of a strong cultivar ('Golden Japan') and other one feeble ('Sapphire') in 2007, 2009 and 2010 is shown in Table 5. No relevant differences were observed in the major macro or micronutrients between the two management systems.
Japanese plum under organic and conventional management 157 Fruit production of the different cultivars in the period 2008-2011 is shown in Table 6. In 2008, ten cultivars had a significantly more fruit production in the CMO and no variety was higher in the organic treatment. This year, the mean fruit yield was three times higher in the conventional treatment.
In 2009, only 'Fortune', 'Red Beaut' and 'Black Amber' produced significantly more fruit in the CMO. The rest of the cultivars did not show significant differences. No significant differences were obtained in mean yields of both management types, even though the organic plot produced around 20% less fruit.   In 2010, five cultivars had significant higher fruit yields in the CMO. Fruit production of 'Sapphire', 'Showtime' and 'Black Amber' was higher in the organic orchard, although only the last cultivar showed a significant difference. Again this year the mean fruit yield did not show a significant difference.
In 2011, 'Souvenir', 'Laetitia' and 'Black Amber' produced significantly more fruit yield in the conventional orchard, while 'Songold' and 'Showtime' produced significantly more in the organic system. Significant differences were not observed in the average yield.
Except for 2008, with a production still erratic, the average fruit production of the organic orchard ranged between 71% and 82% of the conventional one. The 2008-2011 cumulative production showed seven cultivars with a signif icant higher yield in the CMO.

Flowering
The accumulation of chill units during the cold season (from November 1 to February 15), and the temperatures, relative humidities and rainfall in the flowering periods (the average of the months of February, March and April) are shown in Table 2. In 2008, 'Larry Ann', 'Fortune', 'Red Beaut', 'Black Amber' and 'Angeleno' initiated flowering simultaneously in both orchards, whereas the other cultivars showed an earlier flowering in the CMO, specially 'Friar', 'Laetitia' and 'Showtime' (Fig. 1).
In 2009, only 'Fortune' initiated flowering simultaneously in both orchards and the remaining 13 cultivars showed a significant earlier flowering in the CMO, exception made for 'Primetime'.
Japanese plum under organic and conventional management 159 Average data of the flowering period in all the four years revealed that flowering in the conventional orchard started from 2 to 6 days before and lasted for 3 to 5 days more than in the OMO (Table 7).

Leaf fall
Defoliation was evaluated several years from October to December. Fig. 2 shows a detailed defoliation rate for all of the cultivars in both orchards in 2008. At the first sampling in third week of October, 10 cultivars showed significantly less foliage in the OMO. In two subsequent samplings, performed in the f irst and fourth week of November, all the cultivars showed a significantly higher leaf drop in the OMO. In the second week of December, all of the cultivars were completely defoliated in the organic orchard. In the conventional plot, 'Red Beaut', 'Black Amber', 'Friar' and 'Showtime' were also completely defoliated at this time and for the other cultivars 100% of defoliation was achieved by the fourth week of December. An average premature defoliation was detected in the OMO all the years (Fig. 3).
'Showtime', 'Friar', 'Larry-Ann', 'Santa Rosa' and 'Sapphire' showed the higher degree of premature defoliation in the organic plot, while 'Souvenir' and 'Golden Japan' had minor differences in both orchards, and the other cultivars showed an intermediate behaviour. Susceptibility of different Japanese plum cultivars to rust disease caused by Tranzschelia pruni spinosae was evaluated, and as shown in Fig. 4, the most defoliated cultivars were those which showed the greatest sensitivity to rust infection.

Discussion
Organic farming has been regulated in Europe for almost two decades. Among the strengths of this system is a great respect for the environment, food quality and safety (Pfiffner & Niggli, 1996;Hole et al., 2005;Gabriel & Tscharntke, 2007), whereas the weaknesses are often lower yields and the difficulty of controlling pests and diseases (Leake, 1999;Weibel et al., 2007;De Ponti et al., 2011). Organic tree plantations can be completed through a conversion period of three years from a conventional pre-established planting or starting the organic management from 'scratch'. In this study, we evaluated the effect of two types of management, conventional and organic, on tree growth, yield and some phenological events of different Japanese plum cultivars in two newly established orchards.
After the first year of growth in 2005, we observed that several cultivars had larger TCSA in the organic orchard. During 2005, fertilizers were not added to both orchards, but during the summer of 2004 a highly nutritious soybean plant cover was planted and incorporated into the OMO, which could explain these differences. In the next two years, 2006 and 2007, no differences were found between treatments, but from the fourth year, when fruit production began, various cultivars had larger TCSA in the conventional orchard, and differences still persisted seven years after planting. It seems that likely the difference of vigour may be because of the type of fertilizer applied, though a similar foliar content was observed, and even they were within the optimum ranges (Sanz et al., 1991).
The lower tree vigour observed in the OMO might be the cause of the delay in the initiation of bud breaking and flowering and the shorter flowering period. The difference of vigour also led to different yields in both orchards for most of the cultivars.
Another relevant difference observed between both orchards was that defoliation occurred approximately one month earlier in the OMO, and this symptom was especially evident for cultivars suffering high levels of rust disease. García-Galavís et al. (2009) evaluated the susceptibility of the cultivars used in this assay to different pests and diseases and they found that rust fungal disease affected more intensively the OMO. Therefore, in organic plum orchards, in addition to optimizing the application of organic fertilizers, it is important to achieve adequate control mechanisms against rust disease or to use the least susceptible cultivars to this fungus. However, other physiological factors could be involved, because cultivars with low susceptibility to rust disease, such as 'Red Beaut' and 'Souvenir', also showed an earlier leaf drop in the OMO (Fig. 2). As local weather conditions during the years of the study can be considered normal, it is conceivable that they may have not caused an atypical distortion in the observed results. Only specific aspects deserve to be highlighted; for example in 2010 the cultivar 'Friar' only sparsely flowered due to the low number of chilling hours, and this occurred equally in both orchards. Moreover, technical difficulties prevented to determine adequately the yield of the late variety 'Angeleno'. Fruit quality parameters were reported recently (Daza et al., 2012).
We think that this study highlights on some physiological and phenological aspects of the stone fruit trees managed in organic agriculture and could explain the lower yield usually observed in the organic systems (De Ponti et al., 2012).
Thus, it seems clear that even on fertile soil, as was in this study, the OMO leads to plum trees with lower vigour than the CMO. Given that leaf and soil compositions did not show nutrient differences that would adequately explain the differences likely the premature defoliation observed in the organic treatment causes a decreases in the accumulation of nitrogen and carbohydrate reserves in the roots and branches before the winter dormancy phase, which might also influence the vigour of trees and also the productivity. We are currently wor-king on this hypothesis, but this is complex to analyze in mature trees, and our results are still preliminary.
In this work was not conducted a comprehensive study of the economic cost in both types of management, but we can indicate some general aspects. The cost of the products used to combat pests and diseases was about 20% higher in the organic orchard. Irrigation and tillage were similar, and fertilization was somewhat more costly in the organic system, but pruning was more economical. In conclusion, we believe that organic plum production was somewhat more expensive mainly due to its lower yield.