INTRODUCTION

Forest ecosystems are self-maintained via primary production and nutrient cycling(Barnes et al., 1998). Some of the primary production in forest ecosystems is returned to the forest floor via litterfall, and then nutrients in litter are reused by plants after mineralization by decomposition processes carried out by a hugely diverse array of organisms(Lavelle and Spain, 2001). Forest soil provides nutrients, water, and a medium for physical support for plant growth(Kimmins, 1987). Soil nutrients originate primarily from the weathering of soil minerals. However, nutrients released through the decomposition of soil organic matter are fundamental to the mainte-nance of mature forest ecosystems(Daubenmire, 1974, Barbour et al., 1987, Mun et al., 2007).

The litterfall of leaves, branches and other tree organs generally constitutes the principal pathway by which nu-trients and organic matter are transferred to soil(Bray and Gorham, 1964, Wiegert and Monk, 1972, Caritat et al., 2006, Blanco et al., 2008). Litter is the food source for decomposers and detritivores, and the means by which nutrients are returned to the cycling pool(Meentemeyer et al., 1982, Barbour et al., 1987, Baker et al., 2001, Blanco et al., 2008). Nutrient cycling describes the movement within and among various biotic or abiotic entities in which nutrients occur in the environment(Lavelle and Spain, 2001).

As a component of the Korean National Long-Term Ecological Research Program, we conducted a study of litter production and decomposition in Quercus mongolica, Quercus variabilis and Pinus densiflora forests in Mt. Worak National Park. These species are known as the most common species in Korea. The principal objective of this study was to quantify the total amounts of nutrients returned to the forest floor via litterfall in a P. densiflora forest located in Mt. Worak National Park. For this study, seasonal litterfall and the nutrient concentration of litterfall were analyzed for 5 years from January 2009 through December 2013, and the total amounts of nutrient input to the forest floor via litterfall were calculated.

MATERIALS AND METHODS

1.Study area

The Mt. Worak National Park is located between Mt. Soback and Mt. Sogni and stretches over both Gyeongsangbuk-do and Chungcheongbuk-do in Korea. The highest peak in Mt. Worak National Park, Munsubong, is 1,162 m above sea level. The P. densiflora forest studied herein was located 380 m above sea level at Jaecheon-si, Hansu-myeon, Songgye-ri in a south-west direction(N 36° 51' 17", E 128° 64' 41"). Tree density was 1,300 trees ㏊^-1 and average diameter at breast height (DBH) was 17.12±0.38㎝. In the shrub layer, Qurecus variabilis, Fraxinus sieboldiana and Indigofera kirilowii were distributed with low density. In the herb layer, Pteridium aquilinum and Miscanthus sinensis were distributed, It was examined in plot 20㎡. According to the Jecheon meteorological station, about 20 ㎞ from the study area, annual average temperature and precipitation for thirty years from 1980 through 2010 were 10.2°C and 1,387.8 mm, respectively. The annual average temperature and precipitation for four years from 2009 through 2013 were 10.1°C and 2,066.2 mm, respectively.

2.Litterfall collection and chemical analysis

Five circular litter traps, with opening areas of 0.5㎡, were randomly established in the P. densiflora forest in December 2008. Litter traps were leveled approximately 50㎝ above the ground to prevent the input of resuspended windblown materials from the forest floor. Litterfall collections began on January 2009 and continued for 5 years in every month. The litter collected from the litter traps was brought into the laboratory and separated into leaves, branches and bark, reproductive organs and miscellaneous( litterfall from shrubs). Each component weighed after drying in a drying oven at 80°C for 72 hour and ground with a mixer for chemical analysis.

Chemical analysis of each component of litterfall was carried out in 3 replicates. After the litter samples were digested on a block digestor, T-N and T-P were analyzed with a flow injection analyzer(Lachat QuickChem 8000; Zellweger Analytics, Milwaukee, WI, USA). K, Ca, and Mg in litter were determined using an atomic absorption spectrophotometer(Perkin-Elmer 3110; Perkin-Elmer, Waltham, MA, USA) after wet digestion(Allen et al., 1974). All calculations of the nutrient contents of sample materials were based on the means of three replicate analyses. The total amounts of each nutrient returned to the forest floor via litterfall for each year was calculated from each nutrient concentration and the dry weights of each component of litter. The measured values were converted into areas(㏊).

RESULTS AND DISCUSSION

1.Litter production

Litterfall in P. densiflora forest continued throughout the year, peaking in autumn(October in 2009, 2012 and 2013, November in 2010 and 2011)(Fig.1). In late winter and spring, from February to May, the proportion of reproductive organ was relatively higher than the proportion of other litter components. The amounts of litterfall in 2009, 2010, 2011, 2012, and 2013 was 2.810, 3.796, 3.268, 3.284, and 4.045 ton ㏊^-1 yr^-1, respectively(Table 1). The yearly variation of litter production was not great. The average amounts of litterfall for 5 years was 3.441±0.49 ton ㏊^-1 yr^-1.

Lee(2013) reported the litter production of P. densiflora natural forest was 3.29±0.11 ton ㏊^-1 yr^-1. Berg and Laslowski(2006) reported that the litter production of Asian temperate coniferous forests ranged from 910~4,990 ㎏ ㏊^-1 yr^-1. Litter production of P. densiflora forests in this study area fell within those ranges. Jeong et al.(2013) reported that the average amounts of litter production of P. densiflora forest in Mt. Nam was 7.07 ton ㏊^-1 yr^-1. The amounts of litter production in this P. densiflora forest was lower than that of Jeong et al.(2013). Gholz et al.(1985) reported that the differences in litter production might be related to tree density, age and canopy cover among forest(Mun et al., 2007, Namgung and Mun, 2009).

The average proportion of leaf, branch, reproductive organ and miscellaneous of total litterfall for 5 years were 56.4, 14.7, 7.5, and 21.4%, respectively(Fig. 2). The yearly variation in the leaf category was highest among litter components. The proportion of the leaf category in 2009, 2010, 2011, 2012 and 2013 were 57.4, 53.4, 60.7, 58.3, and 53.5%, respectively.

2.Nutrient concentration of litterfall

N concentration in leaf litterfall was low in the winter, and high in the summer and autumn season(Fig. 3A). N concentration in the reproductive organ was high in May through July when the debris of flowers of P. densiflora falls. N concentration in the branch and bark component showed no seasonal trend over five years. Unlike the N concentration, the P concentration in leaf litterfall was high in the spring. P concentration in branch and bark litterfall was lowest among the litterfall components(Fig. 3B). Like P concentration, the K concentration in branch and bark litterfall was lowest among the litterfall components(Fig. 3C). In contrast, Ca concentration in branch and bark litterfall were the highest, and those in reproductive organ were the lowest among the litterfall components(Fig. 3D). Mg concentrations in the leaf litterfall and miscellaneous fractions were higher than those in the branch and bark and reproductive organ fractions(Fig. 3E).

The average concentration of nutrients in each component of litterfall in the P. densiflora forest for 5 years was summarized in Table 2. With the exception of Ca, the nutrient concentration of leaf litterfall was higher than those in other litterfall components.

Lee(2013) reported that the average concentration of N, P, K, Ca and Mg in litterfall in the P. densiflora forest at Mt. Worak National Park over 4 years was 5.61, 0.31, 1.13, 2.27 and 0.51 ㎎/g, respectively. Namgung and Mun(2009) reported that the average concentrations of N, P, K, Ca, and Mg in a P. densiflora forest at the Mt. Worak site over 3 years were 6.34, 0.32, 1.45, 2.44, and 0.37 ㎎/g, respectively; this was similar that those in this study. Namgung(2010) reported that the average concentration of N, P, K, Ca, and Mg in leaf litterfall in the Q. variabilis forest at Mt. Worak over 4 years was 11.53, 0.67, 2.38, 3.15, and 2.14 ㎎/g, respectively. The average concentrations of N, P, and K in leaf litterfall in P. densiflora were lower than those in Q. variabilis. The nutrient concentrations of litterfall in oak forests were higher than those in pine litterfall(Namgung et al., 2008, Namgung, 2010).

3.Nutrient returned to forest floor via litterfall

The amounts of each nutrient returned to the forest floor via litterfall in the P. densiflora forest was summarized in Table 3. As shown in Table 3, the amounts of annual input of N, P, K, Ca and Mg to the forest floor via litterfall in P. densiflora forests in 2009, 2010, 2011, 2012, and 2013 were 20.78, 30.65, 19.21, 18.38, and 24.23 ㎏ ㏊^-1 yr^-1 for N, 0.73, 1.78, 0.80, 0.95, and 1.01 ㎏ ㏊^-1 yr^-1 for P, 4.72, 6.67, 4.61, 2.59, and 2.71 ㎏ ㏊^-1 yr^-1 for K, 8.70, 16.24, 7.39, 4.70, and 5.38 ㎏ ㏊^-1 yr^-1 for Ca, 3.11, 3.68, 2.48, 1.31, and 1.55 ㎏ ㏊^-1 yr^-1 for Mg, respectively. The average amounts of N, P, K, Ca, and Mg returned to the forest floor in this P. densiflora forest for 5 years was 22.73 ± 4.92, 1.05 ± 0.42, 4.26 ± 1.69, 8.48 ± 4.62, and 2.42 ± 1.01 ㎏ ㏊^-1 yr^-1, respectively(Table 3).

Generally, the amounts of nutrients returned to the forest floor via litterfall in pine forests was much lower than those in the oak forests(Vogt et al., 1983). This is because the annual input of litterfall and nutrient concentration of unit weight of litterfall in oak forests were greater than those in the pine forest(Kwak and Kim, 1992, Mun and Kim, 1992, Kim et al., 1997, Mun et al., 2007). Soil nutrient concentrations in pine forests were much lower than those in the oak forest at Mt. Worak(Choi et al., 2006). This may be due to the greater litter production and higher nutrient concentrations of litter in the oak forest relative to those in the pine forest. Another reason for this is that the decay rates of oak litter are higher than the decay rates of pine needle litter(Mun and Joo, 1994, Mun et al., 2007, Namgung et al., 2008, Mun, 2009).