*e-mail: arshuma@yahoo.com
Effect of seed abortion and seed storage on germination and seedling growth in
Aquilaria malaccensis Lamk. (Thymelaeaceae)
Uma Shankar*
Department of Botany, North-Eastern Hill University, Shillong 793 022, India
The fruit of Aquilaria malaccensis is a bilocular dehis- cent capsule, with each locule harbouring a single seed. If this species aborts a developing seed within a fruit, does the remaining seed benefit in terms of enhanced seed weight, and do heavier seeds confer germination and growth advantages over lighter seeds? As seeds mature in the middle of the rainy sea- son when moisture is abundant in the soil, does desic- cation in the storage of seeds reduce germination success? To test these predictions, 821 fruits were examined, including two-thirds fruits that were one- seeded and one-third fruits two-seeded. Fresh seeds varied nearly fivefold in weight, from 28.8 to 134.8 mg.
However, the mean seed weight did not vary signifi- cantly between the seeds from one- and two-seeded fruits. On storage at room temperature (28 ± 2°C), the seeds lost 40 ± 1.2% moisture by the 5th day and 55 ± 1.5% by the 14th day. Yet, the difference in mean seed weight was not significant between the seeds from
one- and two-seeded fruits. Germination percentage of fresh seeds increased with increase in seed weight:
seeds <40 mg failed to germinate, those between 40 and 80 mg resulted in 10.5% germination and those
>80 mg germinated vigorously (54.8%). This trend was steady when seeds stored for 5 and 14 days were sown. Germination percentage of all seeds declined drastically with storage time: from 31 for seeds with- out storage to 25 for 5 days of storage and only 1% for 14 days of storage. Germination was epigeous and it was completed within 40 days from sowing. Time to germination was lesser for heavier than lighter seeds.
Survival of seedlings and their growth from heavier seeds was significantly greater than those from lighter seeds. The study shows that the abortion of one of the two seeds does not enhance the weight of the remaining seed. Yet greater seed weight (>80 mg) is advantageous for better germination and seedling growth in A. malaccensis.
Keywords: Aquilaria malaccensis, germination, seed abortion and storage, seedling growth.
TROPICAL trees most often pack lesser number of seeds within a fruit than the number of ovules within an ovary.
Several factors are responsible for seed abortion, such as available resource1, pollen limitation2,3, seed predation and pathogen attack1,4,5, lethal alleles6,7 and self-organi- zation8. Whether seed abortion has a significant effect on the weight of the remaining seeds is not well known.
Nonetheless, seed size has been shown to influence emer- gence rate and competitive ability of seedlings9–11. Seed weight variation within rainforest species12 is known to have no general effect on the risk of seed removal and speed of seedling emergence13.
In species with only two ovules (such as Aquilaria malaccensis Lamk.), fruits can be either one- or two- seeded. A. malaccensis is an extremely important tropical tree in South East Asia14,15. The species is threatened in the wild15, but survives mainly in plantations and home- steads in India. The main cause of species depletion is the
extraction of an oleoresin from the diseased portions of the heartwood, a source of highly prized ‘agar oil’ used in medicine, perfumery, incense and joss-stick16,17. Of late, conservation of A. malaccensis has assumed importance, particularly after its inclusion in Appendix II of the Con- vention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) in 1995. Thus, the know- ledge of its regeneration ecology is extensively desired for developing protocols for raising large-scale planta- tions18. In nature, A. malaccensis regenerates by seeds that are available in the rainy season. Of late, the genus Aquilaria has drawn attention of researchers worldwide resulting in studies on desiccation sensitivity19, storage behaviour20,21, seedling growth and survival in the forest22–24, and in vitro propagation25.
This study was undertaken to address the following questions. First, since A. malaccensis produces both one- and two-seeded fruits, are seeds from one-seeded fruits heavier than those from two-seeded fruits? Second, since seeds do not store food reserves besides cotyledons, do heavier seeds germinate earlier and produce advan- tageous seedlings than lighter seeds? And third, since seeds mature during the middle of the rainy season when moisture is abundant in the soil, does desiccation in the storage of seeds influence germination success?
Figure 1. Stages of seed germination and seedling development in Aquilaria malaccensis. a, Young plantation; b, Twig bearing flowers in leaf axil; c, Fruit opened into two halves producing two seeds with a tail-like appendage; d, Epigeous germination of seed with hypocotyl and cotyledons emerging from the soil; e, Young seedling in which cotyledons remain covered with testa causing its death; f, g, Normal seedling with emergence of first pair of leaves from the cotyledons and h, Six-day-old seedling, excavated and removed from soil showing radicle pale-pink in colour and as lengthy as the plumule.
Materials and methods The species
A. malaccensis Lamk. (synonym A. agallocha Roxb.) is commonly known as ‘agar’ in India and as ‘gaharu’ in Indonesia and Malaysia. A. malaccensis occurs predomi- nantly in the Indo-Burma hotspot of biodiversity26. In India, it grows in the foothills of Assam and adjoining states. The species prefers a mean annual rainfall of 1500 to 6500 mm, mean maximum temperature from 22°C to
28°C and mean minimum temperature from 14°C to 21°C (ref. 27).
A. malaccensis is a medium-sized deciduous tree attaining 18–30 m height and 1.5–2.5 m girth (Figure 1a). Flowering occurs from March to May, and fruits and seeds develop in two months. Bisexual flowers are green or dull yellow (Figure 1b). Ovary is superior and the two carpels join to form two locules. Flowers exhibit ento- mophilies28. The fresh fruit is a parrot-green capsule (Figure 1c). Each capsule has two locules producing up to one seed in each locule. The capsule dehisces mechani-
cally and seeds remain hanging in their respective locules by a persistent funiculus (Figure 1c). Passive dispersal results in most seedlings landing within 5 m radius of the maternal trees28. The seeds are non-endospermic, 6–8 mm in length and 3–5 mm in width.
Data collection
Fruits of A. malaccensis were collected from Hojai, Assam in August 2000, and transported to the laboratory in polythene bags. The fresh seeds were extracted from the fruits, individually labelled, weighed on an electronic balance to 0.1 mg accuracy and transferred to laboratory trays at room temperature (28 ± 2°C). A total of 1000 seeds were sown in three phases: 200 seeds each from one- and two-seeded fruits immediately after weighing (no storage) and after 5 days of storage, and 125 seeds from one-seeded fruits and 75 seeds from two-seeded fruits after 14 days of storage. Seeds were weighed again after 5 and 14 days of storage prior to sowing. All seeds were sown in root trainers, each with 25 holes. Root trainers were filled with soil collected from the site of seed collection (lateritic sandy loam) and kept under the tree canopy to mimic natural environment. Each root trainer containing 25 seeds represented one replicate and the root trainers were placed in a randomized block design. Day temperature ranged between 25°C and 30°C, and relative humidity between 60% and 80% during the germination period. Root trainers were watered regularly to avoid desiccation of surface soil. Germination tests were conducted according to the international rules for seed testing.
Germination was recorded daily until completion, i.e.
one month after the last seed had germinated. Seeds with a protruding radicle of about 2 mm were considered as germinated (Figure 1d). Seedlings with unsplit seed coats and partially opened or decayed leaves were considered as ‘abnormal’ (Figure 1e) and those with fully opened first pair of leaves and straight stems were considered
‘normal’ (Figure 1f and g). Ungerminated seeds were recovered from the soil and their cotyledons examined by removing the testa to confirm germination failure.
Recovered seeds were placed in 0.1% solution of 2,3,5- triphenyl-2H-tetrazolium chloride (TTC) to test for viabi- lity29; only viable embryos could turn pink. The seedlings were transferred to polythene bags (28 cm height, 12 cm diameter) nearly one month after germination.
Seedling survival was studied periodically at the end of the rainy season in October (first census), at the end of the winter season in March (second census), and at the end of the spring season in May (third census). Seedling growth was assessed in terms of seedling height, leaf number and leaf area. Seedling height was measured by a ruler, the number of leaves was counted, and leaf area was determined using a leaf area meter (LICOR-3000A).
The cause of seedling mortality was ascertained for each dead seedling (Figure 1h).
Data analysis
The data were analysed for three seed-weight classes, viz.
<40 mg (light), ≥ 40–<80 mg (intermediate) and >80 mg (heavy) to interpret the results. Without storage, nearly 50% fresh seeds were in intermediate weight class, 47%
in heavy weight class and only 3% in light weight class.
‘Germination value’, a quantitative index which com- bines the speed and completeness, was calculated by mul- tiplying germination percentage with the ‘peak value’30. The peak value is the highest quotient obtained by divid- ing the cumulative per cent germination on each day by the number of days elapsed since initial imbibitions.
The differences between means of two groups (one- seeded and two-seeded) were individually determined using two-tailed Student’s t-test (if n < 30) or z-test (if n > 30)31. Two-tailed z statistic was used to determine differences in germination percentages between two sam- ples, and the probabilities calculated following StatSoft32. The z statistic is given by the equation:
1 2
1 1 2 2
1 2
ˆ ˆ
ˆ (1 ˆ ) ˆ (1 ˆ ),
p p
z p p p p
n n
= −
− −
+
where pˆ is the weighted mean (critical ratio) of the two sampled proportions and is calculated as pˆ = (x1 + x2)/
(n1 + n2); pˆ1 and pˆ2 the proportions of seeds germinated in the first and second sample; x1 and x2 the number of seeds germinated in the first and second sample, and n1
and n2 are the number of seeds sown in the first and sec- ond sample.
The Kolmogorov–Smirnov (K–S) test was used to determine differences between the observed and expected frequency distribution of seed weights. Germination curves were drawn by plotting cumulative number of seeds germinated everyday (y-axis) against germination time (x-axis) in days. Simple linear regression analysis was used to determine the relationship between seed weight and germination time.
Results Seed packing
In all, 865 fruits were examined. Among these, 821 were ripe and healthy fruits, 568 were one-seeded and 253 two-seeded, yielding a seed-to-ovule ratio of 0.65.
Among the 44 unhealthy and discarded fruits, 26 were diseased and 18 unripe.
Table 1. Minimum, maximum and mean weight of seeds of Aquilaria malaccensis after no storage, and 5 and 14 days of storage. SE indicates standard error of the mean. Two-tailed z test did not show signifi- cant variation in mean seed weight between seeds from one- and two-seeded fruits (values with same superscript) after no storage (df = 398, P = 0.652), 5 days of storage (df = 398, P = 0.206) and 14 days of
storage (df = 198, P = 0.053)
Seed weight (mg)
Seed number Number of
per fruit seeds weighed Minimum Maximum Mean ± SE
No storage
One-seeded 200 31.7 129.8 78.3 ± 1.67a
Two-seeded 200 28.8 134.8 79.4 ± 1.77a
One + two-seeded 400 28.8 134.8 78.9 ± 1.22
5 days of storage
One-seeded 200 18.5 97.7 46.0 ± 1.34a
Two-seeded 200 17.1 91.1 48.5 ± 1.50a
One + two-seeded 400 17.1 97.7 47.2 ± 1.01
14 days of storage
One-seeded 125 12.8 88.4 33.7 ± 1.38a
Two-seeded 75 15.9 83.4 38.6 ± 2.36a
One + two-seeded 200 12.8 88.4 35.5 ± 1.24
Seed weight
Fresh seed weight ranged from 29 to 135 mg, which declined to 17–98 mg after 5 days of storage and 13–88 mg after 14 days of storage (Table 1). Apparently, seeds lost considerable weight in storage at room temperature:
40 ± 1.2% by the 5th day and 55 ± 1.5% by the 14th day.
Mean seed weight showed a significant decline from no storage to 5 days of storage (P < 0.001), and from 5 to 14 days storage (P < 0.001). Two-tailed z-test showed no difference in mean seed weight between seeds from one- and two-seeded fruits after no storage, and 5 and 14 days of storage (Table 1).
The frequency distribution of seed weight of fresh seeds pooled from one- and two-seeded fruits was close to an expected lognormal distribution (Figure 2a). How- ever, the distribution became positively skewed after 5 days (Figure 2b) and highly positively skewed after 14 days of storage (Figure 2c), presumably due to moisture loss. When the distribution was compared between seeds from one- and two-seeded fruits (figures not included), differences were not significant, indicating that one- and two-seeded condition had no effect on seed weight.
Seed viability and germination
Germination started on day-13 and was completed by day-40 for fresh seeds (Figure 3). Similarly, germination started on day-14 and was completed by day-35 for seeds stored for 5 days. In both cases, 90% of total germination was achieved by day-29. Among seeds sown after 14 days of storage, one seed germinated on day-14 and the other on day-15. Germination value (sensu Czabator30)
declined from 97.1 for no storage through 85.2 for 5 days of storage to 6.66 for 14 days of storage.
Seed weight had a strong effect on germination; germi- nation percentages increased with increase in seed weight (Table 2). Light-weight seeds failed in germination with- out storage, and after 5 and 14 days of storage. Heavy seeds showed greater germination percentages than in- termediate seeds without storage as well as after 5 days of storage. After 14 days of storage, only two seeds germi- nated in the heavy-weight class (Table 2).
Germination percentages declined with increase in storage time: from 10.5 through 4.3 to no germination for intermediate seeds, and from 54.8 through 46.5 to only 1 for heavy seeds (Table 2). However, in the heavy-weight class, fresh seeds from one-seeded fruits showed greater germination (P < 0.01) than those from two-seeded fruits, and these differences were not significant after 5 days of storage. Germination percentage of seeds pooled across weight classes declined from 31 for no storage through 25 for 5 days to 1 for 14 days of storage.
Seed weight influenced the time to germination; heavier the seed weight lesser the time needed for germination (Figure 4). Linear regression of germination time on seed weight for a pool of seeds from one- and two-seeded fruits yielded a significant negative correlation after no storage (Figure 4a) and 5 days of storage (Figure 4b).
Because only two seeds germinated after 14 days of stor- age, a correlation could not be drawn.
Seedling survival and growth
Seedlings from seeds of one- and two-seeded fruits showed similar patterns of survival after no storage (r = 0.999,
df = 2, P < 0.001; Figure 5a) and 5 days of storage (r = 0.993, df = 2, P < 0.007; Figure 5b). Percentage of seedling survival at the third census was not different bet- ween seeds of one- and two-seeded fruits after no storage (P = 0.503) and 5 days of storage (P = 0.827). No seed survived by the first census for 14 days of storage (Figure 5c).
Seed weight influenced survival of seedlings (Figure 6). At the third census, survival as a percentage of the to- tal number of seeds germinated was significantly greater
Figure 2. Change in frequency distribution of seed weights of A.
malaccensis from no storage (a) to 5 days of storage (b) and 14 days of storage (c). The Kolmogorov–Smirnov one-sample D statistic showed that the observed frequency distribution (bar) deviated from an expected normal distribution (curve) after no-storage (n = 400, d = 0.0779, p = 0.05), but assumed a lognormal distribution after 5 days (n = 400, d = 0.0393, p = ns) and 14 days of storage (n = 200, d = 0.0451, p = ns). Data for seed weight from one- and two-seeded fruits were pooled for each storage treatment.
in the heavy than the intermediate class after no storage (P = 0.001) and 5 days of storage (P = 0.0005). Survival percentages did not differ between light and intermediate classes after 5 days of storage (P = 0.122).
Figure 3. Germination curves for no storage (circle), 5 days of stor- age (square) and 14 days of storage (cross) after seed collection. Data for seed germination from one- and two-seeded fruits were pooled for each storage treatment.
Figure 4. Linear relationship between seed weight and germination time following (a) no storage (y = 40.32–190.38x, df = 122, r = 0.615), and (b) 5 days of storage (y = 32.32–179.91x, df = 98, r = 0.557) after collection. Data for 14 days of storage are not shown since only two seeds germinated resulting in df = 0. Seeds in light (>0–0.04 g), inter- mediate (>0.04–0.08 g) and heavy (>0.08–0.14 g) weight classes are separated by vertical grids.
Table 2. Germination of seeds from one- and two-seeded fruits of A. malaccensis in three seed-weight classes (based on fresh seed weight) on sowing after no storage, and 5 and 14 days of storage. SE indicates standard error of the mean. Values with different superscripts between one- and
two-seeded categories vary significantly using two-tailed z statistic for two-sample test of percentage (P < 0.05)
Light (>0–40 mg) Intermediate weight (>40–80 mg) Heavy (>80–140 mg) All weights (>0–140 mg) Number Percentage of Number Percentage of Number Percentage of Number Percentage of Seed number of seeds seeds germinated of seeds seeds germinated of seeds seeds germinated of seeds seeds germinated per fruit sown (n) (mean ± SE) sown (n) (mean ± SE) sown (n) (mean ± SE) sown (n) (mean ± SE) No storage
One-seeded 6 0.0 103 10.7 ± 3.0a 91 65.9 ± 5.1a 200 35.5 ± 3.3a
Two-seeded 7 0.0 96 10.4 ± 3.2a 97 44.3 ± 4.9b 200 26.5 ± 3.3a
One + two-seeded 13 0.0 199 10.5 ± 2.2 188 54.8 ± 3.6 400 31.0 ± 2.3
5 days of storage
One-seeded 9 0.0 104 2.9 ± 2.1a 87 41.3 ± 9.8a 200 19.5 ± 3.0a
Two-seeded 8 0.0 81 6.1 ± 1.9a 111 50.5 ± 8.7a 200 30.5 ± 3.0a
One + two-seeded 17 0.0 185 4.3 ± 1.6 198 46.5 ± 7.4 400 25.0 ± 2.2
14 days of storage
One-seeded 4 0.0 51 0.0 45 0.8 ± 0.9a 125 0.8 ± 0.9a
Two-seeded 5 0.0 40 0.0 55 1.3 ± 1.2a 75 1.3 ± 1.2a
One + two-seeded 9 0.0 91 0.0 100 1.0 ± 0.7 200 1.0 ± 0.7
Figure 5. Survival of seedlings of seeds from one-seeded (triangle) and two-seeded (square) fruits following no storage (a), 5 days of stor- age (b) and 14 days of storage (c). G depicts the percentage of seeds germinated, and C1, C2 and C3, depict the percentage of seedlings sur- viving at the end of the rainy, winter and spring season respectively.
Figure 6. Survival of seedlings raised from seeds from light (dia- mond), intermediate (square) and heavy (triangle) classes following no storage (a), 5 days of storage (b) and 14 days of storage (c). G and C1–
C3 are the same as defined in Figure 5.
Seedling growth parameters, viz. seedling height, leaf number and leaf area increased with seedling age, i.e.
from the first to second, and second to third census both in intermediate and heavy seed classes after no storage and 5 days of storage (Figure 7). No seedling from light seeds and no seedling from 14 days of stored seeds survived. One-way analysis of variance showed significant effect of storage time, seed weight and growth period on seedling growth parameters, i.e.
significantly greater growth of seedlings: from fresh than 5 days stored seeds, from heavy than intermediate seeds, and for the second and third censuses than at the first census (Table 3). Seedling growth between the sec- ond and the third censuses did not vary significantly (P > 0.05).
Discussion
Seed packing in tropical plants can vary due to seed abortion33,34. In A. malaccensis, two-third of the fruits produced only one seed from the two ovules housed in a bilocular ovary. The maternal parent is known to favour abortion to gain dispersal advantage if the fruit is a dis- persal unit as in many leguminous pods34,35. In A. malac- censis, dispersal is passive with seeds dropping to the ground after dehiscence of the capsule35, or may be medi- ated by wind as the seeds are tiny though not designed for wind dispersal. Hence, seed abortion in anticipation of dispersal advantage is unlikely.
Seed abortion may occur due to a trade-off between maternal and offspring interests. The most economic use
Table 3. Probabilities obtained from one-way analysis of variance to determine the effect of seed stor- age time (no storage and 5 days of storage), seed weight (intermediate and heavy), and growth period (censuses 1–3) on seedling growth parameters of A. malaccensis. Data were analysed for only those seed- lings that had at least one leaf present. Seedlings from 14 days of storage treatment and from light-weight
class died before the first census; hence these treatments were excluded from the analyses P level
Source of variation Degrees of freedom Seedling height Leaf number Leaf area
Storage time 1 0.195 0.023 0.028
Error 312
Seed weight 1 0.001 0.000 0.000
Error 312
Growth period 2 0.000 0.000 0.000
Error 311
Figure 7. Effect of seed weight and storage time on seedling growth parameters, viz. seedling height (bricks), leaf number (hatch) and leaf area (horizontal line) at the time of the first (C1), second (C2) and third (C3) census. Data were averaged for only those seedlings that had at least one leaf present. Table 3 shows the results of analysis of variance for the data.
of maternal resources is attained by packing as many seeds as possible in order to reduce the cost of packing per seed34. However, seed size may be negatively influ- enced with increasing packing of seeds per fruit as in Mesua ferrea, where heavier seeds germinate better and their seedlings grow and survive better than lighter seeds36. Thus, a trade-off between the packing cost and offspring performance may guide the species to optimize seed number per fruit by inducing abortion, particularly where seed size is dependent on the seed number per fruit. Seed abortion may simply result from the failure of fertilization or pollination. Regardless of the cause of abortion of one of the two seeds in the fruit of A. malac- censis, the range of variation in seed weight, mean seed weight and frequency distribution of seeds in seed-weight classes were not significantly different between the seeds from one- and two-seeded fruits. Further, as seeds develop
in separate locules with equal probability of resource supply due to symmetrical architecture of the locules, seed size and shape (length, width and volume) do not vary significantly between the seeds from one- and two- seeded fruits. Conclusively, the seeds from one-seeded fruits are not heavier than those from two-seeded fruits in A. malaccensis.
Notwithstanding, the heavy seeds germinated faster than the light seeds. Greater and faster germination of heavier seeds has been reported in many tropical spe-
cies37–43. However, contradictory results have been
reported in many other species36,44–46, and germination may be independent of seed weight in some species47,48. Heavy seeds may simply out-compete light seeds, since the latter may house underdeveloped or unhealthy embryos.
Seed weight strongly influenced seedling survival and growth. Seedlings from heavy seeds survived maximally
and produced more height, leaf number and leaf area than those from intermediate seeds. The growth of the seed- lings was faster between the first and the second censuses (interval of 5 months) compared to that between the sec- ond and third censuses (interval of 2 months).
On storage, the frequency distribution of seed weight increases positive skewness, apparently due to a rapid loss of moisture from seeds (40% by the 5th day and 55%
by the 14th day). Although mean seed weight declines on storage due to loss of moisture, the difference between the mean weights of seeds from one- and two-seeded fruits remains insignificant, indicating that seed number per fruit has no effect on the rate of moisture loss.
Seed weight and storage time influenced germination strongly. Freshly sown, heavy seeds showed maximum germination, but light seeds failed to germinate comple- tely. With storage for 5 days, heavy seeds exhibited more germination than intermediate seeds, and light seeds did not germinate again. Finally, with storage for 14 days, all seeds except two failed to germinate. The two germinated seeds were from the heavy-weight class and they died quickly after germination. Hence, overall germination percentage declined from no storage to 5 days of storage and finally to 14 days of storage. Beniwal27 found 65%
germination when selectively picked, ‘healthy’ seeds were sown. It is believed that storing seeds in cool conditions such as in a refrigerator may prolong viability up to 30 days49.
Seed storage time influenced seedling growth, which was better from freshly sown than 5 days of stored seeds.
There was no seedling from light seeds, and all seedlings from intermediate seeds sown after 5 days of storage had died. Thus, it could be recommended that only heavy seeds (preferably over 80 mg fr. wt.) be sown quickly after collection to maximize germination and survival.
In A. malaccensis, non-endospermic seeds mature in the middle of the rainy season when plenty of moisture is available, and predators are abundant to cause damage to the seeds. In this climatic setting, seeds maximize germi- nation by germinating soon after dispersal instead of remaining dormant through the ensuing winter. There- fore, short viability helps this species to capitalize on moisture availability for germination in the same growth season and minimize pressure for investing energy into the seeds to sustain through winter.
Conclusion
This study reveals that seed abortion in A. malaccensis gives rise to one-seeded versus two-seeded fruits, but the varied number of seeds per fruit does not influence seed weight, as one would expect. However, seed weight strongly influenced germination time and percentage, seedling survival and seedling growth. The seeds are sen- sitive to desiccation in storage and may result in loss of
viability and germination as well as in slower growth of seedlings. Short viability enforces germination during the rainy season and excludes the need to store energy in the seeds for maintenance.
1. Janzen, D. H., Variation in seed size within a crop of a Costa Rican Mucuna andreana (Leguminosae). Am. J. Bot., 1977, 64, 347–349.
2. Bierzychudek, P., Pollinator limitation of plant reproductive effort. Am. Nat., 1981, 117, 837–840.
3. Nybom, H., Pollen limited seed set in pseudogamous blackberries (Rubus L. subgenus Rubus). Oecologia, 1987, 72, 562–568.
4. Augspurger, C. K., Seedling survival of tropical trees: interactions of dispersal distance, light gaps, and pathogens. Ecology, 1984, 65, 1705–1712.
5. Herrera, C. M., Selective pressures on fruit seediness: differential predation of fly larvae on the fruits of Berberis hispanica. Oikos, 1984, 42, 166–170.
6. Stanton, M. L., Seed variation in wild radish: effect of seed size on components of seedling and adult fitness. Ecology, 1984, 65, 1105–1112.
7. Wiens, D., Calvin, C. L., Wilson, C. A., Davern, C. I., Frank, D.
and Seavey, S. R., Reproductive success, spontaneous embryo abortion, and genetic load in flowering plants. Oecologia, 1987, 71, 501–509.
8. Uma Shaanker, R., Ganeshaiah, K. N. and Krishnamurthy, K. S., Development of seeds as self-organizing units: testing the predic- tions. Int. J. Plant Sci., 1995, 156, 650–657.
9. Gross, K. L., Effect of seed size and growth form on seedling establishment of six monocarpic perennial plants. J. Ecol., 1984, 72, 369–387.
10. Grime, J. P. and Jeffery, D. W., Seedling establishment in vertical gradients of sunlight. J. Ecol., 1965, 53, 621–642.
11. Piper, J. K., Germination and growth of bird-dispersed plants:
effect of seed size and light on seedling vigor and biomass alloca- tion. Am. J. Bot., 1986, 73, 959–965.
12. Uma Shankar and Synrem, I. L., Variation in morphometric traits of fruits and seeds of Prunus nepaulensis Steud. in Meghalaya, India. Trop. Ecology, 2012, 53(3), 273–286.
13. Paz, H., Mazer, S. J. and Martinez-Ramos, M., Seed mass, seed- ling emergence, and environmental factors in seven rain forest Psychotria (Rubiaceae). Ecology, 1999, 80, 1594–1606.
14. Chang, L. T., Ng, Y. S. and Kadir, A. A., A review on agar (gaharu) producing Aquilaria species. J. Trop. For. Prod., 1997, 2, 272–285.
15. Paoli, G. D., Peart, D. R., Leighton, M. and Sumsoedin, I., An ecological assessment of the nontimber forest product gaharu wood in Gunung Palung National Park, West Kalimantan, Indonesia.
Conserv. Biol., 2001, 15, 1721–1732.
16. Jalaluddin, M., A useful pathological condition of wood. Econ.
Bot., 1977, 31, 222–224.
17. Wollenberg, E. K., Incentives for collecting gaharu (fungal- infected wood of Aquilaria spp.; Thymelaeaceae) in East Kali- mantan. Econ. Bot., 2001, 55, 444–456.
18. Donovan, D. G. and Puri, R. K., Learning from traditional knowl- edge of non-timber forest products: Penan Benalui and the aute- cology of Aquilaria in Indonesian Borneo. Ecol. Soc. (online), 2004, 9(3), 3.
19. Kundu, M. and Kachari, J., Desiccation sensitivity and recalcitrant behavior of seeds Aquilaria agallocha Roxb. J. Seed Sci. Technol., 2000, 28, 755–760.
20. To Luu, D., Chou, M. and Anh, V., Preliminary study of Aquilaria crassna. Trop. For. Tree Seeds Newsl., 2001, 9, 22–23.
21. Aroonrungsikul, C., Kaewsuwan, U. and Wongsatain, W., Pre- liminary study of agarwood (Aquilaria crassna) fruit and seed
storage. In Proceedings of the 47th Kasetsart University Annual Conference, Kasetsart, 17–20 March 2009, pp. 281–288.
22. Lok, E. H. and Ahmad Zuhaidi, Y., The growth performance of plantation grown Aquilaria malaccensis in peninsular Malaysia.
J. Trop. For. Sci., 1996, 8, 573–575.
23. Lok, E. H., Chang, Y. S. and Aziah, M. Y., Early survival and growth in field trials of Aquilaria malaccensis (Karas) and Azadirachta excela (Sentang). J. Trop. For. Sci., 1999, 11, 852–
854.
24. Soehartono, T., Newton, A. C. and Mardiastuti, A., Factors influ- encing the survival and growth of Aquilaria malaccensis seedlings in Indonesia. J. Trop. For. Sci., 2002, 14, 364–378.
25. Meng-ling, He, Shu-yuan, Qi and Lan-juan, Hu, Rapid in vitro propagation of medicinally important Aquilaria agallocha.
J. Zhejiang Univ. Sci. B, 2005, 6(8), 849–852.
26. Whitmore, T. C., Thymelaeaceae. In Tree Flora of Malaya (ed.
Whitmore, T. W.), Longman Press, Kuala Lumpur, Malaysia, 1973, vol. 2, pp. 383–391.
27. Beniwal, B. S., Silvical characteristics of Aquilaria agallocha Roxb. Indian For., 1989, 115, 17–21.
28. Soehartono, T. and Newton, A. C., Reproductive ecology of Aqui- laria spp. in Indonesia. For. Ecol. Manage., 2001, 152, 59–71.
29. Cottrell, H. J., Tetrazolium salt as a seed germination indicator.
Nature, 1947, 159, 748.
30. Czabator, F. J., Germination value: an index combining speed and completeness of pine seed germination. For. Sci., 1962, 8, 386–
396.
31. Hoshmand, A. R., Statistical Methods for Environmental and Agricultural Sciences, CRC Press, New York, 1998, 2nd edn.
32. StatSoft, Inc., STATISTICA for Windows (Computer Program Manual), StatSoft, Inc., USA, 1995.
33. Bawa, K. S. and Buckley, D. P., Seed: ovule ratios, selective seed abortion, and mating systems in Leguminosae. Mo. Bot. Gard.
Monogr. Syst. Bot., 1988, 27, 243–262.
34. Uma Shaanker, R., Ganeshaiah, K. N. and Bawa, K. S., Parent- offspring conflict, sibling rivalry, and brood size patterns in plants. Annu. Rev. Ecol. Syst., 1988, 19, 177–205.
35. Joshi, N. V., Uma Shankar and Ganeshaiah, K. N., The ‘neighbour effect’ and its role in shaping the positional pattern of seed deve- lopment in fruits: an illustration from the pods of Erythrina sube- rosa. Curr. Sci., 1993, 65, 234–238.
36. Khan, M. L., Bhuyan, P., Uma Shankar and Todaria, N. P., Seed germination and seedling fitness in Mesua ferrea L. in relation to fruit size and seed number per fruit. Acta Oecol., 1999, 20, 599–
606.
37. Howe, F. A. and Ritcher, W. M., Effects of seed size on seedling size in Virola surinamensis; a within and between tree analysis.
Oecologia, 1982, 53, 347–351.
38. Dunlop, J. R. and Barnett, J. P., Influence of seed size on germina- tion and early development of loblolly pine (Pinus taeda L.) ger- minants. Can. J. For. Res., 1983, 13, 40–44.
39. Tripathi, R. S. and Khan, M. L., Effects of seed weight and microsite characteristics on germination and seedling fitness in two species of Quercus in a subtropical wet hill forest. Oikos, 1990, 57, 289–296.
40. Barik, S. K., Tripathi, R. S., Pandey, H. N. and Rao, P., Tree regeneration in a subtropical humid forest: effect of cultural dis- turbance on seed production, dispersal and germination. J. Appl.
Ecol., 1996, 33, 1551–1560.
41. Bhuyan, P., Khan, M. L. and Uma Shankar, Trade-off between dispersal efficiency and seedling fitness in Oroxylum indicum, a wind dispersed tropical tree. Int. J. Ecol. Environ. Sci., 2000, 26, 67–73.
42. Khan, M. L. and Uma Shankar, Effect of seed size, light regime, and substratum microsite on seed germination, and seedling growth of Quercus semiserrata Roxb. Trop. Ecol., 2001, 42, 117–125.
43. Uma Shankar, Seed size as a predictor of germination success and early seedling growth in ‘hollong’ (Dipterocarpus macrocarpus Vesque). New For., 2006, 31, 305–320.
44. Maun, M. A. and Cavers, P. B., Seed production and dormancy in Rumex crispus. I. The effects of removal of cauline leaves at anthesis. Can. J. Bot., 1971, 49, 1123–1130.
45. Marshall, D. L., Effect of seed size on seedling success in three species of Sesbania (Fabaceae). Am. J. Bot., 1986, 73, 457–464.
46. Murali, K. S., Pattern of seed size, germination and seed viability of tropical tree species in southern India. Biotropica, 1997, 29, 271–279.
47. Gross, K. L. and Kromer, M. L., Seed weight effects on growth and reproduction in Oenothera biennis L. Bull. Torr. Bot. Club, 1986, 113, 252–258.
48. Perez-Garcia, F., Iriondo, J. M. and Martinez-Laborde, J. B., Germination behaviour in seeds of Diplotaxis erucoides and D.
virgata. Weed Res., 1995, 35, 495–502.
49. Ahmed, M. and Gogoi, P., Agar: History and Scope of Planta- tion – A Perspective. All Assam Agar Traders and Agar Oil Manu- facturers’ Association, Nagaon, India, 2000.
ACKNOWLEDGEMENTS. The experimental work was done at the North-East Unit of the G.B. Pant Institute of Himalayan Environment and Development, Itanagar. I acknowledge partial funding from DBT, New Delhi (grant # BT/PR7928/NDB/52/9/2006), facilities afforded at the Department of Botany, North-Eastern Hill University, Shillong, and help received from Shilpi Agarwal during the experiment and Dr K.
Haridasan in species identification.
Received 26 October 2010; revised accepted 23 January 2012
