Invasion of dwarf bamboo into alpine snow-meadows in northern Japan: pattern of expansion and impact on species diversity

Abstract

Recently, a dwarf bamboo species,Sasa kurilensis; Poaceae, has invaded into alpine snow-meadows in the wilderness area of the Taisetsu Mountains, northern Japan. This dwarf bamboo species has a wide distribution range from lowland to alpine sites of snowy regions. Because of the formation of dense evergreen culms and an extensive rhizome system, other plants are excluded following invasion by this dwarf bamboo, resulting in low species diversity. Dwarf bamboo originally inhabited the leeward slopes of alpine dwarf pine (Pinus pumila) clumps in alpine regions. During the last 32 years, however, dwarf bamboo has expanded its distribution area by up to 47% toward snow-meadows, especially on southeastern facing slopes. This rapid change may be related to the decrease in soil moisture and expansion of the annual growing period caused by the recent acceleration of snowmelt time. A multiyear census revealed that the density of bamboo culms increased 30-150% during 2 years, and the annual expansion of bamboo rhizomes was 39 cm on average. In addition to the expansion of bamboo clumps by vegetative growth, the possibility of migration by seed dispersal was also suggested by a genet analysis. With the increase in culm density, the species richness of snow-meadow vegetation decreased to less than one-quarter of the original level due to intense shading by dwarf bamboo. The rapid vegetation change in these almost pristine alpine environments isolated from the human activity implies that global climate change already influences the alpine ecosystem.

Keywords: Alpine vegetation; Sasa kurilensis; global warming; remote sensing; snowmelt; soil moisture.

Figure 1

Advancing front of dwarf bamboo (Sasa kurilensis) in alpine snow-meadows in the Taisetsu Mountains. Although dwarf bamboo basically inhabited the vicinity ofPinus pumilaclumps in the alpine area, it is expanding toward moist snow-meadow sites. Photograph by G.K. in late September 2009.

Figure 2

Location of the study site in the Taisetsu Mountains. A 500 × 1000 m² plot was set on an east-facing gentle slope that was divided into NE site and SE site.

Figure 3

Distribution of dwarf bamboo clumps in the study site. The yellow area indicates the distribution in 1977 and the red area indicates the additional area occupied by dwarf bamboo between 1977 and 2009. Dark green vegetation isPinus pumilaclumps that are often located on the western side of dwarf bamboo clumps. Blue circles in the SE site indicate the locations in which the belt-transect survey was conducted. m = diffuse margin, s = distinct margin. Refer Table 1 for the increased and decreased area during this period.

Figure 4

Belt-transect across the margin of dwarf bamboo clumps. (a) A 10-m transect was set from the outside to inside of individual dwarf bamboo clumps in which terminal culms were located at the 3-m point, and 1 × 1 m² quadrats were set at 0–1 m, 3–4 m, 6–7 m, and 9–10 m locations for the measurements of culm performance and the environmental survey. (b) Percent coverage of bamboo canopy in each quadrat. (c) Maximum height of bamboo culms in each quadrat. (d) Soil moisture in volume-metric water content at a depth of 1–12 cm in each quadrat. Open boxes indicate diffuse margins and closed boxes indicate distinct margins. The sample size was 6 for diffuse margins and 4 for distinct margins. Box and whisker plot represents the 75th, 50th, and 25th percentile (boxes) with whiskers from the 90th to 10th percentile.

Figure 6

Distribution of clonal patches of dwarf bamboo in the SE site discriminated by the SSR analysis. Separated clumps of the same genotype are connected by broken lines. Dots indicate the sampling points for SSR analysis. For the genetic composition of individual clones (a–p), refer to Appendix 1.

Figure 5

Relationships between species richness and culm height of dwarf bamboo (a), between species richness and relative PPFD (b), and between species richness and soil moisture (c) obtained in the belt-transect survey. Open circles indicate data from diffuse margins and closed circles from distinct margins. ***P< 0.0001, ^†P< 0.10 by Spearman's correlation test.