From two to three dimensions : improvement of the traditional beating sheet method

P. Paquin (1), N. Dupérré (2)
1. Scienceinfuse Inc., 12 Saxby Sud, Shefford, QC, J2M 1S2, Canada. Courriel :
2. Collection manager, Arachnology Department, Zoological Museum. Martin-Luther-King Platz 3, D-20146, Hamburg, Germany. Courriel :

Abstract. Beating sheets are widely used to sample invertebrates associated with vegetation. This technique also constitutes an important component of well-known bio-inventory protocols. The traditional beating sheet is compared here to a modified version that uses a funnel-like pyramidal collecting apparatus which includes a sieving mesh. With identical sampling efforts, the pyramidal beating sheet collects more species (p=0.05) and performs better for smaller taxa (p=0.0106). The pyramidal beating sheet provides a more accurate assessment of spider assemblages associated with the vegetation and presents appreciable methodological options and advantages such as subsequent extraction.
Keywords. sampling protocol, spider fauna, bio-inventories, quantitative sampling.

Résumé. Le battoir est un instrument d'échantillonnage couramment utilisée pour la récolte des invertébrés associés à la végétation. Cette technique est également une composante importante de protocoles de bio-inventaires bien connus. La technique traditionnelle du battoir est comparée avec une variante de la méthode qui utilise plutôt une structure pyramidale munie d'un grillage, formant un entonnoir. Avec un effort identique, le battoir pyramidal récolte un plus grand nombre d'espèces (p=0.05) et performe mieux pour les petits taxons (p=0.0106). Le battoir pyramidal permet une évaluation plus juste des assemblages d'espèces d'araignées associées à la végétation que le battoir traditionnel. Cette variante de la technique traditionnelle permet aussi quelques avantages méthodologiques appréciables, comme l'extraction subséquente.
Mots clés. protocole d'échantillonnage, aranéofaune, bio-inventaire, échantillonnage quantitatif.


The need for data in areas that are significant for conservation has become increasingly in demand in the last decades. Species occurrence data is being used in many parks, reserves, islands, specialized habitats and other geographically defined areas to assess conservation priorities based on distinctive fauna, rare species and endemism (Mittermeier et al. 1998, 2003, Buhlmann 2001, Skelley & Kovarick 2001, Rosenzweig et al. 2003). The rarefaction of natural and undisturbed habitats – particularly pristine forests – also justifies the urgency to conduct inventories as many taxa are thought to face extinction before being described (Dobyns 1997), especially in the tropics (Mittermeier et al. 2003, Ronquist & Gardenfors 2003). Even in areas that have been previously studied, thorough inventories usually reveal unsuspected occurrences, range extensions and undescribed taxa (Paquin & LeSage 2001, Dupérré & Paquin 2007, Lopardo et al. 2008, Paquin et al. 2008a).

In order to conduct invertebrate inventories [biodiversity assessment, structured inventories, bio-inventories, bio-surveys, faunal surveys], sampling protocols have been developed to maximize efficiency and provide samples suitable for the statistical treatment of the collections (Coddington et al. 1991, Colwell & Coddington 1994, Longino & Colwell 1997, Toti et al. 2000). Such protocols use a combination of techniques that each aim at a particular portion of the habitat (vegetation / ground, below the knees / above the knees, etc.) to ensure an adequate effort (Coddington et al. 1991, Marshall et al. 1994, Longino & Colwell 1997). The overall performance of these protocols however relies on the individual efficiency of the techniques that are used.

Among collecting techniques, beating sheets have been extensively used for arthropods associated with the vegetation (e.g. Coddington et al. 1991, Cardoso et al. 2008). The technique consists of beating bushes, scrubs, accessible tree branches, etc., with a stick over a sheet extended by a rigid structure (see Martin 1977). Organisms detached from their substrate are seen on the sheet and collected by the user. Despite widespread use, little has been done to improve its efficiency or adapt the technique to extensive sampling programs. A modification of the traditional beating sheet has been developed by Argiope Entomological Supplies and is currently available through Scienceinfuse inc., model BAT-105. It consists of an inverted pyramid collecting apparatus allowing the attachment of a collecting jar (fig. 1). A field test with this pyramidal beating sheet suggested a much better efficiency when compared with the traditional version of the method.

In this paper, we compare the traditional beating sheet with the modified pyramidal version. A better performance of the pyramidal version would suggest an easy way to improve efficiency and accuracy of invertebrate inventories. Spiders are selected here as test organisms for four reasons: they are abundant on the vegetation, many species can be collected in a short period of time, they represent a wide range of sizes from small (~1 mm) to fairly large (~20 mm), and they can be confidently identified to species.


Description of beating devices and techniques

The traditional beating sheet used here consisted of a 90 cm2 white surface made of resistant material held in tension by a wood cross. The pyramidal beating sheet had a comparable collecting surface of 90 cm2, but formed a pyramidal-shaped funnel that concentrated collected material into a jar screwed at the bottom of the device (Fig. 1). The funnel contained a metal meshed sieve (mesh size 0.63 cm2 [1/4 inch]) attached inside approximately a quarter of the way up from the bottom part. For the experiment, we used a beating stick, an aspirator and large vials containing alcohol with both methods.

Figure 1. The pyramidal beating sheet. A metal meshed sieve is attached inside the funnel. Finer debris and specimens are collected in a dry jar.

A sample is defined by a full hour spent collecting with either one of the techniques. With the traditional technique, the spiders seen on the sheet are collected with the aspirator or by hand when encountered, and transferred to a collecting vial. With the pyramidal sheet, vegetation debris accumulates on the mesh located inside the funnel, and movements done by the user ensure that finer particles and organisms go through the mesh and are collected in the dry jar. After a short period of time, the vegetation accumulated on the mesh becomes an obstacle; it is rapidly inspected by eye to locate potential spiders too big to go through the mesh (which are transferred to alcohol), and the debris discarded. With this method, we used 50 minutes of beating, and a final ten minutes to gather the spiders by examining the content of the jar poured on the side of the beating sheet. Live spiders are easily seen by their movements, collected with an aspirator, and transferred into the vial. This procedure results in a full hour of work, which is comparable to the time spent with the traditional beating sheet to provide one sample.

Study area

In order to compare the two techniques in different conditions, we sampled four habitats (maple forest, mixed forest, regenerating deciduous forest, and open field) that are representative of the Yamaska National Park (45.4ºN; 72.6ºW, Québec, Canada, see Paquin et al. 2008a). We sampled three times during the season (end of May, mid-July and mid-October), which, at these latitudes, includes all the snow-free periods and the range of conditions observed within a growing season. The experiment gathered a total of 24 samples in which each sample represented one hour of beating.


The collections were sorted under a stereoscope and only adult spiders were retained. All spiders were identified to species using Paquin & Dupérré (2003) and subsequent additions (Paquin & Dupérré 2006, Dupérré & Paquin 2007, Paquin et al. 2008b).

Statistical treatment

In order to compare species richness, we used EstimateS 8.0.0 (Colwell 2004) and a total of 500 randomizations with the [RareInfreqCut] parameter adjusted to 10, as suggested by the author. For each number of samples, two values were retained: observed richness (Obs), and predicted richness (ACE, Abundance Coverage Estimation). The values obtained for the traditional beating sheet and the pyramidal beating sheet (and their standard deviation generated by the analysis) were compared using a t-test (Scherrer 1984). We also compared the size of the spiders collected by the two techniques. Using the size values of each species provided by Paquin & Dupérré (2003), we compiled comparative lists of sizes of spiders collected by each technique. The average size was statistically compared in Statview 4.5, using U (Mann-Whitney test), a non-parametric mean test that accounts for a different number of specimens collected with each technique (Scherrer 1984).

Figure 2. Rarefaction analysis comparing the two versions of the beating sheet sampling method (t-test, p= 0.05). A) Observed richness (Obs). B) Predicted richness (ACE). The number of species collected is significantly higher for the pyramidal beating sheet with 3 samples and more, using observed richness (Obs) and 2 samples or more with predicted richness (ACE).

Figure 3. U test (Mann-Whitney) comparing the size of the spiders collected with A: traditional beating sheet, B: pyramidal beating sheet. Test significant, p= 0.0106.


In total, both methods yielded 79 species represented by 873 adult spiders (table 1). The traditional beating sheet yielded 51 species (286 specimens), and the pyramidal beating sheet 72 species (587 specimens). The comparison of the amount of species collected by the two methods carried out by a rarefaction analysis showed that the number of species collected by the pyramidal sheet was greater (t-test, p=0.05) after three samples for the observed richness (Obs, fig. 2a) and two samples for the predicted richness (ACE, fig. 2b). The average size of the spiders collected by the pyramidal beating sheet was significantly smaller (p=0.0106) than the average size of those collected by the traditional beating sheet (fig. 3).

Table 1. Species list and abundance for the two versions of the beating sheet method : the traditional beating sheet and the pyramidal version.


With an identical sampling time and effort, the pyramidal beating sheet collects a significantly higher number of species after only a few samples. The significant difference in the mean size of the spiders between the two methods also demonstrates that the pyramidal beating sheet accounts better for smaller species. The time spent in locating and transferring the spiders one at the time with the traditional technique results in a loss of efficiency when compared with continuous beating and the sorting and transfer in a single and final operation. The pyramidal beating sheet technique provides a better assessment of spider assemblages associated with the vegetation.

The pyramidal beating sheet is slightly heavier than the traditional sheet, which may be a factor to consider for intensive sampling programs. This disadvantage is however compensated by an improved stability that facilitates sampling in windy conditions.

The pyramidal beating sheet also provides other technical advantages. For instance, the jars that can be attached to the pyramidal beating sheet may vary from 500 ml, 1L or 4L and used to quantify the sampling effort. This quantification may be a suitable measurement within a given habitat, but not between different habitat types because the volume of debris (+ organisms) accumulated depends on the nature of the vegetation sampled. A quantification that uses time as a measurement unit is more appropriate for surveys that include different habitat types. For habitats that produce large amounts of debris such as conifer forests, one hour of beating will fill more than one 500 ml jar that can be pooled subsequently to constitute one sample.

Another important advantage of the pyramidal version is that the sorting can be accomplished in several ways. The quickest option is to sort the spiders using detection based on movement. A sample spread on the side of the beating sheet for ten minutes is sufficient for spiders to start moving, which facilitates their detection even for minute species. However, the content of the jar could also be extracted with devices such as Berlese funnels (Southwood 1978), photoecclectors (Masner & Gibson 1979) or a combination of these techniques. Samples may also be temporarily stored in a freezer, or filled with alcohol to be sorted later. However, because of the significant amount of debris, the last options are more laborious and time consuming than using movement as a detection trigger or extraction devices of live specimens.


We would like to thank Alain Mochon for his support and interest, Don Buckle, Gilles Arbour and Tiziano Hurni-Cranston for their comments on an earlier version of the manuscript. We also thank J. Starrett, M. Hedin and S. Lew for their time in an early version of this experiment that allowed us to adjust the effort needed to properly evaluate the pyramidal beating sheet.

Literature cited

Buhlmann KA. 2001. A biological inventory of eight caves in northwestern Georgia with conservation implications. Journal of Cave and Karst Studies 63(3):91–98.

Cardoso P, Scharff N, Gaspar C, Henriques SS, Carvalho R, Castro PS, Schmidt JB, Silva I, Szuts T, de Castro A, Schmidt JB, Silva I, Szuts T, de Castro A, Crespo LC. 2008. Rapid biodiversity assessment of spiders (Araneae) using semi-qualitative sampling: a case study in a Mediterranean forest. Insect Conservation and Diversity 1:1–12.

Coddington JA, Griswold CE, Silva Dávila D, Peñaranda E, Larcher SF. 1991. Designing and testing sampling protocols to estimate biodiversity in tropical ecosystems. Pages 44–60 in Dudley EC (editor), The Unity of Evolutionary Biology: Proceedings of the Fourth International Congress of Systematic and Evolutionary Biology, Volume 1. Dioscorides Press, Portland, Oregon.

Colwell RK 2004. EstimateS: Statistical estimation of species richness and shared species from samples. Version 8. Available from persistent URL: http://purloclcorg/estimates.

Colwell RK, Coddington JA. 1994. Estimating terrestrial biodiversity through extrapolation. Philosophical Transactions Royal Society of London Series B Biological Series 345: 101–118.

Dobyns JR. 1997. Effects of sampling intensity on the collection of spider (araneae) species and the estimation of spider richness. Environmental Entomology 26(2): 150–162.

Dupérré N, Paquin P. 2007. Description of five new spiders from Canada (Araneae: Linyphiidae). Zootaxa 1632:1–20.

Lopardo L, Dupérré N, Paquin P. 2008. Expanding horizons... The first report of the genus Mysmena (Araneae, Mysmenidae) from continental North America, with the description of a new species. Zootaxa 1718:36–44.

Longino JT, Colwell RK. 1997. Biodiversity assessment using structured inventory: capturing the ant fauna of a tropical forest. Ecological Applications 7(4):1263–1277.

Marshall SA, Anderson RS, Roughley RE, Behan-Pelletier V, Danks HV. 1994. Terrestrial arthropod biodiversity: planning a study and recommended sampling technique. Bulletin of the Entomological Society of Canada, Supplement 26:1–33.

Martin JEH. 1977. Collecting, Preparing, and Preserving Insects, Mites, and Spiders. Biosystematics Research Institute, Ottawa, Ontario. Publication 1643. 182 pages.

Masner L, Gibson GAP. 1979 The separation bag - a new device to aid in collecting insects. The Canadian Entomologist 111:1197–1198.

Mittermeier RA, Mittermeier CG, Brooks TM, Pilgrim JD, Konstant, WR, da Fonseca GAB, Kormos C. 2003. Wilderness and biodiversity conservation. Proceedings of the National Academy of Sciences 100:10309–10313.

Mittermeier RA, Myers N, Thomsen JB, da Fonseca GAB, Olibieri S. 1998. Biodiversity and major tropical wilderness areas: approaches to setting conservation priorities. Conservation Biology 12(3):516–520.

Paquin P, Dupérré N. 2003. Guide d'identification des araignées (Araneae) du Québec. Fabreries, Supplément 11. 251 pages.

Paquin P, Dupérré N. 2006. The spiders of Québec: update, additions and corrections. Zootaxa 1133:1–37.

Paquin P, Dupérré N, Mochon A. 2008a. Diversité et liste annotée des araignées (Araneae) du Parc National de la Yamaska (Québec, Canada). Le Naturaliste canadien 132(2):14–29.

Paquin P, Dupérré N, Mochon A, Larrivée M, Simard C. 2008b. Additions to the spider fauna of Québec and Canada (Araneae). Journal of the Entomological Society of Ontario 139:27–39.

Paquin P, LeSage L. 2001 ["2000"]. Diversité et biogéographie des araignées (Araneae) du Parc de Conservation de la Gaspésie, Québec. Proceedings of the Entomological Society of Ontario 131:67–111.

Ronquist F, Gardenfors U. 2003. Taxonomy and biodiversity inventories: time to deliver. Trends in Ecology and Evolution 18:269–270.

Rosenzweig M., Turner WR, Cox JG, Ricketts TH. 2003. Estimating diversity in unsampled habitats of a biogeographical province. Conservation Biology 17:864–874.

Scherrer B. 1984. Biostatistics. Gaétan Morin (éditeur), Chicoutimi, Québec. 847 pages.

Skelley PE, Kovarik PW. 2001. Insect surveys in the southeast: investigating a relictual entomofauna. The Florida Entomologist 84(4):552–555.

Statview 4.5. PPC, S.V. Abacus Concepts Inc. 1992–1995.

Southwood TRE. 1978. Ecological methods. Second edition: Chapman & Hall, University Press, Cambridge. 524 pages.

Toti DS, Coyle FA, Miller, JA 2000. A structured inventory of Appalachian grass bald and heath bald spider assemblages and a test of species richness estimator performance. Journal of Arachnology 28:329–345.

 dans la catégorie