MATERIALS AND METHODS

Seeds were collected from natural C. sativa stands in Duzce, Turkey (42° 54' N, 31°10' E, altitude 350 m) in mid-October 2004. Immediately after collection all seeds were washed and those obviously defective were discarded and the seeds were floated to remove debris and empty or weeviled seeds. The seeds were visually separated into three seed sizes: small (<5 g), medium (5-8 g) and large (>8 g) and the seeds belonging to each of these categories were weighted individually to the nearest 0.01 g (Table 1). Moisture contents were determined on two replicates each containing samples from five seeds, by oven drying for 17 h at 103°C (ISTA, 1996). The moisture content of seeds as a percentage of fresh weight at the beginning of storage was found to be 44%.

Seeds stored for two weeks in polyethylene bags at ±4°C until used were soaked in water for 48 h and the outer coat was removed and seeds were germinated in plastic trays containing sterilized sand, which was watered throughout the entire incubation period. The plastic trays were placed in a germination chamber in the dark at 30/20°C (8/16 h). A number of 240 seeds of each seed size, in 6 replications of 40 each, were used for standard germination test. The germinated seeds were counted every day for 21 days following first signs of germination (ISTA, 1996). Seeds were considered to have germinated when their radicle reached at least 10 mm long.

The germination data were calculated and expressed as Germination Percentage of seeds (GP) that had germinated at the end of the test and Germination Value (GV) which reflects both speed and completeness of germination, which was calculated as follows: GV = (ΣDGS/N)xGPx10, where DGS is the daily germination speed, which is obtained by dividing the cumulative GP by the number of days since the beginning of the test and N is the number of DGS calculated for the germination test (Djavanshir and Pourbeik, 1976).

Seeds were sown in nursery beds to look at the effect of seedling size on emergence, survival and some seedling morphological characteristics after one growing season at Kocaman-Zonguldak Forest Nursery (41°07' N, 30°19' E, altitude 10 m). The nursery soil was sandy loam with a soil pH of 6 in the rooting zone. The climate is mild with humid summers. Annual precipitation averages approximately 1140 mm and the temperature averages 13.7°C.

The study consisted of a randomized complete block design with five replications. Seedbeds were standard 1.2 m wide with four rows and each plot (experimental units) was 4 m long. The seeds were hand sown at 60 seed m^-2 density using marked sticks showing distances within row in December 2004. One seed were sown at each spot in 4-5 cm depth. During the vegetation period, following the sowing date, the only treatment given was irrigation and weed control-hoeing.

Seedling emergence in nursery was determined at the end of the April 2005 (after four weeks following first sign of germination). Seedling survival percentage was determined and a total of 150 seedlings were randomly hand-lifted without harming the roots from each weight class (a total of 450 seedlings) in December. The seedlings were examined in terms of their morphological traits such as root collar diameter, seedling height, seedling root and shoot dry weight and the number of lateral coarse roots (diameter ≥ 2 mm) and fine roots (diameter < 2 mm) which were longer than 5 cm.

Table 1:	Seed weight of C. sativa seed size classes
Means followed by the same letter are not significantly different at p<0.05

Root dry weight and number of roots were determined in roots pruned to 25 cm. Oven dry weights were determined after samples were oven-dried at 70°C for 24 h.

An analysis of variance was used to compare the germination response and morphological data to seed size class. Each of the variables was analyzed using analysis of variance to determine if there were significant differences among the treatments. When significant differences were found, Duncan’ New Multiple Range Test was performed for comparison of the means. Percent data (germination percentage and survival) and count data (root number) were arcsine transformed before analyses.

RESULTS

The effects of seed size on germination showed that germination performance were significantly affected by seed size and maximum germination percentage and germination value were shown by large seeds (GP = 98.8% and GV = 68.7) and the lowest by small seeds (GP = 91.3%; GV = 51.5) (Table 2).

The effects of seed size classes on the emergence, survival, root collar diameter, seedling height, root dry weight, shoot dry weight and total dry of one-year-old sweet chestnut seedlings showed significant differences (Table 3). But shoot dry weight/root dry weight and numbers of roots didn’t show significant differences. There was 71-80% seedling emergence (germination) and no seedlings died during the experiment. Large seed produced the highest survival (80%) and survival of the one-year-old seedlings produced from medium seed size classes was less (79%), but not significantly different. However, survival for the small seed size class was significantly less than the other seed size classes (71%). There was a measurable and significant increase in average first-year seedling height growth associated with larger seed size. Seedling height was the highest in large seed size class (86.4 cm) and seedling height of medium seed size class was less (78.5 cm), but not significantly different. However, seedling height was significantly less than for the other two seed size classes in small seed size class (67.5 cm). Large seeds also produced the highest root collar diameter (13.4 mm).

Table 2:	Germination Percentage (GP) and Germination Value (GV) of C. sativa seed size classes
Means in the same column followed by the same letter are not significantly different at p<0.05

Table 3:	Effect of seed size on survival and various seedling characteristics of one-year-old C. sativa seedlings (RCD: Root Collar Diameter, SH: Shoot Height, RDw: Root Dry weight, SDw: Shoot Dry weight, TDw: Total Dry weight, FRN: Fine Root number, CRN: Coarse Root number)
Means in the same column followed by the same letter are not significantly different at p<0.05

Larger seeds resulted in more root dry weight, shoot dry weight and total dry weight in this study. Although large seeds caused the highest number of fine roots and coarse roots, there were not significant differences among seed size classes (Table 3).

DISCUSSION

The higher seed dimensions could be attributed to better seed size indicates better quality and germination of seeds and genetic potential (Toon et al., 1990; Davidson et al., 1996). In some tree species larger seeds are regarded to produce greater seedlings and seedlings from larger seeds survive better and grow stronger as found in the present study (Dirik, 1993; Long and Jones, 1996; Singh, 1998; Karrfalt, 2004). Bencat and Tokar (1972) found similar results stated in the present study in terms of height growth that the heaviest seed produced the highest height growth in sweet chestnut. The better performance of larger seeds may simply be a reflection of the greater amount of nutrients available to the embryo (Abideen et al., 1993; Khan and Shankar, 2001). In some species large seed did not produce the highest germination (Chaisurisri et al., 1992; Jayasankar et al., 1999; Tilki and Alptekin, 2005), seedling survival and seedling growth (Krishan et al., 1995; Indira et al., 2000).

The production of seedlings that are both large and vigorous is important to the forester who will outplant at the end of the first growing season. Seedling size is critical from the standpoint of survival and the ability to withstand competition from other seedlings, grasses and weeds. Seed sizing appears to be a tool to the nursery manager to produce larger seedlings. However, discarding small seeds should not be practiced unless there is good genetic evidence to demonstrate the practice is not discarding whole families with possible good growth potential as said by Karrfalt (2004). Even the small seeds produced seedlings of acceptable size, although cull rates appeared to be higher in the small seed classes.

The heavier weight seed class performed significantly better than the medium and small seed size class in terms of germination percentage, germination value, seedling emergence, seedling survival and seedling growth in the present study. Thus, it can be stated that large seeds in C. sativa have better germination and survival and the larger the seed, the larger the average seedling would be in seedling diameter, height and dry weight for the one-year-old seedlings. This study suggests that sizing of C. sativa seeds into several categories may result in higher and more uniform germination within in a seedbed and appears to be a tool to the nursery manager to produce larger and more uniform seedlings.