A. Study sites and sampling design

The data were obtained at 42 forest permanent plots in 34 sites. These plots cover major climate zones and biogeographic regions in Japan (Ministry of the Environment 2001, http://www.env.go.jp/press/file_view.php?serial=2872&hou_id=2908) (Table 1, Fig. 1). Furthermore, these plots cover major four forest types found in Japan; evergreen conifer forest, broadleaf and conifer mixed forest, deciduous broadleaf forest, and evergreen broadleaf forest. We classified each plot into one of the four forest types by the dominance of evergreen broadleaf, deciduous broadleaf and conifer tree species as follows:

Evergreen conifer forest (EC): A >0.6

Broadleaf and conifer mixed forest (BC): 0.6 >A >0.4

Deciduous broadleaf forest (DB): A <0.4 & B >0.6

Evergreen broadleaf forest (EB): A <0.4 & B <0.6

, where

A = basal area of conifer tree species / total basal area

B = basal area of deciduous broadleaf tree species / basal area of deciduous and evergreen broadleaf tree species

Most of these forests are old-growth or old secondary forests but some are secondary forests that are younger than 100 years (see SiteList.csv). We classified forests into one of three age category: old-growth (OG), old secondary (OS), and secondary (S).

OG: old-growth forest which was assumed to be >150 years old,

OS: old secondary forest which was assumed to be >100 years old,

S: secondary forest which was assumed to be <100 years old.

Mean annual temperature, mean annual precipitation, and mean maximum snow depth during 1971 to 2000 were extracted from the database Mesh Climate Data 2000, distributed by the Japan Meteorological Agency (2002). The database provides climate variables estimated at the spatial resolution of 1-km. Mean annual temperature was corrected for altitudinal difference between the plot and the 1-km grid mean by the lapse rate of 0.6°C∙100m−1. Snow depth from the database might underestimate or overestimate the snow depth of the plots because snow depth is highly spatially heterogeneous. Therefore, we provide snow depth reported in other publications or according to personal observations in Appendix.pdf.

B. Field method

B-1. Permanent plots

One to four permanent plots were established in each site. The sizes of the plots are usually 1 ha (100×100 m) raging from 0.1 to 1.2 ha. The shapes of the plots are normally squares or rectangle but some plots have different shapes because we avoided edges of the forest that face artificial forest, roads, or buildings (see Appendix.pdf).

Plots were subdivided into 10×10-m grid cells (Fig. 2). Exceptions were HY-EC1 which was subdivided into 10×10 or 10×5-m gird cells (see Appendix.pdf), and CC-DB1 which was subdivided into sixteen 25×25-m grid cells.

B-2. Data collection

Measurements were conducted every year for 20 plots and once for 22 plots. At each plot, we measured mostly at the same season in every census years, normally autumn or winter when the stem growth has ceased. The exceptions were GR and OW (see Appendix.pdf).

Within each grid cell, all live trees with girth at breast height (gbh) >15 cm were tagged with stem identification code, botanically identified, and measured for their gbh to nearest millimeter. For some plots, trees with gbh <15 cm were also measured. Exception is TM-DB2 where only trees with gbh >30 cm were measured. Measurement height on each tree stem was marked by coloring paint for repeated measurements. The position of each tree was recorded to nearest 10 cm as coordinates of the stem within each grid cell for the most of the plots. Exceptions were KM-DB1 where it was measured to nearest 1 m, TM-DB1 and OS-EC1 to nearest 1cm, and CC-DB1, CC-DB2, CC-DB3, OG-DB1, OY-DB1, SD-EB1, TM-DB4 where stem locations were not measured. In some cases, trees which were adjacent to the plot were measured. Such trees can be identified from the X and Y coordinates of the grid cell where the trees were located and those of the stem grid cell position of the trees. For multiple stem trees, all stems with gbh >15 cm were measured, tagged with different stem identification codes, and assigned a common individual identification code. When multiple stems were adhered to each other and each stem was unable to measure, multiple stems were measured as one stem. The gbh of each stem was obtained by dividing the measured gbh of multiple stems by the number of stems.

C. Data verification procedures

Data were manually digitized and checked for typos by the investigators. Furthermore, data were checked for typos by staffs of the Network Center (M. I. Ishihara, S. N. Suzuki and M. Nakamura). Special attentions were paid for:

1) Consistence of individual identification code and tree species or stem location

2) Gbh growth anomalies

3) Recruitments.

When the multiple stems of the same individual (identical individual identification code) were recorded as different species or located apart, we corrected for the typos.

When the stem grew >6.3 cm∙yr−1 in gbh (i.e. >2 cm∙yr−1 in dbh) or shrunk <−3.1 cm∙yr−1 (i.e. < 1 cm∙yr−1 in dbh), we checked for typos and regarded such cases as anomalies. The criteria is based on the fact that 99% of girth growth record between 2004 to 2008 ranged between −1.4 to 3.6 cm∙yr−1, and only 0.14% (104 records) were >6.3 cm∙yr−1 and 0.15% (106 records) were <−3.1 cm∙yr−1. Such anomalies may result from truly anomalous growth, measurement error, measurement at different height, or when the gbh was measured with inclusion of attached vines at the previous census but measured excluding the vines at this census or vice versa.

Although we tried to measure trees with gbh >15 cm to detect accurately the timing of recruitment when the gbh was >15.7 cm (i.e. dbh >5.0 cm), some stems were recorded for the first time when their gbh was >15.7 cm. These trees were treated as newly recruited when the gbh at the first recorded census was <22.0 cm. When the gbh was >22.0 cm, such tree was treated as already recruited by the last census and gbh of the last census was treated as missing values. This is because if the tree was pre-recruitment state at the last census with gbh >15.6 cm and recruited by this census, the gbh at this census should be <21.9 cm (15.6 cm + 6.3 cm∙yr−1).

If any suspicious value remained, the value was recorded as an error (see 11. D.).

D. Taxonomy and Systematics

Most botanical identifications were made in the field and recorded in Japanese common names. Scientific names followed Satake et al. (1989a, b). Both Japanese common names and scientific names are retained in the data (see 11. D). Synonyms in Japanese common names are assigned single scientific name. Nomenclature at family level followed APG III (APG 2009). A large majority of plants were identified to species-level including subspecies and varieties. For some species, the minimum identification level varied between sites. For example, Acer mono was identified to variety-level (e.g. Acer mono var. marmoratum f. dissectum, Acer mono var. glabrum) in Ashoro (AS-DB1, AS-DB2, AS-DB3), Nopporo (NP-DB1), Nibuna-mizusawa (NB-EC1), Otakizawa (OZ-DB1), Kanumazawa (KM-DB1), Ogawa (OG-DB1), Aobayama (AO-EC1), Kanamegawa (KK-DB1), Chichibu (CC-DB1, CC-DB2, CC-DB3), but to species level only in other sites. Eighteen trees (0.03% of all trees) were identified to genus-level only (Salix sp. and Phyllostachys sp.). One hundred thirty nine trees (0.27% of all trees, most of them were in Nibuna-mizusawa (NB-EC1) and Mt. Tsukuba (TB-DB1)) could not be identified (recorded as "NA"). In the original data of Aya (AY-EB1) and Ichinomata (IC-BC1), Castanopsis sieboldii ("sudazii") and Castanopsis cuspidata ("tuburazii") were not distinguished and recorded as "Castanopsis sp." because two species create hybrid (Tanouchi and Yamamoto 1995). In the data of this data paper, "Castanopsis sp." was treated as Castanopsis sieboldii ("sudazii") according to Saito and Sato (2007) for AY-EB1 and to the morphology of seeds and stem (Kuramoto S. unpublished data) and Ishikawa (2009) for Ichinomata.