This study was conducted in the Ginkakuzi-san (also spelled Ginkakuji-san) National Forest, Kyôto City, Japan. This area is located in the warm-temperate zone. The dominant vegetation of the area had been evergreen oak forest approximately 7000 to 1000 years ago, but secondary forests composed of pines and deciduous oaks increased after that (Takahara 2015). In the late 19th century, the forest around the study site was estimated to be covered with small pines, Pinus densiflora Siebold & Zucc., due to human impact (Ogura 2015). After nationalization (in the 1870s), the forest has been protected from felling as a rule. In the 1930s, a mixed forest of pines and deciduous tree species such as Quercus serrata Murray covered the site (Osaka Regional Forest Office 1936). However, the forest has not been completely free from felling; e.g., trees were cut for fuel wood during World War II, and part of the forest was converted to conifer (Cryptomeria japonica (L.f) D.Don and Chamaecyparis obtusa (Siebold & Zucc.) Endl.) plantations in the 1970s. Recently, broadleaved evergreen trees, such as Symplocos prunifolia Siebold & Zucc. and Ilex pedunculosa Miq., were thinned in some parts of the forest, probably to improve light penetration into the forest. Forest diseases also affected the forest. After the 1960s, pine wilt disease strongly affected the national forest and numerous P. densiflora trees died. The responsible pathogen is the nematode Bursaphelenchus xylophilus (Steiner & Buhrer) Nickle, which is thought to have been brought from North America (Mamiya 1988). In the last decade, mass mortality of oak trees affected the forest, as described above. In addition, increased numbers of sika deer have affected the forest (Itô 2015, Itô 2016). Though the population density of the deer in the forest is unknown, camera trap data showed that they inhabited the forest in all seasons (Itô 2015).

A 1.05 ha (210 m × 50 m) study site was established in 1992 (Fig. 1). The site was divided into 420 quadrats of 5 × 5 m. The study site mainly consisted of broadleaved secondary forest stands, and the rest was a conifer plantation. The plantation was thinned in 1996, and some of the evergreen broadleaved trees were thinned in the eastern part of the site in 2005. After mass mortality, dead stems of Q. serrata were felled, and some stems surrounding the dead stems were also felled. The felled trunks were cut into about 1 m lengths and piled on the floor. The infected wood was covered with plastic sheets and disinfected. The cut stems are denoted in the measurement data file.

Figure 1.  

Outline of the study site. A) Contour map. B) Stand type. B: broadleaved, BT: broadleaved (thinned), C: conifer plantation, G: gap (not related to mass mortality of oak trees), GM: gap created or affected by mass mortality of oak trees. A GM in the conifer plantation was generated by a dead oak stem that remained in the plantation.

Two classes of forest layers were defined: the canopy and sub-canopy layer and the understory layer. Stems in the canopy and sub-canopy layer were defined as having a diameter at breast height (dbh) at least 3.0 cm, and stems in the understory layer were defined as those with a dbh smaller than 3.0 cm or a height shorter than 1.3 m (not including those seedlings and shoots sprouted in the current year). During the period from December 1993 to February 1994, all stems in the canopy and sub-canopy layer were marked, identified by species name, and their dbh measured using a measuring tape with 1 mm precision. After that, dbh was measured in 1996, 1999, 2002, 2005 and 2014 during the non-growing season (October to January of the following year) in the same way. Measurement values in 1999 are missing in the most eastern half of the study site because slope collapse prevention works were conducted near the area.

In 1992, the species of woody plants found in the understory layer were recorded for each quadrat to obtain the understory species composition of the forest in autumn (September to November). The same survey was conducted again in 2014 (July to November). The dynamics of the half of the study site that solely consisted of broadleaved forest was reported in Itô (2015) and Itô (2016), but this is the first report of the whole study site.

Ginkakuzi-san National Forest, Kyôto City, Japan

35.028 and 35.029 Latitude; 135.802 and 135.800 Longitude.

The six surveys from 1993 to 2014 found 61 species in the canopy and sub-canopy layer. Table 1 shows their stem density (stems/ha) and basal area (m²/ha) in 1993 and 2014. Clerodendrum trichotomum Thunb. was not included in Table 1, because it had only one stem, which was first found in 2002 but had disappeared in 2005.

The most abundant species was Symplocos prunifolia Siebold & Zucc. in 1993, but Quercus glauca Thunb. surpassed it by 2014. The dominant species in the basal area was consistently Chamaecyparis obtusa, which occupied most of the plantation area of the study site. Quercus serrata decreased in density from 37.1 to 14.3 stems/ha and in basal area from 2.8 to 1.5 m²/ha (Table 1, Fig. 2).

Figure 2.  

Changes in dominance of major species. A: stem density, and B: basal area.

In the understory layer, 88 woody species were found over the two surveys in 1992 and 2014 excluding unidentified species. Table 2 shows the species and the number of quadrats where they were found. Quercus glauca was the most frequent species in both 1992 and 2014. Symplocos prunifolia increased in the number of quadrats from 75 to 274. This may related to gap formation by oak tree deaths and thinning of upper trees as well as the deer-unpalatable trait of the species.

Forestry and Forest Products Research Institute (Matsunosato 1, Tsukuba 305-8687, Japan) has ownership of this data set.