论文部分内容阅读
Abstract [Objectives] This study was conducted to analyze the internal causes of flower reversal in Longan （Dimocarpus longan Lour.） trees. [Methods] With flowering trees and flower reversal trees as experimental materials， the variations in sugar and starch in mature leaves， tender leaves， mature branches， twigs and terminal buds after flower forcing were analyzed. [Results] During flowering process， sugar content showed the greatest difference between flowering and flower reversal trees， and the difference was the greatest in mature leaves. Trees with mature leaves having a sugar content above 44.71 mg/g were found to be more prone to flowering， while those with leaf sugar content lower than 27.80 mg/g were susceptible to flower reversal. In addition， longan trees with a higher sugar content in tender leaves were not prone to flower reversal. [Conclusions] In future， whether off??season flower forcing can be performed on longan trees could be judged through the detection of tree leaves， which is of great significance to prevention of flower reversal in off??season longan production.
Key words Dimocarpus longan； Flower induction； flower reversal； Starch； Sugar
Longan （Dimocarpus longan Lour.） as an important tropical and subtropical evergreen fruit tree has very important economic value. Hainan Island is the only province performing off??season longan production in China. However， during the flowering forcing of off??season longan， flower reversal is caused easily due to the comprehensive influences from improper environment and inherent factors in trees. Poor flowering quality and poor inflorescence development are one of the reasons for instable longan yield， which seriously restricts the development of off??season longan industry.
The supply of carbohydrates plays an important role of regulating flowering and fruit setting. In order to make clear the correlation between flower reversal of off??season longan and the supply of carbohydrates， with flowering and flower reversal off??season longan trees as experimental materials， the differences in carbohydrate accumulation mode in various tree parts were analyzed， so as to find the reasons for flower reversal of off??season longan. This study could provide reference for off??season longan production.
Materials and Methods
Shixia longan trees were selected as the experimental materials. In flowering season， removal of on??season flower clusters and pruning were performed to ensure the accumulation of enough nutrients in trees. After pruning， sprouting happened for three times， and flower forcing was performed when the leaves on the shoots sprouted during the third time had grown sufficiently， but had not reached the mature stage. Flower forcing was performed by external application of potassium chlorate and potassium chloride. The success probability of off??season flower forcing is generally around 80%， so 15 nine??year??old longan trees subjected to off??season flower forcing were selected for independent sampling. The trees should have coincident stem diameter， tree height， crown size and shoot growth. Since flower forcing， mature branches， twigs， mature leaves， tender leaves and terminal buds were collected from the 15 trees every week， until one week after the formation of lateral flower primordium on flowering trees. The samples were divided into a flower reversal group and flowering group. After classification， samples were dried and grinded， followed by preservation for later use. Sugar was extracted with 80% （v/v） ethanol， and determined by HPLC[1]， and strarch content was determined by colorimetric method[2].
Results and Analysis
It could be seen from Fig. 1 that before flower forcing with potassium chlorate， there was a great difference in sugar content in mature leaves between the two groups of trees with similar growth vigor. Specifically， before flower forcing， mature leaves on flowering trees had a sugar content reaching 44.71 mg/g， while flower reversal trees showed a sugar content only of 27.80 mg/g. After flower forcing， the sugar content in mature leaves of flowering trees decreased slowly， and at the time of lateral flower primordium formation， the value was as low as 26.08 mg/g， which was lower than flower reversal trees， while flower reversal trees had no big variation in sugar content in mature leaves， which fluctuated in the range of 25-31 mg/g.
Compared with sugar content， no much starch was accumulated in mature leaves of off??season longan. Before flower forcing， the starch content was 6.22 mg/g in mature leaves of off??season flowering longan tress， and 9.60 mg/g in those of flower reversal trees. After flower forcing， the starch content decreased in mature leaves of flower reversal trees， but slightly increased in mature leaves of flowering trees. Overall， flowering forcing resulted in a higher starch content in mature leaves of flowering trees than flower reversal trees， until the formation of lateral flower primordium （28 d after flower forcing）， the starch content in mature leaves of flowering trees was 8.89 mg/g， and that in mature leaves of flower reversal trees was 6.03 mg/g. However， 7 d after the formation of lateral flower primordium， the starch content in mature leaves of flowering trees decreased greatly， to 5.41 mg/g， which was lower than the level before flower forcing.
For both flowering trees and flower reversal trees， the variation of sugar content in tender leaves differed greatly from that in mature leaves. Flower forcing caused an increase of sugar content in tender leaves of the two groups. Before flower forcing， the sugar content in tender leaves of flowering trees was 32.35 mg/g， and the sugar content in flower reversal trees was 5.66 mg/g， which was 17.50% of that in flowering trees. Then， 21 d after flower forcing， the sugar content in tender leaves of flowering trees increased to 47.56 mg/g， and then decreased slightly， which might be due to the stress response in tender leaves during rapid conversion to mature leaves. After the formation of lateral flower primordium， sugar content in tender leaves on flowering trees decreased rapidly to the level before flower forcing. Twenty eight days after flowering forcing， the sugar content in tender leaves of flower reversal trees was 25.82 mg/g， which was 7.06% of that in tender leaves of flowering trees （Fig. 3）. It could be seen from Fig. 4 that the starch contents in flowering trees and flower reversal trees were both lower than their sugar contents， but there was also a larger difference in starch content in tender leaves between flowering trees and flower reversal trees. Before flowering， the starch content in tender leaves of flowering trees was only 5.87 mg/g， while at this time， the starch content in tender leaves on flower reversal trees was as high as 15.77 mg/g. Flower forcing with potassium chlorate caused a decrease of starch content in tender leaves， and until 28 d after flower forcing when the lateral flower primordium was formed， starch contents in flowering trees and flower reversal trees both decreased to a very low level. At this time， the starch content in tender leaves of flowering trees was 1.40 mg/g， and that in flower reversal trees was 1.54 mg/g.
The starch content exhibited similar variation trends in mature branches on flowering trees and flower reversal trees after flower forcing with potassium chlorate. The starch content increased slightly after flower forcing， decreased then， and increased again 21 d after flower forcing， reaching a peak value on the 28th day after flower forcing. When the lateral flower primordium was formed （28 d after flower forcing）， the sugar content was 24.44 mg/g in mature branches on flower reversal trees， and 18.93 mg/g in mature branches of flowering trees （Fig. 5）.
To determine whether the starch content in mature branches is related to off??season longan flowering process， the variration of starch content in mature branches was analyzed during the flower forcing （Fig. 6）. Before flower forcing， flowering trees and flower reversal trees had close starch contents in mature branches. The starch content showed the same increasing trend for a time， but differed since the 14th day after flower forcing. Specifically， the starch content in mature branches of flower reversal trees decreased rapidly， and reached a very low level on the 28th day after flower forcing， only of 3.79 mg/g， while the starch content in mature branches on flowering trees showed a decrease until the formation of lateral flower primordium， to 9.87 mg/g， which was 2.60 times of that in mature branches of flower reversal trees. The starch content in mature branches of flowering trees was significantly higher than that in those of flower reversal trees during flower forcing.
Agricultural Biotechnology 2018 Twigs play an important role of delivering nutrition to terminal bud， serving as a bridge in flowring process. High surgar storage in twigs is beneficial to flower bud differentiation of terminal buds. Therefore， we detected the variation of sugar in twigs after flowering forcing. Before flower forcing， flowering trees and flower reversal trees had close sugar contents， of 22.8 and 21.80 mg/g， respectively. The values increased slightly 7 d after flower forcing， and then decreased on the 14th day after flower forcing to diffrent degrees. The sugar content decreased to 19.76 mg/g in twigs on flowering trees， and to 10.00 mg/g in twigs on flower reversal trees， on the 14th day after flower forcing. Then， the sugar content in twigs of flower reversal trees increased rapidly to a level appoximate to that in twigs of flowering trees， reaching 19.45 mg/g， followed by a steady varying trend until the formation of lateral flower primordium. It can be seen from the variation of sugar content in twigs on flowering trees that a stable sugar content in twigs is of great significance to flowering of longan， as it not only could ensure a steady flow of sugar to terminal buds， but also is an important garantee for the morphogenesis of twigs （Fig. 7）. Starch is a kind of storage substance， which do not act on flowering process so directly like sugar， but it could convert with sugar mutually under certain conditions， thereby realizing its circulation or storage function.
It could be seen from Fig. 8 that flowering trees and flower reversal trees had close starch contents， which were 3.09 and 3.19 mg/g， respectively. Flower forcing caused the increase of starch content in twigs on flowering and flower reversal trees to different degrees. The starch content in twigs on flowering trees reached its peak value on the 28th day after flower forcing， of 8.23 mg/g. The starch content in twigs on flower reversal trees reached is peak value on the 14th day after flower forcing， of 6.34 mg/g， and on the 28th day afer flower forcing， i.e.， at the time when the lateral flower primordium was formed， the starch content in twigs of flower reversal trees was 3.30 mg/g， which was 40.10% of the starch content in twigs of flowering trees. It could be seen that before the formation of lateral flower primordium， twigs of flowering trees mainly accumulated starch， and after the formation of lateral flower primordium， the decrease of starch content might be because that the formation of flower buds needs more carbohydrates for morphogenesis.
Terminal bubs are the organ where flower bud differentiation happens， and therefore， the variation of sugar content in terminal buds is very important. There was no big difference in sugar content between flowering trees and flower reversal trees before flower forcing， and the sugar content in terminal buds on flowering trees was slightly higher than that in terminal buds on flower reversal trees. Specifically， the sugar contents in flowering and flower reversal trees were 23.52 and 20.52 mg/g， respectively. Flower forcing caused the decrease of sugar content in terminal buds on flowering trees. On the 21th day after flower forcing， the sugar content decreased to 15.43 mg/g； and then， at the time of lateral flower primordium formation， the sugar content in terminal buds on flowering trees increased rapidly to the level before flower forcing， and the value increased continuously with the progress of flower bud differentiation. Flower forcing performed on flower reversal trees also caused the decrease of sugar content in terminal buds. The sugar content in terminal buds of flower reversal trees showed a low value on the 14th day after flower forcing， as low as 12.01 mg/g， and then increased continously， reaching 24.76 mg/g on the 35th day after flower forcing （Fig. 9）. It could be seen from Fig. 10 that within 35 d after flower forcing， the starch content in terminal buds of flower reversal trees was slightly higher than that of flowering trees. Before flower forcing， the starch content in terminal buds on flower reversal trees was 6.78 mg/g， while that in terminal buds on flowering trees was 4.03 mg/g. After flower forcing， the starch content in terminal buds on flower reversal trees increased steadily， reaching 14.72 mg/g on the 35th day after flower forcing. In the whole flower bud differentiation processs， the starch content in terminal buds of flowering trees showed a steady increasing trend， and until the formation of lateral flower primordium （28 d after flower forcing）， the starch content in terminal buds was 6.37 mg/g， which was 1.58 times of that before flower forcing.
Conclusions and Disucssion
More than 70% of longan trees are used for off??season production， and off??season longan has a stable price which is about three times of on??season fruit as well as good sales volume. However， flower reversal after flower forcing is the bottleneck restrciting off??season longan industry， and skilled technicians only could achieve a success probability of 80%[3]. Therefore， the cuases for flower reversal of off??season longan are beneficial to the rescue of industrial loss. When which trees would flower and which trees would not flower are not determined， the sample quantity should be large enough. Therfore， according to the success probability of 80% on off??season flower forcing， more than ten plants are more reasonable， the flower forcing results in this study also verify the rational design of the sample quantity.
Flower reversal refers to the phenomenon that flower buds are aborted and small leaves grow. The nutritional level of leaves at this stage is closely related to flower reversal. This study demonstrated that the supply of carbohydrates before flower forcing is very important to the success of flower forcing （Fig. 1??Fig. 4）， and the supply of enough carbohydrates is the precondition of successful flower forcing. In flower formation process， 29% of carbondrates would be supplied for terminal buds for the preparation of flowering[4]， and leaves provide 91.6% of carbohydrates needed by trees[5]. Therefore， the storage level of carbohydrates in leaves at the physiological differentiation stage of flower buds is an important factor deciding flower formation.
Branches serve as the morphological support and assimilation product circulating channel of trees， and the carbohydrate level in them reflects the strength of their function. It was reported that carbohydrates in branches account for 42% of total assimilation product[4]. Whether for mature branches or for twigs， carbonhydrate level might be related to flower formation from terminal buds. In this study， it was found that the sugar content in mature branches of flowering trees was lower than that in mature branches of flower reversal trees， but flower forcing caused rapid rising of starch content in mature branches of flowering trees， while flower reversal trees exhibited no obvous variation of starch content in mature branches after flower forcing. Wherther for flowering trees or flower reversal trees， flower forcing had no high effect on twigs. A high assimilation product content in terminal buds is beneficial to the retarding of terminal bud aging and remaining of terminal bud activity[5]. Terminal buds are the the organ where flower bud differentiation happens， and the morphogenesis of terminal buds also needs to consume a large quantity of carbohydrates. Therefore， carbohydrate contents in terminal buds before flower forcing is of great significance to flower formation. However， the flowering trees and flower reversal trees both showed an increasing trend of sugar in termnal buds after flower forcing， which might be because that the formation of flower buds or leaf buds is a nutrition absortion and reconstruction process.
To avoid flower reversal， carbohydrate level in leaves beforeflower forcing is of great significance to flower formation. Therefore， this study investigated the carbohydrate contents in varous tree parts before flower forcing， which will be beneficial to the determination of the optimal period for flower forcing of off??season longan and further formation of acceptable tree nutrition standards for flower forcing. In future， whether off??season flower forcing can be performed on longan trees could be judged through the detection of tree leaves， which is of great significance to prevention of flower reversal in off??season longan production.
References
[1] BLANCO GOMIS D， MURO TAMAYO D， et al. Detection of apple juice concentrate in the manufacture of natural and sparkling cider by means of HPLC chemometric sugar analyses[J]. Journal of agricultural and food chemistry， 2004， 52（2）： 201-203.
[2] ROOD SB， LARSEN KM. Gibberellins， amylase， and the onset of heterosis in maize seedlings[J]. Journal of experimental botany， 1988， 39（2）： 223-233.
[3] HONG JW， LI SG， ZHANG L， et al. Carbon analysis on flowering differentiation in off??season longan[J]. Guangdong Agricultural Sciences， 2014， 41（16）： 37-39， 44.
[4] VAILLANT??GAVEAU N， MAILLARD P， WOJNAROWIEZ G， et al. Inflorescence of grapevine （Vitis vinifera L.）： a high ability to distribute its own assimilates[J]. Journal of experimental botany， 2011， 62（12）： 4183-4190.
[5] ANTLFINGER A， WENDEL L. Reproductive effort and floral photosynthesis in Spiranthes cernua （Orchidaceae）[J]. American journal of botany， 1997， 84（6）： 769.
[6] KELLY MO， DAVIES PJ. Photoperiodic and genetic control of carbon partitioning in peas and its relationship to apical senescence[J]. Plant physiology， 1988， 86（3）： 978-982.
Key words Dimocarpus longan； Flower induction； flower reversal； Starch； Sugar
Longan （Dimocarpus longan Lour.） as an important tropical and subtropical evergreen fruit tree has very important economic value. Hainan Island is the only province performing off??season longan production in China. However， during the flowering forcing of off??season longan， flower reversal is caused easily due to the comprehensive influences from improper environment and inherent factors in trees. Poor flowering quality and poor inflorescence development are one of the reasons for instable longan yield， which seriously restricts the development of off??season longan industry.
The supply of carbohydrates plays an important role of regulating flowering and fruit setting. In order to make clear the correlation between flower reversal of off??season longan and the supply of carbohydrates， with flowering and flower reversal off??season longan trees as experimental materials， the differences in carbohydrate accumulation mode in various tree parts were analyzed， so as to find the reasons for flower reversal of off??season longan. This study could provide reference for off??season longan production.
Materials and Methods
Shixia longan trees were selected as the experimental materials. In flowering season， removal of on??season flower clusters and pruning were performed to ensure the accumulation of enough nutrients in trees. After pruning， sprouting happened for three times， and flower forcing was performed when the leaves on the shoots sprouted during the third time had grown sufficiently， but had not reached the mature stage. Flower forcing was performed by external application of potassium chlorate and potassium chloride. The success probability of off??season flower forcing is generally around 80%， so 15 nine??year??old longan trees subjected to off??season flower forcing were selected for independent sampling. The trees should have coincident stem diameter， tree height， crown size and shoot growth. Since flower forcing， mature branches， twigs， mature leaves， tender leaves and terminal buds were collected from the 15 trees every week， until one week after the formation of lateral flower primordium on flowering trees. The samples were divided into a flower reversal group and flowering group. After classification， samples were dried and grinded， followed by preservation for later use. Sugar was extracted with 80% （v/v） ethanol， and determined by HPLC[1]， and strarch content was determined by colorimetric method[2].
Results and Analysis
It could be seen from Fig. 1 that before flower forcing with potassium chlorate， there was a great difference in sugar content in mature leaves between the two groups of trees with similar growth vigor. Specifically， before flower forcing， mature leaves on flowering trees had a sugar content reaching 44.71 mg/g， while flower reversal trees showed a sugar content only of 27.80 mg/g. After flower forcing， the sugar content in mature leaves of flowering trees decreased slowly， and at the time of lateral flower primordium formation， the value was as low as 26.08 mg/g， which was lower than flower reversal trees， while flower reversal trees had no big variation in sugar content in mature leaves， which fluctuated in the range of 25-31 mg/g.
Compared with sugar content， no much starch was accumulated in mature leaves of off??season longan. Before flower forcing， the starch content was 6.22 mg/g in mature leaves of off??season flowering longan tress， and 9.60 mg/g in those of flower reversal trees. After flower forcing， the starch content decreased in mature leaves of flower reversal trees， but slightly increased in mature leaves of flowering trees. Overall， flowering forcing resulted in a higher starch content in mature leaves of flowering trees than flower reversal trees， until the formation of lateral flower primordium （28 d after flower forcing）， the starch content in mature leaves of flowering trees was 8.89 mg/g， and that in mature leaves of flower reversal trees was 6.03 mg/g. However， 7 d after the formation of lateral flower primordium， the starch content in mature leaves of flowering trees decreased greatly， to 5.41 mg/g， which was lower than the level before flower forcing.
For both flowering trees and flower reversal trees， the variation of sugar content in tender leaves differed greatly from that in mature leaves. Flower forcing caused an increase of sugar content in tender leaves of the two groups. Before flower forcing， the sugar content in tender leaves of flowering trees was 32.35 mg/g， and the sugar content in flower reversal trees was 5.66 mg/g， which was 17.50% of that in flowering trees. Then， 21 d after flower forcing， the sugar content in tender leaves of flowering trees increased to 47.56 mg/g， and then decreased slightly， which might be due to the stress response in tender leaves during rapid conversion to mature leaves. After the formation of lateral flower primordium， sugar content in tender leaves on flowering trees decreased rapidly to the level before flower forcing. Twenty eight days after flowering forcing， the sugar content in tender leaves of flower reversal trees was 25.82 mg/g， which was 7.06% of that in tender leaves of flowering trees （Fig. 3）. It could be seen from Fig. 4 that the starch contents in flowering trees and flower reversal trees were both lower than their sugar contents， but there was also a larger difference in starch content in tender leaves between flowering trees and flower reversal trees. Before flowering， the starch content in tender leaves of flowering trees was only 5.87 mg/g， while at this time， the starch content in tender leaves on flower reversal trees was as high as 15.77 mg/g. Flower forcing with potassium chlorate caused a decrease of starch content in tender leaves， and until 28 d after flower forcing when the lateral flower primordium was formed， starch contents in flowering trees and flower reversal trees both decreased to a very low level. At this time， the starch content in tender leaves of flowering trees was 1.40 mg/g， and that in flower reversal trees was 1.54 mg/g.
The starch content exhibited similar variation trends in mature branches on flowering trees and flower reversal trees after flower forcing with potassium chlorate. The starch content increased slightly after flower forcing， decreased then， and increased again 21 d after flower forcing， reaching a peak value on the 28th day after flower forcing. When the lateral flower primordium was formed （28 d after flower forcing）， the sugar content was 24.44 mg/g in mature branches on flower reversal trees， and 18.93 mg/g in mature branches of flowering trees （Fig. 5）.
To determine whether the starch content in mature branches is related to off??season longan flowering process， the variration of starch content in mature branches was analyzed during the flower forcing （Fig. 6）. Before flower forcing， flowering trees and flower reversal trees had close starch contents in mature branches. The starch content showed the same increasing trend for a time， but differed since the 14th day after flower forcing. Specifically， the starch content in mature branches of flower reversal trees decreased rapidly， and reached a very low level on the 28th day after flower forcing， only of 3.79 mg/g， while the starch content in mature branches on flowering trees showed a decrease until the formation of lateral flower primordium， to 9.87 mg/g， which was 2.60 times of that in mature branches of flower reversal trees. The starch content in mature branches of flowering trees was significantly higher than that in those of flower reversal trees during flower forcing.
Agricultural Biotechnology 2018 Twigs play an important role of delivering nutrition to terminal bud， serving as a bridge in flowring process. High surgar storage in twigs is beneficial to flower bud differentiation of terminal buds. Therefore， we detected the variation of sugar in twigs after flowering forcing. Before flower forcing， flowering trees and flower reversal trees had close sugar contents， of 22.8 and 21.80 mg/g， respectively. The values increased slightly 7 d after flower forcing， and then decreased on the 14th day after flower forcing to diffrent degrees. The sugar content decreased to 19.76 mg/g in twigs on flowering trees， and to 10.00 mg/g in twigs on flower reversal trees， on the 14th day after flower forcing. Then， the sugar content in twigs of flower reversal trees increased rapidly to a level appoximate to that in twigs of flowering trees， reaching 19.45 mg/g， followed by a steady varying trend until the formation of lateral flower primordium. It can be seen from the variation of sugar content in twigs on flowering trees that a stable sugar content in twigs is of great significance to flowering of longan， as it not only could ensure a steady flow of sugar to terminal buds， but also is an important garantee for the morphogenesis of twigs （Fig. 7）. Starch is a kind of storage substance， which do not act on flowering process so directly like sugar， but it could convert with sugar mutually under certain conditions， thereby realizing its circulation or storage function.
It could be seen from Fig. 8 that flowering trees and flower reversal trees had close starch contents， which were 3.09 and 3.19 mg/g， respectively. Flower forcing caused the increase of starch content in twigs on flowering and flower reversal trees to different degrees. The starch content in twigs on flowering trees reached its peak value on the 28th day after flower forcing， of 8.23 mg/g. The starch content in twigs on flower reversal trees reached is peak value on the 14th day after flower forcing， of 6.34 mg/g， and on the 28th day afer flower forcing， i.e.， at the time when the lateral flower primordium was formed， the starch content in twigs of flower reversal trees was 3.30 mg/g， which was 40.10% of the starch content in twigs of flowering trees. It could be seen that before the formation of lateral flower primordium， twigs of flowering trees mainly accumulated starch， and after the formation of lateral flower primordium， the decrease of starch content might be because that the formation of flower buds needs more carbohydrates for morphogenesis.
Terminal bubs are the organ where flower bud differentiation happens， and therefore， the variation of sugar content in terminal buds is very important. There was no big difference in sugar content between flowering trees and flower reversal trees before flower forcing， and the sugar content in terminal buds on flowering trees was slightly higher than that in terminal buds on flower reversal trees. Specifically， the sugar contents in flowering and flower reversal trees were 23.52 and 20.52 mg/g， respectively. Flower forcing caused the decrease of sugar content in terminal buds on flowering trees. On the 21th day after flower forcing， the sugar content decreased to 15.43 mg/g； and then， at the time of lateral flower primordium formation， the sugar content in terminal buds on flowering trees increased rapidly to the level before flower forcing， and the value increased continuously with the progress of flower bud differentiation. Flower forcing performed on flower reversal trees also caused the decrease of sugar content in terminal buds. The sugar content in terminal buds of flower reversal trees showed a low value on the 14th day after flower forcing， as low as 12.01 mg/g， and then increased continously， reaching 24.76 mg/g on the 35th day after flower forcing （Fig. 9）. It could be seen from Fig. 10 that within 35 d after flower forcing， the starch content in terminal buds of flower reversal trees was slightly higher than that of flowering trees. Before flower forcing， the starch content in terminal buds on flower reversal trees was 6.78 mg/g， while that in terminal buds on flowering trees was 4.03 mg/g. After flower forcing， the starch content in terminal buds on flower reversal trees increased steadily， reaching 14.72 mg/g on the 35th day after flower forcing. In the whole flower bud differentiation processs， the starch content in terminal buds of flowering trees showed a steady increasing trend， and until the formation of lateral flower primordium （28 d after flower forcing）， the starch content in terminal buds was 6.37 mg/g， which was 1.58 times of that before flower forcing.
Conclusions and Disucssion
More than 70% of longan trees are used for off??season production， and off??season longan has a stable price which is about three times of on??season fruit as well as good sales volume. However， flower reversal after flower forcing is the bottleneck restrciting off??season longan industry， and skilled technicians only could achieve a success probability of 80%[3]. Therefore， the cuases for flower reversal of off??season longan are beneficial to the rescue of industrial loss. When which trees would flower and which trees would not flower are not determined， the sample quantity should be large enough. Therfore， according to the success probability of 80% on off??season flower forcing， more than ten plants are more reasonable， the flower forcing results in this study also verify the rational design of the sample quantity.
Flower reversal refers to the phenomenon that flower buds are aborted and small leaves grow. The nutritional level of leaves at this stage is closely related to flower reversal. This study demonstrated that the supply of carbohydrates before flower forcing is very important to the success of flower forcing （Fig. 1??Fig. 4）， and the supply of enough carbohydrates is the precondition of successful flower forcing. In flower formation process， 29% of carbondrates would be supplied for terminal buds for the preparation of flowering[4]， and leaves provide 91.6% of carbohydrates needed by trees[5]. Therefore， the storage level of carbohydrates in leaves at the physiological differentiation stage of flower buds is an important factor deciding flower formation.
Branches serve as the morphological support and assimilation product circulating channel of trees， and the carbohydrate level in them reflects the strength of their function. It was reported that carbohydrates in branches account for 42% of total assimilation product[4]. Whether for mature branches or for twigs， carbonhydrate level might be related to flower formation from terminal buds. In this study， it was found that the sugar content in mature branches of flowering trees was lower than that in mature branches of flower reversal trees， but flower forcing caused rapid rising of starch content in mature branches of flowering trees， while flower reversal trees exhibited no obvous variation of starch content in mature branches after flower forcing. Wherther for flowering trees or flower reversal trees， flower forcing had no high effect on twigs. A high assimilation product content in terminal buds is beneficial to the retarding of terminal bud aging and remaining of terminal bud activity[5]. Terminal buds are the the organ where flower bud differentiation happens， and the morphogenesis of terminal buds also needs to consume a large quantity of carbohydrates. Therefore， carbohydrate contents in terminal buds before flower forcing is of great significance to flower formation. However， the flowering trees and flower reversal trees both showed an increasing trend of sugar in termnal buds after flower forcing， which might be because that the formation of flower buds or leaf buds is a nutrition absortion and reconstruction process.
To avoid flower reversal， carbohydrate level in leaves beforeflower forcing is of great significance to flower formation. Therefore， this study investigated the carbohydrate contents in varous tree parts before flower forcing， which will be beneficial to the determination of the optimal period for flower forcing of off??season longan and further formation of acceptable tree nutrition standards for flower forcing. In future， whether off??season flower forcing can be performed on longan trees could be judged through the detection of tree leaves， which is of great significance to prevention of flower reversal in off??season longan production.
References
[1] BLANCO GOMIS D， MURO TAMAYO D， et al. Detection of apple juice concentrate in the manufacture of natural and sparkling cider by means of HPLC chemometric sugar analyses[J]. Journal of agricultural and food chemistry， 2004， 52（2）： 201-203.
[2] ROOD SB， LARSEN KM. Gibberellins， amylase， and the onset of heterosis in maize seedlings[J]. Journal of experimental botany， 1988， 39（2）： 223-233.
[3] HONG JW， LI SG， ZHANG L， et al. Carbon analysis on flowering differentiation in off??season longan[J]. Guangdong Agricultural Sciences， 2014， 41（16）： 37-39， 44.
[4] VAILLANT??GAVEAU N， MAILLARD P， WOJNAROWIEZ G， et al. Inflorescence of grapevine （Vitis vinifera L.）： a high ability to distribute its own assimilates[J]. Journal of experimental botany， 2011， 62（12）： 4183-4190.
[5] ANTLFINGER A， WENDEL L. Reproductive effort and floral photosynthesis in Spiranthes cernua （Orchidaceae）[J]. American journal of botany， 1997， 84（6）： 769.
[6] KELLY MO， DAVIES PJ. Photoperiodic and genetic control of carbon partitioning in peas and its relationship to apical senescence[J]. Plant physiology， 1988， 86（3）： 978-982.