Research Article |
Corresponding author: Dávid Bogyó ( davidbogyo@yahoo.co.uk ) Academic editor: Ivan H. Tuf
© 2015 Dávid Bogyó, Tibor Magura, David D. Nagy, Bela Tothmeresz.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Bogyó D, Magura T, Nagy DD, Tóthmérész B (2015) Distribution of millipedes (Myriapoda, Diplopoda) along a forest interior – forest edge – grassland habitat complex. In: Tuf IH, Tajovský K (Eds) Proceedings of the 16th International Congress of Myriapodology, Olomouc, Czech Republic. ZooKeys 510: 181-195. https://doi.org/10.3897/zookeys.510.8657
|
We studied the distribution of millipedes in a forest interior-forest edge-grassland habitat complex in the Hajdúság Landscape Protection Area (NE Hungary). The habitat types were as follows: (1) lowland oak forest, (2) forest edge with increased ground vegetation and shrub cover, and (3) mesophilous grassland. We collected millipedes by litter and soil sifting. There were overall 30 sifted litter and soil samples: 3 habitat types × 2 replicates × 5 soil and litter samples per habitats. We collected 9 millipede species; the most abundant species was Glomeris tetrasticha, which was the most abundant species in the forest edge as well. The most abundant species in the forest interior was Kryphioiulus occultus, while the most abundant species in the grassland was Megaphyllum unilineatum. Our result showed that the number of millipede species was significantly lower in the grassland than in the forest or in the edge, however there were no significant difference in the number of species between the forest interior and the forest edge. We found significantly the highest number of millipede individuals in the forest edge. There were differences in the composition of the millipede assemblages of the three habitats. The results of the DCCA showed that forest edge and forest interior habitats were clearly separated from the grassland habitats. The forest edge habitat was characterized by high air temperature, high soil moisture, high soil pH, high soil enzyme activity, high shrub cover and low canopy cover. The IndVal and the DCCA methods revealed the following character species of the forest edge habitats: Glomeris tetrasticha and Leptoiulus cibdellus. Changes in millipede abundance and composition were highly correlated with the vegetation structure.
Edge effect, soil arthropod, biodiversity, forest ecosystem
Millipedes (Myriapoda, Diplopoda) are detrivores, feeding mainly on decaying plant material and they found usually under leaf litter (
Forest edges have distinctive structure with highly variable environmental conditions.
The edge effect on different arthropod taxa was frequently studied during the last decades (Coleoptera: Magura 2000,
The aim of our study was to test the edge effect hypothesis on millipedes, that predicts higher species diversity and population density in forest edges than in the adjacent habitats (
We tested the edge effect hypothesis on millipedes in forest interior – forest edge – grassland habitat complex in Northeast Hungary. The study area is located in the Hajdúság Landscape Protection Area, about 3.5 km north-east from the centre of Vámospércs, 130–135 m above sea level (47°33'09"N, 21°56'18"E). The Hajdúság Landscape Protection Area is a mosaic of grasslands, forests and wetlands. Average annual temperature of this region is 9.6–9.8 °C, while average annual rainfall is 550–575 mm. Brown forest soils and sandy soils are the main soil types of the study area.
The studied habitat types were as follows: (1) a closed forest dominated by English oak (Quercus robur) and associated with narrow-leafed ash (Fraxinus angustifolia ssp. pannonica), field elm (Ulmus minor) and common alder (Alnus glutinosa) with shrubs and herbs and high percentage of canopy cover, (2) a forest edge with increased ground vegetation and dense shrub cover (dominated by Crataegus monogyna, Prunus spinosa and Rhamnus catharticus) as well as some invasive tree species, like Robinia pseudoacacia, and (3) a mesophilous grassland with dense herbaceous vegetation, dominated by Festuca pratensis, Poa pratensis, Deschampsia caespitosa and Carex species, together with Orchis laxiflora ssp. elegans and Dactylorrhiza incarnata.
This area was a forested area during the last decades. The mesophilous grassland is utilized for grazing (sheep and cattle). The age of trees in the forest interior and forest edge was 45–50 years, while the forest was unmanaged during the last 30 years. The density of trees was 250–300 trees/ha. The width of the forest edge was 6-14m. Non-native species occurred sparsely with single specimens in the forest edge.
We collected millipedes monthly (7 times from April to October in 2009; the year of 2009 was free from extreme weather conditions) during the vegetation period. Sampling of millipedes followed
We selected 13 environmental variables to test the influence on millipede assemblages. We measured soil pH value, soil dehydrogenase activity, soil moisture, soil temperature (in a depth of 2 cm), air temperature and relative humidity on the surface in the study plots. For pH measurement soil solution was prepared from 6.0 g wet soil. Soil samples were put into plastic beakers and after it filled with 50 ml deionized water. The pH was measured with a digital measurement type Testo 206 (Testo AG, Germany). Soil dehydrogenase enzyme activity (indicator of microbiological activity through the oxidative metabolisms in soil) was determined using triphenyltetrazolium chloride method (
Grassland | Edge | Forest | |
---|---|---|---|
Air temperature (°C) | 26.60 | 24.20 | 21.61 |
Canopy cover (%) | 0.00 | 50.30 | 69.80 |
Cover of decaying wood (%) | 0.00 | 23.50 | 23.30 |
Cover of herbs (%) | 93.80 | 25.30 | 23.50 |
Cover of leaf litter (%) | 0.00 | 83.40 | 92.60 |
Dehydrogenase enzyme activity | 0.37 | 0.31 | 0.29 |
Depth of leaf litter (cm) | 0.00 | 2.47 | 2.88 |
Humidity (%) | 69.88 | 66.08 | 71.30 |
Number of woody plant species | 0.00 | 5.40 | 4.90 |
pH | 8.54 | 7.83 | 7.40 |
Shrub cover (%) | 0.00 | 66.10 | 34.80 |
Soil moisture (%) | 33.28 | 28.17 | 16.97 |
Soil temperature (°C) | 22.30 | 17.35 | 16.77 |
Mixed Generalized Linear Model (GLMM) was used to test differences in the millipede abundance, species richness and Shannon diversity among the three habitat types (forest interior, forest edge, grassland). Factorial design was used; habitats and spatial replicates were regarded as factors. The response variables (millipede abundance, species richness, and Shannon diversity) were defined as a quasi-Poisson distribution with log link function (
The composition of the millipede assemblages along the forest interior – forest edge – grassland habitats was compared by hierarchical cluster analysis based on the abundance of millipedes using the Hellinger distance and the Ward fusion algorithm (
There were 999 specimens of millipedes (Diplopoda) identified to species level. In total 9 species of 4 families (Glomeridae, Julidae, Mastigophorophyllidae, Polydesmidae) were recorded from the study area (Table
List of millipede species with their abundances recorded in the studied habitats.
Grassland | Edge | Forest | |
---|---|---|---|
Brachyiulus bagnalli (Broelemann, 1924) | 3 | 5 | |
Brachydesmus superus Latzel, 1884 | 1 | 10 | |
Glomeris tetrasticha Brandt, 1833 | 1 | 382 | 33 |
Julus terrestris Linnaeus, 1758 | 6 | 8 | 8 |
Kryphioiulus occultus (C.L. Koch, 1847) | 7 | 121 | 141 |
Leptoiulus cibdellus (Chamberlin, 1921) | 1 | 34 | 2 |
Mastigona bosniensis (Verhoeff, 1897) | 104 | 38 | |
Megaphyllum unilineatum (C.L. Koch, 1838) | 16 | 21 | 12 |
Polydesmus complanatus (Linnaeus, 1761) | 22 | 23 | |
Total | 31 | 696 | 272 |
The forest edge and forest interior habitats were more species rich (9 species) than the grassland. 5 species of millipedes was found in the grassland habitat. Glomeris tetrasticha, Julus terrestris, Kryphioiulus occultus, Leptoiulus cibdellus and Megaphyllum unilineatum were recorded in all studied habitats. The following millipede species were the most frequent: Glomeris tetrasticha (416 individuals), Kryphioiulus occultus (269 individuals) and Mastigona bosniensis (142 individuals), while Brachyiulus bagnalli (8 individuals) and Brachydesmus superus (11 individuals) had the lowest total abundance in the studied habitats. The most abundant millipede species, Glomeris tetrasticha represented 41.6% of the total millipede catch. The same species was also the most frequent millipede (54.9% of the millipede individuals) in the edge habitat. The highest total abundance of millipedes was found in the edge habitat (696 individuals), while the lowest number of millipede individuals (31) was found in the grassland habitat.
The number of millipede individuals was significantly higher in the forest edge than in the grassland and forest interior. Furthermore, the number of millipede individuals was significantly higher in the forest interior than in the grassland (Chi2=179.275; df=2; p<0.0001; Figure
Millipede abundance, species richness and Shannon diversity at the studied habitats. Mean values (±SD) of the overall millipede abundance (A), species richness (B) and Shannon diversity (C) per samples at the studied habitats. Different letters indicate significant differences by Tukey test.
The millipede assemblages of the studied habitats formed two separated groups by hierarchical cluster analysis (Figure
We identified two groups of quantitative character species by the IndVal method for the studied habitats (Table
DCCA analysis for the millipede species of the study area. Squares represent the sampled habitats (blue squares: samples from the forest interior habitat; red squares: samples from the forest edge habitat; black squares: samples from the grassland habitat).The arrows denote the increase of the value of the environmental variables (airtemp: air temperature on the surface; canopy: canopy cover; DH: soil dehydrogenase enzyme activity; dwood: cover of decaying wood material; herbs: cover of herbs; littcover: cover of leaf litter; littdepth: depth of leaf litter; humidity: relative humidity on the surface; pH: soil pH; shrubs: cover of shrubs; stemp: soil temperature at 2cm depth; smoisture: soil moisture; woodysp: number of woody plant species). Green circles and the four-letter abbreviations indicate the millipede species (BRBA: B. bagnalli; BRSU: B. superus; GLTE: G. tetrasticha; JUTE: J. terrestris; KROC: K. occultus; LECI: L. cibdellus; MABO: M. bosniensis; MEUN: M. unilineatum; POCO: P. complanatus).
Habitat preference and quantitative character values of the millipede species presented with more than 30 individuals altogether. The IndVal column shows the species character value for the corresponding cluster level. Notations: * - p<0.05. A: the number of specimens present, B: the number of samples where the species is present in the sample group.
Species | IndVal | p | Grassland | Edge | Forest | |||
---|---|---|---|---|---|---|---|---|
A | B | A | B | A | B | |||
Forest edge | ||||||||
Leptoiulus cibdellus | 82.7 | * | 1 | 1 | 34 | 9 | 2 | 2 |
Forest edge and forest interior | ||||||||
Glomeris tetrasticha | 99.5 | * | 1 | 1 | 382 | 10 | 33 | 10 |
Mastigona bosniensis | 95 | * | 0 | 0 | 104 | 10 | 38 | 9 |
Kryphioiulus occultus | 90.2 | * | 7 | 7 | 121 | 9 | 141 | 10 |
Polydesmus complanatus | 85 | * | 0 | 0 | 22 | 8 | 23 | 9 |
We found a relatively low number of millipede species in the complex of three habitat types in a protected area of Hungary, which represents approximately 9% of the Hungarian millipede fauna (
The aim of the study was to reveal a relationship between millipedes (abundance, species richness, diversity and assemblage composition) and the vegetation structure along a forest interior-forest edge-grassland gradient. We found that the total millipede abundance was highest in the forest edge which supports the classical edge effect hypothesis (
Our study showed no significant edge effect on the species richness and diversity of millipedes. Millipede species richness and diversity was higher in the forest interior and forest edge than in the grassland, but there was no difference between the two forested habitats (forest interior, and forest edge). A positive edge effect on species richness and/or diversity in forest edges was reported in previous studies on arthropod taxa (
The millipede assemblages of the studied habitats clearly separated from each other, based on the abundance of the millipede species. The millipede assemblages formed two main clusters: the first included the grassland habitat, the second included the two forested habitats. However, forested habitats (forest edge and forest interior) also showed a clear separation from each other. It is known, that saprophagous macroarthropod assemblages are changing with the change of vegetation structure on a landscape scale, or on a smaller scale (
Using the IndVal method we found significant character species for the studied habitats. The grassland habitat had no significant character species. The forest edge was characterized by L. cibdellus, a species preferring humid, woodland habitats with a Northern and Central-East European distribution (
The results of the DCCA showed that forest edge and forest interior habitats are clearly separated from the grassland habitats. Forest edge and forest interior offer more suitable habitat for millipedes with high amount of leaf litter (both cover and depth of leaf litter), dead wood, canopy cover, as well as with more humid microclimate. The studied forest edge habitat was characterized by high air temperature, high soil moisture, high soil pH and low canopy cover (higher openness) which are key factors affecting millipede assemblages in forested habitats (
The results of the study revealed that millipede assemblages altered in a short distance along the grassland-forest edge-forest habitat complex. Our study showed a positive edge effect on millipede abundance and assemblage composition in the studied natural habitats. We found significant edge-associated millipede species by IndVal method. On the other hand there was no edge effect on millipede species richness and diversity. Our results support the former findings that some millipede species may be specialized to natural forest edge habitats. Forest edges may have key habitats in the conservation of millipedes during the next decades.
The authors are thankful for the assistance of Roland Horváth, Levente Lality, Szabolcs Mizser, Edina Simon and Gyula Szabó. We thank the help of the Hortobágy National Park Directorate, especially the help of László Demeter and László Szél. This work was partially supported by the European Union and the European Social Fund through project (grant no.: TAMOP-4.2.2.C-11/1/KONV-2012-0010). Furthermore, the work is supported by the TAMOP 4.2.1/B-09/1/KONV-2010-0007, and TAMOP-4.2.2/B-10/1-2010-0024 projects.The research of Dávid Bogyó was supported by the European Union and the State of Hungary, co-financed by the European Social Fund in the framework of TÁMOP 4.2.4.A/2-11-1-2012-0001 ‘National Excellence Program’.