Trop. Agr. Develop. 60(2):71-80,2016 Dry Matter Production and Distribution after Trunk Formation in Sago Palm (Metroxylon sasu Rottb.) Yoshinori YAMAMOTO 1, *, Kazuki OMORI 1, Youji NITTA 2, Kenichi KAKUDA 3, Yulius Barra PASOLON 4, Fransiscus Suramas REMBON 4, Ray Sadimantara GUSTI 4, Aysyah Anas ARSY 4, Akira MIYAZAKI 1, and Tetsushi YOSHIDA 1 1 Faculty of Agriculture, Kochi University, B-200 Monobe, Nankoku, Kochi 783-8502, Japan 2 Faculty of Agriculture, Ibaraki University, Ami, Ibaraki 300-0393, Japan 3 Faculty of Agriculture, Yamagata University, Tsuruoka, Yamagata 997-8555, Japan 4 Faculty of Agriculture, Haluoleo University, Kendari, Southeast Sulawesi 93232, Indonesia Abstract The differences in dry matter production and its distribution between the two folk varieties of sago palm with higher (Molat) and lower (Rotan) starch productivity were clarified in Kendari, Southeast Sulawesi, Indonesia. The fresh and dry matter weights of whole shoots of the two varieties increased exponentially with age after trunk formation and these differences were 2.7 times higher in Molat than in Rotan at the harvesting stage. Differences were not found between the two varieties after trunk formation in the ratios of whole leaf, whole trunk, and the harvested trunk (from the trunk base to the node of the lowest living leaf) dry weights to the whole shoot weight and the ratio of pith dry weight to the harvested trunk dry weight. The difference in whole shoot weight between the two varieties might be caused by the difference in leaf area per palm (2.9 times difference) observed in the previous report. These results suggested that the difference in starch productivity between Molat and Rotan resulted from the difference in dry matter production, but not from the difference in the distribution of the dry matter to the harvested trunk. Key words: Biomass, Dry matter percentage, Dry matter ratio, Folk variety, Southeast Sulawesi, Starch yield Introduction only a few reports are available on biomass production in sago palm (Flach and Schuiling 1991; Kaneko et al. Although starch productivity in sago palm has been 1996), and no reports described the changes in whole reported to be affected by environmental conditions plant and each organ s or part s biomass production and (Sim and Ahmad, 1978; Kueh et al., 1991; Yamamoto the ratio of each organ s or part s biomass with age. et al., 2003) and by folk varieties (Shimoda and Power, We analyzed the difference in leaf characters of 1986; Yamamoto, 1999; Yamamoto et al., 2006, 2010) sago palm varieties with higher (Molat) and lower (hereafter; varieties), it would be desirable to clarify (Rotan) starch productivity after trunk formation in a the productivity differences depending on dry matter previous report (Yamamoto et al., 2014). The objective of distribution. The amount of starch contained in the pith the present study was to clarify the differences in whole of sago palm is mainly determined by the matter produc- shoot and organ or part weights and the ratios of each tion resulting from the leaf area and the photosynthetic organ or part to the whole shoot to explain the varietal rate of the leaf, and the subsequent distribution of the difference in starch productivity. matter to each organ or part. Several aspects related to dry matter production Materials and Methods in sago palm have been reported, including the char- The study was based on the same palms sampled acters of leaflet and leaf (Nakamura et al., 2004, 2005), in the previous report (Yamamoto et al., 2014), and was the estimation method of leaf area (Omori et al., 2000; carried out in a cultivated sago palm garden at Lakomea Nakamura et al., 2009), the difference in leaf area among Village in Kendari, Southeast Sulawesi Province, Indothe sago palms growing in different districts (Flach and nesia, in mid-september, 2000. The soil type was mineral Schuiling, 1991), and the photosynthetic rate (Flach, soil (Rembon et al., 2010) and the underground water 1977; Uchida et al., 1990; Miyazaki et al., 2007). However, table ranged from 0 to 30 cm in the dry season to 50 cm in the wet season. Weeding and sucker control were Communicated by H. Ehara practiced one to two times per year and no fertilizers Received Feb. 25, 2015 were applied. Accepted Oct. 9, 2015 * Corresponding author Sago palms of varying years after trunk formation yamayosi@kochi-u.ac.jp (YATF) from the two varieties, Molat (local name: Roe)
72 Trop. Agr. Develop. 60(2)2016 and Rotan (ibid.: Rui), were sampled from the area in the research garden where growth conditions were normal. Numbers of sampled palms were 13 for Molat and 9 for Rotan (estimated age of both varieties was 1 7 YATF) (Table 1). The palm at 7 YATF was at the flowering stage for Rotan. The palms at 7 YATF for Molat and at 5.5 6 YATF for Rotan were estimated to be at the flower bud formation stage (Yamamoto et al., 2010). The selected palms were cut at their base with a chainsaw and assessed for total length (from the base of cut trunk to the tip of the top leaflet), numbers of living leaves, and numbers of leaf scars after removal of the debris on the trunk. The leaflet, rachis, and petiole + leaf sheath fresh weights were measured for all the expanded and unexpanded leaves in each sampled palm. In addition, small samples (50 100g) were taken from the central portion of each part of the base, middle, and top leaves for the expanded leaves, as well as in all the leaves for the unexpanded leaves. These samples were measured for their fresh weights with an electronic balance (HL-200 type, Kagakukyoei Co., Ltd). The dry weights of the samples were also measured after drying in an electric oven for 2 days at 85 C, followed by 2 days at 65 C. The dry weight of each part of the expanded leaf was determined by multiplying the fresh weight of each part and its average dry matter percentage (DMP) calculated by the formula; [DMP of the lowest leaf + (DMP of the middle leaf (number of leaves 2)) + DMP of the top leaf] / number of leaves. The dry weight of each part of the unexpanded leaves was calculated by multiplying the fresh weight and the average dry matter percentage. The DMP of each part of the whole leaf was calculated by dividing the total dry weight of expanded and unexpanded leaves by the fresh weight. The trunk length was measured as the length from the base to the node of the lowest living leaf. The lower and upper parts of the trunk relative to the node were referred to as the lower and upper trunks, respectively. The lower trunk was cut into 90-cm logs from the base and all of them were weighed. The trunk diameter was calculated as the average of the distal diameter of each log. The upper trunk was cut into its proper length and weighed. Trunk disk samples for the lower trunk were taken from the central portions of all the logs and that of the bottom, middle, and top logs when the number of logs was less and more than three, respectively. The disks were divided into four equal parts from the center at right angles and one of the quartered disks was used to measure the pith and bark weights after separating them. The ratios of bark and pith weights were also calculated from these data. Additionally, samples of the pith and bark were taken and weighed for fresh weight, and then their dry weight was determined following the same procedures as those of described above for the leaves. For the upper trunk, trunk disks were taken from the middle portion and the fresh and dry weights of the pith were determined following the same procedure as that described above for the lower trunk. The average values of pith and bark dry matter percentage of the lower trunks were calculated from the values of all the logs when the number of logs was three or less and according to the formula: [LL+UL+ML (number of logs 2)] / number of logs, where LL, ML, and UL denote the values of bottom, middle, and top logs when the number of logs exceeded more than three. The dry weights of pith and bark were calculated from the fresh weight and the average dry matter percentages. The dry weights of pith and bark of the upper trunks were calculated by multiplying the fresh weight and dry matter percentage. The average dry matter percentages of pith and bark per trunk were calculated by dividing the total dry weight of pith and bark per trunk by the total fresh weight in each part. Results Changes in whole shoot fresh and dry weights after trunk formation The whole shoot fresh and dry weights increased with YATF in both varieties, ranging from 600 to 3,050 kg in Molat and from 480 to 1,120 kg in Rotan for the fresh weight, and from 100 to 1,030 kg in Molat and from 100 to 440 kg in Rotan for the dry weight (Fig. 1). A conspicuous varietal difference was observed in the weight gain per year after trunk formation, 357.8 kg year in Molat and 92.5 kg year in Rotan for the fresh weight, and 144.3 kg year in Molat and 46.5 kg year in Rotan for the dry weight, with 3.9 and 3.1 times higher values in Molat than in Rotan, respectively. The average whole shoot fresh and dry weights at the harvesting stage were 2,781 and 986 kg for Molat and 1,018 and 363 kg for Rotan. The dry matter percentage of whole shoot ranged from 17 to 41% after trunk formation and the percentage increased with YATF in both varieties (Table 1). The whole leaf dry weight of Molat increased exponentially with YATF, ranging from 76 to 261 kg, while that of Rotan, ranged from 61 to 107 kg and did not show a significant change with YATF (Fig. 2A).
Yamamoto et al.: Dry matter production and distribution in sago palm 73 Table 1. Changes in dry matter percentage in each organ or part in sago palms after trunk formation. Variety YATF 1) Leaflet Leaf (%) Rachis P+LS 2) Whole Pith Trunk (%) Bark Whole Whole shoot (%) Molat 1 45.9 33.2 16.7 25.5 7.3 19.1 9.4 17.4 1 45.1 32.6 17.1 25.7 8.0 24.8 10.4 18.8 1 45.6 32.2 17.0 25.9 7.9 25.8 11.1 19.0 2 47.1 30.9 17.0 27.0 8.9 28.1 12.5 19.3 3.5 44.8 32.2 16.8 25.9 10.2 23.3 12.7 18.0 3.5 48.5 33.1 16.9 27.4 8.7 24.4 11.4 16.6 4 47.3 35.8 17.0 27.5 8.0 23.5 10.6 16.3 4 48.8 32.0 16.9 27.2 8.7 30.0 13.4 18.4 5 45.5 32.5 21.7 30.3 25.2 38.3 27.5 28.2 5.5 50.6 36.6 21.5 31.3 25.0 32.1 26.6 28.2 7 48.4 35.8 21.7 31.2 33.8 39.0 34.7 33.7 7 51.2 35.4 21.6 32.1 34.2 41.2 35.3 34.4 7 52.5 31.5 21.5 31.4 36.2 41.3 37.1 35.5 Rotan 1 45.3 30.3 16.4 24.9 9.1 20.3 10.7 18.1 1 51.7 32.0 17.4 27.6 9.1 24.4 11.9 19.5 1 50.2 31.1 20.0 28.7 9.3 27.1 13.0 20.8 3.5 48.4 31.9 18.6 28.2 12.6 33.7 16.3 20.2 3.5 47.3 33.8 20.7 30.1 19.7 34.6 22.8 25.5 4 48.1 33.8 18.9 28.5 13.7 28.6 17.1 20.9 5.5 47.0 33.0 19.7 29.5 32.3 35.6 33.0 31.9 6.5 48.1 35.0 21.8 31.5 33.4 40.4 34.7 33.7 7 3) 46.9 41.4 27.0 36.3 43.4 43.7 42.2 40.8 1) Years after trunk formation. 2) Petiole + leaf sheath. 3) Flowering stage. The whole trunk dry weight in both varieties increased exponentially with YATF, ranging from 28 to 737 kg in Molat and from 25 to 336 kg in Rotan, and the average value of the whole trunk dry weight in Molat (721 kg) was 2.7 times higher than that in Rotan (267 kg) at the harvesting stage (Fig. 2B). The dry matter percentages of whole leaf and whole trunk in both varieties showed an increasing trend with YATF, ranging from 25 to 36% and from 9 to 42%, respectively (Table 1). The dry matter percentage of whole leaf was higher than that of the whole trunk before the harvesting stage, but it reversed at the harvesting stage. The dry weights of all the leaf components, leaflet, rachis, and petiole + leaf sheath, in Molat increased exponentially with YATF. In contrast, in Rotan, few changes were observed in the dry weights of leaflet and petiole + leaf sheath with YATF, although that of rachis increased linearly. Among the leaf components, the value of the dry weight was higher in the order of petiole + leaf sheath leaflet > rachis in both varieties, with ranges of 30 107, 23 107, and 22 77 kg in Molat Fig. 1. Relationship between years after trunk formation and and 21 38, 22 40, and 15 29 kg in Rotan, respectively. whole shoot fresh and dry weights. The dry matter percentage of each leaf component was Molat, Rotan. Closed and open symbols indicate higher in the order of leaflet > rachis > petiole + leaf fresh and dry weights, respectively. ***: significant at p < 0.001. sheath in both varieties, and little change was observed
74 Trop. Agr. Develop. 60(2)2016 Fig. 2. Relationship between years after trunk formation and whole leaf (A) or trunk dry weight (B). Molat, Rotan ***: significant at p <0.001. Fig. 3. Relationship between years after trunk formation and upper (A) or lower trunk dry weight (B). Molat, Rotan ***: significant at p < 0.001. in the percentages of leaflet and rachis with YATF, while the percentage of the petiole + leaf sheath showed an increasing tendency (Table 1). The dry weights of lower and upper trunk in Molat increased exponentially with YATF, while in Rotan only the lower trunk dry weight increased exponentially with YATF (Fig. 3). The dry weights of bark and pith in both varieties increased exponentially with YATF (Fig. 4). The pith dry weight in Molat and Rotan, ranged from 18 to 600 kg and from 18 to 270 kg, respectively, and markedly differed after 4 years of trunk formation with the average dry weight value of Molat (581 kg) at the harvesting stage being 2.7 times higher than that of Rotan (213 kg). The bark dry weight in Molat and Rotan, ranging from 11 to 153 kg and from 7 to 65 kg, respectively, also markedly differed after 4 years of trunk formation, with the average bark dry weight value at the harvesting stage in Molat being 2.6 times higher than that in Rotan (Fig. 4). The dry matter percentages of bark and pith increased with YATF with both varieties showing higher values in the bark than in the pith (Table 1). The differences between the two parts were conspicuous up to 4 years after trunk formation and became less noticeable closer to the harvesting stage. The dry matter percentage of the lower trunk changed more than that of the upper trunk after trunk formation and the largest difference was observed near the harvesting stage (Table 2). Similar to the whole trunk, the dry matter percentages of bark in both lower and upper trunks changed more than that of the pith after trunk formation. Changes in the ratios of organ or part dry weight after trunk formation The dry weight ratio of the whole leaf and trunk to
Yamamoto et al.: Dry matter production and distribution in sago palm 75 Fig. 4. Relationship between years after trunk formation and pith (A) or bark dry weight (B). Molat, Rotan ***: significant at p < 0.001. Table 2. Changes in dry matter percentages in pith and bark of upper and lower trunks after trunk formation. Lower trunk 2) (%) Upper trunk 3) (%) Variety YATF 1) Pith Bark Whole Pith Bark Whole Molat 1 7.3 19.1 10.1 7.2 11.4 8.1 1 8.0 24.8 11.8 7.8 17.0 9.0 1 7.9 25.8 11.3 7.8 23.9 10.8 2 8.9 28.1 13.7 8.2 22.1 11.0 3.5 10.2 23.3 14.4 8.0 15.3 9.2 3.5 8.7 24.4 11.8 8.3 17.3 9.6 4 8.0 23.5 11.2 7.6 14.6 8.4 4 8.7 30.0 14.6 8.0 20.1 10.4 5 25.2 38.3 29.9 9.7 17.1 10.8 5.5 25.0 32.1 31.3 9.8 18.5 11.8 7 33.8 39.0 39.9 11.9 25.7 14.1 7 34.2 41.2 40.1 12.8 24.1 14.1 7 36.2 41.3 41.1 12.1 23.0 14.3 Rotan 1 9.1 20.3 11.2 8.8 20.8 10.2 1 9.1 24.4 12.1 9.0 24.0 11.7 1 9.3 27.1 14.4 8.6 23.9 11.8 3.5 12.6 33.7 16.9 11.1 27.1 13.8 3.5 19.7 34.6 24.8 13.2 31.2 17.4 4 13.7 28.6 17.3 13.5 27.4 15.9 5.5 32.3 35.6 35.4 16.8 24.4 18.7 6.5 33.4 40.4 36.5 23.5 34.3 25.2 7 4) 43.4 43.7 46.0 33.4 40.9 35.2 1) Years after trunk formation. 2) Trunk portion from the trunk base to the node of the lowest living leaf. 3) Trunk portion from the node of the lowest living leaf to the top end of stem. 4) Flowering stage. the whole shoot dry weight in both varieties changed the harvesting stage (Table 3). similarly after trunk formation. Whole leaf to whole The dry weight ratio of each leaf part to the whole shoot dry weight ratio was 69 75% at trunk formation shoot dry weight in both varieties changed higher in stage and then 20 29% at the harvesting stage, whereas the order of petiole + leaf sheath leaflet > rachis after the trunk to the whole shoot dry weight ratios conversely trunk formation with a gradual decrease of the ratio in were 25 31% at the trunk formation stage and 71 77% at each with YATF. The ratios were approximately 10% in
76 Trop. Agr. Develop. 60(2)2016 Table 3. Changes in dry weight ratio in each organ or part in sago palms after trunk formation. Variety YATF 1) Leaf (%) Trunk (%) Whole shoot Leaflet Rachis P+LS 2) Whole Pith Bark Whole Inflorescence Wt. (kg/palm) Molat 1 23 21 29 73 17 10 27 0 104 1 25 21 30 75 16 8 25 0 139 1 25 20 28 73 16 11 27 0 115 2 25 18 22 65 20 15 35 0 210 3.5 20 16 21 58 28 15 42 0 233 3.5 20 15 18 54 29 17 46 0 216 4 20 17 20 57 27 16 43 0 233 4 21 15 19 54 23 23 46 0 282 5 11 8 11 29 53 17 71 0 396 5.5 13 10 14 37 46 17 63 0 718 7 10 7 10 28 57 15 72 0 1028 7 11 7 10 28 59 13 72 0 943 7 10 6 9 24 61 15 76 0 986 Rotan 1 26 17 28 72 21 8 28 0 86 1 27 17 25 69 20 12 31 0 118 1 25 16 28 69 18 13 31 0 104 3.5 17 12 16 45 35 20 55 0 128 3.5 16 11 16 43 39 18 57 0 211 4 16 12 17 46 33 21 54 0 134 5.5 11 8 10 29 55 16 71 0 274 6.5 11 8 10 28 57 15 72 0 379 7 3) 7 6 7 20 62 15 77 3 437 1) Years after trunk formation. 2) Petiole+leaf sheath. 3) Flowering stage. both the leaflet and the petiole + leaf sheath and about 6 8% in the rachis at the harvesting stage. The dry weight ratio of each leaf part to the whole leaf dry weight also changed similarly in both varieties after trunk formation (30 40% in the leaflet, 25 30% in the rachis, and 35 40% in the petiole + leaf sheath) and the ratio in the rachis changed by about 10% less than that in the other parts. The lower and upper trunk dry weight ratios to the whole trunk dry weight increased and decreased with YATF, respectively, and the ratio of the lower trunk was 92 94% in Molat and 78 92% in Rotan at the harvesting stage (Table 4). The bark and pith dry weight ratios to the whole shoot dry weight in both varieties changed similarly after trunk formation and the pith dry weight ratio increased from about 20% at the trunk formation stage to about 60% at the harvesting stage, in contrast to the small change in the bark dry weight ratio, 10 20% (Table 3). The changes in the dry weight ratios of the whole trunk to the whole leaf and the whole pith to the whole bark are shown in Fig. 5. Both ratios increased exponentially with YATF in both varieties, showing a slightly higher value in Rotan than in Molat and the ratio was about 3.0 and 4.0 at the harvesting stage for each variety, respectively. The changes in the ratios of fresh and dry weights of the whole pith to the whole shoot and the whole trunk and the ratio of the lower trunk pith, which corresponds to the harvested portion for starch extraction, to the lower trunk after trunk formation are shown in Table 5. The ratio of the whole pith to the whole shoot for the dry weight decreased compared to the fresh weights at 5 and 4 years after trunk formation in Molat and Rotan, respectively. The pith ratio for fresh weight increased from about 40% to about 60% with YATF in both varieties. However, the ratios of pith to the whole trunk and the lower trunk in both varieties were about 80%, irrespective of palm age. The ratios of whole pith to the whole shoot and the whole trunk and the ratio of the lower trunk pith to the lower trunk for dry weight did not show significant varietal differences and increased with YATF, with ranges of 15 60%, 60 80%, and 60 80%, respectively. Discussion Although sago palms accumulate significant amounts of starch in the trunk, the amount of starch varies among varieties and is markedly influenced by local environmental conditions (Sim and Ahmed, 1978;
Yamamoto et al.: Dry matter production and distribution in sago palm 77 Shimoda and Power, 1986; Kue et al., 1991; Yamamoto, 1999; Yamamoto et al., 2003, 2006, 2010). In a previous study, the authors described differences in starch productivity among the varieties growing in Kendari, Table 4. Dry weight ratios of trunks above and below the lowest living leaf node to whole trunk dry weight. Variety YATF 1) Lower trunk 2) (%) Upper trunk 3) (%) Molat 1 72 28 1 57 43 1 55 45 2 62 38 3.5 77 23 3.5 82 18 4 82 18 4 78 22 5 95 5 5.5 89 11 7 92 8 7 93 7 7 94 6 Rotan 1 48 52 1 46 54 1 49 51 3.5 86 14 3.5 80 20 4 83 17 5.5 92 8 6.5 88 12 7 4) 78 19 1) Years after trunk formation. 2) Trunk portion from the trunk base to the node of the lowest living leaf. 3) Trunk portion from the node of the lowest living leaf to the top end of stem. 4) Flowering stage. Southeast Sulawesi Province, Indonesia (Yamamoto et al., 2010) and speculated that the variation in starch accumulation might be caused by varietal differences in the amount and/or accumulation rate of photosynthate from the leaf area and its distribution rate to the pith (the harvested portion). In the present study, the difference in starch productivity between the two varieties (Molat, higher and Rotan, lower), in Kendari (Yamamoto et al., 2010) in terms of dry matter production was elucidated. The growth stages of the sampled palms ranged from just after trunk formation to flower bud formation stage (Molat) and flowering stage (Rotan). The whole shoot fresh and dry weights of both varieties increased exponentially with YATF. The varietal differences in both fresh and dry weights increased with palm age and those in Molat were approximately 2.7 times higher than those in Rotan at the harvesting stage. These varietal differences might be ascribed to the differences in leaf area and/or photosynthetic ability between the two varieties. The authors reported previously that the leaf area per palm in Molat was about 2.6 times higher than that in Rotan (Yamamoto et al., 2014). Alternatively, the differences in photosynthetic rate of sago palm varieties might be small, according to the data in previous reports (Miyazaki et al. 2007), although photosynthetic data are not available for Molat and Rotan. These results suggested that the large difference in whole shoot weight between Molat and Rotan was based on the difference in leaf area and the ratio of Molat to Rotan of the leaf area (about 2.6) (Yamamoto et al., 2014) was almost similar to the ratio in whole shoot (about 2.7). The whole leaf and the whole trunk dry weight ratios Fig. 5. Relationship between years after trunk formation and trunk / leaf dry weight ratio (A) or pith / bark dry weight ratio (B). Molat, Rotan *, ** and ***: significant at p < 0.05, p < 0.01 and p < 0.001, respectively.
78 Trop. Agr. Develop. 60(2)2016 Table 5. Changes in whole pith weight ratio to whole shoot and trunk weights, and lower trunk pith weight ratio to the lower trunk weight. Variety YATF 1) Pith fresh weight ratio (%) to whole shoot FW trunk FW trunk FW 2) Pith dry weight ratio (%) to whole shoot DW trunk DW trunk DW 2) Molat 1 41 82 83 17 63 60 1 1 39 38 86 82 84 83 2 43 81 82 20 57 56 3.5 48 81 80 28 65 63 3.5 56 83 82 29 64 61 4 55 83 82 27 63 59 4 50 78 77 24 51 48 5 60 82 82 53 75 75 5.5 52 78 78 46 73 74 7 57 82 81 57 79 80 7 59 85 84 59 82 82 7 60 83 83 61 81 81 Rotan 1 41 86 83 21 73 69 1 42 82 82 20 63 62 1 40 79 80 18 57 57 3.5 56 82 82 35 63 63 3.5 50 80 80 39 69 71 4 51 77 76 33 62 60 5.5 55 80 80 56 78 79 6.5 57 82 82 57 79 80 7 3) 58 76 82 62 78 82 1) Years after trunk formation. 2) Weight ratio of lower trunk (trunk portion from the base to the node of the lowest living leaf) pith to the trunk weight. 3) Flowering stage. 16 16 66 58 59 58 to the whole shoot dry weight in both varieties changed inversely from 70 75 and 30 25% to 25 30 to 75 70%, respectively, from the trunk formation stage to the harvesting stage. The dry weight ratios of whole trunk and leaf at the harvesting stage were almost the same as the values (74% and 26%, respectively) reported by Kaneko et al. (1996) for the sago palms growing in Sarawak, Malaysia. The dry weight ratios were almost identical at 3 4 years after trunk formation and thereafter, the whole trunk ratio increased with YATF, compared to the whole leaf ratio. The whole leaf dry weight in Molat, similar to the whole shoot dry weight, increased exponentially, while that in Rotan did not change apparently with YATF. In Molat, the exponential increase of the whole leaf dry weight might be caused by the small change at 5 years after trunk formation and the sharp increase thereafter. It should be noted that there may have been frequent harvestings of the lower leaves of Molat, which would be suitable for Atap (roofing material) because of its long and wide leaflets up to 5 years after trunk formation, based on the fact that the study area was located near the farmer s house (Yamamoto et al., 2014). The ratio of each leaf part (leaflet, rachis, and petiole + leaf sheath) dry weight to the whole leaf dry weight changed almost identically with YATF in both varieties, with the ratio of rachis being about 10% lower than those of the leaflet and petiole + leaf sheath. These results agreed with those obtained by Flach and Schuiling (1991) who reported that the dry weight ratios of leaflet, rachis, and petiole + leaf sheath were 40.6%, 28.1%, and 31.3%, respectively, in sago palm grown on Seram Island, Indonesia. Moreover, the dry weight ratio of each part of leaf to the whole shoot dry weight decreased with YATF. Petiole + leaf sheath dry weight decreased more rapidly than in the other two parts, due to the gradual shortening of the length with palm age (Yamamoto et al., 2014). The trunk of sago palm is usually harvested from the base to the node of the lowest living leaf for starch extraction (Yamamoto et al., 2003). In the present study, however, the upper portion of trunk from the lowest living leaf, which is not harvested for starch extraction, was also sampled to analyze the whole biomass. The dry weight ratios of the lower and upper trunks to the whole shoot were similar just after trunk formation in both varieties, but thereafter, the ratio for the lower trunk gradually increased with values of 90 95% at the harvest-
Yamamoto et al.: Dry matter production and distribution in sago palm 79 Table 6. Starch yield and dry matter production characters related to the yield. Variety Whole shoot Lower trunk b/a Lower trunk pith Wt. (kg palm ) (a) Wt. (kg palm ) (b) (%) Wt. (kg) (c) Molat 985.5±42.7* 670.4±36.2 68.0 543.6±26.3 Rotan 363.2±82.7 230.0±46.7 63.3 185.1±41.4 c/b (%) 81.1 80.5 Starch (%) 1) 68.4±4.5 69.0±2.4 1) Values quoted from Yamamoto et al. (2010). 2) Calculated from the pith dry weight and the average starch percentage. *Average ± SD. Starch yield (kg palm ) 2) 371.8±18.0 127.7±28.5 ing stage. The dry weight ratio of bark in the trunk to the whole shoot dry weight in both varieties did not differ with palm age, ranging from 10 to 20%, while the ratio of pith increased with YATF and was in the range of 55 60% at the harvesting stage. The pith ratio at the harvesting stage showed almost the same value, 51%, as that reported by Kaneko et al. (1996) for sago palm in Sarawak, Malaysia. Remarkable differences were observed in the pith weight ratios of fresh and dry weights to the whole shoot and the whole trunk and the ratios in the lower trunk and the values of dry weight ratios in both varieties were lower than those of fresh weight ratios up to about 5 years after trunk formation. This finding might be due to the lowest dry matter percentage of pith compared to other parts up to about 5 years after trunk formation (Table 1). In both varieties, the fresh weight ratios of the whole pith to the whole shoot and the pith to trunk in the lower trunk were about 80 85%, irrespective of the palm age. In contrast, the ratios in dry weight increased with YATF and reached a value of about 80% at the harvesting stage. That is, the pith weight ratios to the whole trunk weight in both fresh and dry weights at the harvesting stage were about 80%, irrespective of trunk positions. These results agreed well with those reported by Yatsugi (1977). Starch yield per palm in sago palm is determined by the formula: whole shoot dry weight lower trunk dry weight/ whole shoot dry weight pith dry weight/lower trunk dry weight the starch percentage of pith (dry weight basis), and each value is shown in Table 6. The starch percentages of both varieties corresponded to the average values of three palms at the harvesting stage in the previous report (Yamamoto et al., 2010). The results clearly indicated that the difference in starch yield between the two varieties was related to varietal differences in the whole shoot dry weight, but not in the ratios of pith weight and the starch percentage. Moreover, the value of the varietal difference in the whole shoot dry weight, 2.7 times, was almost the same (2.9 times) as that in the leaf area per palm between the two varieties (Yamamoto et al. 2014). These results suggested that the varietal difference in starch productivity was based on the difference in the dry matter production ability caused by the difference in leaf area per palm, and not on differences in the distribution ratio of photosynthate to the pith. Acknowledgement This research was supported by a Grant-in-Aid for the Tropical Bio-resources Research Program from the Japan Society for the Promotion of Science. References Flach. 1977. Yield potential of the sago palm and its realization. In Sago- 76: Papers of the 1 st international sago symposium The Equatorial swamp as a natural resource. (Tan K. ed.) (Kuching) pp.15777. Flach, M. and D. L. Schuiling 1991. Growth and yield of sago palms in relation to their nutritional needs. In: Towards Greater Advancement of Sago Industry in the 90s. (Proceedings of the 4th international sago symposium) (Ng, T. Y., Y. L. Tie and H. S. Kueh eds.) Lee Ming Press (Kuching) 10310. Kaneko, T., M. Okazaki, N. Kasai, C. Yamaguchi, and A. H. Hassan 1996. Growth and biomass of sago palm (Metroxylon sagu) on shallow peat soils of Dalat District, Sarawak. Sago Palm 4: 63. Kueh, H. S., Y. L. Tie, E. Robert, C. M. Ung, and Hj. O. Jaman 1991. The feasibility of plantation production of sago (Metroxylon sagu) on an organic soil in Sarawak. In: Proceedings of the 4th international sago symposium. (Ng, T. Y., Tie, Y. L. and H. S. Kueh eds.) Lee Ming Press (Kuching) pp.12736. Miyazaki A., Y. Yamamoto, K. Omori, H. Pranamuda, R. S. Gusti, Y. B. Pasolon, and J. Limbongan 2007. Leaf photosynthetic rate in sago palms (Metroxylon sagu Rottb.) grown under field conditions in Indonesia. Jpn. J. Trop. Agr. 51: 54-58. Nakamura, S., Y. Nitta, and Y. Goto 2004. Leaf characteristics and shape of sago palm (Metrxylon sagu Rottb.) for developing a method of estimating leaf area. Plant Prod.Sci. 7: 198-203. Nakamura, S., Y. Nitta, M. Watanabe, and Y. Goto 2005. Analysis of leaflet shape and area for improvement of leaf area estimation method for sago palm (Metroxylon sagu Rottb.). Plant Prod. Sci. 8: 27-31. Nakamura, S., Y. Nitta, M. Watanabe, and Y. Goto 2009. A method for estimating sagopalm (Metroxylon sagu Rottb.) leaf area after trunk formation. Plant Prod. Sci. 12: 58-62. Omori K., Y. Yamamoto, F. S. Jong, and T. Wenston 2000. Relationship between leaflet weight, length or width and leaflet area in sago palm (Metroxylon sagu Rottb.). Jpn. J. Trop. Agr. 44 (Extra issue 1): 156. (in Japanese) Rembon F. S., Y. B. Pasolon, Y. Yamamoto, and T. Yoshida 2010. Physicochemical properties of the soils on sago-palm (Metroxylon sagu Rottb.)-growing areas around Kendari, Province of Southeast Sulawesi, Indonesia. Sago Palm 18: 68.
80 Trop. Agr. Develop. 60(2)2016 Shimoda, H. and A. P. Power 1986. Investigation into development and utilization of sago palm forest in the East Sepik region, Papua New Guinea. In: Sago- 85:Proceedings of the 3rd international sago symposium. (Yamada, N. and K. Kainuma eds.) (Tokyo) pp.9404. Sim, E. S. and M. I. Ahmed 1978. Variation of flour yield in the sago palm, Malaysian Agri. J. 51: 351-358. Uchida, N., S. Kobayashi, T. Yasuda, and T. Yamaguchi 1990. Photosynthetic characteristics of sago palm, Meteroxylon rumuphii Martius. Japan. J. Trop. Agr. 34: 17680. Yamamoto, Y. 1999. The present condition of sago palm cultivation and utilization in Ambon, Seram and Southeast Area in Sulawesi Island in Indonesia. Japan. J. Trop. Agr. 43: 206-212. (in Japanese) Yamamoto, Y., T. Yoshida, Y. Goto, Y. Nitta, K. Kakuda, F. S. Jong, L. B. Hilary, and A. H. Hassan 2003. Differences in growth and starch yield of sago palms (Metroxylon sagu Rottb.) among soil types in Sarawak, Malaysia. Jpn. J. Trop. Agr. 47: 250-259. Yamamoto, Y. 2006. Starch productivity of sago palm (Metroxylon sagu Rottb.) in Indonesia and Malaysia. Japan. J. Trop. Agr. 50: 234-237. (in Japanese) Yamamoto, Y., F. S. Rembon, K. Omori, T. Yoshida, Y. Nitta, Y. B. Pasolon, and A. Miyazaki 2010. Growth characteristics and starch productivity of three varieties of sago palm (Metroxylon sagu Rottb.) in Southeast Sulawesi, Indonesia. Trop. Agr. Develop. 54: 1-8. Yamamoto, Y., K. Omori, Y. Nitta, K. Kakuda, Y. B. Pasolon R. S. Gusti, A. Miyazaki, and T. Yoshida 2014. Changes of leaf characters in sago palm (Metroxylon sagu Rottb.) after trunk formation. Trop. Agr. Develop. 58: 43-50. Yatsugi, T. 1987. Sago starch. In: Starch Science Handbook (Nakamura, M. and S. Suzuki eds.) Asakura Shoten (Tokyo) pp.404-410. (in Japanese)