Decreased Epiphytic Bryophyte Diversity on Mt. Odaigahara, Japan: Causes and Implications

When wildlife populations grow excessively, they affect other flora and fauna within their ecosystems (Fuller & Gill, 2001; Pellerin et al., 2006; Rooney, 2001; Schutz et al., 2003; Stewart & Burrows, 1989; Stockton et al., 2005; Takatsuki, 2009; Webster et al., 2005). For example, the recent increase in the sika deer population in Japan has led to the degradation of ecosys‐ tems in many areas. From 1979 to 2002, the range of this species expanded by as much as 70% (Nakajima, 2007). Although stripping of bark, grazing on grass, and browsing on tree understories are normal foraging behaviors in deer, these activities in excess can cause se‐ vere damage. Excessive bark stripping causes wood decay, leading to a decline in the forest cover (Akashi & Nakashizuka, 1999; Miquelle & van Ballenberghe, 1989; Yokoyama et al., 2001), and excessive browsing and/or grazing may alter the structure and composition of vegetation on the forest floor (Kumar et al., 2006; Rooney & Waller, 2003; Schutz et al., 2003; Stockton et al., 2005; Webster et al., 2005). These environmental changes indirectly affect oth‐ er organisms in the forest ecosystem (Allombert et al., 2005; Feber et al., 2001; Flowerdew & Ellwood, 2001; Rooney, 2001).


Introduction
When wildlife populations grow excessively, they affect other flora and fauna within their ecosystems (Fuller & Gill, 2001;Pellerin et al., 2006;Rooney, 2001;Schütz et al., 2003;Stewart & Burrows, 1989;Stockton et al., 2005;Takatsuki, 2009;Webster et al., 2005). For example, the recent increase in the sika deer population in Japan has led to the degradation of ecosystems in many areas. From 1979 to 2002, the range of this species expanded by as much as 70% (Nakajima, 2007). Although stripping of bark, grazing on grass, and browsing on tree understories are normal foraging behaviors in deer, these activities in excess can cause severe damage. Excessive bark stripping causes wood decay, leading to a decline in the forest cover (Akashi & Nakashizuka, 1999;Miquelle & van Ballenberghe, 1989;Yokoyama et al., 2001), and excessive browsing and/or grazing may alter the structure and composition of vegetation on the forest floor (Kumar et  To protect forest vegetation from further damage by the increased sika deer population, protective management, for example, wrapping tree trunks in wire mesh, have been implemented in addition to deer population control via culling and the erection of deer-proof fences (Ministry of the Environment-Kinki Regional Environment Office, 2009; Takatsuki, 2009). However, although these measures are effective for the protection of vegetation, they sometimes negatively affect other organisms.
This chapter describes the effects of protective management activities on epiphytic diversity at Mt. Odaigahara in central Japan, which is a hotspot for bryophyte diversity. It discusses 2009). This increase led to serious damage to the forest vegetation in this mountain region (Ministry of the Environment, Kinki Regional Environment Office, 2009); for example, extensive bark stripping by the deer resulted in the dieback of damaged trees, and excessive browsing/ grazing led to the loss of vegetation on the forest floor (Fig. 2). The Ministry of the Environment initiated a forest protection program in 1986 to conserve the forest ecosystem in this region (Ministry of the Environment, Kinki Regional Environment Office, 2009). This program was executed in a part (ca. 703 hectare) of Mt. Odaigahara (Ministry of the Environment, Kinki Regional Environment Office, 2009). To prevent bark stripping by the deer, the trunks of around 32,500 trees were wrapped with wire mesh composed of zinc-coated galvanized iron (Ministry of the Environment, Kinki Regional Environment Office, 2009) (Fig. 3).
The middle part of the tree trunk that does not have wire mesh protection has been debarked by deer. Figure 3. Examples of tree trunks without (left) and with (right) wire mesh protection (Oishi, 2011) The photo on the left shows Sasa nipponica Makino et Shibata. browsed by sika deer. In this photo, the upper parts of the plants were browsed by sika deer. The right photo shows a deer fence and the effects of protection from browsing by sika deer: the height of the plants within the deer fence (back) is greater than that of the plants outside the fence.

Changes in bryophyte diversity and the deer population since the 1960s
In the 1980s, epiphytic bryophyte flora in P. jezoensis var. hondoensis were surveyed in 2 parts of Mt. Odaigahara-Masakitoge and Masakigahara Masakigahara (Doei & Nakanishi, 1984) by using 20 × 20 m quadrants (plots A, B, C, and D in Fig. 1), and we examined the changes in these areas that had occurred over in the last 30 years. These areas were dense forests in the 1960s, but by 2008, they had become open forests because the changes in the environmental conditions after a severe typhoon in 1959 gradually resulted in dieback (Ando, 2009). Because many trees were blown down by the typhoon, the light conditions in the forests improved, and S. nipponica grew vigorously on the forest floor (Ando, 2009). This expansion resulted in an increase in the population of sika deer, which in turn resulted in dieback due to debarking (Yokoyama & Shibata, 1998).
The lower parts of many tree trunks (left) and fallen logs (right) are extensively covered with epiphytic bryophytes because of the high humidity.  In 2008, we surveyed the epiphytic bryophyte flora in almost the same places as those examined in a previous study (Doei & Nakanishi, 1984) by using 20 × 20 m quadrants (plots A, B, C, and D in Fig. 1), and we examined the changes in these areas that had occurred over in the last 30 years. Thus, in direct contrast to the increasing deer population, epiphytic bryophyte diversity significantly declined over time.
Possible reasons for the decrease in bryophyte diversity are that (1) the decline in the forest cover indirectly affected bryophyte diversity because of the changes in the environmental conditions (e.g., air humidity), and (2) the protection of trees using wire mesh directly affected bryophyte diversity because of metal pollution.
To determine the reasons for the decline in bryophyte diversity, we examined the correlation between the diversity of epiphytic bryophytes and environmental variables, including wire mesh protection.  The bars represent the mean value of species richness and epiphyte cover on a single tree, and the error bars represent the corresponding standard deviations.

Site selection
A preliminary survey was conducted to identify plots of forest that were dominated by P. jezoensis var. hondoensis trees, including those with and without a protective wire mesh. In total, 9 plots (each 20 × 20 m in size) were selected ( Fig. 1 A-I; Table 2) to examine the influence of wire mesh protection on epiphytic bryophyte diversity The names of plots were changed from those used in previous reports (Oishi, 2011) to adjust to the context of the present study. Plots A-D were identical to those mentioned in section 2.4. We observed that the tree trunks were completely wrapped with wire mesh, from the ground up to a height of 150-180 cm. The mesh was composed of zinc-coated galvanized iron, a commonly used material for wire meshes (Japan Society of Corrosion Engineering, 2000). In each plot, the tree density (m 2 /plot) was measured on the basis of the total basal area of the trunks. Further, the fraction of the trunk area of the P. jezoensis var. hondoensis trees that had been debarked by sika deer was also recorded in each plot.

Bryophyte sampling
The epiphytic bryophyte flora on the trunks of P. jezoensis var. hondoensis trees in the study plots were surveyed from October to November 2008. The bryophyte species covering the tree trunks from ground level to a height of 1.5 m were examined.

Analysis of the correlation between bryophyte diversity and environmental variables
A simple generalized linear model (GLM) using R software for Windows 2.11.0 (R Development Core Team, 2010) was used to identify the correlations between species richness and the bryophyte cover with respect to environmental variables. To identify the most parsimonious model, we performed automated stepwise model selection using the Akaike information criterion (AIC) using the minimum AIC as the best-fit estimator. Bryophytes that had been identified only up to the genus level were not included in the calculation of species richness if any species of that genus was sampled. The environmental variables used in the GLMs were tree density, host tree diameter at breast height (DBH), percentage of debarked area, and percentage of tree trunks with wire mesh protection.

Results & discussion
Associated with the 110 tree trunks in the sampling plots, 68 species were identified in the bryophyte flora survey: 29 mosses and 39 liverworts (Appendix). Fig. 8 shows the species richness and cover in the study plots. The species richness on a single tree ranged from no species to 34 species (mean = 9.1, SD ± 9.0), while the bryophyte cover ranged from 0 to level 5 (mean = 2.0, SD ± 1.8).
The GLMs constructed using the environmental variables are presented in Table 3. These models showed that the species richness and bryophyte cover were significantly correlated with DBH (height, 1.5 m) and tree density (p < 0.01) but negatively correlated with the presence of wire mesh protection (p < 0.01). The GLMs for species richness and bryophyte cover explained 70.1% and 80.4% of the variance, respectively (p < 0.01 for both models).
High tree density and host tree DBH have been suggested to be beneficial for bryophyte diversity, as they provide better microclimates, e.g., humid conditions (Hazell et al., 1998;Ojala et al., 2000;Thomas et al., 2001). Further, the species richness and bryophyte cover may be positively correlated with high host tree DBH because DBH is correlated with bark features (e.g., bark thickness and bark roughness) (Boudreault et al., 2008;Ojala et al.,2000).
The results of this study raise the question of how wire mesh protection negatively affects bryophyte diversity. Considering that the galvanized iron, which is the primary component of the wire mesh, is coated with zinc, it is likely that the zinc affects the bryophytes. In the next section, we compare the zinc concentrations in bryophytes on trees with and without wire mesh protection.

Bryophyte samples
To examine the influence of wire mesh protection on the bryophytes, inductively coupled plasma-mass spectrometry (ICP-MS) was used to compare the concentration of zinc in bryophyte samples, since zinc coats the wire mesh surface. For this evaluation, 2 species of bryophyte that are commonly found on the trunks of P. jezoensis var. hondoensis trees of this region, both with and without wire mesh, were sampled: H.tristo-viride and S. ampliata. For each species, 3 sets of samples each were collected from trees with and without wire mesh.

Analysis of zinc concentration
Dry samples (0.05-0.10 g) were placed in polytetrafluoroethylene vessels and weighed. Subsequently, 5 mL of nitric acid was added to the samples, and they were digested using a microwave system (MLS-1200 MEGA; Milestone General, Tokyo, Japan) before ICP-MS analysis. The samples were then analyzed using a 7500CX ICP-MS system (Agilent Technologies, Wilmington, DE, USA). Spectral interference was minimized or eliminated using the octopole reaction system, with helium as the reaction gas at a flow rate of 2.5 mL/min. The ICP-MS analysis was repeated twice for each sample, and the mean values were used in onesided Student's t-test comparisons of the zinc concentration from bryophyte samples on tree trunks with and without wire mesh. The bars represent the mean value of species richness and epiphyte cover on a single tree, and the error bars represent the corresponding standard deviations. Most tree trunks in plots A, B, C, D, G, and I had wire mesh protection.

Results & discussion
ICP-MS analysis showed a significant 3-to 6-fold higher concentration of zinc in bryophytes inhabiting the bark of trees with wire mesh protection than in those without wire mesh protection (Fig. 9). Previous studies have shown that a considerable amount of zinc is leached from the zinc coating of galvanized iron by rain and dew (Harris, 1946;Seaward, 1974). Research has also shown that zinc is highly toxic to bryophytes (Tyler, 1990). Consequently, from the decreased diversity and increased zinc concentration of bryophytes on trees with wire mesh protection, it is reasonable to conclude that the loss of bryophyte cover and species richness has primarily occurred because of the toxicity of the zinc in the wire mesh. Additionally, other heavy metals in the wire mesh (e.g., iron) may affect bryophytes, with different heavy metals exerting varying levels of toxicity for bryophytes (Tyler, 1990).

Implications for biodiversity conservation
The results show that epiphytic bryophyte diversity is positively influenced by tree density and host tree DBH but negatively influenced by wire mesh protection, because of zinc toxicity (Fig. 10). The decline in bryophyte abundance and diversity on the lower parts of the tree trunks may be a cause for concern for biodiversity conservation on Mt. Odaigahara. This is because bryophytes contribute significantly to the species richness and biomass of tree trunks (Fritz, 2009;Lyons et al., 2000), as well as for ecosystem functions.
Furthermore, in addition to bryophytes, tree bark also provides important habitats for lichens and vascular epiphytes (Williams & Sillett, 2007). However, as heavy metals are toxic to these plants (Tyler et al., 1989), wire mesh protection may also contribute towards decreasing their levels of diversity and ecosystem functions. Unfortunately, considering that wire mesh protection is generally used against mammalian pests due to its direct effectiveness (Salmon et al. 2006;Vercauteren et al. 2006), this negative impact on bryophyte diversity may be widespread. The zinc concentration was significantly higher in bryophytes on trees with wire mesh than in those without wire mesh (p < 0.05; t-test).
Therefore, to establish best practices for biodiversity conservation that includes bryophytes, we should not only protect trees against bark stripping by deer but also focus on the materials used for protection. Alternative techniques for plant protection include the use of tree shelters in which trees are enclosed in plastic tubes (Ward et al., 2000) and forest enclosures using plastic mesh fencing (Vercauteren et al., 2006). However, these alternatives may also affect biodiversity conservation. For example, tree shelters decrease light transmission (Ward et al., 2000), which might alter the composition of bryophyte species on tree trunks. Further, Shibata et al. (2008) reported that forest enclosures sometimes hamper tree regeneration within the fenced areas because of serious seed predation by increased mouse populations.

Conclusion
The difficulties faced in minimizing the effect of plant protection methods on ecosystems that have complex community interactions are shown in this chapter. To establish best practices for biodiversity conservation, adaptive management should be adopted. Within such frameworks, we should examine and revise protective management practices on the basis of scientific data assimilated from regular monitoring of such ecosystems, while also preferentially using metal-free plant protection materials.