- Email: [email protected]
The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan
G Model APSOIL-1879; No. of Pages 8
ARTICLE IN PRESS Applied Soil Ecology xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Applied Soil Ecology journal homepage: www.elsevier.com/locate/apsoil
The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan Motohiro Hasegawa a,∗ , Aino T. Ota b , Daisuke Kabeya a , Toru Okamoto a , Tomoyuki Saitoh a , Yoshihiko Nishiyama c a
Forestry and Forest Products Research Institute, Tsukuba, Ibaraki 305-8687, Japan Yokohama National University, Yokohama, Kanagawa 240-8501, Japan c Kiso Experimental Forest, Forestry and Forest Products Research Institute, Kiso, Nagano 397-0001, Japan b
a r t i c l e
i n f o
Article history: Received 10 December 2012 Received in revised form 3 June 2013 Accepted 17 June 2013 Keywords: Larch Larix kaempferi Broad-leaved trees Collembola Community structure Forest floor plants
a b s t r a c t The effects of broad-leaved trees on the collembolan community in larch plantations were investigated at the foot of Mt. Yatsugatake (1200–1400 m a.s.l.) in Japan. The study sites comprised five pure larch plantation plots (larch dominated more than 95% of the area at breast height) and five mixed forest plots (larch dominated between 50% and 80% of the area at breast height). We compared the collembolan community structures between stand types and related them to the plant community composition and soil properties at each plot. Density and species richness of Collembola were not significantly different between pure larch and mixed plots. Using partial redundancy analysis (pRDA), the variance of collembolan species data in the litter layer was explained by the biomass of grass on the forest floor, and the variance in the soil layer was explained by the biomass of total forest floor plants. These results suggest that the biomass or the composition of forest floor plants influence the collembolan community more than the crown trees in this area. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In Japan, Larix kaempferi, a deciduous conifer species, is planted in dry inland areas typified by light snow cover and cold winters (Nagaike et al., 2003). Most of these plantations were established after the 1940s to replace primary and secondary broad-leaved deciduous forests that had been substantially overharvested, but the contributions of L. kaempferi plantations to regional biodiversity are still unknown (Nagaike, 2002). To achieve ecologically sustainable forest management in this region, studies have been undertaken to determine the status of biodiversity in forests and the effects of forest management practices in the area, i.e., the basic information necessary for ecologically sustainable management (Nagaike, 2002; Ohsawa, 2004, 2007). The study area is covered by Japanese larch plantations, which were established for timber production, and thinning has been conducted twice within a 45year period in this area to enhance the growth of dominant trees. After the planting of larch seedlings, trees species (e.g., birch, oak, and pine) started to regenerate from the stumps or seeds, and grow
∗ Corresponding author. Tel.: +81 298 29 8251; fax: +81 298 73 1543. E-mail addresses: [email protected] (M. Hasegawa), [email protected] (A.T. Ota), [email protected] (D. Kabeya), [email protected] (T. Okamoto), [email protected] (T. Saitoh), [email protected] (Y. Nishiyama).
alongside the larch. During the mowing and thinning stage, mixed trees were cut or left, depending on prevailing silvicultural management policy. As a result, a mosaic of larch plantations with various mixtures of broad-leaved trees has developed over the study area. Plant species identity, vegetation composition, physiology, chemistry, and phenology all influence soil invertebrate community composition (Sylvain and Wall, 2011). The effects of litter species richness on soil fauna have been tested using mixed litter bag experiments (Kaneko and Salamanca, 1999; Takeda, 1987), which have revealed idiosyncratic relationships between plant species and soil animals (St. John et al., 2006; Wardle et al., 2006). The effects of mixing tree species on Collembola have been investigated in a European spruce plantation mixed with beech (Salamon and Alphei, 2009; Salamon et al., 2008; Scheu et al., 2003), and these studies have suggested relatively moderate effects on the collembolan community. Total density and species richness did not change significantly, but species composition was affected (Salamon and Alphei, 2009; Scheu et al., 2003). To determine the effects of mixing tree species on Collembola, its community structure should be evaluated using the response of different functional groups or through multivariate analysis. Salamon et al. (2008) suggest that forest age and stand type are likely to impact Collembola communities via changes in the amount and quality of food resources, including both living plants and herb litter materials. The effects of herbs on Collembola have been emphasized in grassland and meadow studies (Dombos, 2001;
0929-1393/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsoil.2013.06.005
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
2
Greenslade, 1997; Salamon et al., 2004), and the effects of the undergrowth in forest sites were also recently studied (Eisenhauer et al., 2011). The findings have revealed that studies should not only consider the effects of crown trees but also those of undergrowth on collembolan communities. In this study, we investigated the effects of a mixture of broadleaved trees on Collembola in terms of their density, diversity, and species composition. The relationship of the collembolan community with the stand type, forest floor plants, and soil abiotic factors are discussed.
2. Materials and methods The study area was located on the south and west slopes of Yatsugatake mountain range, in Yamanashi and Nagano, central Japan (approximately 36◦ 50–56 N, 140◦ 34–35 E, 1200–1400 m a.s.l.; mean annual air temperature 7.1 ◦ C; mean annual precipitation 1454 mm). Larch (L. kaempferi) was planted at the study area after a previous land cover of grass fields or sparse pine forests. Landscape at 1200–1400 m a.s.l. on the slope of Yatsugatake mountain range was totally at most covered by the larch plantation. The forest of this area was mostly national or prefectural forest and classified into fine compartment (with several ha of area). The management plan for the compartment was often determined as this small unit, therefore the compartment in the vicinity could have different forest structures. Ten study plots were located within a 15 × 20 km area (Fig. 1). The forest age was 40–50 years at the time of sampling (Table 1). The total basal area of each plot ranged from 22.9 to 41.2 m2 ha−1 . Size of a plot was 30 m × 30 m. Each plot was divided in 9 sub-plots: where they 10 m × 10 m. Thinning had been conducted several times until 40 years after planting, but in some plots, all of the trees except larch had been completely removed (Table 1). In other plots, Betula platyphylla, Quercus mongolica, and other tree species had invaded and were left in place during the thinning procedure. As a result, the percentage of larch based on basal area in each plot ranged from 100% to 46.4% (Table 1). We classified these plots as five pure larch plots (>95% larch) and five mixed plots (between 50% and 80% larch). Ideally, in order to consider the effects of the forest category to the soil fauna, these two categories of plots should be set pair-wisely and replicated the pair of plots in the study area. However, because of the limited availability of the study plots, we only fulfilled these design in the part of the plots (Fig. 1). Therefore, we scattered these two categories of plots in the study area as evenly as possible shown in Fig. 1. The average distances (±standard errors) among pure larch plots, among mixed plots and between different categories of plots were 5553 (±977), 5389 (±677) and 5895 (±1195) m respectively. To extract the spatial autocorrelation structure to the collembolan community composition, we use the principal coordinates of neighbor matrices (PCNM) as explained in a statistical analysis. In October 2006 and May 2007, samples from the litter layer and soil (0–5 cm) were collected with a corer (125 mL, 5 cm depth, 25 cm2 area) from each subplot. Most Collembola occur in the litter layer and the upper mineral soil layer, within the top 5 cm of the profile. In total, 180 samples (10 plots × 9 subplots × 2 dates) for each substrate (litter and soil) were collected. Collembola were extracted using a Macfadyen high gradient extractor at a constant temperature of 35 ◦ C for 7 days. The water content of the litter layer was calculated as (wet weight of litter − dry weight of litter)/dry weight of litter). Soil samples were taken from five quadrats in each plot in August 2007 to determine physical and chemical parameters. Cores of 100 mL of the top 5 cm of soil were collected for chemical analysis. For soil pH and EC analysis, 5 g of fresh soil was mixed with 25 mL of deionized water. A glass electrode (HM14P; DKK-Toa
Corp., Tokyo, Japan) was used to measure pH and a conductivity cell electrode (SC82; Yokogawa Electric Corp., Tokyo, Japan) was used to measure soil electrical conductivity (EC). Total carbon and nitrogen concentrations in soil samples were measured with an NC analyzer (Sumigraph NC-900; SCAS Ltd., Tokyo, Japan). The airdried litter layer was classified as broad-leaved, needles (larch), and bamboo grass (Sasa nipponica), and the weight of each category was estimated. At each plot, the diameter at breast height (DBH) above 5 cm was measured for all trees in each plot. We set five 0.5 m2 circles randomly in each plot, and cut out the aboveground forest floor vegetation in each circle. The forest floor plant cuttings were classified into trees, grass, herbs, and Sasa, weighed for each category and dried to constant mass at 70 ◦ C for three days. The number of plant species in each circle was counted. 2.1. Statistical analysis 2.1.1. Detection of spatial autocorrelation, RDA, variation partitioning and GLMM To extract the spatial autocorrelation structure, we used principal coordinates of neighbor matrices (PCNM) developed by Borcard and Legendre (2002) because it is flexible for expressing various spatial scales and can be used in redundancy analysis (RDA). This method creates a set of explanatory variables (PCNMs: eigenvectors) that represent the structure at all spatial scales from the distance matrix among the plots. Ten PCNMs were calculated in our study, and we referred to them as PCNM-1 to PCNM-10. We used the library spacemakeR (Dray et al., 2006) in the statistical environment R (R Core Team, 2012), to calculate the PCNMs. For the RDA, environmental variables were selected by a forward-selection procedure to detect meaningful factors and so reduce the large number of potential factors. Variables were incorporated stepwise into the model according to their increasing effect on the variance (contribution to the eigenvalues of the model), and their significance was tested by comparing them to Monte Carlo permutations (5000 times) of the null model (which does not have the variable to be incorporated). We set a standard selection criterion (p < 0.05) to retain parameters for use in the final models. All of the RDA and forward selections were performed with CANOCO for Windows, version 4.5 (ter Braak and Smilauer, 2002). Initially, we conducted RDA using only the PCNMs as explanatory variables, and the significant variables were selected as relevant spatial autocorrelations in the species composition by a forward-selection procedure. We wanted to determine the effect of the environment variables after partialling out the effect of spatial autocorrelation. Therefore, we executed a partial RDA using the selected PCNMs as covariables and other environmental variables as explanatory variables. We investigated the effects of forest category of plots (pure larch and mixed forest plots, as the fixed effect) to the environmental variables and collembolan community structures. Because the sampling protocol in this study formed nested structure (i.e., the samplings were conducted in 5 subplots within each plot for 2 seasons), the effect of the plot difference and that of seasonal difference were also considered as the random effects. Hence, generalized linear mixed models (GLMMs) (nlme library and glmmLM library on R version 2.13.0; (R Core Team, 2012) were used for the analyses. For comparison of the litter weight, soil chemical and physical properties, and the biomass of forest floor plants between pure larch and mixed forest plots we used a GLMM with the assumption of a Gaussian distribution. We also used a GLMM for the density of Collembola with the assumption of a negative binomial distribution and for the species richness with the assumption of a Poisson distribution due to discrete variables. Significance tests were based on the t-statics for each parameter. Spearman rank correlation
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
ARTICLE IN PRESS
G Model APSOIL-1879; No. of Pages 8
M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
3
Fig. 1. Maps of study plots. Filled circle: pure larch plots; open circle: mixed plots.
coefficients were calculated between environmental variables and various indices of the collembolan community. 3. Results The biomass of tree broad-leaves in the forest floor vegetation biomass was significantly larger in pure larch plots than in mixed plots (Table 2; GLMM, p < 0.05). Other biomass and average species richness of forest floor vegetation were not significantly different between pure larch and mixed plots (GLMM, p > 0.05). All soil properties (Table 3) were not significantly different between pure larch and mixed plots (GLMM, p > 0.05). The weight of larch needles in the litter layer was significantly larger in the pure larch plots than in the mixed plots (Table 4; GLMM, p < 0.05). In contrast, the weight of leaves of broad-leaved species in the litter layer was significantly larger in the mixed plots than in the pure larch plots (p < 0.05). The weight of Sasa in the litter layer was not significantly different. The density of Collembola in the litter and soil layers was not significantly different between pure larch plots and mixed plots (GLMM, p > 0.05; Fig. 2a and b). Species richness and the H’ and J’ indices in the litter and soil layers were also not significantly different between pure larch plots and mixed plots (Figs. 3–5a and b). In the litter layer, the density of Collembola was positively correlated with litter weight or thickness (Table 5). In contrast, diversity and evenness of Collembola were negatively correlated with litter thickness
(probably due to the increased density of dominant species). In the soil layer, the density of Collembola was positively correlated with the total basal area. Species richness was positively correlated with forest age. Diversity and evenness in the soil layer were positively correlated with pH and the species richness of forest plants. Partial RDA ordination diagrams for the collembolan community in the litter layer (eigenvalue for the 1st axis: 0.17) selected grass biomass to explain the variance of species in forward selection (p < 0.05) (Fig. 6a). Densities of Folsomia hidakana and Arrhopalites alticolus were positively correlated, and those of Isotoma viridis and Pseudosinella sp. 2 were negatively correlated with grass biomass (p < 0.05). The spatial factor extracted from PCNM and grass biomass solely explained 28.6% and 17.3% of the variance of species data in the litter layer, and 1.2% was duplicated by both factors. Partial RDA ordination diagrams for the collembolan community in the soil layer (eigenvalue for the 1st axis: 0.21) selected total forest floor plant biomass to explain the variance of species in forward selection (p < 0.05) (Fig. 6b). Densities of Protaphorura sp. 1, Xenylla brevispina, and Arrhopalites minutus were positively correlated, and that of Desoria notabilis cf. pallida was negatively correlated with the total forest floor plant biomass. Spatial factors extracted from PCNM and grass biomass solely explained 20.5% and 20.8% in the variance of species data in the litter layer, and 0.6% was duplicated by both factors.
Table 1 Basal area, percentage of larch and species richness of tall trees in each plot. Pure larch plots were labeled as L1–L5, and mixed plots as M1–M5. Site code
Total basal area (m2 ha−1 )
Basal area of larch (m2 ha−1 )
Basal area of trees except larch (m2 ha−1 )
Percentage of larch based on basal area (%)
Density of tall trees (ha−1 )
Density of larch trees (ha−1 )
Species richness of tall trees
Forest age
L1 L2 L3 L4 L5 M1 M2 M3 M4 M5
32.8 38.0 28.9 28.0 22.9 41.2 28.4 27.4 39.8 28.8
32.8 38.0 28.8 26.9 22.0 33.1 20.0 15.0 19.1 13.4
0 0 0.2 1.1 1.0 8.1 8.5 12.4 20.7 15.4
100 100.0 99.5 96.1 95.8 80.3 70.2 54.8 48.0 46.4
833 978 1067 1156 467 3611 822 1233 1944 1233
833 967 889 533 456 922 489 367 500 356
1 1 4 7 2 15 3 6 13 14
46 38 38 57 55 51 57 38 52 52
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
ARTICLE IN PRESS
G Model APSOIL-1879; No. of Pages 8
M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
4
Table 2 Biomass (g m−2 ) of each category of forest floor plant and average species richness (in 0.5 m2 ) of forest floor plants on each plot. Wood materials of tree seedlings were excluded.
L1 L2 L3 L4 L5 M1 M2 M3 M4 M5
Fern
Grass
Forb
Sasa (current year)
Trees (leave plus needles)
Total biomass
Average species richness (in 0.5 m2 )
0.0 0.8 0.0 1.2 16.3 3.1 39.8 0.0 1.2 0.5
0.1 21.3 0.0 11.6 56.8 7.1 6.8 10.2 4.0 13.3
0.1 11.6 0.0 1.0 9.6 0.5 0.4 1.2 0.1 0.4
6.3 5.3 18.5 11.0 1.4 26.8 8.9 1.9 5.5 7.7
37.6 158.7 157.7 40.2 158.6 77.2 75.0 6.4 22.2 36.7
163.7 372.5 245.3 248.0 317.9 117.7 200.6 244.6 69.4 111.6
2.1 15.2 0.2 6.9 9.7 6.8 5.8 6.6 8.1 11.1
Table 3 Weight (average of values at two dates) of each category of forest floor litter.
L1 L2 L3 L4 L5 M1 M2 M3 M4 M5
Larix (g m−2 )
Broad-leaves (g m−2 )
Sasa (g m−2 )
1160 1070 765 538 474 664 528 454 478 569
19 6 11 26 33 102 135 258 206 224
222 10 78 117 218 21 211 80 32 38
4. Discussion Important factors determining soil fauna composition include vegetation type, soil structure, soil chemistry, organic matter, soil microflora, and soil moisture and temperature (Coulson et al., 2003; Hasegawa, 2001; Klironomos and Kendrick, 1995; Sousa et al., 2003; Wallwork, 1976). Collembola generally display little response to changes in vegetation structure (Hågvar, 1982; Salamon and Alphei, 2009). Our study revealed small differences
in total density and species richness between pure larch plots and mixed plots. Effects of mixing spruce forests with beech on the collembolan community in Europe have been investigated (Salamon and Alphei, 2009; Salamon et al., 2008; Scheu et al., 2003). Scheu et al. (2003) also reported that the mixing of beech and spruce had no significant effect on collembolan biomass. Salamon and Alphei (2009) suggested that the total density and species richness in mixed forests were no larger than in spruce forests. Hasegawa et al. (2009) concluded that the total density and species richness of Collembola were not affected by plant species richness in conifer plantations and broad-leaved forests. Vanbergen et al. (2007) suggested that little evidence supported the idea that localscale habitat variables (habitat richness, tree cover, plant species richness, litter cover, soil pH, depth of organic horizon) affected soil fauna diversity, with Collembola diversity being independent of all variables. Therefore, total density or species richness was only minimally affected by the mixing of broad-leaved trees and conifer plantations. The difference of crown trees did not explain also the difference of species composition of Collembola, but that of the forest floor plants did in the present study. Hasegawa et al. (2009) reported that a conversion from broad-leaved forests to artificial cedar stands affect the community composition. They showed
Table 4 Litter weight, water content and soil physical and chemical properties (pH, EC, C, N and C/N) in soil on each site.
L1 L2 L3 L4 L5 M1 M2 M3 M4 M5
Forest floor litter weight (kg m−2 )
Thickness of litter (cm)
Water content in litter
Water content in soil (based on weight)
Water content in soil (based on volume)
pH
EC (mS/m)
C (%)
N (%)
C/N
3.50 2.58 2.98 2.89 3.18 3.17 2.77 3.04 3.19 2.59
3.76 2.77 3.27 2.98 3.16 3.38 3.49 2.96 3.20 2.80
2.13 1.18 1.91 1.03 1.61 1.14 1.29 1.20 1.52 1.48
1.32 0.97 1.15 0.90 1.12 1.02 0.99 0.83 1.17 1.17
0.28 0.24 0.22 0.25 0.25 0.25 0.26 0.24 0.26 0.29
5.0 4.9 4.9 5.1 5.1 4.7 4.7 5.0 4.7 5.1
1.9 2.8 2.8 2.5 3.0 1.8 2.0 2.0 1.7 1.3
23.9 19.8 18.9 15.5 20.3 16.2 21.4 13.6 21.5 22.0
1.2 1.1 1.1 0.9 1.2 0.9 1.1 0.8 1.1 1.2
19.5 17.8 17.8 17.8 17.4 18.1 19.3 16.6 20.1 19.1
Table 5 Spearman rank correlation coefficients between collembolan community indices [density, species richness, diversity index (H ) and evenness index (J )] and environmental variables. Significant combinations at p < 0.05 were shown. Litter Density Total basal area Density of larch Forest age Sasa biomass (current year) Average species richness of forest floor plants Forest floor litter biomass Forest floor litter thickness Soil pH
Soil Species richness
Diversity (H )
Evenness (J )
Density
Species richness
Diversity (H )
Evenness (J )
−0.66 −0.68
0.68 0.93
0.70 0.78
−0.81
0.71 0.67
0.67 0.66
0.78
0.80
−0.83
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
Fig. 2. Mean density (individuals m−2 ) of total Collembola in the litter layer (a) and soil layer (b) in pure larch plots (L1–L5) and in mixed plots (M1–M5). Bars indicate standard errors. The average of nine samples in spring (open column) and autumn (screen column) are shown.
Fig. 3. Species richness of Collembola in the litter layer (a) and soil layer (b) in pure larch plots s (L1–L5) and in mixed plots (M1–M5). Bars indicate standard errors. Totals from nine samples in spring (open column), autumn (screen column), and the total through the study period are shown.
5
Fig. 4. Diversity index (H ) of Collembola in the litter layer (a) and soil layer (b) in pure larch plots (L1–L5) and in mixed plots (M1–M5). Bars indicate standard errors. Totals from nine samples in spring (open column), autumn (screen column), and the total through the study period are shown.
that the species richness of fungal feeders and sucking feeders was positively correlated with the species richness of forest floor plants; thus, differences in the collembolan species composition may result from changes in the ground flora. Salamon and Alphei (2009) showed that the density of fungal feeders increased in mixed stands, whereas the density of the epedaphic and partly herbivorous group was larger in the spruce stands. They speculated that this was presumably due to the higher diversity of ground vegetation in the spruce stands. The previous studies which showed the significant effects of crown trees on collembolan communities also showed the effects of forest floor plants (Hasegawa et al., 2009; Salamon and Alphei, 2009). The changes in communities in crown trees and those in forest floor plants seemed to occur simultaneously in the previous studies, because their studies discussed the mixing of deciduous broad-leaved and evergreen conifer trees and light condition on the forest floor were different between pure evergreen conifer plantation and mixed forest. In contrast, our study has discussed the mixing of deciduous conifer (larch) and deciduous broad-leaved trees (oak, birch etc.). This difference of crown trees in our study seemed be hard to cause the different light condition on forest floor between pure larch and mixed plots. Therefore, communities of crown trees may not be well related to communities of forest floor plants, and collembolan communities. The development of the forest floor plant communities seemed to be determined by the silvicultural schedules, including a closed canopy and thinning. Conifer plantation management may therefore indirectly affect the collembolan community. Hasegawa et al. (2012) suggested that soil mite communities in the conifer plantation plots were related to various indices of understory plants. Hishi et al. (2009) reported that the species composition of oribatid mites in different sites was influenced by the understory vegetation type
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8 6
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
(a)
(b)
Fig. 5. Evenness index (J ) of Collembola in the litter layer (a) and soil layer (b) in pure larch plots (L1–L5) and in mixed plots (M1–M5). Bars indicate standard errors. Totals from nine samples in spring (open column), autumn (screen column), and the total through the study period are shown.
in conifer plantations (deciduous broad-leaved, evergreen broadleaved, Sasa). The effects of the different undergrowth composition on soil mites could be attributable to the addition of diverse leaf litter. Hishi et al. (2009) suggested that in monocultural plantations, the resource diversity on the forest floor resulting from the undergrowth could be important to the beta diversity of oribatid mites. In our study, the limited variation of habitat and food resources for Collembola in larch plantations may have led to a bigger response to understory plants because the effect seemed to be more than solely a response to the crown trees. F. hidakana, which had a positive correlation with grass biomass in this study, has been recorded in forest ecosystems (Hasegawa et al., 2006). This species is also known to be able to suppress damping-off disease in cabbage and Chinese cabbage in the rhizosphere by feeding on the Rhizoctonia solani fungus (Shiraishi et al., 2003). Therefore, F. hidakana may be able to feed on fungi and survive in the rhizosphere of grassland. I. viridis, which had a negative correlation with grass biomass in the litter layer, is known to be a species favoring open sites (Gisin, 1943; Ponge, 1993). Therefore I. viridis here favored more open space in the plots. Densities of X. brevispina were correlated with the forest floor plant biomass. The herb layer in forest ecosystems plays an important role for epigeic species (Salamon and Alphei, 2009), probably as a food or habitat resource. X. brevispina has a life history involving migration from the soil layer to the tree canopy (Itoh, 1991; Yoshida and Hijii, 2006), and the biomass of forest floor plants may therefore affect their survival. Soil properties effects on collembolan communities have been investigated (Hågvar and Abrahamsen, 1984; Loranrger et al., 2001; Cassagne et al., 2003). In the European studies, the abundance of relatively dominant hemiedaphic Collembola species such as
Fig. 6. (a) Partial redundancy analysis (pRDA) ordination diagrams for the collembolan community in the litter layer. Solid arrows show the direction of increase for grass biomass on the forest floor, which was selected by forward selection (p < 0.05). Dotted arrows show the direction in which species with a significant correlation with grass biomass increased. Fh: Folsomia hidakana, Aa: Arrhopalites alticolus, Iv: Isotoma viridis, P2: Pseudosinella sp. 2. Filled circle: collembolan communities in pure larch plots; open circle: collembolan communities in mixed plots. (b) pRDA ordination diagrams for the collembolan community in the soil layer. Solid arrows show the direction of increase for total forest floor plant biomass, which was selected by forward selection (p < 0.05). Dotted arrows show the direction in which species with a significant correlation with total forest floor plant biomass increased. Pr1: Protaphorura sp. 1, Xb: Xenylla brevispina, Am: Arrhopalites minutus, Dnp: Desoria notabilis cf. pallida.
Parisotoma notabilis and Isotomiella minor appear to depend predominantly on abiotic factors, most importantly soil pH (Salamon and Alphei, 2009). In our study, Isotomiella tamurai and D. notabilis cf. pallida were relatively abundant species, with P. notabilis and I. minor dominating. The dominant species were similar to those dominating at the study sites in European forests, and therefore similar results to those found from European studies may be expected. However, soil properties (including soil pH) were not very different between stand types. The small variation in soil characteristics between plots may result in similar collembolan community structures even between different forest types in the study area. The soils in our study area have very dark A horizons (melanic epipedon) and contain up to 10% organic carbon (Table 4). This is consistent with the properties of melanic Andisols.
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
Melanic Andisols, which have a thick black horizon at or near the soil surface, are generally found under grassland vegetation, and Japanese pampas grass (Miscanthus sinensis) vegetation is believed to be responsible for the development of these black-colored soils (Hiradate et al., 2004; Shoji et al., 1994). Although the present vegetation of our study area is larch plantation forest, meadows with patchy forest covered all around the area about 90 years ago at least (Misawa, 1929). The present soils in our study area seems to be influenced by the surface soil characteristics in the era of meadows (black soils) and reduces the variation in soil characteristics between plots. The factors discussed before are often similar between plots located close to each other, and changes occur synchronously. Therefore, the distance between study plots is likely to have influenced the similarity of the community structures of the soil fauna in previous investigations (Lindo and Winchester, 2009). Studies on Collembola and Oribatida communities in four recently emerged nunataks (ice-free land in glacial areas) in Iceland has suggested that isolation of a few kilometers does not affect colonization by soil invertebrates, but that geographical distance does influence the species composition, indicating that the community is assembled by dispersal, e.g., the mass effect (Ingimarsdóttir et al., 2012). Borcard and Legendre (2002) suggested that studies using PCNM efficiently detected spatial factors in oribatid mite distribution data. In our study for both litter and soil collembolan communities, about 20% of the variance of species data was explained by the spatial factors extracted by the PCNM. The percentage was equivalent to the values obtained for the forest floor plants. Therefore, the process of partialling out the spatial factors was effective for determining the factors influencing community structures. 5. Conclusions The density and species richness of Collembola were not significantly different between pure larch plots and mixed plots. Using pRDA, the variance of collembolan species data in the litter layer was explained by the biomass of grass on forest floor, and variance in the soil layer was explained by the biomass of total forest floor plants. These results suggested that the biomass or composition of forest floor plants influenced the collembolan community more than the crown trees in this area. In contrast, soil properties were not so different between plots, and not affected to the collembolan communities. Acknowledgments We thank the staff of the Insect Ecology Laboratory and the Kiso Experimental Station at the Forestry and Forest Products Research Institute for their help and guidance in the field research and laboratory analysis. This study was supported by a Grant-in-aid for Scientific Research sponsored by the Ministry of Education, Culture, Sports, Science and Technology, Japan [(C) (22580175)]. References Borcard, D., Legendre, P., 2002. All scale spatial analysis of ecological data by means of principal coordinates of neighbour matrices. Ecol. Model. 153, 51–68. Cassagne, N., Gers, C., Gauquelin, T., 2003. Relationships between Collembola, soil chemistry and humus types in forest stands (France). Biol. Fertil. Soils 37, 355–361. Coulson, S.I., Hodkinson, I.D., Webb, N.R., 2003. Microscale distribution patterns in high Arctic soil microarthropod communities: the influence of plant species within the vegetation mosaic. Ecography 26, 801–809. Dombos, M., 2001. Collembola of loess grassland: effects of grazing and landscape on community composition. Soil Biol. Biochem. 33, 2037–2045. Dray, S., Legendre, P., Peres-Neto, P.R., 2006. Spatial modelling: a comprehensive framework for principal coordinate analysis of neighbour matrices (PCNM). Ecol. Model. 196, 483–493.
7
Eisenhauer, N., Yee, K., Johnson, E.A., Maraun, M., Parkinson, D., Straube, D., Scheu, S., 2011. Positive relationship between herbaceous layer diversity and the performance of soil biota in a temperate forest. Soil Biol. Biochem. 43, 462–465. Gisin, H., 1943. Ökologie und Lebensgemeinschaften der Collembolen im schweizerischen Excursionsgebiet Basels. Rev. Suisse Zool. 50, 94. Greenslade, P., 1997. Are Collembola useful as indicators of the conservation value of native grasslands? Pedobiologia 41, 215–220. Hågvar, S., 1982. Collembola in Norwegian coniferous forest soils. I. Relations to plant communities and soil fertility. Pedobiologia 24, 255–296. Hågvar, S., Abrahamsen, G., 1984. Collembola in Norwegian coniferous forest soils. III. Relation to soil chemistry. Pedobiologia 27, 331–339. Hasegawa, M., 2001. The relationship between the organic matter composition of a forest floor and the structure of a soil arthropod community. Eur. J. Soil Biol. 37, 281–284. Hasegawa, M., Fukuyama, K., Makino, S., Okochi, I., Goto, H., Mizoguchi, T., Sakata, T., Tanaka, H., 2006. Collembolan community dynamics during deciduous forests regeneration in Japan. Pedobiologia 50, 117–126. Hasegawa, M., Fukuyama, K., Makino, S., Okochi, I., Tanaka, H., Okabe, K., Goto, H., Mizoguchi, T., Sakata, T., 2009. Collembolan community in broad-leaved forests and in conifer stands of Cryptomeria japonica in central Japan. Pesq. Agropec. Bras. 44, 881–890. Hasegawa, M., Okabe, K., Fukuyama, K., Makino, S., Okochi, I., Tanaka, H., Goto, H., Mizoguchi, T., Sakata, T., 2012. Community structures of Mesostigmata, Prostigmata and Oribatida in broad-leaved regeneration forests and conifer plantations of various ages. Exp. Appl. Acarol., http://dx.doi.org/10.1007/s10493-012-9618-x. Hiradate, S., Nakadai, T., Shindo, H., Yoneyama, T., 2004. Carbon source of humic substances in some Japanese volcanic ash soils determined by carbon stable isotopic ratio, ␦13C. Geoderma 119, 133–141. Hishi, T., Ikezaki, S., Enoki, T., 2009. Effect of understory vegetations on species diversity of oribatid mites in an abandoned Chamaecyparis obtusa plantation forest within Ochozu watershed. Edaphologia 84, 11–20 (in Japanese with English summary). Ingimarsdóttir, M., Caruso, T., Ripa, J.R., Magnúsdóttir, Ó.B., Migliorini, M., Hedlund, K., 2012. Primary assembly of soil communities: disentangling the effect of dispersal and local environment. Oecologia, http://dx.doi.org/10.1007/s00442-012-2334-8. Itoh, R., 1991. Growth and life cycle of an arboreal Collembola, Xenylla brevispina KINOSHITA with special reference to its seasonal migration between tree and forest floor. Edaphologia 45, 33–48. Kaneko, N., Salamanca, F., 1999. Mixed leaf litter effects on decomposition rates and soil microarthropod communities in oak–pine stand in Japan. Ecol. Res. 14, 131–138. Klironomos, J.N., Kendrick, W.B., 1995. Relationships among microarthropods, fungi, and their environment. Plant Soil 170, 183–197. Lindo, Z., Winchester, N.N., 2009. Spatial and environmental factors contributing to patterns in arboreal and terrestrial oribatid mite diversity across spatial scales. Oecologia 160, 817–825. Loranrger, G., Bandyopadhyaya, I., Razzaka, B., Ponge, J.F., 2001. Does soil acidity explain altitudinal sequences in collembolan communities? Soil Biol. Biochem. 33, 381–393. Misawa, K., 1929. Landscape types of the foot of Yatsugatake Volcano. Geogr. Rev. Jpn. 5, 790–821, 872–899 (in Japanese). Nagaike, T., 2002. Differences in plant diversity between conifer (Larix kaempferi) plantations and broad-leaved (Quercus crispula) secondary forests in central Japan. For. Ecol. Manage. 168, 111–123. Nagaike, T., Hayashi, A., Abe, M., Arai, N., 2003. Differences in plant diversity in Larix kaempferi plantations of different ages in central Japan. For. Ecol. Manage. 183, 177–193. Ohsawa, M., 2004. Species richness of Cerambycidae in larch plantations and natural broad-leaved forests of the central mountainous region of Japan. For. Ecol. Manage. 189, 375–385. Ohsawa, M., 2007. The role of isolated old oak trees in maintaining beetle diversity within larch plantations in the central mountainous region of Japan. For. Ecol. Manage. 250, 215–226. Ponge, J.F., 1993. Biocenoses of Collembola in Atlantic temperate grass–woodland ecosystems. Pedobiologia 37, 223–244. R Core Team, 2012. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, ISBN 3-900051-07-0, URL http://www.R-project.org/ Salamon, J.A., Alphei, J., 2009. The Collembola community of a Central European forest: influence of tree species composition. Eur. J. Soil Biol. 45, 199–206. Salamon, J.A., Schaefer, M., Alphei, J., Schmid, B., Scheu, S., 2004. Effects of plant diversity on Collembola in an experimental grassland ecosystem. Oikos 106, 51–60. Salamon, J.A., Scheu, S., Schaefer, M., 2008. The Collembola community of pure and mixed stands of beech (Fagus sylvatica) and spruce (Picea abies) of different age. Pedobiologia 51, 385–396. Scheu, S., Albers, D., Alphei, J., Buryn, R., Klages, U., Migge, S., Platner, C., Salamon, J.A., 2003. The soil fauna community in pure and mixed stands of beech and spruce of different age: trophic structures and structuring forces. Oikos 101, 225–238. Shiraishi, H., Enami, Y., Okano, S., 2003. Folsomia hidakana (Collembola) prevents damping-off in cabbage and Chinese cabbage by Rhizoctonia solani (AG-4) in a greenhouse. Pedobiologia 47, 33–38. Shoji, S., Nanzyo, M., Dahlgren, R.A., 1994. Volcanic Ash Soils: Genesis, Properties and Utilization. Elsevier, Tokyo, Japan, 312 pp.
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8 8
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
Sousa, J.P., da Gama, M.M., Ferreira, C.S., 2003. Effects of replacing oak-woods by eucalyptus on edaphic Collembola communities: does the size and type of plantation matter? Acta. Entomol. Iber. Macaro. 1, 1–10. St. John, M.G., Wall, D.H., Behan-Pelletier, V., 2006. Does plant species co-occurrence influence soil mite diversity? Ecology 87, 625–653. Sylvain, Z.A., Wall, D.H., 2011. Linking soil biodiversity and vegetation: implications for a changing planet. Am. J. Bot. 98, 517–527. Takeda, H., 1987. Dynamics and maintenance of collembolan community structure in a forest soil ecosystem. Res. Popul. Ecol. 29, 291–346. ter Braak, C.J.F., Smilauer, P., 2002. CANOCO Reference Manual and Canodraw for Windows User’s Guide: Software for Canonical Community Ordination (version 4.5). Microcomputer Power, Ithaca, NY, USA.
Vanbergen, A.J., Watt, A.D., Mitchell, R.J., Truscott, A.M., Palmer, S.C.F., Ivits, E., Eggleton, P., Jones, T.H., Sousa, J.P., 2007. Scale-specific correlations between habitat heterogeneity and soil fauna diversity along a landscape structure gradient. Oecologia 153, 713–725. Wallwork, J.A., 1976. The Distribution and Diversity of Soil Fauna. Academic Press, London, UK. Wardle, D.A., Yeates, G.W., Barker, G.M., Bonner, K.I., 2006. The influence of plant litter diversity on decomposer abundance and diversity. Soil Biol. Biochem. 38, 1052–1062. Yoshida, T., Hijii, N., 2006. Seasonal distribution of Xenylla brevispina (Collembola) in the canopy and soil habitat of a Cryptomeria japonica plantation. Pedobiologia 50, 235–242.
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
ARTICLE IN PRESS Applied Soil Ecology xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Applied Soil Ecology journal homepage: www.elsevier.com/locate/apsoil
The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan Motohiro Hasegawa a,∗ , Aino T. Ota b , Daisuke Kabeya a , Toru Okamoto a , Tomoyuki Saitoh a , Yoshihiko Nishiyama c a
Forestry and Forest Products Research Institute, Tsukuba, Ibaraki 305-8687, Japan Yokohama National University, Yokohama, Kanagawa 240-8501, Japan c Kiso Experimental Forest, Forestry and Forest Products Research Institute, Kiso, Nagano 397-0001, Japan b
a r t i c l e
i n f o
Article history: Received 10 December 2012 Received in revised form 3 June 2013 Accepted 17 June 2013 Keywords: Larch Larix kaempferi Broad-leaved trees Collembola Community structure Forest floor plants
a b s t r a c t The effects of broad-leaved trees on the collembolan community in larch plantations were investigated at the foot of Mt. Yatsugatake (1200–1400 m a.s.l.) in Japan. The study sites comprised five pure larch plantation plots (larch dominated more than 95% of the area at breast height) and five mixed forest plots (larch dominated between 50% and 80% of the area at breast height). We compared the collembolan community structures between stand types and related them to the plant community composition and soil properties at each plot. Density and species richness of Collembola were not significantly different between pure larch and mixed plots. Using partial redundancy analysis (pRDA), the variance of collembolan species data in the litter layer was explained by the biomass of grass on the forest floor, and the variance in the soil layer was explained by the biomass of total forest floor plants. These results suggest that the biomass or the composition of forest floor plants influence the collembolan community more than the crown trees in this area. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In Japan, Larix kaempferi, a deciduous conifer species, is planted in dry inland areas typified by light snow cover and cold winters (Nagaike et al., 2003). Most of these plantations were established after the 1940s to replace primary and secondary broad-leaved deciduous forests that had been substantially overharvested, but the contributions of L. kaempferi plantations to regional biodiversity are still unknown (Nagaike, 2002). To achieve ecologically sustainable forest management in this region, studies have been undertaken to determine the status of biodiversity in forests and the effects of forest management practices in the area, i.e., the basic information necessary for ecologically sustainable management (Nagaike, 2002; Ohsawa, 2004, 2007). The study area is covered by Japanese larch plantations, which were established for timber production, and thinning has been conducted twice within a 45year period in this area to enhance the growth of dominant trees. After the planting of larch seedlings, trees species (e.g., birch, oak, and pine) started to regenerate from the stumps or seeds, and grow
∗ Corresponding author. Tel.: +81 298 29 8251; fax: +81 298 73 1543. E-mail addresses: [email protected] (M. Hasegawa), [email protected] (A.T. Ota), [email protected] (D. Kabeya), [email protected] (T. Okamoto), [email protected] (T. Saitoh), [email protected] (Y. Nishiyama).
alongside the larch. During the mowing and thinning stage, mixed trees were cut or left, depending on prevailing silvicultural management policy. As a result, a mosaic of larch plantations with various mixtures of broad-leaved trees has developed over the study area. Plant species identity, vegetation composition, physiology, chemistry, and phenology all influence soil invertebrate community composition (Sylvain and Wall, 2011). The effects of litter species richness on soil fauna have been tested using mixed litter bag experiments (Kaneko and Salamanca, 1999; Takeda, 1987), which have revealed idiosyncratic relationships between plant species and soil animals (St. John et al., 2006; Wardle et al., 2006). The effects of mixing tree species on Collembola have been investigated in a European spruce plantation mixed with beech (Salamon and Alphei, 2009; Salamon et al., 2008; Scheu et al., 2003), and these studies have suggested relatively moderate effects on the collembolan community. Total density and species richness did not change significantly, but species composition was affected (Salamon and Alphei, 2009; Scheu et al., 2003). To determine the effects of mixing tree species on Collembola, its community structure should be evaluated using the response of different functional groups or through multivariate analysis. Salamon et al. (2008) suggest that forest age and stand type are likely to impact Collembola communities via changes in the amount and quality of food resources, including both living plants and herb litter materials. The effects of herbs on Collembola have been emphasized in grassland and meadow studies (Dombos, 2001;
0929-1393/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsoil.2013.06.005
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
2
Greenslade, 1997; Salamon et al., 2004), and the effects of the undergrowth in forest sites were also recently studied (Eisenhauer et al., 2011). The findings have revealed that studies should not only consider the effects of crown trees but also those of undergrowth on collembolan communities. In this study, we investigated the effects of a mixture of broadleaved trees on Collembola in terms of their density, diversity, and species composition. The relationship of the collembolan community with the stand type, forest floor plants, and soil abiotic factors are discussed.
2. Materials and methods The study area was located on the south and west slopes of Yatsugatake mountain range, in Yamanashi and Nagano, central Japan (approximately 36◦ 50–56 N, 140◦ 34–35 E, 1200–1400 m a.s.l.; mean annual air temperature 7.1 ◦ C; mean annual precipitation 1454 mm). Larch (L. kaempferi) was planted at the study area after a previous land cover of grass fields or sparse pine forests. Landscape at 1200–1400 m a.s.l. on the slope of Yatsugatake mountain range was totally at most covered by the larch plantation. The forest of this area was mostly national or prefectural forest and classified into fine compartment (with several ha of area). The management plan for the compartment was often determined as this small unit, therefore the compartment in the vicinity could have different forest structures. Ten study plots were located within a 15 × 20 km area (Fig. 1). The forest age was 40–50 years at the time of sampling (Table 1). The total basal area of each plot ranged from 22.9 to 41.2 m2 ha−1 . Size of a plot was 30 m × 30 m. Each plot was divided in 9 sub-plots: where they 10 m × 10 m. Thinning had been conducted several times until 40 years after planting, but in some plots, all of the trees except larch had been completely removed (Table 1). In other plots, Betula platyphylla, Quercus mongolica, and other tree species had invaded and were left in place during the thinning procedure. As a result, the percentage of larch based on basal area in each plot ranged from 100% to 46.4% (Table 1). We classified these plots as five pure larch plots (>95% larch) and five mixed plots (between 50% and 80% larch). Ideally, in order to consider the effects of the forest category to the soil fauna, these two categories of plots should be set pair-wisely and replicated the pair of plots in the study area. However, because of the limited availability of the study plots, we only fulfilled these design in the part of the plots (Fig. 1). Therefore, we scattered these two categories of plots in the study area as evenly as possible shown in Fig. 1. The average distances (±standard errors) among pure larch plots, among mixed plots and between different categories of plots were 5553 (±977), 5389 (±677) and 5895 (±1195) m respectively. To extract the spatial autocorrelation structure to the collembolan community composition, we use the principal coordinates of neighbor matrices (PCNM) as explained in a statistical analysis. In October 2006 and May 2007, samples from the litter layer and soil (0–5 cm) were collected with a corer (125 mL, 5 cm depth, 25 cm2 area) from each subplot. Most Collembola occur in the litter layer and the upper mineral soil layer, within the top 5 cm of the profile. In total, 180 samples (10 plots × 9 subplots × 2 dates) for each substrate (litter and soil) were collected. Collembola were extracted using a Macfadyen high gradient extractor at a constant temperature of 35 ◦ C for 7 days. The water content of the litter layer was calculated as (wet weight of litter − dry weight of litter)/dry weight of litter). Soil samples were taken from five quadrats in each plot in August 2007 to determine physical and chemical parameters. Cores of 100 mL of the top 5 cm of soil were collected for chemical analysis. For soil pH and EC analysis, 5 g of fresh soil was mixed with 25 mL of deionized water. A glass electrode (HM14P; DKK-Toa
Corp., Tokyo, Japan) was used to measure pH and a conductivity cell electrode (SC82; Yokogawa Electric Corp., Tokyo, Japan) was used to measure soil electrical conductivity (EC). Total carbon and nitrogen concentrations in soil samples were measured with an NC analyzer (Sumigraph NC-900; SCAS Ltd., Tokyo, Japan). The airdried litter layer was classified as broad-leaved, needles (larch), and bamboo grass (Sasa nipponica), and the weight of each category was estimated. At each plot, the diameter at breast height (DBH) above 5 cm was measured for all trees in each plot. We set five 0.5 m2 circles randomly in each plot, and cut out the aboveground forest floor vegetation in each circle. The forest floor plant cuttings were classified into trees, grass, herbs, and Sasa, weighed for each category and dried to constant mass at 70 ◦ C for three days. The number of plant species in each circle was counted. 2.1. Statistical analysis 2.1.1. Detection of spatial autocorrelation, RDA, variation partitioning and GLMM To extract the spatial autocorrelation structure, we used principal coordinates of neighbor matrices (PCNM) developed by Borcard and Legendre (2002) because it is flexible for expressing various spatial scales and can be used in redundancy analysis (RDA). This method creates a set of explanatory variables (PCNMs: eigenvectors) that represent the structure at all spatial scales from the distance matrix among the plots. Ten PCNMs were calculated in our study, and we referred to them as PCNM-1 to PCNM-10. We used the library spacemakeR (Dray et al., 2006) in the statistical environment R (R Core Team, 2012), to calculate the PCNMs. For the RDA, environmental variables were selected by a forward-selection procedure to detect meaningful factors and so reduce the large number of potential factors. Variables were incorporated stepwise into the model according to their increasing effect on the variance (contribution to the eigenvalues of the model), and their significance was tested by comparing them to Monte Carlo permutations (5000 times) of the null model (which does not have the variable to be incorporated). We set a standard selection criterion (p < 0.05) to retain parameters for use in the final models. All of the RDA and forward selections were performed with CANOCO for Windows, version 4.5 (ter Braak and Smilauer, 2002). Initially, we conducted RDA using only the PCNMs as explanatory variables, and the significant variables were selected as relevant spatial autocorrelations in the species composition by a forward-selection procedure. We wanted to determine the effect of the environment variables after partialling out the effect of spatial autocorrelation. Therefore, we executed a partial RDA using the selected PCNMs as covariables and other environmental variables as explanatory variables. We investigated the effects of forest category of plots (pure larch and mixed forest plots, as the fixed effect) to the environmental variables and collembolan community structures. Because the sampling protocol in this study formed nested structure (i.e., the samplings were conducted in 5 subplots within each plot for 2 seasons), the effect of the plot difference and that of seasonal difference were also considered as the random effects. Hence, generalized linear mixed models (GLMMs) (nlme library and glmmLM library on R version 2.13.0; (R Core Team, 2012) were used for the analyses. For comparison of the litter weight, soil chemical and physical properties, and the biomass of forest floor plants between pure larch and mixed forest plots we used a GLMM with the assumption of a Gaussian distribution. We also used a GLMM for the density of Collembola with the assumption of a negative binomial distribution and for the species richness with the assumption of a Poisson distribution due to discrete variables. Significance tests were based on the t-statics for each parameter. Spearman rank correlation
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
ARTICLE IN PRESS
G Model APSOIL-1879; No. of Pages 8
M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
3
Fig. 1. Maps of study plots. Filled circle: pure larch plots; open circle: mixed plots.
coefficients were calculated between environmental variables and various indices of the collembolan community. 3. Results The biomass of tree broad-leaves in the forest floor vegetation biomass was significantly larger in pure larch plots than in mixed plots (Table 2; GLMM, p < 0.05). Other biomass and average species richness of forest floor vegetation were not significantly different between pure larch and mixed plots (GLMM, p > 0.05). All soil properties (Table 3) were not significantly different between pure larch and mixed plots (GLMM, p > 0.05). The weight of larch needles in the litter layer was significantly larger in the pure larch plots than in the mixed plots (Table 4; GLMM, p < 0.05). In contrast, the weight of leaves of broad-leaved species in the litter layer was significantly larger in the mixed plots than in the pure larch plots (p < 0.05). The weight of Sasa in the litter layer was not significantly different. The density of Collembola in the litter and soil layers was not significantly different between pure larch plots and mixed plots (GLMM, p > 0.05; Fig. 2a and b). Species richness and the H’ and J’ indices in the litter and soil layers were also not significantly different between pure larch plots and mixed plots (Figs. 3–5a and b). In the litter layer, the density of Collembola was positively correlated with litter weight or thickness (Table 5). In contrast, diversity and evenness of Collembola were negatively correlated with litter thickness
(probably due to the increased density of dominant species). In the soil layer, the density of Collembola was positively correlated with the total basal area. Species richness was positively correlated with forest age. Diversity and evenness in the soil layer were positively correlated with pH and the species richness of forest plants. Partial RDA ordination diagrams for the collembolan community in the litter layer (eigenvalue for the 1st axis: 0.17) selected grass biomass to explain the variance of species in forward selection (p < 0.05) (Fig. 6a). Densities of Folsomia hidakana and Arrhopalites alticolus were positively correlated, and those of Isotoma viridis and Pseudosinella sp. 2 were negatively correlated with grass biomass (p < 0.05). The spatial factor extracted from PCNM and grass biomass solely explained 28.6% and 17.3% of the variance of species data in the litter layer, and 1.2% was duplicated by both factors. Partial RDA ordination diagrams for the collembolan community in the soil layer (eigenvalue for the 1st axis: 0.21) selected total forest floor plant biomass to explain the variance of species in forward selection (p < 0.05) (Fig. 6b). Densities of Protaphorura sp. 1, Xenylla brevispina, and Arrhopalites minutus were positively correlated, and that of Desoria notabilis cf. pallida was negatively correlated with the total forest floor plant biomass. Spatial factors extracted from PCNM and grass biomass solely explained 20.5% and 20.8% in the variance of species data in the litter layer, and 0.6% was duplicated by both factors.
Table 1 Basal area, percentage of larch and species richness of tall trees in each plot. Pure larch plots were labeled as L1–L5, and mixed plots as M1–M5. Site code
Total basal area (m2 ha−1 )
Basal area of larch (m2 ha−1 )
Basal area of trees except larch (m2 ha−1 )
Percentage of larch based on basal area (%)
Density of tall trees (ha−1 )
Density of larch trees (ha−1 )
Species richness of tall trees
Forest age
L1 L2 L3 L4 L5 M1 M2 M3 M4 M5
32.8 38.0 28.9 28.0 22.9 41.2 28.4 27.4 39.8 28.8
32.8 38.0 28.8 26.9 22.0 33.1 20.0 15.0 19.1 13.4
0 0 0.2 1.1 1.0 8.1 8.5 12.4 20.7 15.4
100 100.0 99.5 96.1 95.8 80.3 70.2 54.8 48.0 46.4
833 978 1067 1156 467 3611 822 1233 1944 1233
833 967 889 533 456 922 489 367 500 356
1 1 4 7 2 15 3 6 13 14
46 38 38 57 55 51 57 38 52 52
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
ARTICLE IN PRESS
G Model APSOIL-1879; No. of Pages 8
M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
4
Table 2 Biomass (g m−2 ) of each category of forest floor plant and average species richness (in 0.5 m2 ) of forest floor plants on each plot. Wood materials of tree seedlings were excluded.
L1 L2 L3 L4 L5 M1 M2 M3 M4 M5
Fern
Grass
Forb
Sasa (current year)
Trees (leave plus needles)
Total biomass
Average species richness (in 0.5 m2 )
0.0 0.8 0.0 1.2 16.3 3.1 39.8 0.0 1.2 0.5
0.1 21.3 0.0 11.6 56.8 7.1 6.8 10.2 4.0 13.3
0.1 11.6 0.0 1.0 9.6 0.5 0.4 1.2 0.1 0.4
6.3 5.3 18.5 11.0 1.4 26.8 8.9 1.9 5.5 7.7
37.6 158.7 157.7 40.2 158.6 77.2 75.0 6.4 22.2 36.7
163.7 372.5 245.3 248.0 317.9 117.7 200.6 244.6 69.4 111.6
2.1 15.2 0.2 6.9 9.7 6.8 5.8 6.6 8.1 11.1
Table 3 Weight (average of values at two dates) of each category of forest floor litter.
L1 L2 L3 L4 L5 M1 M2 M3 M4 M5
Larix (g m−2 )
Broad-leaves (g m−2 )
Sasa (g m−2 )
1160 1070 765 538 474 664 528 454 478 569
19 6 11 26 33 102 135 258 206 224
222 10 78 117 218 21 211 80 32 38
4. Discussion Important factors determining soil fauna composition include vegetation type, soil structure, soil chemistry, organic matter, soil microflora, and soil moisture and temperature (Coulson et al., 2003; Hasegawa, 2001; Klironomos and Kendrick, 1995; Sousa et al., 2003; Wallwork, 1976). Collembola generally display little response to changes in vegetation structure (Hågvar, 1982; Salamon and Alphei, 2009). Our study revealed small differences
in total density and species richness between pure larch plots and mixed plots. Effects of mixing spruce forests with beech on the collembolan community in Europe have been investigated (Salamon and Alphei, 2009; Salamon et al., 2008; Scheu et al., 2003). Scheu et al. (2003) also reported that the mixing of beech and spruce had no significant effect on collembolan biomass. Salamon and Alphei (2009) suggested that the total density and species richness in mixed forests were no larger than in spruce forests. Hasegawa et al. (2009) concluded that the total density and species richness of Collembola were not affected by plant species richness in conifer plantations and broad-leaved forests. Vanbergen et al. (2007) suggested that little evidence supported the idea that localscale habitat variables (habitat richness, tree cover, plant species richness, litter cover, soil pH, depth of organic horizon) affected soil fauna diversity, with Collembola diversity being independent of all variables. Therefore, total density or species richness was only minimally affected by the mixing of broad-leaved trees and conifer plantations. The difference of crown trees did not explain also the difference of species composition of Collembola, but that of the forest floor plants did in the present study. Hasegawa et al. (2009) reported that a conversion from broad-leaved forests to artificial cedar stands affect the community composition. They showed
Table 4 Litter weight, water content and soil physical and chemical properties (pH, EC, C, N and C/N) in soil on each site.
L1 L2 L3 L4 L5 M1 M2 M3 M4 M5
Forest floor litter weight (kg m−2 )
Thickness of litter (cm)
Water content in litter
Water content in soil (based on weight)
Water content in soil (based on volume)
pH
EC (mS/m)
C (%)
N (%)
C/N
3.50 2.58 2.98 2.89 3.18 3.17 2.77 3.04 3.19 2.59
3.76 2.77 3.27 2.98 3.16 3.38 3.49 2.96 3.20 2.80
2.13 1.18 1.91 1.03 1.61 1.14 1.29 1.20 1.52 1.48
1.32 0.97 1.15 0.90 1.12 1.02 0.99 0.83 1.17 1.17
0.28 0.24 0.22 0.25 0.25 0.25 0.26 0.24 0.26 0.29
5.0 4.9 4.9 5.1 5.1 4.7 4.7 5.0 4.7 5.1
1.9 2.8 2.8 2.5 3.0 1.8 2.0 2.0 1.7 1.3
23.9 19.8 18.9 15.5 20.3 16.2 21.4 13.6 21.5 22.0
1.2 1.1 1.1 0.9 1.2 0.9 1.1 0.8 1.1 1.2
19.5 17.8 17.8 17.8 17.4 18.1 19.3 16.6 20.1 19.1
Table 5 Spearman rank correlation coefficients between collembolan community indices [density, species richness, diversity index (H ) and evenness index (J )] and environmental variables. Significant combinations at p < 0.05 were shown. Litter Density Total basal area Density of larch Forest age Sasa biomass (current year) Average species richness of forest floor plants Forest floor litter biomass Forest floor litter thickness Soil pH
Soil Species richness
Diversity (H )
Evenness (J )
Density
Species richness
Diversity (H )
Evenness (J )
−0.66 −0.68
0.68 0.93
0.70 0.78
−0.81
0.71 0.67
0.67 0.66
0.78
0.80
−0.83
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
Fig. 2. Mean density (individuals m−2 ) of total Collembola in the litter layer (a) and soil layer (b) in pure larch plots (L1–L5) and in mixed plots (M1–M5). Bars indicate standard errors. The average of nine samples in spring (open column) and autumn (screen column) are shown.
Fig. 3. Species richness of Collembola in the litter layer (a) and soil layer (b) in pure larch plots s (L1–L5) and in mixed plots (M1–M5). Bars indicate standard errors. Totals from nine samples in spring (open column), autumn (screen column), and the total through the study period are shown.
5
Fig. 4. Diversity index (H ) of Collembola in the litter layer (a) and soil layer (b) in pure larch plots (L1–L5) and in mixed plots (M1–M5). Bars indicate standard errors. Totals from nine samples in spring (open column), autumn (screen column), and the total through the study period are shown.
that the species richness of fungal feeders and sucking feeders was positively correlated with the species richness of forest floor plants; thus, differences in the collembolan species composition may result from changes in the ground flora. Salamon and Alphei (2009) showed that the density of fungal feeders increased in mixed stands, whereas the density of the epedaphic and partly herbivorous group was larger in the spruce stands. They speculated that this was presumably due to the higher diversity of ground vegetation in the spruce stands. The previous studies which showed the significant effects of crown trees on collembolan communities also showed the effects of forest floor plants (Hasegawa et al., 2009; Salamon and Alphei, 2009). The changes in communities in crown trees and those in forest floor plants seemed to occur simultaneously in the previous studies, because their studies discussed the mixing of deciduous broad-leaved and evergreen conifer trees and light condition on the forest floor were different between pure evergreen conifer plantation and mixed forest. In contrast, our study has discussed the mixing of deciduous conifer (larch) and deciduous broad-leaved trees (oak, birch etc.). This difference of crown trees in our study seemed be hard to cause the different light condition on forest floor between pure larch and mixed plots. Therefore, communities of crown trees may not be well related to communities of forest floor plants, and collembolan communities. The development of the forest floor plant communities seemed to be determined by the silvicultural schedules, including a closed canopy and thinning. Conifer plantation management may therefore indirectly affect the collembolan community. Hasegawa et al. (2012) suggested that soil mite communities in the conifer plantation plots were related to various indices of understory plants. Hishi et al. (2009) reported that the species composition of oribatid mites in different sites was influenced by the understory vegetation type
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8 6
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
(a)
(b)
Fig. 5. Evenness index (J ) of Collembola in the litter layer (a) and soil layer (b) in pure larch plots (L1–L5) and in mixed plots (M1–M5). Bars indicate standard errors. Totals from nine samples in spring (open column), autumn (screen column), and the total through the study period are shown.
in conifer plantations (deciduous broad-leaved, evergreen broadleaved, Sasa). The effects of the different undergrowth composition on soil mites could be attributable to the addition of diverse leaf litter. Hishi et al. (2009) suggested that in monocultural plantations, the resource diversity on the forest floor resulting from the undergrowth could be important to the beta diversity of oribatid mites. In our study, the limited variation of habitat and food resources for Collembola in larch plantations may have led to a bigger response to understory plants because the effect seemed to be more than solely a response to the crown trees. F. hidakana, which had a positive correlation with grass biomass in this study, has been recorded in forest ecosystems (Hasegawa et al., 2006). This species is also known to be able to suppress damping-off disease in cabbage and Chinese cabbage in the rhizosphere by feeding on the Rhizoctonia solani fungus (Shiraishi et al., 2003). Therefore, F. hidakana may be able to feed on fungi and survive in the rhizosphere of grassland. I. viridis, which had a negative correlation with grass biomass in the litter layer, is known to be a species favoring open sites (Gisin, 1943; Ponge, 1993). Therefore I. viridis here favored more open space in the plots. Densities of X. brevispina were correlated with the forest floor plant biomass. The herb layer in forest ecosystems plays an important role for epigeic species (Salamon and Alphei, 2009), probably as a food or habitat resource. X. brevispina has a life history involving migration from the soil layer to the tree canopy (Itoh, 1991; Yoshida and Hijii, 2006), and the biomass of forest floor plants may therefore affect their survival. Soil properties effects on collembolan communities have been investigated (Hågvar and Abrahamsen, 1984; Loranrger et al., 2001; Cassagne et al., 2003). In the European studies, the abundance of relatively dominant hemiedaphic Collembola species such as
Fig. 6. (a) Partial redundancy analysis (pRDA) ordination diagrams for the collembolan community in the litter layer. Solid arrows show the direction of increase for grass biomass on the forest floor, which was selected by forward selection (p < 0.05). Dotted arrows show the direction in which species with a significant correlation with grass biomass increased. Fh: Folsomia hidakana, Aa: Arrhopalites alticolus, Iv: Isotoma viridis, P2: Pseudosinella sp. 2. Filled circle: collembolan communities in pure larch plots; open circle: collembolan communities in mixed plots. (b) pRDA ordination diagrams for the collembolan community in the soil layer. Solid arrows show the direction of increase for total forest floor plant biomass, which was selected by forward selection (p < 0.05). Dotted arrows show the direction in which species with a significant correlation with total forest floor plant biomass increased. Pr1: Protaphorura sp. 1, Xb: Xenylla brevispina, Am: Arrhopalites minutus, Dnp: Desoria notabilis cf. pallida.
Parisotoma notabilis and Isotomiella minor appear to depend predominantly on abiotic factors, most importantly soil pH (Salamon and Alphei, 2009). In our study, Isotomiella tamurai and D. notabilis cf. pallida were relatively abundant species, with P. notabilis and I. minor dominating. The dominant species were similar to those dominating at the study sites in European forests, and therefore similar results to those found from European studies may be expected. However, soil properties (including soil pH) were not very different between stand types. The small variation in soil characteristics between plots may result in similar collembolan community structures even between different forest types in the study area. The soils in our study area have very dark A horizons (melanic epipedon) and contain up to 10% organic carbon (Table 4). This is consistent with the properties of melanic Andisols.
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
Melanic Andisols, which have a thick black horizon at or near the soil surface, are generally found under grassland vegetation, and Japanese pampas grass (Miscanthus sinensis) vegetation is believed to be responsible for the development of these black-colored soils (Hiradate et al., 2004; Shoji et al., 1994). Although the present vegetation of our study area is larch plantation forest, meadows with patchy forest covered all around the area about 90 years ago at least (Misawa, 1929). The present soils in our study area seems to be influenced by the surface soil characteristics in the era of meadows (black soils) and reduces the variation in soil characteristics between plots. The factors discussed before are often similar between plots located close to each other, and changes occur synchronously. Therefore, the distance between study plots is likely to have influenced the similarity of the community structures of the soil fauna in previous investigations (Lindo and Winchester, 2009). Studies on Collembola and Oribatida communities in four recently emerged nunataks (ice-free land in glacial areas) in Iceland has suggested that isolation of a few kilometers does not affect colonization by soil invertebrates, but that geographical distance does influence the species composition, indicating that the community is assembled by dispersal, e.g., the mass effect (Ingimarsdóttir et al., 2012). Borcard and Legendre (2002) suggested that studies using PCNM efficiently detected spatial factors in oribatid mite distribution data. In our study for both litter and soil collembolan communities, about 20% of the variance of species data was explained by the spatial factors extracted by the PCNM. The percentage was equivalent to the values obtained for the forest floor plants. Therefore, the process of partialling out the spatial factors was effective for determining the factors influencing community structures. 5. Conclusions The density and species richness of Collembola were not significantly different between pure larch plots and mixed plots. Using pRDA, the variance of collembolan species data in the litter layer was explained by the biomass of grass on forest floor, and variance in the soil layer was explained by the biomass of total forest floor plants. These results suggested that the biomass or composition of forest floor plants influenced the collembolan community more than the crown trees in this area. In contrast, soil properties were not so different between plots, and not affected to the collembolan communities. Acknowledgments We thank the staff of the Insect Ecology Laboratory and the Kiso Experimental Station at the Forestry and Forest Products Research Institute for their help and guidance in the field research and laboratory analysis. This study was supported by a Grant-in-aid for Scientific Research sponsored by the Ministry of Education, Culture, Sports, Science and Technology, Japan [(C) (22580175)]. References Borcard, D., Legendre, P., 2002. All scale spatial analysis of ecological data by means of principal coordinates of neighbour matrices. Ecol. Model. 153, 51–68. Cassagne, N., Gers, C., Gauquelin, T., 2003. Relationships between Collembola, soil chemistry and humus types in forest stands (France). Biol. Fertil. Soils 37, 355–361. Coulson, S.I., Hodkinson, I.D., Webb, N.R., 2003. Microscale distribution patterns in high Arctic soil microarthropod communities: the influence of plant species within the vegetation mosaic. Ecography 26, 801–809. Dombos, M., 2001. Collembola of loess grassland: effects of grazing and landscape on community composition. Soil Biol. Biochem. 33, 2037–2045. Dray, S., Legendre, P., Peres-Neto, P.R., 2006. Spatial modelling: a comprehensive framework for principal coordinate analysis of neighbour matrices (PCNM). Ecol. Model. 196, 483–493.
7
Eisenhauer, N., Yee, K., Johnson, E.A., Maraun, M., Parkinson, D., Straube, D., Scheu, S., 2011. Positive relationship between herbaceous layer diversity and the performance of soil biota in a temperate forest. Soil Biol. Biochem. 43, 462–465. Gisin, H., 1943. Ökologie und Lebensgemeinschaften der Collembolen im schweizerischen Excursionsgebiet Basels. Rev. Suisse Zool. 50, 94. Greenslade, P., 1997. Are Collembola useful as indicators of the conservation value of native grasslands? Pedobiologia 41, 215–220. Hågvar, S., 1982. Collembola in Norwegian coniferous forest soils. I. Relations to plant communities and soil fertility. Pedobiologia 24, 255–296. Hågvar, S., Abrahamsen, G., 1984. Collembola in Norwegian coniferous forest soils. III. Relation to soil chemistry. Pedobiologia 27, 331–339. Hasegawa, M., 2001. The relationship between the organic matter composition of a forest floor and the structure of a soil arthropod community. Eur. J. Soil Biol. 37, 281–284. Hasegawa, M., Fukuyama, K., Makino, S., Okochi, I., Goto, H., Mizoguchi, T., Sakata, T., Tanaka, H., 2006. Collembolan community dynamics during deciduous forests regeneration in Japan. Pedobiologia 50, 117–126. Hasegawa, M., Fukuyama, K., Makino, S., Okochi, I., Tanaka, H., Okabe, K., Goto, H., Mizoguchi, T., Sakata, T., 2009. Collembolan community in broad-leaved forests and in conifer stands of Cryptomeria japonica in central Japan. Pesq. Agropec. Bras. 44, 881–890. Hasegawa, M., Okabe, K., Fukuyama, K., Makino, S., Okochi, I., Tanaka, H., Goto, H., Mizoguchi, T., Sakata, T., 2012. Community structures of Mesostigmata, Prostigmata and Oribatida in broad-leaved regeneration forests and conifer plantations of various ages. Exp. Appl. Acarol., http://dx.doi.org/10.1007/s10493-012-9618-x. Hiradate, S., Nakadai, T., Shindo, H., Yoneyama, T., 2004. Carbon source of humic substances in some Japanese volcanic ash soils determined by carbon stable isotopic ratio, ␦13C. Geoderma 119, 133–141. Hishi, T., Ikezaki, S., Enoki, T., 2009. Effect of understory vegetations on species diversity of oribatid mites in an abandoned Chamaecyparis obtusa plantation forest within Ochozu watershed. Edaphologia 84, 11–20 (in Japanese with English summary). Ingimarsdóttir, M., Caruso, T., Ripa, J.R., Magnúsdóttir, Ó.B., Migliorini, M., Hedlund, K., 2012. Primary assembly of soil communities: disentangling the effect of dispersal and local environment. Oecologia, http://dx.doi.org/10.1007/s00442-012-2334-8. Itoh, R., 1991. Growth and life cycle of an arboreal Collembola, Xenylla brevispina KINOSHITA with special reference to its seasonal migration between tree and forest floor. Edaphologia 45, 33–48. Kaneko, N., Salamanca, F., 1999. Mixed leaf litter effects on decomposition rates and soil microarthropod communities in oak–pine stand in Japan. Ecol. Res. 14, 131–138. Klironomos, J.N., Kendrick, W.B., 1995. Relationships among microarthropods, fungi, and their environment. Plant Soil 170, 183–197. Lindo, Z., Winchester, N.N., 2009. Spatial and environmental factors contributing to patterns in arboreal and terrestrial oribatid mite diversity across spatial scales. Oecologia 160, 817–825. Loranrger, G., Bandyopadhyaya, I., Razzaka, B., Ponge, J.F., 2001. Does soil acidity explain altitudinal sequences in collembolan communities? Soil Biol. Biochem. 33, 381–393. Misawa, K., 1929. Landscape types of the foot of Yatsugatake Volcano. Geogr. Rev. Jpn. 5, 790–821, 872–899 (in Japanese). Nagaike, T., 2002. Differences in plant diversity between conifer (Larix kaempferi) plantations and broad-leaved (Quercus crispula) secondary forests in central Japan. For. Ecol. Manage. 168, 111–123. Nagaike, T., Hayashi, A., Abe, M., Arai, N., 2003. Differences in plant diversity in Larix kaempferi plantations of different ages in central Japan. For. Ecol. Manage. 183, 177–193. Ohsawa, M., 2004. Species richness of Cerambycidae in larch plantations and natural broad-leaved forests of the central mountainous region of Japan. For. Ecol. Manage. 189, 375–385. Ohsawa, M., 2007. The role of isolated old oak trees in maintaining beetle diversity within larch plantations in the central mountainous region of Japan. For. Ecol. Manage. 250, 215–226. Ponge, J.F., 1993. Biocenoses of Collembola in Atlantic temperate grass–woodland ecosystems. Pedobiologia 37, 223–244. R Core Team, 2012. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, ISBN 3-900051-07-0, URL http://www.R-project.org/ Salamon, J.A., Alphei, J., 2009. The Collembola community of a Central European forest: influence of tree species composition. Eur. J. Soil Biol. 45, 199–206. Salamon, J.A., Schaefer, M., Alphei, J., Schmid, B., Scheu, S., 2004. Effects of plant diversity on Collembola in an experimental grassland ecosystem. Oikos 106, 51–60. Salamon, J.A., Scheu, S., Schaefer, M., 2008. The Collembola community of pure and mixed stands of beech (Fagus sylvatica) and spruce (Picea abies) of different age. Pedobiologia 51, 385–396. Scheu, S., Albers, D., Alphei, J., Buryn, R., Klages, U., Migge, S., Platner, C., Salamon, J.A., 2003. The soil fauna community in pure and mixed stands of beech and spruce of different age: trophic structures and structuring forces. Oikos 101, 225–238. Shiraishi, H., Enami, Y., Okano, S., 2003. Folsomia hidakana (Collembola) prevents damping-off in cabbage and Chinese cabbage by Rhizoctonia solani (AG-4) in a greenhouse. Pedobiologia 47, 33–38. Shoji, S., Nanzyo, M., Dahlgren, R.A., 1994. Volcanic Ash Soils: Genesis, Properties and Utilization. Elsevier, Tokyo, Japan, 312 pp.
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005
G Model APSOIL-1879; No. of Pages 8 8
ARTICLE IN PRESS M. Hasegawa et al. / Applied Soil Ecology xxx (2013) xxx–xxx
Sousa, J.P., da Gama, M.M., Ferreira, C.S., 2003. Effects of replacing oak-woods by eucalyptus on edaphic Collembola communities: does the size and type of plantation matter? Acta. Entomol. Iber. Macaro. 1, 1–10. St. John, M.G., Wall, D.H., Behan-Pelletier, V., 2006. Does plant species co-occurrence influence soil mite diversity? Ecology 87, 625–653. Sylvain, Z.A., Wall, D.H., 2011. Linking soil biodiversity and vegetation: implications for a changing planet. Am. J. Bot. 98, 517–527. Takeda, H., 1987. Dynamics and maintenance of collembolan community structure in a forest soil ecosystem. Res. Popul. Ecol. 29, 291–346. ter Braak, C.J.F., Smilauer, P., 2002. CANOCO Reference Manual and Canodraw for Windows User’s Guide: Software for Canonical Community Ordination (version 4.5). Microcomputer Power, Ithaca, NY, USA.
Vanbergen, A.J., Watt, A.D., Mitchell, R.J., Truscott, A.M., Palmer, S.C.F., Ivits, E., Eggleton, P., Jones, T.H., Sousa, J.P., 2007. Scale-specific correlations between habitat heterogeneity and soil fauna diversity along a landscape structure gradient. Oecologia 153, 713–725. Wallwork, J.A., 1976. The Distribution and Diversity of Soil Fauna. Academic Press, London, UK. Wardle, D.A., Yeates, G.W., Barker, G.M., Bonner, K.I., 2006. The influence of plant litter diversity on decomposer abundance and diversity. Soil Biol. Biochem. 38, 1052–1062. Yoshida, T., Hijii, N., 2006. Seasonal distribution of Xenylla brevispina (Collembola) in the canopy and soil habitat of a Cryptomeria japonica plantation. Pedobiologia 50, 235–242.
Please cite this article in press as: Hasegawa, M., et al., The effects of mixed broad-leaved trees on the collembolan community in larch plantations of central Japan. Appl. Soil Ecol. (2013), http://dx.doi.org/10.1016/j.apsoil.2013.06.005