Orthopteran communities in the conifer-broadleaved woodland zone of the Russian Far East

We investigate orthopteran communities in the natural landscape of the Russian Far East and compare the habitat requirements of the species with those of the same or closely related species found in the largely agricultural landscape of central Europe. The study area is the 1,200 km Lazovsky State Nature Reserve (Primorsky region, southern Russian Far East) 200 km east of Vladivostok in the southern spurs of the Sikhote-Alin Mountains (134°E/43°N). The abundance of Orthoptera was recorded in August and September 2001 based on the number present in 20 randomly placed 1 m2 quadrates per site. For each plot (i) the number of species of Orthoptera, (ii) absolute species abundance and (iii) fifteen environmental parameters characterising habitat structure and microclimate were recorded. Canonical correspondence analysis (CCA) was used first to determine whether the Orthoptera occur in ecologically coherent groups, and second, to assess their association with habitat characteristics. In addition, the number of species and individuals in natural and semi-natural habitats were compared using a t test. A total of 899 individuals of 31 different species were captured, with numbers ranging between 2 and 13 species per plot. Species diversity was higher in semi-natural habitats than natural habitats. There was a similar but non-significant pattern in species density. Ordination analysis indicated four orthopteran communities, which were clearly separable along a moisture and vegetation density gradient. The natural sites in the woodland area of the Lazovsky Zapovednik are characterized by species-poor and low-density orthopteran assemblages compared to the semi-natural sites. But, the natural sites have a higher diversity of habitat specialists. Our findings corroborate the hypothesis that intermediate habitat disturbance levels support particularly species-rich animal communities at high densities. Under such regimes, orthopterans presumably mostly profit from the high diversity in plant species, which generates great structural and microclimatic heterogeneity. 673 * Present address: Helmholtz-Centre for Environmental Research Leipzig-Halle UFZ, Department of Community Ecology, Theodor-Lieser-Str. 4, D-06120 Halle, Germany. ments in relation to habitat structure and microclimate and (iii) compare orthopteran habitat preferences there and in Europe where the same or closely related species occur. MATERIAL AND METHODS


INTRODUCTION
While natural forests in Europe were to a great extent transformed by man into agricultural land and settlement, huge areas of the East Palaearctic are still forested (Newell, 2004;Yan & Shugart, 2005) and thus are important reference areas for the study of temperate woodland landscapes.The Far East is one of the three biodiversity hotspots in Russia (Venevsky & Venevskaia, 2005) and a centre of diversity and endemism of Orthoptera in Eurasia (Sergeev, 1998).
Their taxonomy and distribution are well studied, and the ease with which they can be sampled and their functional importance make Orthoptera suitable subjects for ecological and biogeographical studies (Sergeev, 1997;Lockwood & Sergeev, 2000).Habitat selection in Orthoptera is based on a complex combination of different and often interrelated environmental factors.Of these parameters, the microclimate at oviposition sites, which is often affected by vegetation structure, plays a crucial role (Uvarov, 1977;Willott & Hassall, 1998).Sergeev (1997) stressed the suitability of orthopteran communities for ecological and biogeographical investigations.In recent decades many such studies have been done in the northern hemisphere.Especially in North America, where different aspects of rangeland grasshopper communities have been studied in detail (e.g.Kemp et al., 1990;Kemp, 1992a, b;Fielding & Brusven, 1993a, b, 1995;Joern, 2004Joern, , 2005)).Most community studies in the Palaearctic are for central Europe and dry and semi-dry grassland habitats (e.g.Fartmann, 1997;Behrens & Fartmann, 2004).Information on the Asian part of the Palaearctic is restricted to biogeographic data (Stebaev et al., 1989;Sergeev, 1998) and detailed studies of orthopteran assemblages are lacking.Since woodlands are usually not considered to be an orthopteran habitat (Theuerkauf & Rouys, 2006) and old forests are rare in central Europe, little is known about habitat selection and community structure of Orthoptera in natural woodland areas in the Palaearctic.
We therefore investigated orthopteran communities in the natural landscape of the Russian Far East and compared the results with observations from the human and agriculturally dominated landscape of central Europe, because many taxa occur throughout the Palaearctic (Sergeev, 1992(Sergeev, , 1997)).Hence, orthopteran assemblages in the Lazovsky State Nature Reserve (Primorsky region, Russian Far East) were studied to (i) determine their species composition and abundance in different natural and seminatural habitats, (ii) analyse orthopteran habitat require-ments in relation to habitat structure and microclimate and (iii) compare orthopteran habitat preferences there and in Europe where the same or closely related species occur.

Study area
The study area is located in the Primorsky region (southern Russian Far East, Fig. 1).The landscape is formed by the Sikhote-Alin Mountains, stretching from the southwest to the northeast, parallel to the coastline (average altitude about 1,000 m a. s. l.).Woodland covers 80% of the Primorsky region -taiga in the north, conifer-broadleaved woodlands in the south (Newell, 2004).
The 1,200 km 2 sized Lazovsky State Nature Reserve (i.e.Lazovsky "Zapovednik", the official Russian category for this protected area) is situated about 200 km east of Vladivostok in the southern spurs of the Sikhote-Alin Mountains (134°E/43°N).It mainly consists of woodlands dominated by Mongolian oak (Quercus mongolica) with an admixture of Korean pine (Pinus koraiensis) and various other tree species.The species richness of the Zapovednik is impressive (1212 species of vascular plants, 57 mammals and 318 birds), with many rare and highly endangered species including the Amur tiger (Panthera tigris altaica) (Chochrjakow & Schochrin, 2002;Newell, 2004).Open habitats are very rare, and include natural coastal dunes and swamps, parts of the floodplains, screes and mountain peaks, anthropogenic meadows in woodland-clearings near the three ranger camps and a few set-aside fields at the reserve border.
A monsoon climate with warm, humid summers and cold, dry winters is characteristic of the study area.The average annual precipitation is 750-850 mm, decreasing from the coast inland.Due to a greater influence of continental climate the mean temperature inland is -20°C in January and 20°C in July-August.

Sampling of Orthoptera
Sampling was carried out on 18 plots representing all the typical orthopteran habitats of the Lazovsky Zapovednik, except floodplains, which were studied by Specht (2004).Nine natural (coastal dunes, semi-dry coastal grasslands, swamps) and nine semi-natural habitats (fallows dominated by Artemisia spp.and meadows) were investigated.The area of the plots was > 2,000-10,000 m 2 with a homogenous vegetation structure at every site.
Except for the small Nemobiinae species, which live hidden under stones or in litter on the ground, sampling provided reliable quantitative data.Most of the specimens were determined in the field and then released.Individuals that could not be identified in the field (Tetrix spp., some Chorthippus spp.) and voucher specimens of each species were collected and identified later.For determination the keys of Bey-Bienko & Mishchenko (1951a, b) and Storozhenko (1986) were used.Nomenclature is based on Storozhenko (1986) and for species that also occur in Europe on Heller et al. (1998).

Habitat structure
For each plot we measured/estimated fifteen environmental parameters: inclination, exposure, heights of one (minimum) up to three different vegetation layers (e.g.turf -tall grass -Artemisia) and % cover of the following habitat components: total vegetation, field layers, Cyperaceae, Poaceae, herbs, mosses, litter, bare soil, stones and hollows (in swamps).

Data analysis
Canonical correspondence analysis (CCA) (using CANOCO 4.51; ter Braak & Šmilauer, 2002), a direct gradient ordination technique, was used to determine the organization of orthopteran species into distinct communities and the relations between habitat structure and species composition (Fielding & Brusven, 1993b, 1995;Palmer, 1993;Szövényi, 2002;Torrusio et al., 2002).Environmental and species data were log arithmically transformed [y' = ln (y + 1)] to obtain approximately normal distributions and homogenous variances.Species of Orthoptera that occurred only on one plot and/or of which < 10 specimens in total were found were not included in the data set (Table 1).Inclination and exposure were not used as variables in the CCA because only two plots were slightly inclined (< 5°); cover (%) of bare soil and stones made up one variable; out of the three field layer heights the maximum vegetation height was used in CCA.The statistical validity of the ordination was tested using a Monte Carlo permutation test (null model: 9,999 unrestricted permutations).This was carried out for every environmental variable and all canonical axes (i.e. the complete model).Only significant variables were included stepwise in the model, and at each step only the variable that explained most of the remaining error variance (manual-forward selection of CANOCO) was chosen.Non-significant variables were those that explained little of the additional variance at the time they could be added to the model.They also may intercorrelate with other environmental variables (Storch et al., 2003); which was examined using Spearman's rank correlation analysis.T tests (using SPSS 11.5) were used to assess significant differences in orthopteran density and species number for the plots at the natural and semi-natural sites.Prior to the analyses, variables were tested for normal distribution using Kolmogorov-Smirnov test.

Orthopteran assemblages
Five of the ten environmental variables significantly contributed to the CCA ordination model.The model explained 60% of the variance in the number of species and 96% of the variance in the species-environment relation (total inertia: 3.08, sum of all canonical eigenvalues: 1.93; Monte Carlo test: F = 4.04, P 0.001).Each of the four canonical axes was highly correlated with one particular environmental variable, facilitating the ecological interpretation of each axis (see Table 2 for details).All non-significant variables intercorrelated with other environmental variables (Table 3) and explained < 10% additional variance at the stage when they would have been included in the model.
Based on the first two canonical axes four distinct communites can be separated (Fig. 2a): Cover of hollows strongly correlated with the first canonical axis.Note that there is no "real" environmental moisture gradient underlying this ordination pattern, because there are only two plots (swamps) with hollows.Along the first canonical axis species are divided into two distinct main groups: Taxa that were recorded only in swamps (natural habitat) and those only in mesic to dry habitats (semi-natural and natural habitats).The swamps have a species-poor orthopteran community.However, two species, Oxya maritima and Pteronemobius nitidus, were restricted to the wetlands.
Based on the arrangement of species along the second canonical axis it is possible to further discriminate between of the orthopteran communities in the remaining habitats.The axis is negatively correlated with herb cover.

Sum of specimens
Fidelity (no. of occupied plots) Abbreviation Species TABLE 1.The 14 most common orthopteran species on the 18 plots in order of their fidelity.Species that occurred only on one plot and/or with < 10 specimens in total are not included (see Data analysis).Distribution in Europe: species that also occur in Europe are indicated by an "+".Exlusiveness: species that are restricted to natural (n) and semi-natural (sn) habitat types are indicated.
community found in herb-rich fallows and meadows (semi-natural habitat).
Clearly separated from all other species and strongly negatively correlated with herb cover is Chorthippus maritimus.On the most extreme dune sites (natural habitat) with sparse vegetation and short swards Chorthippus maritimus is typical of the third species-poor community.Sporadically a few other species occur at low density.
Also, but to a lower extent Omocestus haemorrhoidalis and Ruspolia nitidula are negatively correlated with herb cover.Both species occur syntopically in coastal semi-dry grasslands (natural habitat) when they form part of another orthopteran community.
Further insights in the relations between species of Orthoptera and vegetation structure are indicated by the third and fourth axes (Fig. 2b).The folllowing species are positively correlated with the cover of field layers and Poaceae: Ruspolia nitidula, Phaneroptera falcata and Polionemobius taprobanensis.R. nitidula was most abundant in Poaceae-rich and dense coastal grasslands.P. falcata and P. taprobanensis were found in nearly all habitats, except the coastal dunes with extensive areas of bare ground and swamps.
the axis, is strongly negatively correlated with cover of field layers and Poaceae.C. hammarstroemi, D. fascipes and T. infernalis reach their highest densities in fallows with bare soil.The first two are restricted to these sites.

Species richness and abundance
According to Stebaev et al. (1989) many of the observed orthopteran species are associated with the southern border of the forest zone and limited to the Pacific Ocean districts of the Russian Far East with a monsoon climate.Numerous species are concentrated in the study area, in regions of conifer-broadleaved forest and forest-steppe in the south of the Russian Far East (Stebaev et al., 1989;Sergeev, 1992).
The natural sites in the woodland area of the Lazovsky Zapovednik are characterized by species-poor and lowdensity orthopteran assemblages compared to the seminatural sites.At the semi-natural sites the highest densities and species numbers were found on young fallows.The young set-aside fields studied were characterized by the highest structural heterogeneity.
What are the reasons for a high orthopteran diversity at the semi-natural sites?Several studies (Fartmann & Mattes, 1997;Kruess & Tscharntke, 2002;Gebeyehu & Samways, 2003) indicate that orthopteran species richness is highest at sites subject to intermediate disturbance (e.g.grazing) and high structural heterogeneity.Joern (2005) found that grasshopper species richness in grassland is positively correlated with plant species richness and heterogeneity of vegetation structure, and negatively with vegetation height and grass biomass.Sites subject to an intermediate level of disturbance (like young fallows) have more plant species (Grime 1973a, b) and therefore a greater structural heterogeneity.Thus, greater resources are available for the coexistence of more orthopteran species (Dennis et al., 1998).In our view oviposition sites and food resources are the most important, especially the presence of bare ground for oviposition can be a limiting factor.Many, or even the majority of orthopteran species lay their eggs in the ground and prefer bare or sparsely covered ground for oviposition (Richards & Waloff, 1954).In this study the number of orthopteran species exclusive to natural and semi-natural sites is similar (5 vs. 4 species).Including the orthopteran data of Specht (2004) for the floodplains of the Lazovsky Zapovednik and thus all the typical natural orthopteran habitats in the study area, five further species are restricted to natural habitats [Bryodemella tuberculata (Fabricius, 1775), Dianemobius csikii (Bolivar, I., 1901), Eirenephilus longipennis (Shiraki, 1910), Oedaleus infernalis Saussure, 1884 and Tetrix tenuicornis (Sahlberg, 1893)].Orthopteran assemblages of the natural sites of the Lazovsky Zapovednik are species-poor, but have a greater diversity of highly specialised species.
In accordance with the results of Joern (2005) there is a positive relationship between overall Orthopteran density and species richness, indicating that there are at least some parameters that promote both species diversity and density.High orthopteran densities are often the result of a trade-off between optimal microclimatic conditions on the one hand and sufficient food and low predation pressure on the other (Fielding & Brusven, 1992;Gottschalk, 1996;Fartmann & Mattes, 1997;Behrens & Fartmann, 2004).Of crucial importance for Orthoptera abundance are the microclimatic conditions during egg and larval development (Ingrisch, 1979(Ingrisch, , 1980)).In this period most species benefit from high temperatures (van Wingerden et al., 1991a).In the sparse vegetation of the coastal dunes microclimate may be favourable, but food shortage and easy access for avian predators seem to result in a low orthopteran density.An increase in vegetation density results in lower temperatures in the egg habitat, mostly located near the soil surface (van Wingerden et al., 1991a).Although, the dense stands of the old set-aside fields should provide enough food and enemy-free space, the microclimate seems to be unfavourable for many orthopteran species.Because the young open Artemisia fallows in the Lazovsky Zapovednik provide the best combination of requirements, the orthopteran abundance there is the highest.

Orthopteran assemblages
Presence and assemblage of orthopteran species are distinctly different among habitat types.This depends on orthopteran habitat preference, which is determined by species adaptation to habitat structure, microclimate and disturbance intensity (Joern, 1982;Fielding & Brusven, 1995;Samways, 1997;Szövényi, 2002).The distribution pattern of eurytopic and stenotopic species differ -due to their specific habitat requirements with the latter species characteristic of different habitat types in the Lazovsky Zapovednik.Where possible, the following compares the present results (Fig. 2, Table 1) with ecological observations from Europe.
The eurytopic and thermophilous species Phaneroptera falcata and Polionemobius taprobanensis are not restricted to one orthopteran community and occur in both natural and semi-natural habitats in the study area.However, their habitat requirements differ.The terricolous, flightless cricket P. tabrobanensis is found in high densities in short-turf vegetation.It is a thermophilous species distributed throughout the Indo-Malayan region and the southern Far East (Schmidt 1999).The phytophilous/arbusticolous and very mobile (capable of flying) bush-cricket P. falcata occurs on those habitats at low densities.It prefers tall herbaceous vegetation and has a Transpalaearctic distribution (Detzel, 1998a).
The following groups of species have a higher habitat specifity and form four distinct orthopteran communities: 1. Community of coastal dunes with bare ground The geophilous Chorthippus maritimus, a far eastern sibling species of C. biguttulus (Linnaeus, 1758) (Krivolutskaya, 1997), occurs in the study area only in natural and dynamic habitats with extensive areas of bare ground or stones, such as coastal dunes, floodplains or screes (own observation).In the floodplains of the Lazovsky Zapovednik it typically co-occurs with Bryodemella tuberculata and Eirenephilus longipennis (Specht, 2004).C. biguttulus, a Eurosiberian and common species in central Europe, shows different habitat requirements: it is phytophilous and eurytopic with a preference for semi-dry grassland (Fartmann, 1997;Behrens & Fartmann, 2004).

Community of semi-dry coastal grassland
Ruspolia nitidula, a Palaeotropic/Mediterranean species is thermophilous, phytophilous/graminicolous and especially the immature stages are hygrophilous (Detzel, 1998b).The restriction of this bush-cricket to coastal grassland with a moderate maritime climate in the Lazovsky Zapovednik seems to result from these adaptations.R. nitidula prefers vertical structures and in the study area it is most abundant (up to 7 individuals/10 m 2 ) in ecotones of dense, tall-grass stands, which possibly offer a suitable microclimate and structure.
The stenotopic Omocestus haemorrhoidalis, a Transpalaearctic species, has similar vegetation structure preferences in the Lazovsky Zapovednik and in central Europe.It is geo-/phytophilous, graminicolous, xero-and thermophilous.Microhabitat preferences in the Russian Far East correspond to those observed by Fartmann (1997) in northeast Germany: open, short-turf semi-dry grassland with structural heterogeneity and bare soil.

Community of herb-rich meadows and fallows
Mecostethus parapleurus and Oecanthus longicaudus are herbicolous and thermophilous species of the herbrich fallows and meadows in the study area.M. parapleurus is distributed along the southern boundary of the nemoral woodland zone from the Atlantic Ocean to Japan; in central Europe it is a thermophilous species, which prefers wet meadows (Detzel, 1998c).The habitat requirements of O. longicaudus seem to resemble those of the Mediterranean O. pellucens (Scopoli, 1763) in central Europe: According to Detzel (1998d) O. pellucens is a typical species of tall forb-rich plant communities.
Chorthippus hammarstroemi is an eastern Palaearctic, meso-xerophilous steppe species (Sergeev, 1997) that is abundant in anthropogenic habitats, such as fields (Sergeev, 1998).In the Lazovsky Zapovednik the cricket Dianemobius fascipes nigrofasciatus occurs syntopically with C. hammarstroemi exclusively in recently set-aside fields, but reaches its highest density in open, dry fallows with much bare soil/stones.Hence, this terricolous cricket is considered to be thermo-and xerophilous.On the Kuril Islands it colonises soils heated by volcanic activity, close to solfataras and fumaroles (Krivolutskaya, 1997).It is widely distributed in the temperate region of East Asia; on the Japanese Islands the southern range limit is about 30°N (Masaki, 1996).

CONCLUSIONS
The majority of Orthoptera in the boreal and temperate forest zones of the Palaearctic are associated with open habitats (Sergeev, 1992(Sergeev, , 1997)).These grassland habitats often are created and maintained by disturbance (Theuerkauf & Rouys, 2006), which is especially important at the microhabitat scale.At this level, disturbance creates the microclimatic conditions and microhabitats necessary for the survival of eggs and larval stages.This applies to Orthoptera and other thermophilous insects (e.g.Lepidoptera : Fartmann, 2004: Fartmann, , 2006)).As for the whole temperate zone of the Palaearctic (Birks, 2005;Mitchell, 2005), closed forests are the natural vegetation of most parts of the Lazovsky Zapovednik.In the absence of man, open habitats are restricted to sites with extreme environmental conditions.Only sites that are highly disturbed (dunes, floodplains), too dry (dunes, screes) or wet (swamps) and too cold (mountain tops) are free of trees, all the intermediate sites are covered with woodland.Due to the origin of the landscape one would expect that the hot spots of orthopteran diversity and density will be sites created by natural disturbance.But, this is not the case as semi-natural sites subject to anthropogenic disturbance (recently set-aside fields) have the highest species numbers and abundance.The orthopteran communities of the natural habitats are not as species-rich as the anthropogenic sites, but there are more stenotopic species restricted to those habitats.
Comparable observations exist for central Europe; but disturbance intensity as well as the spatial and temporal pattern in the cultural European landscape are very different from a natural landscape.Human activity in central Europe creates either extensive and continuously high disturbance (e.g.intensive cropland) or approximately stable conditions with low disturbance frequency (e.g.high forests with long cutting intervals), which both lead to habitat loss for insects of open habitats.Natural, dynamic disturbance patterns and low-intensity anthropogenic management have positive effects on habitats and populations of stenotopic insects (Sergeev, 1998;Di Giulio et al., 2001), but today such practices are rare in central Europe (Fartmann, 2006).Hence, our findings in the natural landscape of the Russian Far East confirm a strategy of nature conservation that has recently attracted interest in central Europe.That is, the conservation and restoration of natural landscape dynamic/disturbance (Finck et al., 1998) and support of anthropo-zoogenic management with low intensity (van Wingerden et al., 1991b;Wallis De Vries & Raemakers, 2001;Theuerkauf & Rouys, 2006) for a lasting conservation of insect biodiversity.
The strongest positive correlations (in order of arrangement) with herb cover are for Mecostethus parapleurus, Chorthippus hammarstroemi, Oecanthus longicaudus and Dianemobius fasciatus.They form a second