Elsevier

Applied Soil Ecology

Volume 45, Issue 3, July 2010, Pages 293-297
Applied Soil Ecology

Combining pitfall traps and soil samples to collect Collembola for site scale biodiversity assessments

https://doi.org/10.1016/j.apsoil.2010.05.005Get rights and content

Abstract

Collembola are rarely included in landscape-level biodiversity assessments, large-scale surveys and monitoring projects because huge numbers of specimens would accumulate even in moderately sized programmes. Budgets are always limited, so sampling methods and identification need to be optimized. As no single sampling method collects all collembolan species equally well, we tested the efficiency of a combination of six pitfall traps and five soil subsamples in 30 oil seed rape fields in Eastern Austria. Work effort in man hours for sampling, sorting and identification was quantified for each method and related to the species richness of the collected fauna. Total identification effort was four times higher for the soil subsamples than the pitfall traps, however, soil samples also yielded more species (53 and 34, respectively). Out of the 70 species collected in total, an average of thirteen species per site was found in the pitfall samples, seventeen in the soil subsamples and 25 when combining the two methods. Using more than six pitfall samples alone would not have collected considerably more species. For the soil subsamples, still more species can be expected with the processing of more than five subsamples, but this would also result in higher costs. When including Collembola in large scale biodiversity assessments, surveys or monitoring projects, we therefore recommend combining the two methods. In combination, the identification of the catch from only two pitfalls and two soil subsamples already collected more than the average number of species in five soil subsamples or six pitfalls, respectively. Thus, combining the methods yielded a more complete picture of the collembolan community of a site than either method alone.

Introduction

Landscape-level biodiversity assessments, large-scale surveys and monitoring projects involve the processing of large numbers of animal or plant specimens. While arthropods particularly are quickly collected in large numbers, the identification of most taxa is very time consuming and therefore costly. As budgets are always limited, sampling and identification need to be optimized, in order to collect a sufficient number of sample units with the lowest possible effort. Collembola have rarely been considered in large-scale projects (Black et al., 2003, Sousa et al., 2006, Vanbergen et al., 2007), as they are found in high abundance in most terrestrial habitats, and their identification is especially intricate and labour and cost intensive.

As Collembola live in the soil pores, on the soil surface, in the litter layer and the vegetation, no single sampling method collects all species appropriately. The eu- and hemiedaphic species are normally extracted from soil and litter samples using Berlese-Tullgren or Macfadyen extractors. Between one to ten soil samples are usually collected per site and processed individually (see for example Deharveng, 1996, Black et al., 2003, Palacios-Vargas et al., 2007, Cole et al., 2008, Salmon et al., 2008). Alternatively, Bruckner et al. (2000) recommended collecting a large number of soil samples in the field and identifying subsamples (aliquots) after pooling all extracted animals. This reduces the large variation in abundance and species richness found in most sample units which is due to their aggregated spatial distribution in soil (Hopkin, 1997; p. 163).

The epedaphic or surface active species are best collected in pitfall traps; suction sampling (Stewart and Wright, 1995) is less frequently applied. Up to ten pitfall traps are placed per site and exposed for a few days to a few months (see for example Durbešić et al., 2006, Fountain et al., 2007). Comparing results from pitfall catches from different places or periods is difficult, because trap number, size, material, and conservation fluid all influence the catch considerably (Adis, 1979), and sampling is not yet standardized.

Few ecological studies combine coring and trapping (Jakel and Roth, 1998, Fountain and Hopkin, 2004, Bitzer et al., 2005) to adequately represent all life forms. However, combining the methods with a lower total number of sampling units may yield a better estimate of community composition of sites than one method alone. To test this hypothesis, we sampled the Collembola assemblages from 30 agricultural fields and compared the (i) pitfall trap, (ii) soil sample and (iii) pooled species richness. Additionally, we quantified the man hours spent for collecting, extracting, sorting and identifying the catch of each method to identify the most efficient sampling strategy. The results presented here are especially relevant for large scale biodiversity assessments, where the collembolan assemblages of a great number of sites are to be characterized as efficiently as possible.

Section snippets

Site description

As part of a study of on the landscape ecology of plant parasites and their predators (Drapela et al., 2008, Zaller et al., 2008a, Zaller et al., 2008b, Zaller et al., 2009), we designated an agricultural study region of 240 km2 size approximately 40 km east of Vienna, Austria (central coordinates: 16°57′E, 48°04′N). The main soil type of the region is chernozem; the climate is pannonian (continental). Within this region, 30 winter oilseed rape fields were selected, embedded in differently

Results

We collected 35,981 Collembola in total, 8042 from soil subsamples and 27,939 from pitfall traps. The activity density in the pitfalls varied between 82 and 3048 individuals and the abundance in the soil subsamples between 55 and 1200 individuals per site (five subsamples pooled). 70 species were recorded in total; 34 in the traps and 53 in the subsamples (Table 1). The site frequency of most species varied greatly between the two methods. Eighteen species were collected combining the methods;

Discussion

In this study, most epedaphic species were confined to the pitfall traps and euedaphic species to the soil subsamples from the pooled soil samples. The epedaphic species occurring in the subsamples were mostly juveniles or adults, probably migrating into deeper soil layers during periods of unfavorable conditions at the surface (for example during drought; Hopkin, 1997; p. 168). Similarly, euedaphic species coincidentally appeared in pitfall traps. These were possibly attracted by the

Acknowledgements

We thank Alex Bandion and Bettina Ibera for their help in the field. This project was funded by a scholarship of the University of Natural Resources and Applied Life Sciences, Vienna and by the Austrian Science Fund (grant no. P16972).

References (43)

  • J. Adis

    Problems of interpretating arthropod sampling with pitfall traps

    Zool. Anz.

    (1979)
  • A.B. Babenko et al.

    Collembola of Russia and Adjacent Countries: Family Hypogastruridae

    (1994)
  • R.J. Bitzer et al.

    Biodiversity and community structure of epedaphic and euedaphic springtails (Collembola) in transgenic rootworm Bt corn

    Environ. Entomol.

    (2005)
  • Bretfeld, G., 1999. Synopses on Palaearctic Collembola. Vol. 2. Symphypleona. Abh. Ber. Naturkundemus. Görlitz 71, 318...
  • Colwell, R.K., 1994–2008. EstimateS, Version 8: Statistical Estimation of Species Richness and Shared Species from...
  • R.K. Colwell et al.

    Interpolating, extrapolating, and comparing incidence-based species accumulation curves

    Ecology

    (2004)
  • L. Deharveng

    Soil Collembola diversity, endemism, and reforestation: a case study in the Pyrenees (France)

    Conserv. Biol.

    (1996)
  • T. Drapela et al.

    Spider assemblages in winter oilseed rape affected by landscape and site factors

    Ecography

    (2008)
  • D. Druce et al.

    Sampling strategies for millipedes (Diplopoda), centipedes (Chilopoda) and scorpions (Scorpionida) in savanna habitats

    Afr. Zool.

    (2004)
  • P. Durbešić et al.

    Abundance and seasonal dynamics of arthropods in the meadow community Arrhenatheretum elatioris near Varazdin, Croatia

    Period. Biol.

    (2006)
  • A.M. Ellison et al.

    Rapid inventory of the ant assemblage in a temperate hardwood forest: species composition and assessment of sampling methods

    Environ. Entomol.

    (2007)
  • Cited by (43)

    • Importance of grassy and forest non-crop habitat islands for overwintering of ground-dwelling arthropods in agricultural landscapes: A multi-taxa approach

      2022, Biological Conservation
      Citation Excerpt :

      Pitfall traps reflect arthropod activity rather than abundance and species-specific trap efficiency is highly influenced by species traits (Knapp et al., 2020). Soil-dwelling arthropod records can be biased in emergence traps as some species are able to penetrate the trap during post-overwintering period (McCravy, 2018), whereas soil samples frequently suffer from a small sample size compared to other sampling techniques (Querner and Bruckner, 2010). The combination of different sampling techniques can provide a robust solution to this issue.

    • Responses of Collembola communities to mixtures of wheat varieties: A trait-based approach

      2021, Pedobiologia
      Citation Excerpt :

      The biomass of fungi and fungal-invaded area of wheat available to Collembola should increase in the more fungal-sensitive wheat varieties. Most of the species collected in our study are frequently observed in agricultural crops, e.g. Entomobrya multifasciata, Folsomides parvulus, Heteromurus nitidus, Isotomurus palustris, Lepidocyrtus cyaneus, L. lanuginosus, Megalothorax minimus, Neotullbergia ramiscuspis, Parisotoma notabilis, Proisotoma minuta, Pseudosinella alba, Sminthurinus aureus, Sminthurinus elegans (Axelsen and Kristensen, 2000; Frampton et al., 2001; Fratello et al., 1985; Ponge et al., 2003; Querner and Bruckner, 2010; Rebecchi et al., 2000; Sousa et al., 2004). The number of wheat varieties, the number of functional groups and functional group composition (c1, c2, c3 or c4) did not have any effects on collembolan species and trait assemblages when tested alone.

    View all citing articles on Scopus
    View full text