Oecologia DOI 10.1007/s00442-015-3342-2
COMMUNITY ECOLOGY - ORIGINAL RESEARCH
Concordance and discordance between taxonomic and functional homogenization: responses of soil mite assemblages to forest conversion Akira S. Mori1 · Aino T. Ota1 · Saori Fujii1 · Tatsuyuki Seino2 · Daisuke Kabeya3 · Toru Okamoto3 · Masamichi T. Ito1,4 · Nobuhiro Kaneko1 · Motohiro Hasegawa3
Received: 10 November 2014 / Accepted: 5 May 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract The compositional characteristics of ecological assemblages are often simplified; this process is termed “biotic homogenization.” This process of biological reorganization occurs not only taxonomically but also functionally. Testing both aspects of homogenization is essential if ecosystem functioning supported by a diverse mosaic of functional traits in the landscape is concerned. Here, we aimed to infer the underlying processes of taxonomic/functional homogenization at the local scale, which is a scale that is meaningful for this research question. We recorded species of litter-dwelling oribatid mites along a gradient of forest conversion from a natural forest to a monoculture larch plantation in Japan (in total 11 stands), and collected data on the functional traits of the recorded species to quantify functional diversity. We calculated the taxonomic and functional β-diversity, an index of biotic homogenization. We found that both the taxonomic and functional Communicated by Liliane Ruess. Electronic supplementary material The online version of this article (doi:10.1007/s00442-015-3342-2) contains supplementary material, which is available to authorized users. * Akira S. Mori [email protected]; [email protected]‑net.ne.jp 1
Graduate School of Environment and Information Sciences, Yokohama National University, 79‑7 Tokiwadai, Hodogaya, Yokohama 240‑8501, Japan
2
Faculty of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305‑8577, Japan
3
Forestry and Forest Products Research Institute, 1 Matsunosato, Tsukuba, Ibaraki 305‑8687, Japan
4
Present Address: Faculty of Economics and Management, Surugadai University, Azu 698, Hanno, Saitama 357‑8555, Japan
β-diversity decreased with larch dominance (stand homogenization). After further deconstructing β-diversity into the components of turnover and nestedness, which reflect different processes of community organization, a significant decrease in the response to larch dominance was observed only for the functional turnover. As a result, there was a steeper decline in the functional β-diversity than the taxonomic β-diversity. This discordance between the taxonomic and functional response suggests that species replacement occurs between species that are functionally redundant under environmental homogenization, ultimately leading to the stronger homogenization of functional diversity. The insights gained from community organization of oribatid mites suggest that the functional characteristics of local assemblages, which support the functionality of ecosystems, are of more concern in human-dominated forest landscapes. Keywords Community reassembly · Ecosystem functioning · Forest restoration · Oribatid mites · Soil arthropods · Taxonomic and functional homogenization
Introduction Ecological communities are experiencing changes in the current Anthropocene era. The emergence of novel communities is expected as a result of environmental changes such as biological invasion, climate change, and land-use intensification (Hobbs et al. 2009). The responses of individual species to these changes are not identical, thus resulting in the de-assembling, shuffling and reassembling of species at various spatial and temporal scales (Zavaleta et al. 2009). Although community reorganization can be variable, intensified stress to biota often results in the predominance of
13
Oecologia
common species (habitat generalists) across locations (Clavel et al. 2011). As a result, biological assemblages can be simplified in terms of their compositional characteristics; this process is referred to as “biotic homogenization” (McKinney and Lockwood 1999; Olden et al. 2004; Olden and Rooney 2006). The homogenization of biological communities has been inferred from the reductions in between-site dissimilarity in terms of species composition (β-diversity) along a focal temporal and spatial gradient (e.g., Karp et al. 2012). However, assessing the processes of biotic homogenization may not be straightforward, because β-diversity is a complex measure that reflects the different factors that generate compositional dissimilarity between locations. To face this inherent difficulty for β-diversity evaluation, there is an emergent methodological development (Anderson et al. 2010). This is particularly the case for the modern method of β-diversity partitioning (Baselga 2010, 2012); β-diversity can be decomposed into “turnover” and “nestedness” components. Because the two components emerge as a result of different processes of community organization (e.g., Baselga 2010; Gutiérrez-Cánovas et al. 2013), this method can potentially provide a comprehensive picture of the underlying processes of biotic homogenization. For example, two cases of the changes in community compositional characteristics, in which one has significant changes in the number of species and no species replacement and the other the opposite pattern, can generate similar values of the overall β-diversity (Baselga 2010). While the overall β-diversity changes in the former case primarily result from species loss or gain in each local community, those in the latter case reflect the simultaneous gain and loss of species in each locality leading to the replacement of one species by another (Leprieur et al. 2011). The resultant nestedness component is also referred to as “richness difference,” which represents how differences in species richness that are not due to species turnover contribute to patterns of the overall β-diversity (Legendre 2014). Notably, several recent studies have shown that biotic homogenization sometimes occurs despite no net change in local species richness (α-diversity) (Bühler and Roth 2011; McCune and Vellend 2013; Rodrigues et al. 2013), which is in contrast to the well-documented patterns of biotic homogenization associated with a net loss of biodiversity resulting from the response of sensitive specialist species to habitat modifications (Clavel et al. 2011). Considering such variable effects of local species richness on the spatial variations of communities among localities, separating the effects of α-diversity from the overall β-diversity will be an important step to further help understand the causes and consequences of biotic homogenization. Another methodological advancement is that the method of β-diversity partitioning is now also applicable
13
to functional β-diversity as well as taxonomic β-diversity (Villéger et al. 2013). It is increasingly recognized that the functional characteristics of species (functional traits) have a more direct mechanistic linkage with the functionality of ecosystems than does the taxonomic identity of species (Cadotte et al. 2011; Mori et al. 2013a). That is, functional diversity may be more important than species diversity if the primary concern is the conservation of the vital functionality of ecosystems (e.g., Gagic et al. 2015; Valencia et al. 2015). In this regard, evaluating the functional aspects of biotic homogenization is of importance, as taxonomic and functional homogenization may be linked but are not synonymous (Pool and Olden 2012; Tobias and Monika 2011). Taxonomic homogenization may thus only be the tip of the iceberg, in which the more serious effects of biotic homogenization can be inferred by evaluating functional homogenization. Here, with the above issues in mind, we quantified and partitioned the taxonomic and functional β-diversity for the soil mesofauna (oribatid mites) along a gradient of primary forests to monoculture plantations in central Japan. In this landscape where forests have been managed for timber production, there is a possible homogenization of litterdwelling oribatid communities (Mori et al. 2015). This is because the heterogeneity of leaf litterfall, which is one of the most important factors for determining the abundance and distribution of litter microarthropods (Hansen 2000b; Kaneko and Salamanca 1999; Mori et al. 2013b), was substantially reduced in monoculture plantations (Mori et al. 2015). Considering the importance of soil microarthropods as an essential group of organisms to sustain ecosystem functions such as decomposition and nutrient cycling (Bardgett and Wardle 2010; Wall et al. 2012), there is an urgent need to further advance our understanding of biotic homogenization. To achieve this goal, we aimed to evaluate multiple facets of biotic reorganization in oribatid communities so as to provide information helpful for the safeguard of biodiversity at multiple biological and spatial scales.
Materials and methods Study site The study area is located in the Yatsugatake Mountains, central Japan (approximately 15 × 20 km in area, 36.50– 56°N, 140.34–35°E) with an elevation range of approximately 1200–1400 m a.s.l. The mean annual temperature and precipitation are approximately 7.1 °C and 1454 mm, respectively. The land use of this region is characterized by plantations of larch (Larix kaempferi) trees. The age of the larch stands was approximately 40–50 years at the time of the sampling events.
Oecologia
The present study is based on the oribatid community inventory of Mori et al. (2015). In the following sections, we provide a brief explanation of the study design. We established 11 study plots with a size of 30 × 30 m (in total 0.99 ha) within the study region. In each plot, we recorded the species and size (diameter at breast height; DBH) for all trees larger than 5 cm in DBH. Thinning had been conducted several times in the 40 years after planting; however, all the trees except larch trees had been completely removed in several plots. In the other plots, trees of other species (mostly hardwood species such as Betula platyphylla and Quercus mongolica) were present and were left in the stands during the thinning procedure. In addition, one plot was established in a natural forest that had not been converted to conifer plantation. Although this forest may not have been completely free from human interference in the past, it has been primarily maintained under natural regeneration processes for a certain time (at least since the beginning of the last century). As a result, the percentage of larch based on the basal area of each plot (relative basal area; rBA) ranged from 0.0 (natural forest) to 100.0 % (pure larch plantation). The total basal area of each plot ranged from 22.9 to 41.2 m2 ha−1; also see Hasegawa et al. (2013) for study plot details. In our study plots, there were no plot-specific differences in any of the measured soil properties (water content, pH, electric conductivity, and C and N concentration) (Hasegawa et al. 2013); the only differences were the amount and composition of plant litter on the forest floor, thus reflecting the dominance of tree species in each plot. Our preliminary test showed that the rBA values of larch trees reflected the amount of larch litterfall (Mori et al. 2015). Considering the importance of the composition and amount of litter for soil microarthropods (Takeda 1987; Zaitsev et al. 2014), the rBA values of larch trees for each plot were primarily used as the index representing the degree of environmental homogenization (Mori et al. 2015). The rBA values were significantly positively correlated with the values of basal area and stem density of larch trees in each stand (Pearson’s correlation; r = 0.90, P
COMMUNITY ECOLOGY - ORIGINAL RESEARCH
Concordance and discordance between taxonomic and functional homogenization: responses of soil mite assemblages to forest conversion Akira S. Mori1 · Aino T. Ota1 · Saori Fujii1 · Tatsuyuki Seino2 · Daisuke Kabeya3 · Toru Okamoto3 · Masamichi T. Ito1,4 · Nobuhiro Kaneko1 · Motohiro Hasegawa3
Received: 10 November 2014 / Accepted: 5 May 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract The compositional characteristics of ecological assemblages are often simplified; this process is termed “biotic homogenization.” This process of biological reorganization occurs not only taxonomically but also functionally. Testing both aspects of homogenization is essential if ecosystem functioning supported by a diverse mosaic of functional traits in the landscape is concerned. Here, we aimed to infer the underlying processes of taxonomic/functional homogenization at the local scale, which is a scale that is meaningful for this research question. We recorded species of litter-dwelling oribatid mites along a gradient of forest conversion from a natural forest to a monoculture larch plantation in Japan (in total 11 stands), and collected data on the functional traits of the recorded species to quantify functional diversity. We calculated the taxonomic and functional β-diversity, an index of biotic homogenization. We found that both the taxonomic and functional Communicated by Liliane Ruess. Electronic supplementary material The online version of this article (doi:10.1007/s00442-015-3342-2) contains supplementary material, which is available to authorized users. * Akira S. Mori [email protected]; [email protected]‑net.ne.jp 1
Graduate School of Environment and Information Sciences, Yokohama National University, 79‑7 Tokiwadai, Hodogaya, Yokohama 240‑8501, Japan
2
Faculty of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305‑8577, Japan
3
Forestry and Forest Products Research Institute, 1 Matsunosato, Tsukuba, Ibaraki 305‑8687, Japan
4
Present Address: Faculty of Economics and Management, Surugadai University, Azu 698, Hanno, Saitama 357‑8555, Japan
β-diversity decreased with larch dominance (stand homogenization). After further deconstructing β-diversity into the components of turnover and nestedness, which reflect different processes of community organization, a significant decrease in the response to larch dominance was observed only for the functional turnover. As a result, there was a steeper decline in the functional β-diversity than the taxonomic β-diversity. This discordance between the taxonomic and functional response suggests that species replacement occurs between species that are functionally redundant under environmental homogenization, ultimately leading to the stronger homogenization of functional diversity. The insights gained from community organization of oribatid mites suggest that the functional characteristics of local assemblages, which support the functionality of ecosystems, are of more concern in human-dominated forest landscapes. Keywords Community reassembly · Ecosystem functioning · Forest restoration · Oribatid mites · Soil arthropods · Taxonomic and functional homogenization
Introduction Ecological communities are experiencing changes in the current Anthropocene era. The emergence of novel communities is expected as a result of environmental changes such as biological invasion, climate change, and land-use intensification (Hobbs et al. 2009). The responses of individual species to these changes are not identical, thus resulting in the de-assembling, shuffling and reassembling of species at various spatial and temporal scales (Zavaleta et al. 2009). Although community reorganization can be variable, intensified stress to biota often results in the predominance of
13
Oecologia
common species (habitat generalists) across locations (Clavel et al. 2011). As a result, biological assemblages can be simplified in terms of their compositional characteristics; this process is referred to as “biotic homogenization” (McKinney and Lockwood 1999; Olden et al. 2004; Olden and Rooney 2006). The homogenization of biological communities has been inferred from the reductions in between-site dissimilarity in terms of species composition (β-diversity) along a focal temporal and spatial gradient (e.g., Karp et al. 2012). However, assessing the processes of biotic homogenization may not be straightforward, because β-diversity is a complex measure that reflects the different factors that generate compositional dissimilarity between locations. To face this inherent difficulty for β-diversity evaluation, there is an emergent methodological development (Anderson et al. 2010). This is particularly the case for the modern method of β-diversity partitioning (Baselga 2010, 2012); β-diversity can be decomposed into “turnover” and “nestedness” components. Because the two components emerge as a result of different processes of community organization (e.g., Baselga 2010; Gutiérrez-Cánovas et al. 2013), this method can potentially provide a comprehensive picture of the underlying processes of biotic homogenization. For example, two cases of the changes in community compositional characteristics, in which one has significant changes in the number of species and no species replacement and the other the opposite pattern, can generate similar values of the overall β-diversity (Baselga 2010). While the overall β-diversity changes in the former case primarily result from species loss or gain in each local community, those in the latter case reflect the simultaneous gain and loss of species in each locality leading to the replacement of one species by another (Leprieur et al. 2011). The resultant nestedness component is also referred to as “richness difference,” which represents how differences in species richness that are not due to species turnover contribute to patterns of the overall β-diversity (Legendre 2014). Notably, several recent studies have shown that biotic homogenization sometimes occurs despite no net change in local species richness (α-diversity) (Bühler and Roth 2011; McCune and Vellend 2013; Rodrigues et al. 2013), which is in contrast to the well-documented patterns of biotic homogenization associated with a net loss of biodiversity resulting from the response of sensitive specialist species to habitat modifications (Clavel et al. 2011). Considering such variable effects of local species richness on the spatial variations of communities among localities, separating the effects of α-diversity from the overall β-diversity will be an important step to further help understand the causes and consequences of biotic homogenization. Another methodological advancement is that the method of β-diversity partitioning is now also applicable
13
to functional β-diversity as well as taxonomic β-diversity (Villéger et al. 2013). It is increasingly recognized that the functional characteristics of species (functional traits) have a more direct mechanistic linkage with the functionality of ecosystems than does the taxonomic identity of species (Cadotte et al. 2011; Mori et al. 2013a). That is, functional diversity may be more important than species diversity if the primary concern is the conservation of the vital functionality of ecosystems (e.g., Gagic et al. 2015; Valencia et al. 2015). In this regard, evaluating the functional aspects of biotic homogenization is of importance, as taxonomic and functional homogenization may be linked but are not synonymous (Pool and Olden 2012; Tobias and Monika 2011). Taxonomic homogenization may thus only be the tip of the iceberg, in which the more serious effects of biotic homogenization can be inferred by evaluating functional homogenization. Here, with the above issues in mind, we quantified and partitioned the taxonomic and functional β-diversity for the soil mesofauna (oribatid mites) along a gradient of primary forests to monoculture plantations in central Japan. In this landscape where forests have been managed for timber production, there is a possible homogenization of litterdwelling oribatid communities (Mori et al. 2015). This is because the heterogeneity of leaf litterfall, which is one of the most important factors for determining the abundance and distribution of litter microarthropods (Hansen 2000b; Kaneko and Salamanca 1999; Mori et al. 2013b), was substantially reduced in monoculture plantations (Mori et al. 2015). Considering the importance of soil microarthropods as an essential group of organisms to sustain ecosystem functions such as decomposition and nutrient cycling (Bardgett and Wardle 2010; Wall et al. 2012), there is an urgent need to further advance our understanding of biotic homogenization. To achieve this goal, we aimed to evaluate multiple facets of biotic reorganization in oribatid communities so as to provide information helpful for the safeguard of biodiversity at multiple biological and spatial scales.
Materials and methods Study site The study area is located in the Yatsugatake Mountains, central Japan (approximately 15 × 20 km in area, 36.50– 56°N, 140.34–35°E) with an elevation range of approximately 1200–1400 m a.s.l. The mean annual temperature and precipitation are approximately 7.1 °C and 1454 mm, respectively. The land use of this region is characterized by plantations of larch (Larix kaempferi) trees. The age of the larch stands was approximately 40–50 years at the time of the sampling events.
Oecologia
The present study is based on the oribatid community inventory of Mori et al. (2015). In the following sections, we provide a brief explanation of the study design. We established 11 study plots with a size of 30 × 30 m (in total 0.99 ha) within the study region. In each plot, we recorded the species and size (diameter at breast height; DBH) for all trees larger than 5 cm in DBH. Thinning had been conducted several times in the 40 years after planting; however, all the trees except larch trees had been completely removed in several plots. In the other plots, trees of other species (mostly hardwood species such as Betula platyphylla and Quercus mongolica) were present and were left in the stands during the thinning procedure. In addition, one plot was established in a natural forest that had not been converted to conifer plantation. Although this forest may not have been completely free from human interference in the past, it has been primarily maintained under natural regeneration processes for a certain time (at least since the beginning of the last century). As a result, the percentage of larch based on the basal area of each plot (relative basal area; rBA) ranged from 0.0 (natural forest) to 100.0 % (pure larch plantation). The total basal area of each plot ranged from 22.9 to 41.2 m2 ha−1; also see Hasegawa et al. (2013) for study plot details. In our study plots, there were no plot-specific differences in any of the measured soil properties (water content, pH, electric conductivity, and C and N concentration) (Hasegawa et al. 2013); the only differences were the amount and composition of plant litter on the forest floor, thus reflecting the dominance of tree species in each plot. Our preliminary test showed that the rBA values of larch trees reflected the amount of larch litterfall (Mori et al. 2015). Considering the importance of the composition and amount of litter for soil microarthropods (Takeda 1987; Zaitsev et al. 2014), the rBA values of larch trees for each plot were primarily used as the index representing the degree of environmental homogenization (Mori et al. 2015). The rBA values were significantly positively correlated with the values of basal area and stem density of larch trees in each stand (Pearson’s correlation; r = 0.90, P
